FOURTH EDITION

BIOLOGICAL
PERFORMANCE
of
MATERIALS
Fundamentals of
Biocompatibility
FOURTH EDITION

BIOLOGICAL
PERFORMANCE
of
MATERIALS
Fundamentals of
Biocompatibility

Jonathan Black

Boca Raton London New York

A CRC title, part of the Taylor & Francis imprint, a member of the
Taylor & Francis Group, the academic division of T&F Informa plc.
Copyright Jonathan Black.

CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742

© 2006 by Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Version Date: 20131106

International Standard Book Number-13: 978-1-4200-5784-3 (eBook - PDF)

This book contains information obtained from authentic and highly regarded sources. Reasonable efforts
have been made to publish reliable data and information, but the author and publisher cannot assume
responsibility for the validity of all materials or the consequences of their use. The authors and publishers
have attempted to trace the copyright holders of all material reproduced in this publication and apologize to
copyright holders if permission to publish in this form has not been obtained. If any copyright material has
not been acknowledged please write and let us know so we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmit-
ted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented,
including photocopying, microfilming, and recording, or in any information storage or retrieval system,
without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www.copyright.
com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood
Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and
registration for a variety of users. For organizations that have been granted a photocopy license by the CCC,
a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used
only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com

and the CRC Press Web site at

http://www.crcpress.com
Preface

Biocompatibility of materials increasingly occupies the consciousness of

engineers dealing with medical and biological problems. The engineer has
long been accustomed to dealing with materials, limits on design. These
limits, such as yield stress, endurance limit, and rupture life, are reflected in
design margins tailored to the criticality of the specific application. In situ-
ations involving biological interactions as a portion of the design problem,
the additional materials limit of biocompatibility must be considered.
Failure of compatibility (that is, incompatibility) is proving to be the ulti-
mate limit to the engineering solution of many biomedical problems. As a
result, it is necessary to incorporate a thorough grounding in the aspects of
biocompatibility — or, as I prefer to term it more generally, biological per-
formance of materials — into the training of bioengineers.
When this book was first conceived, in the early 1970s, no suitable text-
books dealing with broad aspects of biomedical materials or, as the field
rapidly came to be called, biomaterials, were available. Today, as this field
has matured into biomaterials science and engineering (BSE), many edited
collections and topical monographs are available for students and workers
at many different levels. However, none seems to suit the neophyte: the
former are invariably written by a panel of experts and thus tend to be
uneven in attempting to be comprehensive and the latter are the work of a
single investigator or research group focusing on relatively narrow and
parochial interests. Both of these types of books have a place and many are
extremely valuable to the advanced worker, but they all fail to meet the
needs of the student or the professional without a background in the field.
Thus, it appears that the current work is still needed; it focuses primarily
on principles of biological performance at a relatively fundamental level:
interactions between living and nonliving materials whose consideration
sets BSE apart as a distinct field of investigation and knowledge.
Biological Performance of Materials: Fundamentals of Biocompatibility was orig-
inally intended for use as an undergraduate text for a one-term, jun-
ior–senior-level bioengineering course on biological performance. I and
others have used it in this role. However, with the assignment of selected
articles as reading and study sources, it has also proven useful as the central
text in undergraduate survey courses on biomaterials and on artificial
organs. With additional reading material from the scientific and clinical
literature and from materials science texts, it has also been used as the focus
of a first-year graduate course in biomaterials for students with engineering
(but not biological or medical) backgrounds and, conversely, as a supple-
mentary text for courses on implants for nursing students with little or no
engineering training. Finally, engineers working in medical device develop-
ment and evaluation in industrial as well as governmental settings have
found it a useful reference book. The scarcity of reference to actual materials
and specific applications has apparently made this diversity of use possible;
this revision attempts to maintain the versatility of the work. Primary train-
ing in materials science and biology is useful, but not totally essential,
because this book is intended for use in conjunction with undergraduate
texts in materials science and biology, as needed, so as to accommodate
variations in individual degrees of preparation.
We begin with an examination of the concept of “biocompatibility” and
arguments for the broader concept of biological performance. Two major
sections are devoted to the effect of biological systems on materials (“bio-
degradation” = material response) and of materials on biological systems
(“biocompatibility” = host response), respectively. Selected additional read-
ings are provided at the end of each chapter.
The reader will note an emphasis on methods for determination of biolog-
ical performance, throughout and especially in Chapter 17 and Chapter 18.
This reflects the centrality of material and host response in the clinical per-
formance of medical devices and surgical implants as well as the continued
need to select new and modified materials for specific applications. These
questions become even more challenging and complex as increasing num-
bers of viable and nonviable untraditional materials come under consider-
ation for clinical use. The practicing engineer will find this book a useful
source of references, test methods, and approaches to the problem of estab-
lishing biological performance of materials. Generic materials properties are
tabulated in Interpart 1; Interpart 2 is an example of diagnostic approaches
to detection of clinical issues associated with biomaterials in animal models
and in patients. The final four chapters deal with design, qualification, stan-
dardization, and regulation of implant materials and will be of special assis-
tance to the professional. In response to comments on earlier editions, an
extensive glossary is also included.
Due to the fundamental nature of this examination, I have elected in this
revision to retain many earlier examples and studies, providing updated
material and more current references only when needed. The reader is
advised to make use of the online resources of the National Library of
Medicine (PubMed*) to provide additional, more recent, and more special-
ized information.
I wish to thank the many undergraduate and graduate students and col-
leagues whose ideas, questions, and discussions have contributed signifi-
cantly to the scope and content of this work. Special thanks are due to G.K.
Smith and J.L. Woodman for their seminal contributions to Chapter 14 and
Chapter 15. An appeal for corrections and suggestions was issued to readers
of two listserves (BIOMAT-L and BIOMCH-L) and considerable useful feed-
back was received.

* http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed.
 In the preface to the second edition (1992), I suggested that inappropriate
host response to implants and premature device failure secondary to mate-
rials degradation continue to impose unwanted limits on engineering solu-
tions to biological and medical problems. This is still the case today. It can
only be hoped that ideas and information contained in this revised work
will contribute to the further improvement of biomaterials in their applica-
tion to the alleviation of human disability, disorder, and disease.

Jonathan Black
Abstract

Biological Performance of Materials: Fundamentals of Biocompatibility presents

an organized approach to examining and understanding the interactions
between materials used in medical devices and implants and living organ-
isms. After an introductory section addressing definitions and aspects of
biological environments, the work is divided into three principal sections.
These deal with material response to biological systems, host response to
biomaterials, and test methods for determining biological response in vitro
as well as in animal models and clinical settings. Interparts provide summa-
ries of physical properties of commonly used metallic, polymeric, and
ceramic biomaterials as well as a guide to understanding clinical perfor-
mance of implanted biomaterials. In addition to numerous references to the
literature, each chapter includes an additional bibliography; an extensive
glossary completes the work. Now in its fourth edition, this work draws on
Black’s more than 35 years experience as a teacher, researcher, and consultant
in biomaterials science and engineering.
The Author

Jonathan Black is professor emeritus of bioengineering at Clemson Univer-

sity in Clemson, South Carolina. He holds degrees in physics (Cornell Uni-
versity), engineering science (Pennsylvania State University), and
metallurgy (biomaterials) (University of Pennsylvania). Before his appoint-
ment as the first occupant of the Hunter Chair of Bioengineering at Clemson
in 1988, he was a member of the Department of Orthopaedic Surgery at the
University of Pennsylvania for 17 years with a secondary appointment in
the Department of Bioengineering. From 1992 to 1995, he was a senior vis-
iting fellow in the IRC for biomaterials at Queen Mary and Westfield College
(London), with support from an SERC fellowship.
Black has been active in research and teaching in several areas of bioma-
terials, with special reference to the biological performance of metallic
implants and to the needs of orthopaedic clinical practice. He is the author
of many articles and several textbooks, including Biological Performance of
Materials (1981, 1992, 1999, 2005), Orthopaedic Biomaterials in Research and
Practice (1988), and, with G. Hastings, Handbook of Biomaterial Properties
(1998). He has a long-term interest in implant retrieval and analysis and is
the author of a major 1992–1993 study of the field for the USFDA.
Black has been involved in professional activities in biomaterials for more
than 30 years and is a charter fellow of biomaterials science and engineering
(FBSE). He is a charter member and past president of the Society for Biom-
aterials (U.S.) and has been a frequent presenter and session chair at the
Gordon Research Conferences on Biomaterials and an organizer of the tri-
ennial Biointeractions conference series in the United Kingdom. He has
served on a number of advisory and editorial boards and was an assistant
editor of the Journal of Biomedical Materials Research from 1978 to 1995. Black
is an associate member of the American Academy of Orthopaedic Surgeons
and recipient of the presidential gold medal from the British Orthopaedic
Association.
In 1992, Black established and served as principal of IMN Biomaterials, a
professional consultancy in biomaterials and orthopaedic engineering. He
concentrated his efforts in this area after retirement from Clemson in 1993
and closed this enterprise at the end of 1998. He continues to chair the
Scientific Advisory Board for Stryker Orthopaedics.
Contents

Part I General Considerations

Chapter 1 Biocompatibility: Definitions and Issues ....................... 3

1.1 Introduction ....................................................................................................3
1.2 Biological Performance .................................................................................5
1.3 Consensus Definitions ..................................................................................6
1.4 Discussion .......................................................................................................7
1.5 The Discipline of Biomaterials ..................................................................10
1.6 Afterword: Paradigmatic Shift ..................................................................12
References ............................................................................................... 14
Bibliography ........................................................................................... 15

Chapter 2 Introduction to the Biological Environment ................ 17

2.1 General Considerations ..............................................................................17
2.2 Comparison of External and Internal Conditions..................................17
2.3 Problems in Definition of the Biological Environment .........................18
2.4 Elements of the Biological Environment .................................................20
2.5 Implant Life History ...................................................................................22
2.6 Preimplantation Handling Effects ............................................................28
References ............................................................................................... 29
Bibliography ........................................................................................... 30

Part II Material Response: Function and

Degradation of Materials In Vivo

Chapter 3 Swelling and Leaching ................................................... 35

3.1 Introduction ..................................................................................................35
3.2 Fick’s Laws of Diffusion.............................................................................35
3.3 Absorption ....................................................................................................36
3.4 Examples of Undesirable Absorption ......................................................38
3.5 Osmotic Equilibrium...................................................................................42
3.6 Leaching ........................................................................................................43
3.7 Example of Planned Leaching: Drug Release.........................................44
3.8 Effects of Swelling and Leaching..............................................................46
References ............................................................................................... 46
Bibliography ........................................................................................... 47
Chapter 4 Corrosion and Dissolution ............................................. 49
4.1 Chemistry of Corrosion ..............................................................................49
4.2 Classification of Reactions..........................................................................50
4.3 The Pourbaix Diagram................................................................................51
4.4 The Electrochemical Series.........................................................................54
4.5 Corrosion Rate..............................................................................................56
4.6 Potential-Current Relationships in Corrosion ........................................57
4.7 Forms of Corrosion .....................................................................................58
4.8 Corrosion in Implant Applications...........................................................64
4.9 Engineering Variables Affecting Corrosion Rates ..................................66
4.10 Corrosion Factors Peculiar to Biological Environments .......................67
4.11 Ceramic Dissolution ....................................................................................68
4.12 Polymer Dissolution....................................................................................69
4.13 Final Remarks...............................................................................................70
References ............................................................................................... 70
Bibliography ........................................................................................... 71

Chapter 5 Reactions of Biological Molecules with

Biomaterial Surfaces..................................................................... 73
5.1 Introduction ..................................................................................................73
5.2 Denaturation.................................................................................................74
5.3 Organometallic Compounds......................................................................74
5.4 Mechanical Aspects of Interfaces ..............................................................77
5.5 Results of Interfacial Adhesion of Molecules .........................................80
5.6 Effects of Charged Interfaces and Ions ....................................................82
5.7 Final Comments ...........................................................................................83
References ............................................................................................... 84
Bibliography ........................................................................................... 84

Chapter 6 Mechanics of Materials: Deformation and Failure ..... 87

6.1 Introduction ..................................................................................................87
6.2 Mechanics of Materials ...............................................................................87
6.3 Elastic Modulus ...........................................................................................90
6.4 Yield Strength ...............................................................................................96
6.5 Fracture Strength .........................................................................................97
6.6 Final Comment...........................................................................................104
References ............................................................................................. 104
Bibliography ......................................................................................... 105

Chapter 7 Friction and Wear .......................................................... 107

7.1 Introduction ................................................................................................107
7.2 Friction.........................................................................................................107
7.3 Lubrication..................................................................................................109
7.4 Wear ............................................................................................................. 114
7.5 Conclusions.................................................................................................122
References ............................................................................................. 122
Bibliography ......................................................................................... 123

Interpart 1 Implant Materials: Properties .................................... 125

I1.1 Introduction ................................................................................................125
I1.2 Metals...........................................................................................................126
I1.3 Polymers......................................................................................................129
I1.4 Ceramics......................................................................................................131
I1.5 Composites .................................................................................................132
References ............................................................................................. 134
Bibliography ......................................................................................... 135

Part III Host Response: Biological Effects of Implants

Chapter 8 The Inflammatory Process ............................................ 139

8.1 Introduction ................................................................................................139
8.2 The Inflammatory Response ....................................................................139
8.3 Infection.......................................................................................................150
8.4 Effects of Implant Degradation Products ..............................................155
8.5 A Final Comment ......................................................................................160
References ............................................................................................. 160
Bibliography ......................................................................................... 162

Chapter 9 Coagulation and Hemolysis ......................................... 165

9.1 Introduction ................................................................................................165
9.2 The Coagulation Cascade.........................................................................165
9.3 Approaches to Thromboresistant Materials Development.................170
9.4 Hemolysis ...................................................................................................176
9.5 Final Comments .........................................................................................179
References ............................................................................................. 180
Bibliography ......................................................................................... 181

Chapter 10 Adaptation .................................................................... 183

10.1 Introduction ................................................................................................183
10.2 Tissue Growth Strategies..........................................................................183
10.3 Examples of Adaptation in Implant Applications ...............................186
10.4 A Final Comment on Adaptation ...........................................................197
References ............................................................................................. 199
Bibliography ......................................................................................... 201
Chapter 11 In Vitro Tissue Growth and Replantation ................ 203
11.1 General Considerations ............................................................................203
11.2 What Is Tissue Engineering? ...................................................................204
11.3 The Cell–Receptor Paradigm ...................................................................206
11.4 Matrices and Cell Sources ........................................................................210
11.5 Thinking Twice about Tissue Engineering ............................................214
11.6 Some Final Comments ..............................................................................220
References ............................................................................................. 222
Bibliography ......................................................................................... 223

Chapter 12 Allergic Foreign Body Response ............................... 225

12.1 Specific vs. Nonspecific Response ..........................................................225
12.2 Mechanisms of Immune Response .........................................................226
12.3 Classes of Hypersensitivity Reactions ...................................................230
12.4 Hypersensitivity Reactions Associated with Implants........................230
12.5 Final Comment...........................................................................................240
References ............................................................................................. 241
Bibliography ......................................................................................... 243

Chapter 13 Chemical and Foreign-Body Carcinogenesis ........... 245

13.1 Definitions...................................................................................................245
13.2 Chemical Carcinogenesis..........................................................................246
13.3 Foreign Body Carcinogenesis ..................................................................257
13.4 Nonspecific Carcinogenesis .....................................................................262
13.5 Evidence for Implant Carcinogenesis in Humans ...............................263
References ............................................................................................. 268
Bibliography ......................................................................................... 270

Chapter 14 Mineral Metabolism.................................................... 273

14.1 Introduction ................................................................................................273
14.2 Iron Metabolism.........................................................................................274
14.3 Chromium Metabolism.............................................................................283
14.4 Human Dietary Metal Intake ..................................................................285
References ............................................................................................. 288
Bibliography ......................................................................................... 289

Chapter 15 Systemic Distribution and Excretion ........................ 291

15.1 Introduction ................................................................................................291
15.2 Movement of Solid Bodies .......................................................................291
15.3 Transport of Dissolved Species ...............................................................296
15.4 Distribution and Excretion of Dissolved Species .................................301
15.5 Final Comment........................................................................................... 311
References ............................................................................................. 312
Bibliography ......................................................................................... 315

Chapter 16 Effects of Degradation Products on Remote

Organ Function ........................................................................... 317
16.1 Introduction ................................................................................................317
16.2 Examples of Systemic Effects ..................................................................318
16.3 A Review of Systemic Aspects of Host Response................................321
16.4 A Final Comment ......................................................................................323
References ............................................................................................. 325
Bibliography ......................................................................................... 326

Interpart 2 Implant Materials: Clinical Performance ................. 327

I2.1 Introduction ................................................................................................327
I2.2 An Example: Total Hip Replacement.....................................................330
I2.3 A Final Word ..............................................................................................332
References ............................................................................................. 332
Bibliography ......................................................................................... 333

Part IV Methods of Testing for Biological Performance

Chapter 17 In Vitro Test Methods ................................................. 337

17.1 Test Strategies .............................................................................................337
17.2 In Vitro Test Types .....................................................................................338
17.3 Tissue Culture Tests...................................................................................338
17.4 Blood Contact Tests ...................................................................................347
17.5 Final Comments .........................................................................................350
References ............................................................................................. 351
Bibliography ......................................................................................... 353

Chapter 18 In Vivo Implant Models.............................................. 355

18.1 Introduction ................................................................................................355
18.2 Test Types....................................................................................................358
18.3 A Final Comment ......................................................................................370
References ............................................................................................. 370
Bibliography ......................................................................................... 373
Appendices .........................................................................................................374
Chapter 19 Clinical Testing of Implant Materials ...................... 383
19.1 Goal of Clinical Trials ...............................................................................383
19.2 Design of Clinical Trials ...........................................................................384
19.3 Conclusions from Clinical Trials .............................................................392
19.4 Aspects of the Decision for General Clinical Use ................................394
19.5 Final Comments .........................................................................................398
References ............................................................................................. 400
Bibliography ......................................................................................... 401

Chapter 20 Standardization and Regulation of

Implant Materials ....................................................................... 403
20.1 Historical Perspective ...............................................................................403
20.2 Drug Standardization Activities..............................................................404
20.3 Biomaterials Standardization Activities.................................................406
20.4 U.S. Federal Regulation of Medical Devices and Biomaterials .........415
20.5 Regulation of Materials for Implants .....................................................419
20.6 The Biomaterials Supply “Crisis” ...........................................................422
References ............................................................................................. 423
Bibliography ......................................................................................... 424

Chapter 21 Design and Selection of Implant Materials............. 427

21.1 Introduction ................................................................................................427
21.2 The Design Process....................................................................................429
21.3 The Value of Prospective Design ............................................................438
References ............................................................................................. 439
Bibliography ......................................................................................... 440

Chapter 22 Clinical Performance of Biomaterials ....................... 441

22.1 Historical Aspects ......................................................................................441
22.2 Procedures for Device Retrieval and Analysis .....................................443
22.3 Common Concerns about Device Retrieval and Analysis .................447
22.4 Proposed National Implant Data Retrieval and Analysis
Program (NIDRA) .....................................................................................450
22.5 Elements of a NIDRA System .................................................................451
22.6 Autopsy Retrieval Studies........................................................................455
22.7 Concluding Remarks.................................................................................456
References ............................................................................................. 457
Bibliography ......................................................................................... 457

Glossary ...............................................................................................................459
G.1 Introduction ................................................................................................459
G.2 Glossary.......................................................................................................460
G.3 Deprecated Terms ......................................................................................469
References ............................................................................................. 470

Index .....................................................................................................................471
 Part I

General Considerations
1
Biocompatibility: Definitions and Issues

1.1 Introduction
The issue of biocompatibility arises from recognition of the profound differ-
ences between living tissues and nonliving materials. In an historical and a
practical perspective, a wide range of interactive behavior occurs between
tissues and materials. In any of these interactions, beneficial and adverse effects
may be observed. Thus, materials considered foods and beverages can be
nutritious or non-nutritious. From another viewpoint, they can be considered
toxic or nontoxic. Such judgments are relative to use or abuse rather than to
an absolute scale. Although it is a central nervous system depressor, alcohol
has a positive virtue as a disinhibiting stimulant and social drug in small doses.
Internally, large doses are toxic, and still larger doses are lethal. However, even
in toxic doses, alcohol is a useful external antiseptic.
It is desirable to extend this sort of relativism of dose and type of use to
examination of the interactions between biomaterials and living systems.
Biomaterials are materials of natural or man-made origin that are used to
direct, supplement, or replace the functions of living tissues. When these
materials evoke a minimal biological response, they have come to be termed
biocompatible. As it is typically used, the term biocompatible is inappropri-
ate and defective of content. Compatibility is strictly the quality of harmo-
nious interaction. Thus, the label biocompatible suggests that the material
described displays universally “good” or harmonious behavior in contact
with tissue and body fluids. It is an absolute term without any referent.
Furthermore, the traditional ideas of biocompatibility refer essentially to the
effect of the material on the biological system. Effects of biological processes on
materials are rarely included in the meaning, unless the results of material
changes elicit a change in biological response. The effects of the biological
system on the material are usually lumped in the term biodegradation, which
implies “bad” behavior — again without a referent. However, in the case of a
deeply implanted suture, biodegradation may be a sought-after result.
One can protest that this is a semantic discussion without content. On the
contrary, I think that the terminology used and the assumptions inherent in

3
4 Biological Performance of Materials: Fundamentals of Biocompatibility

that terminology tend to condition the approach taken in experiment and

analysis. Thus, the most common approach to establishing the biocompati-
bility of a material is still to establish the absence of deleterious effects
associated with its use in biological applications. Once such tests are com-
pleted, the material is regarded as qualified. I believe that the absolute nature
of the language employed has led to the use of absolute, and thus inappro-
priate, criteria.
The real issues in the use of materials in medical and surgical devices are
not any more absolute than is the choice of a material for any other engi-
neering application. The choice of materials for construction of a device or
machine is made early in the design process. The properties of the candidate
materials, particularly those properties that bear on the intended function
of the complete assembly, then interact strongly with the design. The ultimate
test of the appropriateness of the choice of materials is the performance of
the completed device in its intended application. In this performance, the
interaction between design choices (shape, size, linkage, etc.) and materials
properties (strength, density, composition, etc.) can be seen. Chapter 21 deals
with some of these points more fully.
The real issue of biocompatibility is not whether there are adverse biolog-
ical reactions to a material, but whether that material performs satisfactorily
(that is, in the intended fashion) in the intended biomedical application and
thus can be considered a successful biomaterial. This goal should lead
directly to a traditional engineering design process of considering the advan-
tages and disadvantages inherent in the selection of a particular material for
a design in a specific application. Among the factors considered must be the
interaction of the material with the biological processes in its intended site
of operation.
One of the consequences of this relative approach to material performance
in contact with living tissues must also be the rejection of the idea that any
material in any selected application can be categorically safe or unsafe. In
dealing with food additives, it is possible to draw up a list of materials that
are “generally regarded as safe” (GRAS).* This results from the situation in
which each dye, sweetener, flavoring agent, etc. is serving the same function,
no matter what the apparent application, and is consistently used in a low
but relatively uniform amount or “dose.” Food is always ingested and never
implanted, and all food is subjected to the same succession of physiological
degradation, storage, and excretion processes. By contrast, a material used
successfully in one medical device, and thus considered a biomaterial, may
then be used in a different form in a different location with another intended
response, with an unsatisfactory result. Thus, it should be no surprise that
there is no GRAS list for biomaterials.**

* This register or list is maintained by the US-FDA’s Center for Food Safety and Applied Nutri-
tion (CFSAN) and, as of May 2005, contained 158 completed entries. See: http://
www.cfsan.fda.gov/~rdb/opa-gras.html.
** However, see Section 20.4.1 for an attempt to create such a list.
Biocompatibility: Definitions and Issues 5

1.2 Biological Performance

I adopted the term biological performance as a descriptor of materials in
order to replace the historical or classical idea of biocompatibility. Biological
performance and two closely related terms are defined as follows:

Biological performance: the interaction between materials and living

systems
The two aspects of this performance are:
Host response: the local and systemic response, other than the in-
tended therapeutic response, of living systems to the material
Material response: the response of the material to living systems

The generality of these terms is obvious. Their definitions have no inherent

value judgments, and they do not suggest absolute qualities.
However, these terms are not sufficient for a full discussion. I have stressed
the need for consideration of interactions between materials and living sys-
tems on a relative, rather than an absolute, basis. This suggests the need for
a system of grading or ranking based upon the results of tests. Additional
terms are needed to implement such a concept. The first three definitions
are therefore supplemented with several others:

Reference (or control) material: a material that, by standard test, has

been determined to elicit a reproducible, quantifiable host or ma-
terial response*
Level of host (or material) response: the nature of the host (or material)
response in a standard test with respect to the response obtained
with a reference material
A standard test, as referred to in these two definitions, is simply any
well-defined, repeatable test. The requirements for such tests are
discussed in Chapter 17 and Chapter18.

Finally, I suggest that the use of the term biocompatibility be retained for
historical reasons, but with a narrow and careful redefinition:

Biocompatible (-ility): biological performance in a specific application

that is judged suitable to that situation

* Note that this definition carries no implication of “good” or “bad” behavior on the part of the
reference material. A reference material might be a material with minimal host response (a neg-
ative reference) or an extreme host response (a positive reference). Reference materials may or
may not be selected from those conventionally used for implant fabrication. The use of a material
as a reference material does not qualify it as a biomaterial.
6 Biological Performance of Materials: Fundamentals of Biocompatibility

So, at the end of the consideration, when host and material responses are
known and the particular device application is examined, a final value judg-
ment can then be made that leads to the acceptance or rejection of the
material for that application. Such a selection and a resulting record of
adequate performance do not “qualify” a material. Rather, they increase the
confidence in the use of the material as a biomaterial in that particular
application and point to possible successful use in similar applications.
This last point is extremely important. The final arbiter of biocompatibility,
as defined here, can only be satisfactory clinical performance, insofar as
material properties affect the outcome of the treatment. Thus, if a biomaterial
has been in use for a long time, it makes no sense to go back to the materials
science laboratory, the tissue culture laboratory, and the animal colony in an
attempt to determine its biocompatibility. The experiment has already been
performed during human clinical use; what is required is careful observation
to determine and interpret the outcome. This point is discussed at length in
Chapter 22.

1.3 Consensus Definitions

A major attempt has been made to reach consensus concerning some defi-
nitions related to biocompatibility. A working consensus conference, spon-
sored by the European Society for Biomaterials, was held in 1986 to discuss
these matters in an international setting (Williams 1987). Of the 13 terms that
gained consensus definitions; the following are relevant to this discussion:

Biomaterial: a nonviable material used in a medical device, intended to

interact with biological systems
Host response: the reaction of a living system to the presence of a
material
Biocompatibility: the ability of a material to perform with an appropri-
ate host response in a specific situation

Although they are terse, these are not bad definitions. They preserve the
idea of interaction and of relative, rather than absolute, attributes. They are
limited definitions in that they specifically exclude living tissues from the
spectrum of biomaterials. In addition to exhibiting active physiological pro-
cesses, tissues are materials with definable physical structure and properties;
thus, their exclusion seems unwarranted and unwise. It is unfortunate that
the conferees chose to deprecate the terms biological performance and
material response because I consider them to embody concepts that lend
clarity to the discussion of biocompatibility.
Biocompatibility: Definitions and Issues 7

The success of this conference and, in particular, the widespread accep-

tance of biocompatibility as a relative rather than an absolute attribute of
biomaterials led to the convening in 1991 of a second consensus conference,
again in Chester, U.K., under the same sponsorship (Williams et al. 1992).
After intensive discussion, five additional terms related to biomaterials and
biological performance were agreed to by consensus:

Biomaterial: a material intended to interface with biological systems to

evaluate, treat, augment, or replace any tissue, organ, or function of
the body
Bioactive material: a biomaterial designed to elicit or modulate biolog-
ical activity
Bone bonding: the establishment, by physicochemical processes, of con-
tinuity between implant and bone matrix
Biodegradation: the breakdown of a material mediated by a biological
system
Inherent thrombogenicity: thrombus formation controlled by the ma-
terial surface

The definition of biomaterials clearly matured in the interval between these

two conferences. The concept of bioactivity will be discussed in the next
section in the context of interactivity. Biodegradation now was given a special
meaning: not merely the passive response of a material to the physicochem-
ical conditions found in living systems, but also involving actual cellular
effects on the pericellular environment (see Section 2.3). Finally, the com-
pound term inherent thrombogenicity represents the first small, solid
achievement in defining the very complex interactions between blood and
biomaterials (see Chapter 9 for further discussion).
However, words become part of language by the repetition of their use or
they are abandoned, in the same way that paths broken through the wild
may or may not become superhighways. Thus, it remains to be seen how
these and other terms discussed here will be conventionally understood and
used in the scientific literature.*

1.4 Discussion
The use of the definitions discussed here has gradually redirected the study
of interactions between biological systems and materials. It has moved from
efforts to obtain qualification and blanket assurances of safety to description
and grading of biological performance based upon the careful development

* See the Glossary for additional consensus definitions.

8 Biological Performance of Materials: Fundamentals of Biocompatibility

of standard tests and the characterization of reproducible response relative

to reference materials. This is a central idea that will guide considerations
throughout this work.
I feel strongly that absolute qualification of an artificial or processed
material in biological applications is not possible now or in the foreseeable
future. It is necessary to establish minimum requirements for performance
at various stages of materials development. Strategies for setting such levels
are discussed in Chapter 21.
It has been suggested that no foreign material is biocompatible (old mean-
ing) and that the best that can be expected is that the results of its use are
“physiologically tolerable.” This is a somewhat negative view that overlooks
the essentially benign responses elicited by many materials in living systems.
However, such a suggestion should attract attention to another important
point. Living systems differ most from machines in respect to the constant
flux and change of their components — that is, in their physiology. Biological
performance, particularly host response, ought not to be defined in terms of
tissue structure and pathology but primarily in terms of local, organ, and
systemic physiology. Deviations from usual physiological conditions may
lead to changes in the structure and function of living tissues.
The key to understanding host response and, to a lesser degree, material
response, is knowledge of the participation of the material in the physiology
of the host. Extrapolation of results obtained in tissue culture and animal
models rests significantly on knowledge of how normal and abnormal phys-
iological processes in these systems differ from those in humans, in health
and in disease. Thus, in addition to concentrating on making relative rather
than absolute determinations, care will also be taken to attend to physiology
(normal and abnormal) and its interspecies variations.
Osborn (1979) attempted to take such physiological considerations into
account by classifying all biomaterials as biotolerant, bioinert, or bioactive,
conveying respectively the sense of negative (but tolerable) local host
response, absence of local host response, and positive (desired) local host
response. Slutskii and Vetra (1996) revived Osborn’s approach by defining
a term, reactogenicity, as the intrinsic property (or combination of properties)
of a biomaterial that produces a certain tissue reaction. This approach essen-
tially converts Osborn’s categories into a continuous scale — presumably
running from normal to some degree of inflammation (see Chapter 8) —
because Slutskii prefers to restrict reactogenicity to the description of the
intrinsic properties of “biocompatible,” that is, successful biomaterials. Both
of these treatments mirror an earlier approach by Steinemann (1975), in
which he explicitly classified metallic implant materials, by reference to the
nature of the local encapsulization response, as toxic, scar tissue, or vital.
Steinemann’s approach is of particular interest in that he tried, somewhat
unsuccessfully, to correlate these categories with an intrinsic material prop-
erty, polarization resistance (see Section 4.6), thus presaging Slutskii and
Vetra.
Biocompatibility: Definitions and Issues 9

Collectively, these approaches provide little illumination because they fail

to look beyond the immediate biomaterial–tissue interface to the larger
requirements of specific clinical applications. However, it is possible to
extend these ideas, taking into account the development of the field of
biomaterials and eliminating the undesirable use of the prefix bio- (as used
by Osborn). Examining the historical development of biomaterials, I have
identified four phases (or types) of biomaterials, based upon changing con-
cepts of host response:

• Phase 1. Inert (biomaterials): implantable materials that elicit little

or no host response
• Phase 2. Interactive (biomaterials): implantable materials designed
to elicit specific, beneficial responses, such as ingrowth, adhesion,
etc.
• Phase 3. Viable (biomaterials)*: implantable materials, incorporating
or attracting live cells at implantation, that are treated by the host
as normal tissue matrices and are actively resorbed and/or remod-
eled
• Phase 4. Replant (biomaterials)*: implantable materials consisting of
native tissue cultured in vitro from cells obtained previously from
the specific implant patient

It is now recognized that searches for type 1** materials are as pointless
as the historical pursuit of the Philosopher’s Stone — the talisman that would
turn any base material into gold. Many biomaterials in present clinical use,
such as porous structures and bioactive coatings, as well as ones in devel-
opment are properly called type 2 materials. Type 3 materials are the subject
of active research and commercial interest, with initial examples in limited
clinical use (see Warnke et al. 2004). Advances in control and manipulation
of the genetic code in mammals suggest that no intellectual barrier exists to
prevent the broad future realization of type 4 materials at the tissue and
organ levels. In fact, a true type 4 material (implantable, live tissue with the
identical genetic code and immunological determinants of the recipient
patient) represents the ultimate fulfillment of the original search for biocom-
patibility: implantable materials demonstrating harmonious interaction.
Medical practice today utilizes great numbers of artificial devices and
implants. As long ago as 1988, by some estimates (Moss et al. 1991), as
many as one in 22 Americans had at least one implant and the proportion
clearly continues to increase. For any one application, a wide variety of

* In the 15 years since I conceived these definitions, the field has advanced rapidly. Thus, I now
implicitly include the implantation of DNA, as plasmids or in transfected cells, in the definition
of phase 3 and 4 biomaterials and similarly broaden the phase 4 definition to read “in vitro or in
vivo.”
** I have generally used the term phase to discuss the historical development and type to address
the actual biomaterials.
10 Biological Performance of Materials: Fundamentals of Biocompatibility

similar designs exist with different degrees of efficacy. In contrast to this

profusion of design, materials suitable as biomaterials continue to be
scarce. Perhaps no more than a few dozen of the millions of available
metal, polymer, and ceramic compositions have proven useful in medical
devices and implants.
The limiting factor in the use of materials as biomaterials continues to be
achieving appropriate biological performance. Better understanding of bio-
logical performance and the factors affecting it will lead to a variety of useful
new materials options. This in turn will lead to substantial expansion of the
role that artificial devices can play in the prevention and treatment of human
disability and disease. At the end of this progression, when the technology
for preparation of type 4 materials is readily, widely, and affordably avail-
able, artificial devices will be called upon to serve only as “bridges” to
replantation and the field of biomaterials will emerge in its rightful place as
one of the healing arts.

1.5 The Discipline of Biomaterials*

The concluding comments in the preceding section presuppose that a field
of study or an intellectual discipline — biomaterials — exists. Such a field
exists when a definition that includes all valid aspects and excludes all other
aspects is recognized. I define the field of biomaterials thus:

Biomaterials: the organized study of the materials properties of the

tissues and organs of living organisms; the development and char-
acterization of pharmacologically inert materials to measure, restore,
and improve function in such organisms; and the interaction be-
tween viable and nonviable materials

This definition can be seen in graphic form in Figure 1.1. The bases for the
field, its foundation disciplines, are the traditional intellectual fields of engi-
neering, medicine, and the physical and biological sciences. The body of the
field is the materials science approach to manufactured (artificial) and nat-
ural (or native) biomaterials (including tissues). The apex or distinguishing
feature of the field is the study of biological performance of materials as
defined by their host and material responses. This last area is unique to the
discipline of biomaterials and leads to a further definition:

Biomaterials science: the study and knowledge of the interaction be-

tween living and nonliving materials

* Parts of this section were previously published, in different form, in Black et al. (1992).
Biocompatibility: Definitions and Issues 11

Biological Performance
of Materials:
Host Response Inter-
Material Response actions

Structure/Composition/
Function Relationships Living
Nonliving
in Manufactured Materials
Materials
and Natural (Patient)
Materials

Foundation Engineering Medicine

Disciplines Physical Sciences Biological Sciences

FIGURE 1.1
The structure of the discipline of biomaterials.

A complementary definition is:

Biomaterials engineering: the application of the principles of biomate-

rials science and its foundation sciences to the solution of practical
problems of human health, disability, and disease

Taking into account the definitions of biomaterials as objects and as a field

of study, I advance three propositions about the nature of the intellectual
field of biomaterials:

Biomaterials is a materials science: the central issue is the dependence

of physical properties on composition and structure.
Biomaterials is an interdisciplinary science: its unique feature is consid-
eration of the interactions between living and nonliving materials.
Biomaterials is a medical science: its ultimate goal is the improvement
of human health and quality of life.

Biomaterials science and engineering are evolving fields of study, investi-

gation, and development. However, their meets and bounds are now under-
stood well enough to assert safely that they are parts of an established
discipline, the field of biomaterials.
12 Biological Performance of Materials: Fundamentals of Biocompatibility

1.6 Afterword: Paradigmatic Shift

As the field of biomaterials began to be organized as a field of research and
an academic discipline in the early 1970s,* a recognizable paradigm emerged
that dominated the efforts in the field (Figure 1.2). In this analysis, the central
issue was seen as the interaction between implant and patient at the interface:
the effect of the patient on the implant (“biodegradation” = material
response) and the converse effect of the implant on the patient (“biocompat-
ibility” = [local] host response). The principal foci of investigation of material
response were fracture, wear, corrosion, and dissolution, with lesser interests
in transport, storage, and excretion of degradation products. Host response
was viewed primarily in a local or interface context, with emphasis on
ingrowth, inflammation, fibrosis (encapsulation), and, when in chronic con-
tact with blood, coagulation and hemolysis.
As the field began to mature and interest moved away from the apparently
impossible and fruitless search for inert (type 1) materials to active pursuit
and design of bioactive (type 2: interactive) materials, some trends could be
noted:

• Biological models began to focus less on animals and patients and

more on cells and molecules.

IMPLANT PATIENT

INTERFACE

Material Degradation Local Host Response

(fracture; wear; corrosion (ingrowth; inflammation;
[transport, storage, excretion]) fibrosis; coagulation/hemolysis
[remote; systemic])

FIGURE 1.2
The old biomaterials paradigm.

* A key milestone was the organization of the Society for Biomaterials (SFB) in 1973.
Biocompatibility: Definitions and Issues 13

• Investigations and design of materials began to emphasize three-

dimensional structures over two-dimensional interfaces.
• Effects of material-living systems’ interactions were increasingly
examined in terms of systemic biology as well as local biological
response.
• Initial efforts at replacement of absent, damaged, or diseased parts
of the human body broadened into concern for diagnosis, repair,
and regeneration.
• Overall, the field moved from a more applied engineering approach
of providing pragmatic solutions to clinical problems to a concern
for more science-based fundamental understanding of clinical prob-
lems and of potential routes to their solution.

These shifts in emphasis are contained in the modern reference to the field
as “biomaterials science and engineering,” rather than as “biomaterials,” and
are reflected in a new paradigm (Figure 1.3). Here, material is replaced by
matrix and patient by cell, and a third factor, signal, is added. Each of these
factors interacts with the other two: the cells are affected by their matrix and
by bound and free signals; the signals are modified by cellular activity and
by association with matrix while the matrix can be remodeled by cells, based
upon its native structure and upon interaction with signaling molecules. The
emphasis is now larger than merely a study of the interface between body
and biomaterial, as in the old paradigm; now it is on interactions and their
local, systemic, and remote consequences.

Structure and Properties Type and Behavior

(design; synthesis; assembly; (source; culture;
cellular effects) stimulation; regulation
[genetic alteration])

MATRIX CELL

Nature and Effect SIGNAL

(production; distribution
[gene expression; novel])

INTERACTIONS

FIGURE 1.3
The emerging new biomaterials paradigm.
14 Biological Performance of Materials: Fundamentals of Biocompatibility

The foci of interest have also shifted:

• Matrix: here interest still closely resembles traditional materials sci-

entists’ concerns for the relationships between structure and prop-
erties; however, rather than mere selection and adaptation of
materials, the activities tend to involve prospective design, synthe-
sis, and assembly and investigation of cellular effects on the resulting
material.
• Cells: replacing earlier attention centered on the biological response
(of tissues) at the interface, the studies now address the cells, which
can interact with the matrix as well as local and systemic signals.
The focus is on cell source, in vitro culture, stimulation, and regula-
tion, as well as the possibility of transient or permanent genetic
alteration.
• Signals: the final member of the triad is a class of molecules of
biological and synthetic origins that can affect cellular behavior,
especially insofar as cell–matrix interactions are concerned. Here,
the focus of interest is on characterization, production, and distri-
bution of known and novel molecular sequences and molecules
and on the possibility of their production in situ through genetic
expression.

An obvious early exemplar of this new or modern paradigm is the type 3

or biohybrid biomaterial: a matrix capable of being remodeled by implanted
or attracted cells, cells cultured in vitro or recruited upon implantation, and
signals to regulate cell–matrix and construct–host interactions. This para-
digm makes it clear that the newly emerging bioengineering fields of genetic,
cellular, and tissue engineering are, in fact, not new but simply logical
extensions of the historic and maturing field of biomaterials science and
engineering.
Thus, as a basic science field, as an engineering field allied to clinical
medicine, and as a class of materials, biomaterials continues to grow and
diversify. The new paradigm questions our earlier assertions of knowledge
and suggests that biomaterials science and engineering will continue to be
an exciting and viable field for the foreseeable future.

References
Black, J., Shalaby, S.W. and LaBerge, M., Biomaterials education: an academic view-
point, J. Appl. Biomater., 3, 231, 1992.
Moss, A.J. et al., Use of selected medical device implants in the United States, 1988,
Advance Data No. 191, February 26, 1991, Centers for Disease Control, National
Center for Health Statistics, Washington, D.C.
Biocompatibility: Definitions and Issues 15

Osborn, J.F., Biomaterials and their use in implants, Schw. Mschr. Zahnheilk., 89, 1138,
1979.
Slutskii, L.I. and Vetra, J.J., Letter to the editor: biocompatibility and reactogenicity
of materials: a semantic and logical analysis of definitions and their practical
significance, Cells Mater., 6, 137, 1996.
Steinemann, S., [Corrosion, compatibility, and mechanical properties of metallic im-
plants] (Ger.), Fortschr. Kiefer Gesichtschir., 19, 50, 1975, discussed in Steinemann,
S., Introduction, in Winter, G.D., Leray, J.L., de Groot, K.(Eds), Evaluation of
Biomaterials, John Wiley & Sons, Chichester, U.K., 1, 1980.
Warnke, P.H. et al., Growth and transplantation of a custom vascularised bone graft
in a man, Lancet, 364, 766, 2004.
Williams, D.F. (Ed.), Definitions in Biomaterials: Proceedings of a Consensus Conference
of the European Society for Biomaterials, Chester, England, March 3–5, 1986. Else-
vier, Amsterdam, 1987.
Williams, D.F., Black, J. and Doherty, P.J., Second consensus conference on definitions
in biomaterials, in Doherty, P.J. et al. (Eds.), Biomaterial–Tissue Interfaces. Ad-
vances in Biomaterials Vol. 10, Elsevier, Amsterdam, 525, 1992.

Bibliography
Bush, R.B., Biomaterials: an introduction for librarians, Sci. Technol. Lib., 15(4), 3, 1996.
Feldman, D.S. et al., A biocompatibility hierarchy: justification for biomaterial en-
hanced regeneration, in Wise, D.L., Trantolo, D.J., Altobelli, D.E. et al. (Eds.),
Encyclopedic Handbook of Biomaterials and Bioengineering, Part A: Materials, Vol.
1, Marcel Dekker, New York, 223, 1995.
Peppas, N.A. and Langer, R., New challenges in biomaterials, Science, 264, 25 March
1994, 1715.
2
Introduction to the Biological Environment

2.1 General Considerations

The central idea developed in the previous chapter is that biological perfor-
mance should be defined in terms of interaction between materials and their
operational setting, the biological environment. This is not qualitatively dif-
ferent from the normal consideration given to the material aspects of per-
formance and durability during any engineering design process. However,
two relative quantitative aspects set biological performance apart and create
the need for an independent study of material and host responses:

• High demand: the biological environment, especially internal to liv-

ing systems, is a remarkably aggressive one, resembling tropical
marine conditions. It is a milieu of high chemical activity combined
with a highly variable spectrum of combined mechanical stresses.
• Invariant conditions: despite its aggressive aspects, the biological
environment displays an extraordinary quality of constancy in phys-
ical conditions and composition. Complex control systems exist to
maintain that constancy; thus, deviations from established condi-
tions attendant to the presence of materials may be expected to incite
restoring responses.

The latter portions of this work deal with many aspects of this peculiar
environment during examination of typical material and host responses. At
this point, it is advisable to examine general points that will serve as guides
for discussion.

2.2 Comparison of External and Internal Conditions

The aggressive aspects of the biological environment may be understood if
the differences between conditions external and internal to living systems

17
18 Biological Performance of Materials: Fundamentals of Biocompatibility

are examined. Externally, the familiar aspects of the physical world can be
found. Most materials are inorganic and are partially or fully oxidized.
Although physical processes are interrelated, there is an absence of active
environmental control systems. Time constants for change are long, deter-
mined by processes of chemical reaction and diffusion, and driven by sources
that supply energy primarily through radiation, conduction, and convection.
A wide variety of atomic species are present. Structure and chemical content
vary greatly and little evidence of compositional or structural optimization
can be found.
By contrast, the internal environment arises from a system in which mate-
rials (molecules and tissues) are largely organic and are partially or fully
reduced. Most changes are mediated by active, energy-requiring control
systems. In many cases, multiple parallel systems with different time con-
stants and extensive intersystem interactions control a single transformation
or process. Time constants are orders of magnitude shorter than for most
inorganic reactions due to mediation by specialized organic catalysts
(enzymes) and the derivation of energy from chemical sources through cou-
pled reactions. Although a great variety of chemical content and structure
exists, arrangements and combinations of a few elements — primarily car-
bon, oxygen, hydrogen, and nitrogen — comprise the vast majority of this
complexity. Elements that are present are generally utilized and structures
display a parsimonious efficiency, providing an overall impression of design
optimization.
Whatever one’s views on the origin and development of biological sys-
tems, one must be impressed by the complexity of these systems and their
economy of action. They obtain their objectives by excluding, through acci-
dent, design, or active process, materials that are unnecessary or harmful to
the function of their individual processes. These phenomena act to exclude
all materials other than healthy, autologous (belonging to the same organism)
tissue. Furthermore, the systems interact locally as well as on a regional and
global (whole organism) scale. Thus, a constant aspect of the biological
environment is that the introduction of a foreign material will elicit a host
response, which may have local, systemic, and remote aspects.

2.3 Problems in Definition of the Biological Environment

Until now, I have referred to “a” biological environment. However, a variety
of sets of conditions is associated with life processes, and it is difficult to
define the actual environment in which a material or device is called upon
to function. More will be said about this later. The difficulty arises from a
lack of detailed knowledge of in vivo conditions and the local variations that
can occur in the face of overall maintenance of conditions, termed homeo-
stasis, necessary for life. Also, there is ambiguity in defining the region that
Introduction to the Biological Environment 19

is coupled with an implant. Implants in isolated locations can interact with

the rest of the system through diffusion of ions and fluids, circulation of
blood, and drainage of the lymphatics. Even the definition of absolute vol-
umes of material in communication with an implant is difficult. As Chapter
15 will show, the volume of water in which an implant is immersed in the
human body may be taken as 10–15, 8.4, or 1000 l, depending upon the details
of consideration.
A last general point has to do with the maintenance of homeostasis. In a
particular location, temperature, pH, pO2, equivalent electrical potential,
hydrostatic and osmotic pressure, and tissue/fluid composition are closely
controlled. However, the observation of such active control should not lead
to unwarranted conclusions concerning its adaptability. The control systems
most in evidence are those that control for the usual situation and small
deviations. Superimposed upon these are “emergency” protection systems,
such as coagulation, inflammation, and nonspecific immune response. These
initiate programmed deviations that attempt to restore normal conditions
locally, as long as systemic integrity is maintained. Taking a control systems
viewpoint, one can foresee challenges that can overwhelm the restorative
capabilities of local and systemic control. Only challenges that occur within
the “design” spectrum of the system can be expected to elicit satisfactory
responses, except by chance. Thus, when viewing host response, it is wise
to recognize the limited environmental variations that occur in the absence
of outside intrusion. It is also prudent to consider the qualitative and quan-
titative differences between chance intrusion and deliberate functional
implantation of biomaterials.
Materials must be tested in vitro before implantation, even in animals. It
is desirable to attempt, in large or small part, to replicate the operating
environment that the material may encounter after implantation. Here it is
useful to distinguish among four classes of exposure environments:

• Physiological: chemical (inorganic) and thermal conditions con-

trolled to normative mammalian values for the intended application
• Biophysiological: physiological conditions with the addition of
appropriate types and concentrations of initially nondenatured
(active) cell products (serum proteins, enzymes, etc.)
• Biological: biophysiological conditions with the addition of appro-
priate viable, active cells
• Pericellular (circumcellular): a special case of “biological”: the con-
ditions in the immediate vicinity of appropriate, viable, active cells

I term these “classes” of environments because the exact value of parameters

within each depends upon the specific details of the location within tissue
or organ and, in the case of materials (rather than device) testing, upon the
design details and functional goals of the device.
20 Biological Performance of Materials: Fundamentals of Biocompatibility

In vitro testing is usually carried out under physiological or biophysiological

conditions only. The problems associated with in vitro testing and its compar-
ison to in vivo conditions will be discussed further in Chapter 17 and Chapter
18. In this chapter, “biological environment” is taken to mean, most generally,
the combination of conditions that an implanted material will encounter acutely
and chronically in actual service: the combination of biological and pericellular
conditions, as well as the instantaneous requirements placed upon the design
and function of the device in which it is incorporated. The combination of these
intrinsic and extrinsic environmental effects with the overall patient require-
ments during the proposed period of implantation produces what is properly
termed the implant life history: the total combination of requirements that the
biomaterial must meet to be successful in a specific application.

2.4 Elements of the Biological Environment

The human body is generally considered in terms of a standard or reference
configuration: this is the 70-kg man.* This standard has the macroscopic
parameters given in Table 2.1. Wide individual variations from these param-
eters exist. Age, type and level of activity, disease state, national origin, and
genetic factors will also affect the absolute values. Furthermore, the values
given here, as in the following tables in this chapter, are mean normative
values for a male Anglo–Saxon individual aged in his mid-30s. They repre-
sent an average expectation and do not account for any variations within
physiological limits. It is common to describe such variations under the
overall term “biological variation.” The physicochemical and mechanical
conditions encountered in the body can also be defined (Table 2.2).
When the effects of release of material from the implant into the body are
considered, it is necessary to know the starting or nominal inorganic chemical
composition of the body. Although the concentrations of major elements have
been known for some time, those of trace elements, present in very low con-
centrations, are just beginning to be appreciated fully. Table 2.3 presents nom-
inal or reference mean values; see Chapter 15 for a more detailed discussion.
Taken together, these parameters define the intrinsic physicochemical and
mechanical parameters appropriate to generic biological environments. Exact
dimensional and functional details of a particular anatomical site may be
required when designing tests for particular materials or devices. In another
area, more detailed information is desirable. Blood is a delicate and pervasive

* It is usual practice to speak of the standard man rather than the standard human. Recent com-
ments in the lay literature concerning the focus of federal funding for biomedical research have
drawn attention, once again, to important differences between men and women. Thus, the val-
ues in Table 2.1 should be taken as guidelines; other sources, such as Lentner (1981), should be
addressed for values applicable to gender-specific applications for northern hemisphere
Anglo–Saxon subjects.
Introduction to the Biological Environment 21

TABLE 2.1
Macroscopic Parameters of the Reference Humana
Weight: 70 kg Height (medium frame): 1.80 m
Surface area: 1.88 m2 Volume: 0.065 m3
Composition: Density:
Water: 60% (42 l) Fat: 0.9 g/cm3
Solid: 40% (28 kg) Whole body: 1.07 g/cm3
Distribution of tissue types (as percentages of body weight)
Muscle: 43
Bone: 30
Internal organs:
Heart: 0.4
Liver: 2
Kidneys (2): 0.5
Spleen: 0.2
Lungs: 1.6
Brain: 2.3
Viscera: 5.6
Skin: 7
Blood: 7.2 (5 l)
Basal metabolic rate: 37/kcal.m2/h
a Values given for a male individual in his mid-30s.
Source: Lentner, C. (Ed.), Geigy Scientific Tables, Ciba–Geigy,
Basle, 1981.

tissue. It is essential to understand its makeup and normal values, especially

for applications involving blood contact on a chronic basis. Information describ-
ing the composition and cellular distribution of blood is given in Table 2.4.
It is difficult to obtain engineering properties of biological materials for
use in design processes. With a group of contributors, Black and Hastings
(1998) took a major step by attempting to collate reliable properties of normal
human tissues and fluids. However, effects of age and disease processes on
engineering properties of tissues are still not well reported. Reference should
be made to contemporary literature or individual experimental studies may
need to be undertaken to obtain design data for specific applications.
Beginning with the information given in Table 2.1 to Table 2.4 and from
other sources, it is possible to develop a picture of the thermal, mechanical,
and chemical environment that an implant will encounter when it is
implanted in a specific anatomical site. Some of the material responses dur-
ing implant service life will be described in Chapter 3 to Chapter 7 of this
work. This defined environment may be changed acutely and chronically
by the presence of an implant.* Chapter 8 to Chapter 15 deal with some

* Although it is possible to define four types of “biological” environments, it should be appreci-

ated that the fourth — the pericellular environment — is the least known and understood. On
this scale of consideration, it is clear that cellular events produce dynamic and continuing
changes (Konttinen et al. 2005). These are difficult to measure and to replicate in vitro. In the gen-
eral case, there can probably be no substitute for in vivo studies in intact animals to examine
host–material interactions at this level.
22 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 2.2
Physicochemical and Mechanical Conditions in Humans
Value Location
pH 1.0 Gastric contents
4.5–6.0 Urine
6.8 Intracellular
7.0 Interstitial
7.15–7.35 Blood
pO2 (mmHg) 2–40 Interstitial
12 Intramedullary
40 Venous
100 Arterial
160 Atmospheric
pCO2 (mmHg) 40 Alveolar
2 Atmospheric
Temperature (°C) 37 Normal core
20–42.5 Deviations in disease
28 Normal skin
0–45 Skin at extremities

Mechanical Stress (MPa) Tissues

0–0.4 Cancellous bone
0.08–0.1 Across aortic valve (ventricular diastole)
0.12–0.16 Across mitral valve (ventricular systole)
0–4 Cortical bone
4 Muscle (peak stress)
40 Tendon (peak stress)
80 Ligament (peak stress)

Stress Cycles
(per year) Activity
3 × 105 Peristalsis
3 × 106 Swallowing
0.5–4 × 107 Heart contraction
0.1–1 × 106 Finger joint motion
1–2 × 106 Walking

variations in the biological environment that arise, locally and systemically,

as a result of implantation — that is, the host response.

2.5 Implant Life History

The thermal, mechanical, and chemical parameters described in previous
sections are sufficient to predict, in general, the acute or instantaneous bio-
logical environment encountered by an implant. These acute values differ
little from patient to patient; differences have only small effects on acute host
and material responses. Phenotypic and genotypic biological differences that
Introduction to the Biological Environment 23

TABLE 2.3
Inorganic Composition of the Human Body
Total Body Conc.
Burden (aver.)
Basic Oxygen 43,000 g 61.4%
elementsa Carbon 16,000 g 22.9%
Hydrogen 7,000 g 10.0%
Nitrogen 1,800 g 2.6%
Total 67,800 g 96.9%

Physiological Calcium 1000 g 1.43%

elementsa Phosphorus 780 g 1.11%
Potassium 140 g 0.20%
Sulfur 140 g 0.20%
Sodium 100 g 0.14%
Chlorine 95 g 0.14%
Total 2255 g 3.22%

Trace Magnesium 19 g 271 ppm

elementsa Iron 4.2 g 61.4 ppm
Zinc 2.3 g 33 ppm
Iodine 130 mg 1.9 ppm
Copper 72 mg 1.0 ppm
Aluminum 61 mg 0.9 ppm
Vanadium 18 mg 260 ppb
Selenium <13 mg <190 ppb
Manganese 12 mg 170 ppb
Nickel 10 mg 140 ppb
Molybdenum <9.5 mg <136 ppb
Titanium 9 mg 130 ppb
Chromium <6.6 mg <94 ppb
Cobalt <1.5 mg <21 ppb
Total <25.84 g <0.37%
a Total of body burdens exceeds 70,000 g and 100% due to
variety of primary sources and experimental error in in-
dividual values.
Source: Data from Lentner, C. (Ed.), Geigy Scientific Tables,
Ciba–Geigy, Basle, 1981.

affect chronic host response to materials do exist. These may only be dis-
cernible by clinical testing of a specific patient; analyses of body fluids and
tissues are probably inadequate for a full understanding of individual dif-
ferences. It is unfortunate that technology for determination of the functional
behavior of implants and implant–patient interactions is weak compared
with that available to biological scientists for the study of natural organs in
situ.
Beyond these obvious similarities and possible individual biological dif-
ferences, the demands and expectations of individuals vary considerably. A
total hip replacement prosthesis for a 40-year-old head of a family presents
24 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 2.4
Components and Composition of Human Blood
Blood
Packed cell volume 38.5%
Serum volume 61.5%
Serum composition (mean values)
Cations mEq/l Anions mEq/l
Sodium 142 Chlorine 101
Potassium 4 Bicarbonate 27
Calcium 5 Phosphate 2
Magnesium 2 Sulfate 1
Total 153 Organic acids 6
Proteins 16
Total 153
Other elements
Iron 0.75–1.75 mg/l (ppm)
Nickel 1.0–5.0 μg/l (ppb)
Titanium 3.3 μg/l
Aluminum 2.0 μg/l
Copper 0.8–1.4 μg/l
Chromium 0.3 μg/l
Manganese 0.4–1.0 μg/l
Vanadium <0.2 μg/l
Cobalt 0.15 μg/l

Serum proteins
Total 65–80 g/ l
Distribution (%):
Albumin 61.5
Globulins (total) 34.5
α 8.2
β 10.3
δ 12.6
Fibrinogen 4.0

Cellular Distribution Blood Typical Dimension

Type Concentration (μm)
Erythrocyte 4–5.6 × 106/μl 8–9
Platelet 1.5–3 × 105/μl 2–4
Leukocyte 2.8–11.2 × 103/μl —

Leukocyte Distribution Typical Dimension

Type (%) (μm)
Neutrophils 59 10–15
Eosinophils 2.4 10–15
Basophils 0.6 10–15
Monocytes 6.5 12–20
Lymphocytes 31 7–18
Sources: Data from Lentner, C. (Ed.), Geigy Scientific Tables, Ciba–Geigy, Basle, 1981,
and author’s research.
Introduction to the Biological Environment 25

TABLE 2.5
Implant Life History
Implant: anterior cruciate ligament replacement
Type: permanent
Patient indications:
Post-traumatic replacement, age: 35–48
(est. mean life expectancy: 40 years)
pH = 7 ± 0.3
pO2 = <40 mmHg
pCO2 = <40 mmHg
25°C ≤ T ≤ 37°C
Mechanical conditionsa
Strain (range of maximum): 5–10%
Loads: (moderate activity level, including recreational jogging)
Peak Load
Activity (N) Cycles/Year Total Cycles
Stairs:
Ascending 67 4.2 × 104 1.7 × 106
Descending 133 3.5 × 104 1.4 × 106
Ramp walking:
Ascending 107 3.7 × 103 1.5 × 105
Descending 485 3.7 × 103 1.5 × 105
Sitting and rising 173 7.6 × 104 3.0 x 106
Undifferentiated <210 9.1 × 105 3.6 × 107
Level walking 210 2.5 × 106 1.0 × 108
Jogging 630 6.4 × 105 2.6 × 107
Jolting 700 1.8 × 103 7.3 × 105
Totals 4.2 × 106 2.9 × 108
a Adapted from Table III in Chen, E.H. and Black, J., J. Biomed.
Mater. Res., 14, 567, 1980.

a quite different engineering problem from such a device for an 80-year-old

nursing home resident. Accounting for these functional differences com-
pletes the description of the service environment; the full picture thus formed
is termed, as previously defined, the implant life history.
Implant life histories vary considerably from application to application
and, of necessity, involve a high degree of engineering estimation. Within
given target (patient) groups, the choice and intensity of work and leisure
activities will vary widely (e.g., Schmalzried et al. 2000). As a result, implant
life histories can only be regarded as predictive guides in the development,
evaluation, and study of implantable biomaterials.
Table 2.5 gives an example of an implant life history, in this case for a
permanent anterior cruciate ligament replacement. The supposed patient
presents an example of an individual with moderate demands: presumably
a full-time worker with evening and/or weekend physical recreational inter-
ests. If this individual were disabled in some manner or employed in a setting
with unusual physical demands, such as mining or construction, or took
part in other more demanding physical activities, such as wind surfing,
mountain climbing, or parachute jumping, these facts should be noted and
26 Biological Performance of Materials: Fundamentals of Biocompatibility

the physical consequences accounted for, in terms of different estimated peak

loads and numbers of repetitions.
Asymmetry may also be a factor. Whether it is intrinsic or acquired, “hand-
edness” is a possible source of laterality in physical properties of tissues,
such as bone (Dane et al. 2001), subject to dynamic remodeling (see Chapter
10). Less controversial and more obvious are the visible effects of repetitive
asymmetric physical activities, such as certain sports (soft- and hard-ball
pitching, golf, archery, etc.) or work tasks (using a sledge hammer, painting,
etc.). In some individual cases, laterality in tissue properties or material
response (Joshi et al. 2001) is observed without any obvious source.
Within a particular application, the selection of particular materials and
designs incorporating them has come to be termed demand matching.
Demand matching cannot account for changes in a patient’s life postimplan-
tation, but it can be used to guide selection of technologies preimplantation.
In the best of all possible words, one would try to design the longest lived,
most durable materials that evoked optimum responses. However, present
concerns about the cost of medical care and the way in which biomaterials
and biomedical devices contribute to these costs result in cost containment,
or more properly cost minimization, as a stimulus for “good enough” pro-
vided devices (and their constituent biomaterials) — that is, that will meet
the patient’s needs and expectations without excessive cost.
There is no general agreement on what patient features contribute to
demand and how these features should be balanced in deriving a predictive
formula. Demand matching tends to be individualized for the medical prac-
titioner and to depend upon subjective as well as objective measures. Of the
objective measures available, age (as measured by life expectancy) and gen-
der (as a predictor of body weight and activity level) are the most widely
accepted.
The mean U.S. life expectancy as a function of age is shown graphically
in Figure 2.1 (NCHS 2002). Life expectancy now declines nearly linearly from
early childhood (~6 months of age) to around age 50. Thereafter, however,
the curve has a positive upward bend. Thus, mean life expectancy at age 60
is 21.6 years and, by age 70, it has declined by only 7.2 years to 14.4 years.
Table 2.6 shows the clear effect of gender and national origin on life expect-
ancy. It should be remembered that these data are averages.* Therefore,
although mean life expectancy of U.S. residents at 75 is 11.3 years, only 24
of every 1000 that turn 75 will die before their next birthday.
In an example of demand matching (Black 1997), I have discussed selection
of alternate bearing technologies in implants for total hip replacement arthro-
plasty; patient age at surgery, as well as other, less well defined demand
components, has been taken into account.

* It is interesting to note that U.S. life expectancy at birth continues to increase, although it rose
by more than 50% (49.2 to 76.9 years) during the 20th century. Thus, although Figure 2.1 and
Table 2.6 reflect data only to age 100, U.S. life insurance underwriters have recently begun to use
tables that extend to 110 years of life.
Introduction to the Biological Environment 27

80

70

60

Life 50
Expectancy
(years) 40

30

20

10

0
0 10 20 30 40 50 60 70 80 90 100
Age (years)

FIGURE 2.1
Average U.S. life expectancy 2000 mean (average of male and female, all national origins).
(National Center for Health Statistics (NCHS), National Vital Statistics Reports, 51(3), 29, 2002.)

TABLE 2.6
U.S. Life Expectancy (Years)a
All Races White Black
Age Total Persons Male Female Male Female Male Female
0 76.9 74.1 79.5 74.8 80.0 68.2 74.9
20 57.8 55.2 60.3 55.7 60.7 49.9 56.3
35 43.6 41.3 45.8 41.7 46.1 36.6 42.1
50 30.0 27.9 31.8 28.2 32.0 24.2 28.9
65 17.9 16.3 19.2 16.3 19.2 14.2 17.4
70 14.4 13.0 15.5 13.0 15.5 11.7 14.1
75 11.3 10.1 12.1 10.3 12.1 9.4 11.2
80 8.6 7.6 9.1 7.6 9.1 7.3 8.6
85 6.3 5.6 6.7 5.5 6.6 5.7 6.5
90 4.7 4.1 4.8 4.0 4.7 4.5 4.8
100 2.6 2.4 2.7 2.2 2.4 2.9 2.7
a Year 2000.
Source: National Center for Health Statistics (NCHS), National Vital Statistics Reports,
51(3), 29, 2002.
28 Biological Performance of Materials: Fundamentals of Biocompatibility

2.6 Preimplantation Handling Effects

One tends to think of the biological environment as that into which the
implant passes after manufacture and storage. This assumption overlooks
two intermediate processes common to all implant applications. In the first
place, the implant may become contaminated, accidentally or as a side
effect of planned processing and handling during manufacture, storage,
and insertion. It is usually assumed that the implant surface is a pure, clean
one with the composition of the bulk material. The truth may be far dif-
ferent. Organic films introduced during manufacture or by inadvertent
handling may persist. Oxidation or other attack may occur during preop-
erative storage. Materials may be picked up from packaging used for
storage or during sterilization. Contaminants may be transferred from
surgical instruments.
For this reason, experimental studies of biological performance should
include surface characterization of actual implant specimens selected from
the full group fabricated in a particular study in the condition just before
surgical insertion. Furthermore, care should be taken when materials are
incorporated into devices for clinical trials and use, to see that the surface
conditions are the same as those found during earlier developmental studies.
Second, all implants must be cleaned and sterilized before use; the man-
ufacturer may supply some in sterile, double-wrapped packages, but others
must be sterilized in the laboratory or hospital before use. The common
forms of sterilization used in implant practice are:

• Cold solution
• Dry heat
• Moist heat (steam)
• Gas
• Gas plasma
• Gamma irradiation

Some typical sterilization parameters for each of these common methods are
listed in Table 2.7. The particular method and parameters used must be
suited to the individual implant type to provide maximum safety with
minimum cost and implant degradation. Newer methods include electron
beam irradiation and radio frequency plasma gas sterilization (Chau et al.
1996; Feldman and Hui 1997), which have the virtue of cleaning implant
surfaces as well as sterilizing them.
The process of sterilization, if overlooked, may affect perception of the
material and the host response. It is possible for some sterilization processes,
such as irradiation, to change material properties, particularly of polymers,
Introduction to the Biological Environment 29

TABLE 2.7
Methods and Typical Parameters of Sterilization
Method Temperature Time Notes
Cold solution RT 1–3 h Commercial solutions; usually include
formaldehyde or gluteraldehyde
Dry heat 160–175°C (max.) 0.5–2 h Time/temperature vary inversely
Moist heat 120–130°C (max.) 2–15 min Time/temperature vary inversely
Gas RT — 55°C 1–24 h Gas is usually ethylene oxide,
400–1200 mg/l; 48-h degassing
required
Plasma discharge 45–55°C 1–2 h RF discharge (var. frequencies) in <0.5
torr gas; hydrogen peroxide or
peracetic acid most common
Irradiation RT 2–24 h 60Co gamma irradiation, 10–40 kGy

dose; time/dose rate vary inversely

immediately (Nuutinen et al. 2002) and/or during subsequent preimplanta-

tion storage (Edidin et al. 2002). This might be interpreted, in error, as a
material response effect if it is detected after implantation, or it might pro-
duce changes in host response secondary to the changes in the materials’
properties (Stanford et al. 1994). It is also possible for traces of liquid or
gaseous sterilants to be carried into the implant site, thus modifying the host
response. Finally, sterilization of an unclean implant may render it sterile
but not clean or pyrogen free (Gorbet and Sefton 2005), thus affecting the
host response (see Section 8.2.2). Therefore, in any examination of material
and host responses to implanted materials, it is necessary to pay close atten-
tion to actual surface conditions and sterilization effects as a prologue to
exposure to the biological environment.

References
Black, J., Prospects for alternate bearing surfaces in total replacement arthroplasty of
the hip, in Performance of the Wear-Couple BIOLOX Forte in Hip Arthroplasty,
Puhl, W. (Ed.), Enke Verlag, Stuttgart, 1997, 2.
Black, J. and Hastings, G.W. (Eds.), Handbook of Biomaterial Properties, Part 1, Chapman
& Hall, London, 1998.
Chau, T.T. et al., Microwave plasmas for low-temperature dry sterilization, Bio-
materials, 17, 1273, 1996.
Chen, E.H. and Black, J., Materials design analysis of the prosthetic anterior cruciate
ligament, J. Biomed. Mater. Res., 14, 567, 1980.
Dane, S. et al., Differences between right- and left-femoral bone mineral densities in
right- and left-handed men and women, Int. J. Neurosci., 111(3–4), 187, 2001.
Edidin, A.A. et al., Accelerated aging studies of UHMWPE. I. Effect of resin, process-
ing, and radiation environment on resistance to mechanical degradation, J.
Biomed. Mater. Res., 61(2), 312, 2002.
30 Biological Performance of Materials: Fundamentals of Biocompatibility

Feldman, L.A. and Hui, H.K., Compatibility of medical devices and materials with
low-temperature hydrogen peroxide gas plasma, Med. Dev. Diagn. Ind., 19(12),
57, 1997.
Gorbet, M.B. and Sefton, M.V., Endotoxin: The uninvited guest, Biomaterials, 26, 6811,
2005.
Joshi, A., Ilchmann, T. and Markovic, L., Socket wear in bilateral simultaneous total
hip arthroplasty, J. Arthrop., 16(1), 117, 2001.
Konttinen, Y.T. et al., The microenvironment around total hip replacement prostheses,
Clin. Orthop. Rel. Res., 430, 28, 2005.
Lentner, C. (Ed.), Geigy Scientific Tables, Ciba–Geigy, Basle, 1981.
National Center for Health Statistics (NCHS), National Vital Statistics Reports, 51(3),
29, 2002.
Nuutinen, J.P. et al., Effect of gamma, ethylene oxide, electron beam, and plasma
sterilization on the behavior of SR-PLLA fibers in vitro, J. Biomater. Sci. Polym.
Ed., 13(12), 1325, 2002.
Schmalzried, T.P. et al., Wear is a function of use, not time, Clin. Orthop. Rel. Res.,
381, 36, 2000.
Stanford, C.M., Keller, J.C. and Solursh, M., Bone cell expression on titanium surfaces
is altered by sterilization treatments, J. Dent. Res., 73(5), 1061, 1994.

Bibliography
Altman, P.L. and Dittmer, D.S. (Eds.), Biological Handbooks: Blood and Other Body Fluids,
1961; Biology Data Book, 1964. Federation of American Societies for Experimental
Biology (FASEB), Bethesda, MD.
Åstrand, P.-O. et al., Textbook of Work Physiology: Physiological Bases of Exercise, 4th ed.,
Human Kinetics Pub., Champaign, IL, 2003.
Baier, R.E. et al., Radiofrequency gas plasma (glow discharge) disinfection of dental
operative instruments, including handpieces, J. Oral Implantol., 18(3), 236, 1992.
Block, S.S. (Ed.), Disinfection, Sterilization and Preservation, 5th ed., Lippincott, Will-
iams & Wilkins, Philadelphia, 2000.
Cooney, D.O., Biomedical Engineering Principles, Marcel Dekker, New York, 1976.
Ganong, W.F., Review of Medical Physiology, 21st ed., McGraw–Hill, New York, 2003.
Gaughran, E.R.L. and Kereluk, K. (Eds.), Sterilization of Medical Products, Johnson &
Johnson, New Brunswick, NJ, 1977.
Kurtz, S.M. et al., Advances in the processing, sterilization, and crosslinking of ultra-
high molecular weight polyethylene for total joint arthroplasty, Biomaterials, 20,
1659, 1999.
LeVeau, B. (Ed.), Williams and Lissner: Biomechanics of Human Motion, 2nd ed., W.B.
Saunders, Philadelphia, 1977.
Matthews, I.P., Gibson, C. and Samuel A.H., Sterilization of implantable devices, Clin.
Mater., 15, 191, 1994.
Nair, P.D., Currently practiced sterilization methods — some inadvertent consequenc-
es, J. Biomater. Appl., 10, 121, 1955.
Nordin, M., Frankel, V.H. and Frankel, V.H., Basic Biomechanics of the Skeletal System,
3rd ed. Lippincott Williams & Wilkins, Philadelphia, 2001.
Northrip, J.W., Introduction to Biomechanic Analysis of Sport, 2nd ed., Wm. C. Brown,
Dubuque, IA, 1979.
Introduction to the Biological Environment 31

Snyder, W.S. (Ed.), Report of the Task Group on Reference Man, International Commission
on Radiological Protection, No. 23, Pergamon, Oxford, 1975.
Staff, Plastics Design Library (Eds.), The Effect of Sterilization Methods on Plastics and
Elastomers. W.A. Morris, Inc. (for Plastics Design Library), Norwich, NY, 1994.
Wise, D.L. et al. (Eds.), Encyclopedic Handbook of Biomaterials and Bioengineering, Part
B: Applications. (Vols. 1, 2), Marcel Dekker, New York, 1995.
 Part II

Material Response:
Function and Degradation of
Materials In Vivo
 3
Swelling and Leaching

3.1 Introduction
The simplest form of interaction between implant materials and the biolog-
ical environment is the transfer of material across the material–tissue inter-
face in the absence of reaction. If the substance — ions or fluid — moves
from the tissue into the biomaterial, the result in a fully dense material will
be swelling due to conservation of volume. Even in the absence of fluid
uptake, the biomaterial may absorb some component or solute from the
surrounding fluid phase. If the fluid moves into the tissue, or if one compo-
nent of the biomaterial dissolves in the fluid phase of the tissue, the resulting
material porosity is said to be due to leaching. Both of these effects have
profound influences on the behavior of materials despite the absence of
externally applied mechanical stresses and obvious shape changes. Swelling
and leaching result from the process of diffusion. Before considering the
effects on materials’ properties, the fundamentals of diffusion and the dif-
fusion models appropriate to each situation will be examined.

3.2 Fick’s Laws of Diffusion

The fundamental relationship that governs diffusion in isotropic materials is

∂C
F = −D (3.1)
∂x

where
F = rate of transfer per unit area of cross section
D = diffusion constant
x = coordinate normal to cross section
C = concentration of diffusing material

35
36 Biological Performance of Materials: Fundamentals of Biocompatibility

Equation 3.1 is called Fick’s first law; diffusion that obeys this relationship
is termed Fickian or type I diffusion. In this simple case, the diffusion con-
stant, D, depends only upon the material diffusing (the solute) and the matrix
through which it moves. Thus, D is independent of concentration, position,
and time. However, D does depend upon the type of diffusion process taking
place, which includes surface, grain boundary, and volume diffusion. This
dependence is given by

D = Do e[ − Q/KT ] (3.2)

with dimensions of cm2 sec–1,

where
Q = energy of activation of the diffusion process
(Qvolume > Qgrain boundary > Qsurface [4:3:2 or 4:2:1])
Do = intrinsic diffusion constant (Do [surface] >
Do [grain boundary] > Do [volume])

Thus, it is easy to see that, within a given material system (a substance

diffusing through a biomaterial) Dsurface > Dgrain boundary > Dvolume and surface
diffusion is favored at typical biological temperatures.
When Equation 3.1 is applied to the problem of one-directional flow in an
infinite medium, a differential equation of this form can be obtained:

∂C ∂2 C
=D 2 (3.3)
∂t ∂x

Equation 3.3 is usually called Fick’s second law. The application of Equation
3.1 and Equation 3.3 to different geometries, and to the initial and boundary
conditions, of specific situations is sufficient to determine the distribution
and mass transfer rate of diffusing materials in all cases.

3.3 Absorption
The simplest case that results in swelling is that of diffusion from a fluid
with a fixed concentration, in the presence of perfect mixing, into an infinite
medium. This is the case for the early period of absorption in any geometric
arrangement; when the diffusing material is mostly near the surface, geo-
metric factors have little effect. The arrangement, initial conditions, and
change of concentration in the solid (biomaterial) phase with time are shown
in Figure 3.1.
Swelling and Leaching 37

Co

Solute
concentration
with inc. time

(fully stirred)

x=0
SOLID LIQUID

t = 0: C = 0 , x < 0 C = C o, x > = 0

FIGURE 3.1
Absorption from a liquid into a solid biomaterial.

The exact solution for the concentration at a given point as a function of

time is
1
⎛ x ⎞2
C = Co ⎜ ⎟ (3.4)
( )
⎜⎝ 2 Dt ⎟⎠

where
Co = external concentration
x = distance perpendicular to interface

⎡ 2 ⎤ ∞
2
()
erfc a = ⎢ ⎥ ∫ e − y dy (3.5)
⎣ π⎦ a

The function erfc(a) is related to the error function erf(a) and will be found
tabulated in texts on diffusion and heat transfer.
Integration of Equation 3.4 over distance for two values of time leads to
the following relationship for the total mass transfer, Mt, across the boundary:

1
⎡ Dt ⎤ 2
Mt = 2Co ⎢ ⎥ (3.6)
⎣ π ⎦

The following conclusions follow directly from Equation 3.4 and Equation
3.6:

• The distance of penetration of any given concentration (“the diffu-

sion front”) increases in proportion to the square root of time.
38 Biological Performance of Materials: Fundamentals of Biocompatibility

• The time required for any point to reach a given concentration is

proportional to the square of the distance from the surface and is
inversely proportional to the diffusion constant.
• The total amount of diffusing material entering the biomaterial
through a unit area of interface increases as the square root of time
and is proportional to the diffusion constant.

The situation shown in Figure 3.1 is correct for volume or grain bound-
ary diffusion, with appropriate values for the diffusion constant, when
the liquid phase is well mixed, as in the case of fluid or solute uptake
from arterial blood. If the liquid phase is not well mixed, as in the case
of interstitial fluid surrounding a soft tissue implant site, then the situation
is more complex. The simplest case occurs when a stagnation layer exists
at the solid surface: this may be modeled as a third phase with a “resis-
tance” to diffusion, often expressed by reducing the fluid phase concen-
tration by a multiplier, k, which is less than unity. Thus, the adjusted
concentration, Co′, is given by:

Co′ = k Co (3.7)

A second complication arises when the solid phase is able to absorb water
(or another solvent) as well as the solute under study. This produces a
steadily thickening solvated surface layer in which the solute can diffuse
more readily than in the unsolvated solid. This has the effect of increasing
the 1/2 power exponent in Equation 3.4 and Equation 3.6 to values closer
to one (and similarly of altering the preceding conclusions). Such a process
is termed non-Fickian or type II diffusion.

3.4 Examples of Undesirable Absorption

One special case will be examined: diffusional pickup by a sphere in a
medium of constant concentration. The application is the problem of the
“variant” heart valve poppet. The poppet-type of heart valve consists of a
polymeric sphere (the “poppet”) usually fabricated from a silicone elastomer,
a metal constraining cage, and valve seat with a fabric sewing ring (Figure
3.2). This was an early and usually successful design for human aortic valve
replacement; however, for a small percentage of patients, the ball jammed
in the cage or, in some cases, the ball actually escaped. Discoloration of such
recovered balls, termed “variant,” suggested that some material had been
absorbed, leading to swelling.
Swelling and Leaching 39

FIGURE 3.2
Starr–Edwards “poppet” type aortic heart valve. Note polyester fabric sewing ring, metal (co-
base alloy) cage, and seat and silicone ball (“poppet”). (Courtesy R. Baier.)

The mathematical solution for this case is reached by transforming Equa-

tion 3.3 into spherical coordinates and solving for the appropriate boundary
conditions. One finds that Mt is proportional to Dt/a2, where a is the radius
of the sphere. That is,

Mt ⎛ Dt ⎞
= k⎜ 2 ⎟ (3.8)
M∞ ⎝a ⎠

where
M∞ = maximum amount of material absorbed
k = constant

The solution suggests that simple absorption should produce a weight

gain that is linear with time. Data obtained by Carmen and Kahn (1968) from
a study of ten variant poppets are given in Table 3.1. A linear regression (r
= 0.91) yields a linear rate of weight gain of 0.27% per month in good
functional agreement with our calculation. This rate presumably equals kD/
a2 from Equation 3.8.
The absorption of materials from blood can have a variety of consequences.
McHenry et al. (1970) studied poppets retrieved from five patients with
clinical symptoms of aortic valve malfunction and reported discoloration
(yellowing), fatty smell, and strut (cage) grooves. These poppets, which were
40 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 3.1
Weight Gain by “Variant” Poppets
Weight Gain Period of Implantation
(%) (weeks)
0.1 1
0.1 2
0.9 11
2.1 18
4.15 34
3.4 39
2.5 48
3.1 52
3.3 52
5.5 80
Source: Adapted from Carmen, R. and
Kahn, P., J. Biomed. Mater. Res., 2, 457,
1968.

implanted for longer times than those studied by Carmen and Kahn (1968),
contained up to 16% simple and complex lipids by weight. The simple lipids
(1.5 to 2% by weight) were composed of 60 to 65% cholesterol esters, 15 to
20% triglycerides, 5% fatty acids, and 10% cholesterol. The material absorbed
(in these two reports) is a portion of the “fatty” component of blood serum,
usually called lipid on a collective basis. It can be presumed that the shape
and physical property changes observed were a result of lipid absorption.
Lipids are present in other body fluids and may be absorbed by implants
that are not in the blood flow.
Swanson et al. (1973) reported on a study of 49 silicone elastomer finger
joint prostheses that were retrieved, for various reasons, after implantation
for up to 36 months in patients. All implants were examined for lipid content
by chloroform extraction, the resulting weight change was determined, and
a direct analysis of the extractant fluid was performed. Of these implants,
12 had fractured in use. A summary of the findings is given in Table 3.2. The
question was raised as to whether lipid absorption had any relationship to
the fracture in service. The authors concluded that there is no relationship
between time of implantation, or lipid content, and risk of fracture.
It is surprising in the light of the Carmen and Kahn study that no rela-
tionship was seen between lipid content and time. However, two points must
be made:

• The boundary conditions for an implant within the tissue capsule in

a finger joint are quite different from those present in flowing blood;
Co, Cs, and mass transfer are all different in this case. Thus, the previous
solution not predicting the results is not surprising.
• The maximum weight gain reported in Table 3.2 is 0.515%. This is
the value that would be expected after less than 2 months if one used
Swelling and Leaching 41

TABLE 3.2
Lipid Content of Retrieved Finger Joint Prostheses
Intact Implants Fractured Implants
Implantation Cumulative Cumulative Estimated
Duration Lipids Lipids Lipids
(months) No. (wt.%)a No. (wt.%) No. (wt.%)b
12 15 0.475 ± 0.05 8 0.515 ± 0.2 8 0.515
24 24 0.461 ± 0.04 10 0.447 ± 0.2 2 0.175
36 37 0.469 ± 0.03 12 0.443 ± 0.1 2 0.423
a Cumulative: 24 month group includes data from 12 months; 36 month group
includes data from 24 months.
b Estimated: 24 months = estimated lipid (wt.%) for implants failing between 12
and 24 months; 36 months = estimated lipid (wt.%) for implants failing between
24 and 36 months.
Source: Adapted from Swanson, A.B. et al., Orthop. Clin. N. Am., 4(4), 1097, 1973.

the rate computed from Table 3.1 and is one-tenth of the maximum
value reported for the heart valve poppets. Therefore, an alternate
explanation for the failure to observe a linear increase in weight with
time might be that the equilibrium level is far lower in the finger
joint and is reached before 12 months, the earliest summary date
given in Table 3.2.

Examination of the data in Table 3.2, with an attempt to estimate mean

lipid values of the failed implants within 12-month intervals, suggests
another possible conclusion concerning the relationship between lipid con-
tent and device fracture. It is possible that there are two populations of
failed implants: those that take up lipid with abnormal rapidity (repre-
sented by the failures in the first 12-month period) and those that take it
up quite slowly (represented by the later failures). In the first case, one
might invoke weakening due to swelling, perhaps secondary to reaction
between lipids and the implant (Pfleiderer et al. 1995); in the second,
decreased tear resistance associated with the absence of a plasticizing effect
might be suggested. (See Chapter 6 for a more complete discussion of these
phenomena.)
The effects of absorption in these series can be summarized as change
in color, change in volume (swelling), and possible change in mechanical
properties (as evidenced by the presence of strut grooves). Mechanical
property changes secondary to absorption, in the general case, can include
reduction in modulus of elasticity, increase in internal viscosity, increase
in ductility (or possible decrease in the presence of reaction with the
absorbed species), change in the coefficient of friction, and reduction of
wear resistance.
42 Biological Performance of Materials: Fundamentals of Biocompatibility

3.5 Osmotic Equilibrium

Problems arise even in the absence of obvious property changes. The increase
in volume due to absorption results in the application of a negative internal
hydrostatic stress to the material. This effect can be understood through the
following calculation. If a material is immersed in a solution so that the
solute is soluble in the liquid phase (solvent) and the solid phase (material),
there will be an initial interfacial pressure:

Π = MRT (3.9)

where
M = molarity of solute in liquid phase
R = gas constant = 8.31 × 107 erg °K–1 mol–1
T = absolute temperature

This is van’t Hoff’s law and the pressure, Π, is called the osmotic pressure.
It is an expression of the fact that the activity of the solute in the liquid phase
is greater than that in the solid phase.
Driven by the osmotic pressure, the solid phase will absorb solute. When
this happens, the solid phase expands as if subject to a negative hydrostatic
pressure. This pressure, P, is called the swelling pressure and

P=Π–p (3.10)

where p = hydrostatic restoring force of the swollen material. That is, as the
material swells, an elastic restoring force is generated through the strain of
the material that counteracts the osmotic pressure. As the material swells,
this restoring force rises, the activity of the solute in the solid phase rises,
and the swelling pressure approaches zero as equilibrium conditions are
obtained. Obviously, changes of solute concentration in the liquid phase
disturb this equilibrium and can produce further expansion or contraction
of the solid phase.
Advantage is taken of this phenomenon in the production of hydrophilic
gels for biomedical applications such as contact lenses. These materials have
low elastic moduli and high affinity for water. Because water has a molarity
of 55.6, high osmotic pressure coupled with low negative hydrostatic restor-
ing forces permit the fabrication of compositions with an equilibrium water
content approaching 98%; volume expansions may exceed 5000% (Refojo
1976). Even more modest swelling affects mechanical properties; see Section
6.3.2.
Swelling and Leaching 43

3.6 Leaching
The simplest case that results in leaching is that of removal of diffusing
material from the surface at a constant rate. This is approximately the case
of elution by moving blood; it is closely parallel to the in vitro situation of
evaporation from a surface. The arrangement, initial conditions, and change
of concentration of the solute in the solid (biomaterial) phase with time are
shown in Figure 3.3.
In most real situations, an additional condition is present. If the fluid
medium is in motion but not fully stirred (as shown), some rate of transfer
must be assumed. The simplest case is to take this rate as proportional to
the surface concentration at any moment. In particular, this rate is taken to
be linearly dependent upon the difference between a surface concentration,
Cs, and bulk concentration, Co. Thus, the boundary condition is:

∂C
−D
∂x
(
= α Co − Cs ) at x = 0 (3.11)

If h = α/D, Equation 3.10 leads to the following solution:

C − Cs ⎛ x ⎞ ⎛ ⎡⎢⎣hx+ h2Dt ⎤⎥⎦ ⎞ ⎛ x ⎞

= erfc ⎜ ⎟ ⎜ −e ⎟ erfc ⎜ + h Dt ⎟ (3.12)
Co − Cs ⎝ 2 Dt ⎠ ⎝ ⎠ ⎝ 2 Dt ⎠

For a particular value of h, Mt is proportional to (Dt)1/2. Other conditions,

such as concentration-dependent diffusion constants, surface and bulk reac-
tions, etc., can affect the former of these relationships.

Co

Solute
concentration
with inc. time
(unstirred)

x=0
SOLID LIQUID
t = 0: C = Co, x < 0 C = 0, x > = 0

FIGURE 3.3
Leaching from a solid biomaterial into a liquid.
44 Biological Performance of Materials: Fundamentals of Biocompatibility

3.7 Example of Planned Leaching: Drug Release

When drugs are administered by intramuscular injection, their release into
other compartments of the body is governed by diffusion. The effect of this
may be seen in Figure 3.4. On the left-hand portion of this figure, the saw-
toothed trace shows the effective drug concentration, at some point distant
from the point of administration, resulting from three divided doses. After
the first dose, the concentration at that point rises through the suboptimal
into the optimal therapeutic range, then begins to decline rapidly due to the
combined effects of storage in other tissues, metabolic degradation, and
excretion. The declining dashed line segment shows the continued concen-
tration decline that would occur if the second (and third) doses were not
administered. Unfortunately, the selected temporal spacing of the doses, for
their size, was too short: the concentration after the third dose is sufficient
to produce toxic side effects. The relationship between drug concentration
in the target tissue or organ and its effect is shown on the right side of Figure
3.4; this relationship is referred to as a “dose–response” curve, sometimes
called a Bertrand curve.
The curvilinear dashed line in the left-hand portion of Figure 3.4 represents
an optimum strategy: the drug concentration rises smoothly and rapidly to
an optimal level and then is maintained at constant. With care, such an effect
can be obtained by continuous intravenous injection. However, it is possible
to utilize diffusional effects to produce a device that can produce the same
results.

Toxic
Drug concentration

Optimal

Sub-optimal

Dose: 1 2 3
Time Therapeutic
effect

FIGURE 3.4
Drug release strategies. (Adapted from Chien, Y.W., Med. Prog. Technol., 15, 21, 1989.)
Swelling and Leaching 45

Air side
B2

B M M

B1
A Skin side
IDEAL PRACTICAL

FIGURE 3.5
Sustained (reservoir) drug release device.

Figure 3.5 schematically depicts (left side) an ideal implantable drug

release device that would accomplish this goal. It consists of two design
elements: a barrier membrane (B) and a drug reservoir or matrix (M). The
drug is dissolved as a solute in the filler or may be present simply as an
imiscible dispersed, but diffusable, phase. If the intrinsic diffusion properties
for the drug in question are such that the conditions

DB << DM, Dtissue (3.13)

can be maintained, then there will be a near constant release rate until Mt
becomes a significant fraction of M∞. This secondary condition can be met
by making the drug content of the matrix phase large compared with the
desired total dose (and removing the device when its function has been
served) or by repeated injection of drug into the device matrix. The latter
arrangement is termed a reservoir or sustained release device.
However, actual or practical devices involve more complexity than this
simple example. Figure 3.5 also schematically depicts (right side) a practical
design for an external skin (transdermal) release device, such as the familiar
nicotine patches for assistance in quitting smoking. (Note that the dimension
perpendicular to the skin is greatly expanded for clarity.) Fundamentally,
the same two elements are present: a barrier membrane (B1) and a matrix
with the drug (M). However, two other elements are necessary:

• B2: a protective barrier to prevent the loss of the drug or filler from
the backside or the diffusion of air or water into the device. For
cosmetic reasons, this barrier may also be colored and include print-
ing, such as a trademark or instructions for use.
• A: an adhesive layer to secure the device on the patient’s skin.
Because this layer is between the reservoir and the tissue, its diffu-
sional properties must be taken into account in the selection and
dimensioning of B1.

This is a schematic example of a device that utilizes diffusional leaching

to produce controlled drug release. Simpler designs are possible, such as
application of a drug-impregnated single- or multilayer coating to a struc-
tural implant. In addition, for complex devices such as the example given,
46 Biological Performance of Materials: Fundamentals of Biocompatibility

adjustment of geometries and the use of different forms of drugs can produce
a wide range of release rate (and resultant tissue concentration) profiles with
time. Combining the effects of absorption (which may alter diffusion rates
by changing reservoir osmotic pressure) with changes in the composition of
the solvent or the filler phase produces still more options for the drug release
device designer. Finally, the designer might consider introducing active
design elements, such as invoking sonopheresis by utilizing ultrasonic cav-
itation to alter the diffusional properties of the adjacent tissue (Mitragotri et
al. 1995) or iontopheresis by taking account of ionic charge of certain drugs
at or near neutral pHs and imposing an electrical gradient across the
device–tissue interface (Singh et al. 1994).

3.8 Effects of Swelling and Leaching

There are adverse aspects to large deformations that may be caused by
swelling and leaching. It is possible to exceed the creep stress in the material.
This will produce continual deformation and absorption, rather than the
attainment of an equilibrium solute concentration. Even if this does not
happen, swelling reduces the elastic limit of a material (that is, the available
strain to the elastic limit) and may lead to a mode of failure termed static
fatigue or “crazing,” especially in brittle materials. Crazing is the develop-
ment of microcracks that merge and can eventually result in fracture.
Leaching, the reverse of swelling, usually has a less pronounced effect on
properties. The primary undesirable aspects of unplanned leaching are the
local and systemic biological reactions to the released products. Excessive
leaching (for instance, intergranular leaching in metals) can result in a reduc-
tion in fracture strength. The defects produced by leaching can coalesce into
macroscopic voids. If these begin to constitute a significant percentage of the
volume of rigid materials, the elastic modulus will decline. The decrease is
proportional to the second power of the volume percent of voids (see Section
6.3.2).

References
Carmen, R. and Kahn, P., In vitro testing of silicone rubber heart-valve poppets for
lipid absorption, J. Biomed. Mater. Res., 2, 457, 1968.
Chien, Y.W., Rate-control drug delivery systems: controlled release vs. sustained
release, Med. Prog. Technol., 15, 21, 1989.
McHenry, M.M. et al., Critical obstruction of prosthetic heart valves due to lipid
absorption by Silastic, J. Thorac. Cardiovasc. Surg., 59, 413, 1970.
Mitragotri, S. et al., A mechanistic study of ultrasonically enhanced transdermal drug
delivery, J. Pharmaceutical Sci., 84(6), 697, 1995.
Swelling and Leaching 47

Pfleiderer, B. et al., Study of aging of silicone rubber biomaterials with NMR, J. Biomed.
Mater. Res., 29, 1129, 1995.
Refojo, M.F., Vapor pressure and swelling pressure of hydrogels, in Hydrogels for
Medical and Related Applications, Andrade, J.D. (Ed.), American Chemical Soci-
ety, Washington, D.C., 1976, 37.
Singh, P. and Maibach, H.I., Transdermal iontophoresis. Pharmacological consider-
ations, Clin. Pharmacol., 26(5), 327, 1994.
Swanson, A.B. et al., Durability of silicone implants — an in vivo study, Orthop. Clin.
N. Am., 4(4), 1097, 1973.

Bibliography
Berti, J.J. and Lipsky, J.J., Transcutaneous drug delivery: a practical review, Mayo Clin.
Proc., 70, 581, 1995.
Brophy, J.H., Rose, R.M. and Wulff, J., Thermodynamics of Structure, Vol. II of The
Structure and Properties of Materials, Wulff, J. (Ed.), John Wiley & Sons, New
York, 1964.
Crank, J., The Mathematics of Diffusion, 2nd ed., Oxford University Press, London, 1975.
Daugherty, A.L. and Mrsny, R.J., Emerging technologies that overcome biological
barriers for therapeutic protein delivery, Expert Opin. Biol. Ther., 3(7), 1071, 2003.
Edwards, D.A. and Langer, R., A linear theory of transdermal transport phenomena,
J. Pharmaceutical Sci. 83(9), 1315, 1994.
Kost, J., Biomaterials in drug delivery systems, in Encyclopedic Handbook of Biomaterials
and Bioengineering, Part A: Materials, Vol. 2, Wise, D.L. et al. (Eds.), Marcel
Dekker, New York, 1995, Chapter 34.
Purdon, C.H. et al., Penetration enhancement of transdermal delivery — current
permutations and limitations, Crit. Rev. Ther. Drug Carr. Sys., 21(2), 97, 2004.
4
Corrosion and Dissolution

4.1 Chemistry of Corrosion

Everyone has an intuitive understanding of corrosion. The purpose of this
chapter is to consider the principles of chemistry that underlie this under-
standing, to characterize and classify the types of corrosion that may occur,
and to see how these considerations apply to the use of metals as implants.
The layperson tends to equate corrosion with the “rusting” of iron. How-
ever, the term has a far broader application. The word corrosion derives from
the Latin rodere, to gnaw. The chemical processes that contribute to the
phenomenon of corrosion can be strictly termed processes of reaction and/
or dissolution in the presence of water. However, it has come to be used to
describe any chemical attack on solid materials, especially metals. In the case
of metals, reaction tends to dominate; for ceramic and polymeric biomateri-
als, dissolution dominates. The bulk of this chapter will address the corrosion
of metals; Section 4.11 and Section 4.12 deal briefly with dissolution of
ceramics and polymers, respectively.
In the case of metals, the following four generic reactions are the most usual
(note: typical valence states and changes are given; other values are possible):

Ionization: the direct formation of cations (positively charged metallic

ions) generally under acidic or reducing (oxygen-poor) conditions:

M → M+ + e–
0 +1 (= valence of metallic atom) (4.1)

Oxidation: the direct reaction of metal with gaseous or dissolved oxy-

gen, without the participation of water. In the extreme examples,
oxidation is recognized as burning:

M + O2 → MO2
0 +4 (4.2)

49
50 Biological Performance of Materials: Fundamentals of Biocompatibility

Hydroxylation: the reaction of metal with water under alkaline (basic)

or oxidizing conditions to yield a hydrated oxide or hydroxide.
Because most hydroxides are only sparingly soluble in alkaline con-
ditions, this process often leads to the formation of a passivating
film, of which more will be said later:

2M + O2(diss.) + 2H2O → 2M(OH)2

0 +2 (4.3)

Reaction: the combination of metal or metallic ions with other cations

and anions (negatively charged ions); this is often termed complex
formation:

MO22– + HCl → MOCl– + OH–

+2 +2 (4.4)

In biomaterials’ applications, the presence of specific and nonspecific organic

binding molecules results in the formation of organometallic complexes as
a general rule.
Each of these processes has the effect of decreasing the amount of pure
metal present and of producing metal-bearing ions and compounds. Con-
sideration of the effects of corrosion must take note of the attack on the
parent metallic component as well as the formation of reaction products.

4.2 Classification of Reactions

It is necessary to make some order out of these various chemical processes
that contribute to corrosion so that one can approach the prediction of cor-
rosive attack in a systematic manner. All of these and other processes
involved in corrosion can be classified by answering two questions:

• Does the reaction depend upon pH?

• Does the reaction depend upon local electrical potential?

For the purpose of discussion, the answers to these questions will be

designated with “+” or “–.” Thus, a reaction such as dissolution of gas, which
is independent of pH and potential, would be an example of a “– –” reaction.
There may be as many as several dozen possible reactions for the interaction
of an elemental metal with pure water in the presence of oxygen.
A further observation simplifies these efforts of organization. For any
particular combination of pH and potential, there will be a single dominant
reaction for a specific metal in a specific solution; that is, of all of the possible
Corrosion and Dissolution 51

corrosion
1.6
b
0.8
EM
(volts) a
0
passivation

- 0.8

- 1.6 immunity

0 4 8 12
pH

FIGURE 4.1
Pourbaix diagram for chromium in pure water.

reactions, one will be the maximum or principal contributor to the degrada-

tion of a metallic part and to the concentration of metal in solution.

4.3 The Pourbaix Diagram

4.3.1 Reactions of Chromium in Pure Water
Classification of all possible reactions between a metallic element and water
(and its constituents) by their pH and potential dependence, combined with
the determination from various types of experiments of those combinations
that favor particular reactions provides the information needed for a graphic
representation of the overall reaction system (see Figure 4.1). This is called
a pH-potential or, more usually, a Pourbaix (poor BAY) diagram, after Marcel
Pourbaix (Pourbaix 1966) who popularized its use. This diagram is for pure
chromium in pure water and summarizes the reactions among five primary
species (Cr, O2, H2, H+, and OH–).
A Pourbaix diagram has three major regions. Each of these represents
combinations of ranges of pH and potential for which a reaction or a related
group of reactions is dominant; each region corresponds to one of the three
fundamental conditions that describe the response of metals to aqueous
solutions:

• Immunity. In this region, the dominant reaction is ionization. How-

ever, throughout the region the resulting equilibrium concentration
of chromium in solution, in all ionic forms, is less than 10–6 M. This
52 Biological Performance of Materials: Fundamentals of Biocompatibility

concentration is generally taken as the threshold between corrosion

and immunity. If reaction processes yield a total equilibrium con-
centration (of all metal-bearing ions) of less than this value, it is
engineering practice to speak of the metal as immune from corrosion,
or more simply immune, under that particular set of conditions.
Because the level of 10–6 M (typically 50 ppb) greatly exceeds normal
physiological concentrations for most of the less common ions (espe-
cially those containing trace metals), it is good practice to assume
in implant applications that some metal is released for all combina-
tions of pH and potential.
• Passivation. In this region, the dominant reactions lead to the for-
mation of oxides and hydroxides. Because these products for chro-
mium are largely insoluble at a pH above 4, they cling to the interface
between the metal and the solution, reducing and eventually pre-
venting further reaction. This renders the chromium passive.
Throughout this region, the solubility of the oxides and hydroxides
is low enough that the total concentration of chromium in solution
is, as it was in the immune region, less than 10–6 M.
• Corrosion. In this region, a variety of processes can attack metallic
chromium (at low or high values of pH) or its passive coating (if
present, at intermediate values of pH). The result throughout the
region is a total equilibrium concentration of chromium in solution
that equals or exceeds 10–6 M; thus, in engineering terms, the chro-
mium is said to corrode.

Two additional features are of interest. The diagonal dotted lines in Figure
4.1 define the reactions of gaseous oxygen and hydrogen with water. The
upper line, b, is that for oxygen; the lower line, a, is that for hydrogen. Both
of these reactions are of the “++” type, so the lines slope. The region between
the lines is that in which water is stable. Above the oxygen line, b, oxygen
is released, and below the hydrogen line, a, hydrogen is released. Dominant
reactions of the “+ –” type produce vertical region boundary segments; those
of the “– +” type produce horizontal segments. Reactions of the “– –” type
do not produce regions of dominance on this type of diagram.
In biological systems that are pH controlled by buffering, the local oxygen
or hydrogen partial pressure defines an effective local potential. Thus, tissues
perfused with arterial blood and maintained at pH 7.37 have an equivalent
potential of + 0.782 V. This last fact makes it possible to apply the Pourbaix
approach to the prediction of metallic corrosion in vivo.

4.3.2 Reactions of Chromium in the Presence of Aqueous Chloride Ion

Figure 4.2 is again the Pourbaix diagram for chromium, but with two impor-
tant changes:
Corrosion and Dissolution 53

corrosion
1.6 interstitial
b saliva fluid
intracellular
0.8 fluid

passivation
EM
0 a
(volts)
gastric
fluid bile,
urine
- 0.8

immunity
- 1.6

0 4 8 12
pH

FIGURE 4.2
Pourbaix diagram for chromium in water (1 N Cl–).

• The solution is now water with 1.0 N chloride ion (Cl–) to simulate
the situation in vivo more closely. The principal effect of this addi-
tion is to shrink the passive region radically. This results from
reaction of the chloride anion with free metal ions and the passive
layer to form soluble complexes, thus raising the effective solubility
of chromium.
• The areas of pH and potential (as defined by pO2) for various body
fluids have been superimposed. Note that the areas for interstitial
and intracellular fluids actually lie closer to pH 7.0; they are plotted
as more alkaline for clarity. From this, it can be seen that pure
chromium would perform satisfactorily in neutral conditions in the
bile duct or urinary tract, but would be unsatisfactory in the stom-
ach, where pH may approach 1. General tissue applications for pure
chromium might be considered to be marginal because they lie on
the boundary between the passivation and corrosion regions.

It should be clear that a Pourbaix diagram is useful in predicting corrosion

in only a general way. The following limitations should be realized:

• As determined by local ionic conductivity and by limitations in the

diffusion of oxygen, hydroxide, and hydrogen ions, the local micro-
conditions dictate the exact equivalent potential that may be
expected. Thus, the regions shown on this diagram for different
physiological situations may be considered as reflecting average
conditions. In particular, pericellular conditions may be radically
different.
54 Biological Performance of Materials: Fundamentals of Biocompatibility

• The areas of dominance and other details of the diagrams are those
that prevail after all reactions have come to equilibrium. Reactions
may be very slow, as is the case for many involving chromium
compounds, leading to prolonged nonequilibrium conditions.
• Reactions and their kinetics depend upon the history of the metal
to some degree. Thus, a bare piece of chromium placed under con-
ditions that lie within the passive region of the Pourbaix diagram
will undergo reactions leading to formation of hydroxides; however,
if its surface is pretreated* to produce an oxidized (passive) layer,
the dominant reactions under the same set of conditions will be
hydration and partial dissolution of this surface layer. Such reactions
may be very slow; in the case of prepassivated chromium-containing
alloys, they are so slow that such passive films are termed metastable
and the materials may be used for some applications with conditions
within the corrosion region of the diagram.
• Pourbaix diagrams are available for most elemental metals in pure
water, but they do not exist for the vast majority of alloys or for
other aqueous solutions. However, because the corrosion resistance
of chromium-containing stainless steels and cobalt-base “super
alloys” depends to a great degree on the presence of chromium
hydroxide passivation films, the diagram for chromium in the pres-
ence of chloride ion (Figure 4.2) is quite useful in understanding the
chemical aspects of their material response in vivo.
• Finally, as will be discussed in Section 4.10, the presence of active
cell products in vivo may modify the rate of reactions and the nature
of their products.

4.4 The Electrochemical Series

4.4.1 Ideal Series
It is possible to plot a complete Pourbaix diagram for any real metal or alloy
in a defined solution. However, a further simplification may be made if one
is only interested in the relative corrosion resistance of metals. One can begin
by noting that the boundary between the immunity region and the corrosion
region for acid and neutral pH is horizontal. The left-hand intercept (or more
correctly, the potential for pH = 0) is a single potential value. This potential
will be different for each metal alloy.

* Passivation by acid treatment or by anodic polarization (see Section 4.7.2) is common practice
for engineering application within this pH-potential region. Many proprietary surface treat-
ments also take advantage of control of structural and compositional features of the passive
layer to gain maximum kinetic protection from dissolution.
Corrosion and Dissolution 55

TABLE 4.1
Ideal and Practical Electrochemical Series
Potential Ideal Practical
Noble or cathodic
Gold Platinum
Platinum Gold
Silver Stainless steel (passive)
Copper Titanium
E=0 -------------- Hydrogen Silver
Lead Nickel
Tin Stainless steel (unpassivated)
Nickel Copper
Cobalt Tin
Iron Lead
Chromium Cast iron
Aluminum Wrought iron
Titanium Aluminum
Magnesium Magnesium
Base or cathodic

On the left side of Table 4.1, a number of metals are ranked by this potential
in an ideal (sometimes termed absolute) electrochemical series. This electro-
chemical series is arranged with the most noble or cathodic potentials (with
respect to the H/H+ half cell reaction) at the top and increasingly base or
anodic potentials as one proceeds down the list. Note that the apparent sign
of an electrode depends upon whether it is self-polarized (as in corrosion)
or externally polarized (as in electroplating). A self-polarized anode is neg-
ative and an externally polarized one is positive and vice versa for the
cathode. The oxidation-reduction nature of the reactions is identical in both
situations: reduction takes place at the cathode and oxidation at the anode.
Despite the use of potentials obtained under acidic, oxygenated conditions,
the ideal series is a reasonable measure of the relative corrosion resistance
of metals under a variety of conditions in pure water. The higher the place
in the list (the more cathodic the potential), the more noble or corrosion
resistant the metal is. Section 4.7.2 will show that the relative position in an
electrochemical series determines which of a coupled pair of dissimilar met-
als may undergo corrosion.

4.4.2 Practical Series

In the right-most column of Table 4.1, many of the same metals, and some
alloys, are listed in a practical electrochemical series. The use of a practical
series, in this case for the exposure of these metals to seawater, begins to
take into account the situation peculiar to a specific application. Seawater
exposure, particularly in tropical climates, is the engineering condition that
most closely simulates environments encountered by implants, except for
the general lack of soluble organic species. Comparison of the practical to
56 Biological Performance of Materials: Fundamentals of Biocompatibility

the ideal series demonstrates some interesting differences related to the

differences in environment:

• The two noblest metals in both series, platinum and gold, change
their relative positions. This reflects the fact that, although neither
is strongly attacked by seawater, gold forms chlorides more readily
than does platinum.
• Titanium moves well up the list, reflecting the highly insoluble
nature of most of its compounds, particularly its TiO2 passivation
layer that forms spontaneously in air.
• Unpassivated stainless steel, a class of alloys of iron, nickel, and
chromium (as well as other minor elements), is not particularly high
on this list. The choice of stainless steels as implant materials (pri-
marily in temporary applications, such as in internal fixation devices
for fractures) depends primarily upon their mechanical properties
and machinability in the presence of an acceptable level of corrosion
resistance (when passivated before use) and moderate local host
response to its corrosion products.

A practical series begins to take factors of corrosion other than equilibrium

thermodynamics into account. That is, it reflects not only the possibility of
corrosion, but also details of actual attack in specific environments. Pourbaix
diagrams and electrochemical series tell something about the likelihood of
corrosion. They define equilibrium conditions and imply rates of corrosion
as proportional to deviations from equilibrium. It is important to know
something about the rates of corrosion. These rates will help to determine,
for instance, the rate of release of metallic ions from an implant.

4.5 Corrosion Rate

In engineering applications, corrosion rates are expressed as rates of surface
dissolution or recession per year. The common unit is the mpy (mil [0.001
in.] per year). This unit is far too big to be used to examine corrosion rates
in the biological environment for the same reasons that an equilibrium con-
centration of 10–6 M is too high to be considered indicative of immunity.
A more direct measure may be obtained by examining Equation 4.1
through Equation 4.3. These are characteristic of metallic corrosion. Note
that in each case the valence of the metal is reduced, with a required transfer
of charge or electrons to another species. Because the sites of reduction
(cathode) and oxidation (anode) are separated in space, an equivalent current
must flow.
Corrosion and Dissolution 57

Corrosion (at the anode) of one molecular weight of metal, with an accom-
panying valence change of +1 (for instance, from 0 to +1), will result in the
transfer of 1 faraday (F) of charge. Thus, for a corroding anode, one may
determine the corrosion rate by measuring the net current flow and dividing
by the area. The unit of corrosion, defined in this way, is then amp/cm2.

4.6 Potential-Current Relationships in Corrosion

Briefly consider an ideal case of corrosion, taking the reaction at the anode
to be Equation 4.1.

M → M+ + e–

At the cathode, assume that the reaction is the reduction of dissolved oxygen:

O2(d) + 2 H2O + 4 e– → 4 OH– (4.5)

An alternative reaction — the reduction of hydrogen ion with the release of

gaseous hydrogen — is also possible:

2 H+ + 2 e– → H2(g) (4.6)

The latter reaction will occur preferentially if the oxygen potential is very
low or the metal is extremely active (base or non-noble). In this case, consider
Equation 4.5 to be the cathodic reaction.
Now look at Figure 4.3. The initial potentials at the anode and cathode are
EAo and ECo. Due to the differences in potential, current begins to flow (cor-
rosion takes place). This may be represented by moving to the right of the
diagram. Due to the familiar Ohm’s law relationship among current, poten-
tial difference, and resistance, the effective potential of the cathode drops
and that of the anode rises. If nothing happens to intervene, the potentials
become equal. This mixed potential, EM, is then maintained, and the current,
i3, is defined by

EM
= i3 (4.7)
Total resistance

The current (i3) divided by the area of the anode yields the corrosion rate.
If an external resistance, Rex, that is large compared to the previous resis-
tance is inserted between anode and cathode, the resulting current is i2 and
is defined by
58 Biological Performance of Materials: Fundamentals of Biocompatibility

O2 +
ECo 2H
2O + 4 -
e
4 OH -
Potential EM
0 -
+ e Current (log)
M +
M
EAo
i1 i2 i3

FIGURE 4.3
Potential–current (log) relationships in corrosion.

ECo − EAo − i2 Rex

= i2 (4.8)
Total resistance

The current i2 is less than i3; thus, less corrosion takes place in a given
period of time. This is the situation when a passivation or insulating layer
can be maintained on a metal in a pH-potential region that would normally
promote corrosion. If the supply of oxygen is limited by diffusion, for
instance, the potentials of anode and cathode may remain more widely
separated, and a still smaller current i1, resulting in less corrosion, may flow.
In either case (i = i1 or i2), the effective potential will be somewhere between
ECo and EAo, depending upon the relative areas of the cathode and anode.
Thus, it should be clear that the actual rate of corrosion may vary widely
for a given set of equilibrium conditions. Local oxygen supply, conductivity
of the bathing electrolyte, and the extent of the electrode surfaces, as well
as the presence of various inhibitors and enhancers of corrosion, may affect
the result.
Corrosion in real environments is not usually detected by measurement
of potentials and currents or of concentrations of ions in solutions. Rather,
it is recognized by the evidence of attack on the bulk material, the chemical
gnawing away of the fabricated part.

4.7 Forms of Corrosion

I am indebted to Mars Fontana and Norbert Greene (Fontana 1985), who
have collected many diverse descriptions of the physical appearance of
corrosion and grouped them into eight categories or forms, depending upon
mechanism and common features of the result of attack. The eight forms of
Corrosion and Dissolution 59

corrosion will be briefly examined and some comments made on their mech-
anisms.

4.7.1 Uniform Attack

Uniform attack, or general overall corrosion, is a self-explanatory term. This
is the process that is taking place in the corrosion region and, by oxide/
hydroxide dissolution, in the passivation region of the Pourbaix diagram. It
is the most common form of corrosion. In the absence of equilibrium con-
centrations of their constituent ions in the bathing solution, it will occur for
all metals. Even in the immunity region, uniform attack will result in a slow
removal of metal from implants. Thus, it is fair to state that, due to uniform
attack, all metals have a finite corrosion rate in vivo. However, uniform attack
may not be noticed until, or unless, significant amounts of metal are lost.
Because all metals currently used in implants are relatively highly resistant
to uniform attack, little evidence of such attack is ever seen in implant
applications.
Uniform attack is usually measured in terms of surface recession. An
approximation to this rate may be obtained from Equation 4.9 if the surface
area of an implant (A, in.2), the density of the alloy used (D, g/cm3), the
exposure time (T, hours), and the total weight loss (w, mg) are known:

534 w
mpy = (4.9)
DAT

For a typical implant alloy with a density near 8 g/cm3, 1 mpy ≈ 0.7 mg/
cm2/day. In vivo uniform corrosion rates for well-passivated alloys are
thought to be about 1/100 of this value (Steinemann 1980), reinforcing the
prior statement of the low utility of this measure in biomedical applications.

4.7.2 Galvanic Corrosion

Galvanic (or two-metal) corrosion takes place when two different metals are
in physical (electronic) contact and are immersed in an ionic conducting fluid
medium such as serum or interstitial fluid. This is also referred to as couple
corrosion. An example of a situation that may lead to this is shown in Figure
4.4. The “difference” between the screw and plate, responsible for the gal-
vanic process, may be due to different compositions (major and/or minor
constituents) and/or processing.
For a particular combination of pH and potential (defined by pO2), the
two metals will have different electrochemical potentials. The one that is less
noble than the other — that is, below it on a suitable practical electrochemical
series — becomes anodic. The surface of the less noble metal that is in contact
with the solution will experience attack, of a uniform nature, with a release
60 Biological Performance of Materials: Fundamentals of Biocompatibility

ACTUAL

Interstitial fluid

Metallic (electronic) contact

SCHEMATIC Plate Screw

Interstitial fluid (ionic) contact

FIGURE 4.4
Conditions for galvanic corrosion.

of metallic ions. The other metal becomes cathodic. Electrons move to it

through the physical connection, driven by the intermetallic potential dif-
ference. The electrons may then reduce dissolved oxygen or hydrogen ions,
depending upon local conditions. Although the less noble metal of the pair
may not corrode because it is in a passive region in its respective Pourbaix
diagram, the more noble metal cannot corrode under any conditions. Thus,
because it is cathodic, it is said to have cathodic protection.
The actual details of galvanic corrosion depend upon a large number of
complicating factors, including the relative size of the areas of electronic and
ionic contact, as well as on the actual metal pair involved. However, although
some exceptions occur in actual practice, it is safe to assume that galvanic
corrosion can occur in any metal pair in acid pH. Note that all three condi-
tions must be met for galvanic corrosion to take place; thus, the presence of
two or more different compositions of metallic implants within an animal
or patient will not produce galvanic effects unless the implants are in direct
physical contact so that an electron current may flow between them.

4.7.3 Crevice Corrosion

Crevice corrosion is one of a number of forms of corrosion related to struc-
tural details. The basic requirement for the occurrence of this process is the
presence of a crevice (a narrow, deep crack): an interface between parts of a
device, such as between plate and screw head, or a defect such as an incom-
plete fatigue crack. The details of the initiation of crevice corrosion are not
yet clear. Once begun, however, it is characterized by oxygen depletion in
the crevice, anodic metallic corrosion along the crevice faces, and cathodic
protective conditions on the metal surface around its mouth.
Corrosion and Dissolution 61

Static nonflowing conditions in the solution seem to favor crevice corro-

sion, perhaps because of the formation of a metallic ion concentration gra-
dient away from the open end of the crevice. Because the areas of attack are
concentrated, evidence of crevice corrosion can easily be seen in the mating
areas in multipart devices, such as between screw and plate in retrieved
fracture fixation devices (Colangelo and Greene 1969). This is a frequently
observed effect; the majority of multipart fracture fixation devices retrieved
from patients show crevice corrosion and/or pitting corrosion (see Section
4.7.4) (Cook et al. 1985). High local concentrations of corrosion products may
also result in precipitation of oxides, hydroxides, or phosphates in adjacent
tissues (Jacobs et al. 1995). In conjunction with stress corrosion, crevice cor-
rosion may change the mechanical behavior of metals subjected to cyclic
loading (see Section 6.5.2).

4.7.4 Pitting Corrosion

Pitting corrosion is a special case of crevice corrosion. It is a more isolated,
symmetric form of attack; inclusions, scratches, or handling damage may
initiate it. Pitting corrosion proceeds through processes similar to those for
crevice corrosion, although static conditions and reduced oxygen potential
seem less important and autocatalysis may play an important role (Punckt
et al. 2004). Thus, pits often occur in large numbers, like freckles, and grow
down in the direction of gravity in unstirred solutions.
It would be desirable to avoid pits in highly stressed implants because
they constitute points of stress concentration and may serve as the starting
points for mechanical cracks to develop. Like the effects of crevice corrosion,
pits are easy to see. When they are very small, they change the surface finish,
often producing a “frosted” or matte appearance. Larger, more developed
pits often have accumulations of colored corrosion products in them and
may show up as green, brown, or black spots against the otherwise polished
surface of the implant. For this reason, it is inadvisable to clean implants
vigorously after removal before they have been examined for evidence of
corrosion. In practice, it may be difficult to distinguish between pitting and
crevice corrosion at early stages of attack. Multiple part implants have, in
the past, often shown evidence of crevice, pitting, and fatigue corrosion
(Cohen and Lindenbaum 1968); however, as alloy cleanliness has improved
with time, the prevalence of these effects has declined.

4.7.5 Intergranular Corrosion

Intergranular corrosion is somewhat related to crevice corrosion but has
different origins and produces different effects. It is more common in devices
that are made by casting. Cast metals have multiple crystals or grains, with
impurities preferentially deposited between the grains during solidification.
As a result, the chemistry of a grain boundary will be different from that of
62 Biological Performance of Materials: Fundamentals of Biocompatibility

the grains on either side and will probably have a different, and generally
less noble, electrochemical potential. The consequence in a corrosive envi-
ronment is an intergranular attack resembling crevice corrosion. A part may
appear essentially normal and then suddenly crumble into grains under a
mechanical stress. A less radical effect is sometimes seen in brass doorknobs
in old houses. Because of the perspiration left on the knob by generations
of hands, the intergranular corrosion of zinc precipitates “etches” the surface
and makes the individual grains visible.
Intergranular corrosion is obviously more common in alloys than in pure
metals and is favored by high levels of impurities and inclusions. If not
properly heat treated, stainless steel may corrode by this mechanism due to
a relative depletion of chromium from the grain boundaries. Welding of
alloys that results in local melting and resolidification can also lead to a
variant of this called knife-edge attack. The name derives from the appear-
ance of the failure: a straight crack through the metal parallel to and near
the weld. Again, proper heat treatment after welding will restore the right
compositional distribution and reduce or prevent this type of attack.

4.7.6 Leaching
Leaching is similar to intergranular corrosion. However, in this case the
components of a particular alloy are sufficiently weakly bound to each other
and differ enough in chemical reactivity so that there is a large difference in
the rate of loss of the alloy components by uniform attack. Thus, leaching
as a form of corrosion is a special case of leaching (discussed in Chapter 3),
with an accompanying chemical reaction. Attack of this kind will remove
metal with a regular periodic variation of effect at a microscopic level. It is
peculiar to certain alloy systems but can be induced by two conditions:

• The introduction into the solution around the metal of an agent that
attacks one component of the alloy in preference to another. For
instance, fluoride ion (F–) will selectively remove aluminum from
copper–aluminum alloys.
• The presence of more than one phase in the alloy. Usually, all of the
grains in an alloy have the same composition. The alloy is then said
to have a single phase. However, it is possible for individual grains
to be of two or more different, discrete compositions. Such an alloy
is said to possess multiple phases. Because electrochemical potential
varies as chemical composition, these phases may have a different
susceptibility to various forms of corrosive attack. Note that heat
treatment to reduce the size of the grains will not change this situ-
ation. For this reason, multiphase alloys are not usually used in
corrosive applications. Thus, considerable academic concern arose
when ASTM F-562 — a multiphase alloy of cobalt, nickel, chromium,
and molybdenum containing 35% nickel — was introduced for use
Corrosion and Dissolution 63

in implants. However, experience suggests that, despite the differ-

ences in the phase compositions, all of the phases are sufficiently
passive under the conditions experienced in soft and hard tissue
implant sites so that leaching does not occur in this alloy.

Leaching produces surface appearances similar to those produced by pitting

or intergranular attack.

4.7.7 Erosion Corrosion

Erosion corrosion is a rare form of corrosion. This is an acceleration of attack
on a metal because of relative movement between the surrounding fluid and
the metallic surface. It is not a unique process, but it serves to increase the
rate of attack by several other mechanisms. This happens because many
corrosion processes tend to be self-limiting. That is, the accumulation of the
products of corrosion at the interface between metal and solution tends to
reduce the rate of reaction. Flowing solution will sweep away these corrosion
end products as well as provide new amounts of dissolved reactants, such
as chloride ion and oxygen. In extreme cases, the solution may physically
erode the passive layer in regions of passivity. The reformation of this layer
and its removal by continued flow produces progressive attack on the metal
and renders that region of the pH-potential diagram corrosive rather than
passive as predicted.
The physical damage resembles pitting, except that these pits are elongated
in the direction of flow and are generally larger and less symmetric than
those seen in pitting corrosion under static conditions. The peculiarities of
flow, especially if it is a stable pattern, will often result in etching a clear
picture of the course of the flow on the surface of the metal.

4.7.8 Stress and Fatigue Corrosion

Stress corrosion is the last of the eight forms of corrosion. Simply stated,
tensile stress increases the chemical activity of metals. A flexed metal com-
ponent will sustain a tensile stress on one side and a compressive stress on
the other side. This produces a difference in electrochemical potential that
renders the convex surface anodic with respect to the concave one. As an
acceleration of uniform attack or, perhaps, secondary to tensile rupture of
the passive film, corrosion may then attack the convex surface. Local corro-
sion rates (as measured by corrosion current) may be two- to threefold
elevated over the uniform corrosion rate (Bundy et al. 1991)
Because the formation of even a small crack in a loaded structure, such as
a flexed plate, will concentrate stress, this attack tends to initiate cracks
that grow rapidly, leading to possible structural failure. The cracks
extend between grains; however, this process can be differentiated from
64 Biological Performance of Materials: Fundamentals of Biocompatibility

intergranular cracking due to the small number, relative isolation, and

branched structure of stress corrosion cracks.
Many metals display an endurance limit to cyclic loading. In the presence
of crevice corrosion in physical cracks or in crack-like defects in a passive
surface layer produced by single or repeated cyclic loading, this limit may
be abolished. That is, the maximum stress that can be reached without failure
continuously decreases as the number of load cycles increases instead of
reaching a lower limit. This phenomenon, which is a dynamic form of stress
corrosion, is termed fatigue corrosion and may be an important limit on the
life of metallic implants undergoing cyclic mechanical deformation.

4.8 Corrosion in Implant Applications

Which of these eight forms of corrosive attack are important in implant
applications? As a general rule, corrosive attack is more common on multi-
part implants than single part devices. Some studies indicate that a majority
of multipart orthopaedic fracture fixation devices show evidence of corrosion
after recovery at the end of treatment (Cook et al. 1985). Uniform attack
occurs on these as on all other implants. The primary physical evidence
suggests that crevice and pitting corrosion are the next most important forms
(Cohen and Lindenbaum 1968). Crevice corrosion occurs in the gap between
the screw and the plate in screw–plate assemblies. The attack is most often
seen on the plate — within the hole, but near the longitudinal surfaces.
Occasionally, crevice attack will be noted on the portions of the screw oppo-
site these areas. Due to the reduction in the cross section of the plate at the
hole, these areas have a high stress concentration. Frank mechanical fracture
of the plate through a screw hole can often be associated with microscopic
evidence of crevice corrosion. The introduction of modularity into joint
replacement devices has produced a range of apparent crevice and related
corrosion effects, with occasional but rare component failure, which have
not previously been seen (Gilbert et al. 1993).
Pitting most often occurs on the underside of screw heads. Despite the
characteristic freckle appearance of the pits, they may be hard to distinguish
from mechanical scoring of the screw head and shaft during insertion and/
or removal. Such scoring may result from rubbing against a burr in the hole
in the plate or against a fragment of bone caught between plate and screw
during removal.
Galvanic corrosion may also occur between plates and screws. There is a
slight tendency for this naturally because the plates and screws are fabricated
by different processes; thus, if they are not properly heat treated, they may
have slightly different electrochemical potentials. Mixing screws and plates
from different manufacturers may also produce galvanic effects because each
manufacturer uses a slightly different heat treatment schedule. Screws of a
Corrosion and Dissolution 65

different composition from the plate, as well as metallic foreign bodies such
as drill bit fragments inadvertently introduced into an implant site, may also
cause galvanic corrosion (Fothi et al. 1992).
Attack of this sort is often discovered through reports of persistent, very
localized operative site pain. It may occur, however, as judged by frequent
observations of tissue discoloration during routine device removal proce-
dures, without any apparent sensation. Galvanic corrosion may leave a dis-
coloration with a “burned” or sooty appearance on the screw or the plate in
the area of contact.
Stress corrosion is also possible but extremely rare. Intergranular corrosion,
leaching, and erosion do not occur to any real extent in modern multipart
devices in orthopaedic applications. A solid-state version of erosion corro-
sion may occur if an interface is loose or fixation is poor. Relative motion
between plate and screws may result in physical removal of material, or
fretting. This may disrupt the passive film and produce accelerated corrosion
in much the same way that erosion corrosion takes place. This phenomenon
is difficult to distinguish from simple wear and is called fretting corrosion
(Brown and Merritt 1981).
In single part devices such as cranial plates, intramedullary rods,
endoprostheses, pins, and cerclage wire, the effects are rather more limited.
Uniform attack does occur, as previously noted. Stress corrosion or, more
generally, stress enhancement of fatigue failure (fatigue corrosion) is proba-
bly the most common destructive form. Although rare in prostheses, its
incidence is very high in the highly stressed cerclage wire used for uniting
bone fragments. Intergranular corrosion does occur occasionally and is prob-
ably most often associated with surface inclusions or casting defects in cast
prosthetic sections. It is rarely active enough to lead to mechanical failure
in the absence of cyclic loading.
Corrosion in blood contact areas is much more complex. The abundant
supply of oxygen and the continued flow of electrolytes render most pro-
cesses highly active. In addition, the presence of many small organic mole-
cules influences rates. Sulfur-bearing molecules such as cystine appear to
accelerate corrosion; neutral molecules such as alanine may inhibit corrosion
(Svare et al. 1970) in much the same way as rust inhibitors in engineering
applications. Furthermore, corrosion may profoundly affect surface proper-
ties and thus influence thromobogenic behavior; this will be discussed in
Section 9.3.2.
Corrosion is generally considered to be undesirable. However, in some
biomaterial applications the response to local concentrations of corrosion
products is necessary to the successful function of an implant. For instance,
the copper IUD (intrauterine device) depends for its contraceptive properties
on the release of copper ions by a corrosion process.
It may also be desirable to accept a higher corrosion rate because of other
more critical properties of an alloy. In a surgical setting, the stainless steel
spring clips used for the repair of large cranial aneurismal defects have been
deliberately fabricated from type 301, 416, and 420 steel alloys (McFadden
66 Biological Performance of Materials: Fundamentals of Biocompatibility

1969). These alloys corrode at a more rapid rate than the more usual grade
of stainless steel used in implants, 316L (ASTM F 138). However, these alloys
are superior to 316L for spring fabrication. One feature of local response to
these products is undoubtedly an increased fibroplasia — perhaps producing
a more rapid and mechanically sound scarring process — but otherwise they
appear to be adequately tolerated.

4.9 Engineering Variables Affecting Corrosion Rates

Despite the prevalence of corrosive attack on implanted devices, the rate of
failure of structure or of function is quite small. Why is it that a particular
device can perform well in 99 patients and then cause problems in the 100th?
Conversely, why do materials known to be prone to corrosion occasionally
survive for very long periods in vivo (Blackwood and Periera 2004)? The
answers are not simple. To all of the normal biological variables of human
health and disease, at least four “engineering” variables must be added:

• Composition of the implants — in particular, variations within the

implant and extremes of implant-to-implant variation — can affect
corrosion rates in many ways, as has been discussed.
• Manufacturing variables, including casting conditions, metal purity,
amount of cold work, and the degree and type of heat treatment,
have a profound effect on corrosion rates (Sutow et al. 1976). Cor-
rosion rates, at least initially, are markedly affected by the details of
the passivation process used (Browne and Gregson 1994; Callen et
al. 1995).
• Handling in manufacture, delivery, and insertion can affect results.
Occasionally, corrosion initiation can be traced to unintended phys-
ical damage (Gray 1974).
• Positioning of the implant will affect the stresses upon it as well as
the local environment that it experiences. Anatomical location and
small differences in it may have an important effect (Oron and Alter
1984), primarily due to local differences in pH and pO2 (Morita et
al. 1992).
Corrosion and Dissolution 67

4.10 Corrosion Factors Peculiar to Biological Environments

In addition to the factors just discussed, it is apparent that organic molecules
present in implant sites may affect corrosion rates. These are thought to act
in four ways:

• Formation of organometallic complexes. If conditions at any point

in the pH-potential region of interest favor the formation of organo-
metallic complexes, the metallic content of these represents an addi-
tion to the corrosion rate.
• Alteration of charge of corrosion products. Many organic molecules
are potent oxidizing agents, producing the possibility of different
ionic valences than predicted by pH-potential considerations. For
example, in the presence of serum proteins, it is probable that sig-
nificant amounts of chromium may be released from alloys as Cr+6
rather than Cr+3 (Rogers 1984). Despite the high degree of oxidation,
Cr+6 may persist for several minutes (Liu and Shi 2001) before even
partial reduction occurs.
• Modification of the passive layer. Combination of organic species
with the passive layer, in the passive region or in adjacent regions
of metastability (low passive film-dissolution rate), may alter the
nature of the passive film. These may act to stabilize or destabilize
the film or to change its electrical conductivity, thus altering corro-
sion rates (Svare et al. 1970).
• Changes in wear conditions. Although the presence of serum pro-
teins generally elevates corrosion rates, in in vitro experiments, it
markedly reduces fretting corrosion rates of stainless steel (Brown
and Merritt 1981; Merritt and Brown 1988).

Previously, I have emphasized the need to distinguish among physiolog-

ical, biophysiological, and pericellular environments (Section 2.3). The addi-
tion of cells and bacteria to a biological environment produces the possibility
of true “bio” corrosion phenomena. Degradative cells such as macrophages
(Yang et al. 1992), as well as a wide range of bacteria (Wilson et al. 1997),
can directly corrode metals without phagocytosis (see Section 8.2.3), prima-
rily by modification of the pericellular environment. This latter phenomenon
mirrors that encountered in marine environments (Thomas et al. 1988).
68 Biological Performance of Materials: Fundamentals of Biocompatibility

4.11 Ceramic Dissolution

As a class of materials, ceramics are most generally defined as inorganic,
nonmetallic solids. This category contains a wide range of compounds and
mixtures, primarily compounds of metals, such as oxides, carbonates, sul-
fates, etc. They may be crystalline or amorphous; if they contain chain form-
ers, such as elemental carbon or silica (SiO2), amorphous materials may be
glassy. Multiphase physical mixtures and, in the presence of glassy phases,
compound alloys, are also possible.
Metals under immune conditions, whether deliberately passivated or not,
have surface films composed of oxides or hydroxides. In the most common
biomaterial alloy systems, the surfaces of stainless steels and cobalt-base
super alloys are primarily chromium oxide and hydroxide, while the surfaces
of titanium-base alloys are primarily titania (titanium dioxide, TiO2). Thus,
the behavior of metals (with or without prior passivation), under chemical
conditions producing immunity for the underlying alloy, is essentially that
of dissolution of a ceramic. In fact, very small metallic wear debris may be
completely ceramic in nature and their reactions with biological environ-
ments may be better considered in this light.
Two main classes of ceramics are used as biomaterials: structural (or tech-
nical) and resorbable (or soluble); the latter includes so-called “bioceramics.”
Structural ceramics such as alumina (Al2O3) and zirconia (ZrO2) are selected
for, among other properties, their low chemical reactivity and essential insol-
ubility in water. Along with some forms of carbon — especially vitreous
carbon — these materials may be regarded as insoluble in biomaterial appli-
cations. In general, reports of dissolution products from solid structural
ceramic biomaterials should be regarded as artifactual; however, very large
surface area/volume ratio ceramic materials, such as aggregations of small
(submicron) wear debris, may be able to release measurable amounts of
dissolution products. No evidence indicates that biophysiological, biological,
or pericellular conditions affect this conclusion.
Soluble ceramics are an example of type 2 or interactive biomaterials (see
Section 1.4). The most common types are those that resemble calcium-based
minerals that naturally occur in mammalian bodies, such as calcium
hydroxyapatite (Ca10(PO4)6(OH)2), tricalcium phosphate (Ca3(PO4)2), octacal-
cium phosphate (Ca8H(PO4)6•5H2O), etc. However, other resorbable mate-
rials, such as hydrated calcium sulfate (CaSO4•2H20), Bioglass™ (see Section
10.3.4.3), etc., are in use as biomaterials.
The dissolution behavior of these more soluble ceramic materials depends
upon their composition, processing, and final form, as well as on local pH
and pO2 (but not on local applied potential because, as a class [with the
exception of carbons and graphites], these materials are electrical insulators).
In addition, phagocytic cells (see Section 8.2.3) are able to attack many of
these materials, thus raising their solubility in pericellular environments.
Corrosion and Dissolution 69

Although general rules are hard to draw, the following principles may be
useful:
• Crystalline (polycrystalline) forms tend to be less soluble than glassy
or amorphous ones of the same composition.
• Polycrystalline forms tend to be more soluble than single crystal
forms of the same composition.
• Hydrated forms tend to be more soluble than nonhydrated forms
of the same composition.
• Mass loss per unit time depends significantly on specific surface
area; thus, porous or fine particulate materials tend to dissolve more
rapidly than equal weights of the same material in a solid, nonpo-
rous form.
• Cellular attack, when successful, is more rapid on small particles
(<25 μm) than on solid bodies of the same composition.

4.12 Polymer Dissolution

Polymers possess such a wide range of compositions, structures, and molec-
ular weight that it is difficult to make any generalizations concerning their
dissolution behavior. One worthwhile distinction is whether a polymeric
material is hydrophilic or hydrophobic. Dissolution of hydrophilic polymers,
especially low molecular weight, resembles uniform corrosion of metals, in
that the result is a surface recession (Figure 4.5, right). By contrast, hydro-
phobic polymers, which will nevertheless absorb polar fluids such as water,
may undergo a form of internal attack in which amorphous regions dissolve
preferentially to crystalline ones (Figure 4.5, left). The effect is to produce
increased surface area, increasing the effective dissolution rate and leading,
perhaps, to structural effects similar to those of intergranular corrosion, with
sudden loss of integrity and release of small particles.

Hydrophobic Hydrophilic

FIGURE 4.5
Comparison of dissolution of hydrophobic and hydrophilic polymers.
70 Biological Performance of Materials: Fundamentals of Biocompatibility

4.13 Final Remarks

The one certain thing that can be said about corrosion is that it results in the
release of cations from all metallic implants. Cations also form a wide variety
of organometallic complexes. Some of these soluble products, such as the
ferric and ferrous ions, are familiar parts of the internal environment. Some
are trace elements with known biological roles, such as trivalent chromium
ions. Others are rare enough in nature that they do not have known metabolic
roles and are released in the body — even in the absence of abnormal
corrosion processes — at concentrations orders of magnitude above their
normal in vivo occurrence. Under certain conditions, small particles also may
be released or formed by precipitation, locally or at remote sites, of soluble
products. Dissolution of nonmetallic materials is still more complex and
difficult to summarize, but it too results in release of soluble and particulate
materials of a wide range of compositions. The consequences of such release
will be considered at length in later chapters.

References
Blackwood, D.J. and Pereira, B.P., No corrosion of 304 stainless steel implant after 40
years of service, J. Mater. Sci.: Mater. Med., 15, 755, 2004.
Brown, S.A. and Merritt, K., Fretting corrosion in saline and serum, J. Biomed. Mater.
Res., 15, 479, 1981.
Browne, M. and Gregson, P.J., Surface modification of titanium alloy implants, Bio-
materials, 15, 894, 1994.
Bundy, K.J., Williams, C.J. and Luedemann, R.E., Stress-enhanced ion release — the
effect of static loading, Biomaterials, 12, 627, 1991.
Callen, B.W. et al., Nitric acid passivation of Ti6Al4V reduces thickness of surface
oxide layer and increases trace element release, J. Biomed. Mater. Res., 29, 279,
1995.
Cohen, J. and Lindenbaum, B., Fretting corrosion in orthopedic implants, Clin. Orthop.
Rel. Res., 61, 167, 1968.
Colangelo, V.J. and Greene, N.D., Corrosion and fracture of type 316 SMO orthopedic
implants, J. Biomed. Mater. Res., 3, 247, 1969.
Cook, S.D. et al., Clinical and metallurgical analysis of retrieved internal fixation
devices, Clin. Orthop. Rel. Res., 194, 236, 1985.
Fontana, M.G., Corrosion Engineering, 3rd ed., McGraw–Hill, New York, 1985, Chapter
4, 137.
Fothi, U., Perren, S.M. and Auer, J.A., Drill bit failure with implant involvement —
an intraoperative complication in orthopedic surgery, Injury 23 (Suppl 2), S17,
1992.
Gilbert, J.L., Buckley, C.A. and Jacobs, J.J., In vivo corrosion of modular hip prosthesis
components in mixed and similar metal combinations. The effect of crevice,
stress, motion, and alloy coupling, J. Biomed. Mater. Res., 27, 1533, 1993.
Corrosion and Dissolution 71

Gray, R.J., Metallographic examinations of retrieved intramedullary bone pins and

bone screws from the human body, J. Biomed. Mater. Res. Symp., 5(1), 27, 1974.
Jacobs, J.J. et al., Local and distant products from modularity, Clin. Orthop. Rel. Res.,
319, 94, 1995.
Liu, K.J. and Shi, X., In vivo reduction of chromium (VI) and its related free radical
generation, Mol. Cell. Biochem., 222, 41, 2001.
McFadden, J.T., Metallurgical principles in neurosurgery, J. Neurosurg., 31(4), 373,
1969.
Merritt, K. and Brown, S.A., Effect of proteins and pH on fretting corrosion and metal
ion release, J. Biomed. Mater. Res., 22, 111, 1988.
Morita, M. et al., Influence of low dissolved oxygen concentration in body fluid on
corrosion fatigue behaviors of implant metals, Ann. Biomed. Eng., 20, 505, 1992.
Oron, U. and Alter, A., Corrosion in metal implants embedded in various locations
of the body of rats, Clin. Orthop. Rel. Res., 185, 295, 1984.
Pourbaix, M., Atlas of Electrochemical Equilibria, Pergamon Press, Oxford, 1966.
Punckt, C. et al., Sudden onset of pitting corrosion on stainless steel as a critical
phenomenon, Science, 305, 1133, 2004.
Rogers, G.T., In vivo production of hexavalent chromium, Biomaterials, 5, 244, 1984.
Steinemann, S.G., Corrosion of surgical implants — in vivo and in vitro tests, in
Evaluation of Biomaterials, Winter, G.D., Leray, J.L. and deGroot, K. (Eds.), John
Wiley & Sons, Chichester, U.K., 1980, 1.
Sutow, E.J., Pollack, S.R. and Korostoff, E., An in vitro investigation of the anodic
polarization and capacitance behavior of 316-L stainless steel, J. Biomed. Mater.
Res., 10, 671, 1976.
Svare, C.W., Belton, G. and Korostoff, E., The role of organics in metallic passivation,
J. Biomed. Mater. Res., 4, 457, 1970.
Thomas, C.J., Edyvean, R.G.J. and Brook, R., Biologically enhanced corrosion fatigue,
Biofouling, 1, 65, 1988.
Wilson, M. et al., Corrosion of intraoral magnets by multi-species biofilms in the
presence and absence of sucrose, Biomaterials, 18, 53, 1997.
Yang, J. et al., Human neutrophil response to short-term exposure to F-75 cobalt-base
alloy, J. Biomed. Mater. Res., 26, 1217, 1992.

Bibliography
Bundy, K.J., Corrosion and other electrochemical aspects of biomaterials, Crit. Rev.
Biomed. Eng., 22, 139, 1994.
Deltombe, E., De Zoubov, N. and Pourbaix, M., Chromium, in Pourbaix, M., Atlas of
Electrochemical Equilibria, Pergamon Press, Oxford, 1966, 256.
Fraker, A.C. and Griffin, C.D. (Eds.), Corrosion and Degradation of Implant Materials:
Second Symposium, STP 859, American Society for Testing and Materials, Phil-
adelphia, 1985.
Fusayama, T., Katayori, T. and Nomoto, S., Corrosion of gold and amalgam placed
in contact with each other, J. Dent. Res., 42, 1183, 1963.
Hofmann, G.O., Biodegradable implants in traumatology: a review on the state-of-
the-art, Arch. Orthop. Trauma Surg., 114, 123, 1995.
Jacobs, J.J., Gilbert, J.L. and Urban, R.M., Current concepts review: corrosion of metal
orthopedic implants. J. Bone Joint Surg., 80A, 268, 1998.
72 Biological Performance of Materials: Fundamentals of Biocompatibility

Luckey, H.A. and Kubli, F., Jr. (Eds.), Titanium Alloys in Surgical Implants. STP 796.
American Society for Testing and Materials, Philadelphia, 1983.
Marcus, P. and Oudar, J. (Eds.), Corrosion Mechanisms in Theory and Practice, Marcel
Dekker, New York, 1995.
Pohler, O.E.M., Degradation of metallic orthopedic implants, in Biomaterials in Recon-
structive Surgery, Rubin, L.R. (Ed.), C.V. Mosby, St. Louis, 1983,158.
Pourbaix, M., Electrochemical corrosion of metallic biomaterials, Biomaterials, 5, 122,
1984.
Ravaglioli, A. and Krajewski, A. (Eds.), Bioceramics, Chapman & Hall, London, 1992.
Scully, J.C., The Fundamentals of Corrosion, 3rd ed., Pergamon Press, Oxford, 1990.
Shahgaldi, B.F. et al., In vivo corrosion of cobalt-chromium and titanium wear parti-
cles, J. Bone Joint Surg., 77B, 962, 1995.
Schweitzer, P.A. (Ed.), Corrosion Engineering Handbook, Marcel Dekker, New York,
1996.
Syrett, B.C. and Acharya, A. (Eds.), Corrosion and Degradation of Implant Materials, STP
684, American Society for Testing and Materials, Philadelphia, 1979.
Tengvall, P. and Lundström, I., Physicochemical considerations of titanium as a
biomaterial, Clin. Mater., 9, 115, 1992.
Vermilyea, D.A., Physics of corrosion, Physics Today, Sept. 1976, 23.
Williams, D.F., Corrosion of implant materials, Ann. Rev. Mater. Sci., 6, 237, 1976.
Zitter, H. and Plenk, H., Jr., The electrochemical behavior of metallic implant materials
as an indicator of their biocompatibility, J. Biomed. Mater. Res., 21, 881, 1987.
5
Reactions of Biological Molecules with
Biomaterial Surfaces

5.1 Introduction
Strictly speaking, in a biomaterials–tissue system, there are no surfaces; as
Andrade (1973) has pointed out, there are only interfaces. In this chapter, the
solid–liquid interface produced by the contact of a solid biomaterial with body
fluids will be considered briefly. The solid–liquid interface can affect dissolved
species in the surrounding fluid at two levels of characteristic dimension:

• The molecular level (3 to 15 Å): these effects are essentially chemical.

• The macromolecular level (15 to 500 Å): these effects are more of a
mechanical nature.

Chemical effects depend upon the detailed chemistry and ionic charge dis-
tribution of the surface. The effects that local chemistry can have on some of
the events of coagulation (Section 9.3.2) and on adaptation (Chapter 10),
immune response (Section 12.2.1), and carcinogenesis (Section 13.2) will be
considered. These effects can be undesirable side aspects of the biomaterial
selected for other properties (as in the general blood conduit problem) or
deliberately induced effects required to mediate a cellular response. Examples
of induced effects are common in the results of various surface treatments
used to reduce or eliminate thrombus formation on the surface of blood contact
materials. A less common induced effect is the production of surface activity
(in the chemical sense) to stimulate directly adaptive cellular response.
In addition to the direct chemical (inorganic) effect of surface modification
on cellular activity, local changes in composition, pH, and molarity will
produce a variety of physiochemical changes in proteins, including dissoci-
ation and denaturation. The observed cellular response may be secondary
to these physiochemical changes in proteins.
Dissociation is understood in the general chemical sense as the separation
of ions from molecular species. In addition, it refers to the disaggregation of
multimolecular organic complexes, such as enzyme-cofactor complexes.

73
74 Biological Performance of Materials: Fundamentals of Biocompatibility

These associations, like all chemical bonding processes, depend upon free
energy considerations and may be affected by local pH and ionic concentra-
tion. On the other hand, denaturation can be viewed as a purely topological
and mechanical process and will be discussed in the next section.

5.2 Denaturation
Denaturation is a problem peculiar to large organic molecules such as pro-
teins. Four levels (or orders) of structure are recognized in these molecules:

1° The chemical composition as defined by atomic content and primary

bonds between atoms
2° The spatial arrangement of portions of a molecule as defined by the
requirements of bond angulation at each atomic center and by the
intramolecular bonds other than main chain bonds
3° The spatial arrangements determined by strong intramolecular
bonds and secondary folding to produce stable domains
4° The aggregation of three structures by weak associative bonding
(hydrogen or Van der Waals bonds)

The primary and, to some degree, secondary levels of structure are deter-
mined during synthesis. The tertiary may be produced by a self-assembly
process or occur secondarily to a usually extracellular, one-time cleavage of
a portion of the synthesized molecule. Thus, if these structures are disturbed
by heat or local chemical activity, the molecule may not be able to revert to
its original or native structure. Such a molecule is said to be denatured and
may arouse a variety of biological responses despite a normal or near normal
chemical composition (primary structure). Finally, because it depends upon
weak bonds, the quaternary of structure is quite sensitive to pH and con-
centration changes. The results of these changes may be as simple as slight
alterations in configuration or as profound as the dissociation of multimo-
lecular structures, such as those formed by enzymes and cofactors.

5.3 Organometallic Compounds

5.3.1 Definitions
Beyond the effects on structure, surfaces may obviously be chemically reac-
tive. As a class, metals are the most reactive implant materials. They may
Reactions of Biological Molecules with Biomaterial Surfaces 75

provoke a host response due to the formation of corrosion products (dis-

cussed in Section 8.2.5). However, many of these corrosion products are
organometallic complexes or compounds and, as such, have a special behav-
ior of biological importance.
As the name implies, organometallic compounds have two components:
one is an organic moiety and the other is the metallic moiety. The association
between the two can be a weak interaction or, at the other end of the
spectrum, a very strong interaction, as in the case of the Fe-containing por-
phyrin heme in the hemoglobin complex. There also can be a degree of
specificity of structure, again as seen in the heme molecule. If another metal-
lic ion were to replace the iron, the function of oxygen transport would be
impaired. This is apparently the case when elevated chromium levels are
present in heme synthesis sites (Smith 1982).
Three terms are useful in discussions of organometallic compounds:

• Chelation: a type of interaction between an organic compound (hav-

ing two or more points at which it may coordinate with a metal)
and the metal to form a ring-type structure
• Coordination: the joining of an ion or molecule to a metal ion by a
nonionic valence bond to form a complex ion or molecule
• Ligand: any ion or molecule that, by donating one or more pairs of
electrons to a central metal ion, is coordinated with it to form a
complex ion or molecule, as in the cobalt complex [CoCl(NH3)5]Cl2,
in which Cl and NH3 in the bracketed portion are ligands coordi-
nated with Co

5.3.2 Stability
In a free metal cation, all of the five d orbitals have the same energy level.
However, in a chelate or complex, some of the filled orbitals are oriented
toward the chelating atoms. Repulsion between nonbonding electrons in a
d orbital and those of the chelating atom causes electrons in these orbitals
to be less stable with respect to the other orbitals. In addition, bonding can
preferentially stabilize one orbital with respect to the others. The theory
dealing with repulsion from the field produced by the chelating atoms is
called crystal field theory; the total effects are dealt with in ligand field theory.
By preferentially filling low-energy orbitals in organic ions, the metallic d
orbitals can stabilize the molecular system. For example, if three orbitals
have a low energy and two have a higher energy, as in an octahedral complex,
the configuration would be much more stable with six electrons occupying
the low-energy levels than with the electrons spread throughout all five
orbitals. The gain in bonding energy achieved in this manner is referred to
as the crystal field stabilization energy (CFSE).
Complexes with more of the electrons in the lower energy levels are more
stable than those with all the d orbitals equally filled. Trivalent chromium
76 Biological Performance of Materials: Fundamentals of Biocompatibility

and cobalt, for example, with three and six electrons, respectively, will form
very stable complexes or ions. Consequently, the tendency of these ions to
form complexes is very great. A complex of univalent copper, on the other
hand, has zero CFSE because the five orbitals are completely filled. As a
result, cuprous complexes will be less stable than those of trivalent cobalt
or trivalent chromium.
Crystal field effects are important in predicting the rates and mechanisms
of reactions of coordination compounds. The essential feature here is that ions
that are strongly crystal-field stabilized will be slow to react, and nonstabilized
ions will be more liable (reactive). This explains why Co+3, which has consid-
erable CFSE, is so nonreactive. In order for Co+3 to react, the octahedral con-
figuration, which creates the large CSFE, must first be disrupted.

5.3.3 Production
The production of the organometallic compounds by implants is controlled
by a dynamic equilibrium that occurs after implantation of a metallic spec-
imen or device. This equilibrium is established between the alloy and the
intermediate organometallic compound, as well as between the alloy and
the more traditional inorganic ions. The rate of corrosion will then depend
on the removal of the intermediate compound. If more is removed by dep-
osition in tissues, then corrosion could proceed at an increased rate. Because
the removal of the intermediate is the rate-limiting step, differences seen
between biological response to powder and bulk implants probably reflect
different surface (interface) reaction conditions.
The equilibrium will also depend upon three other factors:

• The organometallic complex may be formed on the surface or in

solution. If it is formed on the surface, the ratio of the implant surface
area to the fluid volume available for equilibrium (SA/FLV) will
govern the formation rate and the equilibrium concentration. This
is apparently the case for complexes formed between chromium or
nickel and serum proteins (Woodman et al. 1984). If the complex
forms preferentially in solution, then SA/FLV affects only the rate
of formation, as is apparently the case for cobalt.
• The chemical composition of the surface may affect the strength of
the initial association and the rate of loss (desorption) of complexes
that form at the interface. Table 5.1 presents data for absorption and
desorption of albumin, the most common serum protein, by a variety
of materials. A single monolayer corresponds to 0.2 to 0.7 μg cm–2,
depending upon packing of molecules. “Desorption” in this context
is actually exchange because release in the absence of proteins in
solution (a highly unphysiological condition) may be different. In
particular, under these conditions, polyethylene releases no measur-
able albumin into an albumin-free solution (Brash et al. 1974). Fluid
Reactions of Biological Molecules with Biomaterial Surfaces 77

TABLE 5.1
Albumin Absorption and Desorption from Surfaces
Material Absorption (μ
μg.cm–2/24 h): Desorption (%/24 h):
(Conc.: 2 mg/ml) (Conc.: 1 mg/ml)
Metals
Silver 2.01 ± 0.22 23
Vanadium 0.13 ± 0.06 73
Titanium 0.05 ± 0.02 86
Oxides
TiO2 0.15 ± 0.02 70
Al2O3 0.06 ± 0.01 83
Polymers (Conc.: 3.7 mg/ml) (Conc.: 0.1 mg/ml)
Polyethylene 0.28 42
(Conc.: 1 mg/ml) (Conc.: 0.2 mg/ml)
Cuprophane™ 0.28 ± 0.05 90+
Polyurethane 1.0–2.8 Undetermined
Note: Metals and oxides: absorption/desorption in 0.01 M citrate/phos-
phate buffered saline (pH = 7.4) at 37°C using 125I-labeled human
albumin.
Polymers: absorption/desorption in Tyrodes solution (pH = 7.4) at
25°C using 125I-labeled human albumin.
Source: Williams, R.L. and Williams, D.F., Biomaterials, 9, 206, 1988.
Brash, J.L. et al., Trans. Am. Soc. Artif. Int. Organs, XX, 69, 1974.

flow near the interface also has an effect, with release/exchange rates
increasing with increasing flow rate. The situation at the implant–tis-
sue interface is more complex than this, due to competition between
proteins (see Section 5.5).
• The surface energy, which expresses not only chemical composition,
but also local spatial arrangement of atoms and bonds as well, also
may affect the adsorption/desorption rate (Baszkin and Lyman
1980).

5.4 Mechanical Aspects of Interfaces

Far more important than these chemical effects that take place at atomic
dimensions are the mechanical effects that occur on a larger scale. These are
primarily associated with the fact that the solid–liquid interface is a phase
boundary. As such, it has an interfacial energy proportional to its area asso-
ciated with it.
Consider the following situation (Figure 5.1). Five molecules of identical
composition are shown; however, kinetic and other local affects may produce
transient changes in shape, despite the fact that a sphere presents the lowest
area, and thus the lowest energy state, of the phase interface. For simplicity,
a molecule, P, will be considered to be spherical as it approaches a solid
78 Biological Performance of Materials: Fundamentals of Biocompatibility

FIGURE 5.1
Molecules near a liquid (L)–solid (S) interface.

surface, S, in liquid, L. Suppose that it will stick or adhere to the surface; the
work of adhesion is then (for adhesion of phase A to B):

WAB = γA + γB – γAB (5.1)

where
γA , γB = “free” surface tensions
γAB = interfacial surface tension

For this situation, one can write:

WSP = γSL + γPL – γPS (5.2)

Remember that the surface tensions are negative and that WSP must be
negative for adhesion to take place. For example, suppose:

γSL = 70 dyn/cm

γPL = 40 dyn/cm

γPS = 50 dyn/cm

then,

WSP = –70 – 40 + 50 = –60 (a change from –110 → –50) (5.3)

Thus, adhesion would result.

However, now look more closely at the interface between the molecule
and the solid (Figure 5.2). The normal interfacial equilibrium condition (the
Young–Dupree equation) must be satisfied:
Reactions of Biological Molecules with Biomaterial Surfaces 79

P “Big” θ
θ
L

S
θ

“Little” θ

FIGURE 5.2
Conditions for molecular adhesion at an interface.

γSL = γPS + γPLcosθ (5.4)

Note that, for any degree of adhesion (WSP ≤ 0), θ will be less than 180°.
Because the molecule was initially considered to be a sphere, this condition
can only be achieved by deformation of the natural (free) shape. A careful
analysis combining the Young–Dupree equation with the restoration forces
resulting from this molecular deformation would permit a more exact cal-
culation of θ. However, the simple form can be taken as an estimator of the
deformation. Thus, the larger the value of θ is, the smaller the deforming
force and the likelihood of permanent mechanical damage to the molecule
are.
An interesting point concerns the opposite extreme, i.e., when θ goes to
zero (cos θ = 1). This is called full wetting and requires that:

γPS = γPL; γSL = 2γPL (5.5)

This value of surface tension is called the critical surface tension (γc). It can
be determined by measuring θ for a variety of structurally related liquids
and extrapolating the data to determine the limiting value of surface tension
(γc) as θ approaches zero. A plot of cos θ vs. γLV is called a Zisman plot. The
critical surface tension, γc, is determined by the intercept at cos θ = 1. A
schematic result is shown in Figure 5.3 for a material surface with γc = 28
dyn/cm.*

* See Section 9.3.3 for a discussion of the supposed role of γc in cell–surface interactions; de Palma
et al. (1972) present interesting examples of Zisman plots obtained on metallic implants before
and after blood contact.
80 Biological Performance of Materials: Fundamentals of Biocompatibility

1.0

0.8
Cos θ
γ
0.6 C
(28 dyne/cm)

0.4

0.2

0
0 10 20 30 40 50 60
γ LV (dyne/cm)

FIGURE 5.3
Model Zisman plot (five fluids). (Note: 1 dyn/cm = 1 erg/cm2)

5.5 Results of Interfacial Adhesion of Molecules

A few of the effects that can result from molecular adhesion to biomaterials
or tissues are:

• Enzyme activity and rate constants depend closely upon details of

the 3° and 4° molecular structure of enzymes. Accidental or delib-
erate enzyme adhesion at interfaces can be expected to modify their
functional behavior significantly.
• Organic substrate response to enzymatic action is also structure
specific. Many substrate molecules have a degree of orientational
freedom, possessing features such as saturated bonds about which
free rotation is possible. Association of this type of substrate with a
surface could hinder or prevent such rotation. Depending upon the
configuration in which the molecule is “frozen,” enzymatic attack
might be accelerated or inhibited.
• Many complex organic molecules, as synthesized, have a “tail” por-
tion that serves to inactivate them. In the normal course of events,
enzymatic processes act to strip this small segment and release the
molecule into an active substrate pool. Contact with surfaces and
the forces resulting from adhesion may cause premature activation.
Reactions of Biological Molecules with Biomaterial Surfaces 81

Some molecules are apparently designed specifically to become

activated in this manner by contact with foreign surfaces. An exam-
ple is fibrinogen, which is reduced slightly in molecular weight and
converted to the active protein, fibrin, by surface contact-induced
cleavage.
• Immunological response to proteins also depends strongly upon 2°,
3°, and 4° structure. Contact with surface by native proteins pro-
duces unnatural configurations of the following types:
• Conversion or activation of molecules (as mentioned above)
• Transient deformations during surface contact that are restored
upon subsequent desorption
• Partial or total denaturation due to surface adhesion forces
Considerable evidence indicates that molecular deformations of
each of these three types can excite antibody production and
trigger a variety of immune responses, directly or on a subsequent
challenge.

In an effort to study the possible immunological results of the surface

denaturation of proteins, Stern et al. (1972) exposed a series of polymers,
including epoxies, silicones, and poly(acryl)amide), to fresh rabbit serum.
The serum was then injected into the host animals, and the production of
antibodies was investigated. Unless the in vitro exposure included exposure
to macrophages as well as serum, no antibody titers were developed. How-
ever, in the presence of peritoneal macrophages, a number of these materials
produced positive titers. This is evidence of a cell-mediated response to
denatured serum proteins, recognized as foreign bodies (see Section 12.4.1).
The absence of effect (antibody production) when the serum was directly
injected suggests that the denaturation was reversible and present only when
the serum proteins were adsorbed to the test surfaces.
This experiment reminds one of a further complication. The data in Table
5.1 were obtained from pure albumin (one protein) solutions. The actual
exposure in the biological environment involves many proteins, as in Stern’s
use of serum in vitro. In such a situation, proteins encounter the surface
depending upon the product of their concentration and their self-diffusion
velocity, which is approximately inversely related to the square root of their
molecular weight. (Additional factors, such as molecular shape, also affect
self-diffusion rates.)
Thus, protein–surface interactions in vivo (or in vitro from mixed solutions)
should be thought of as a succession of events; early arrivers (low molecular
weight/high concentration) may be potentially displaced by late arrivers
(high molecular weight/low concentration). This process, first recognized
by Leo Vroman in the blood coagulation process (see Section 9.2) and termed
by others the Vroman effect, is shown schematically in Figure 5.4:
82 Biological Performance of Materials: Fundamentals of Biocompatibility

Total (A+B+C)

Concentration (µg˙cm-2)
A C

Time

FIGURE 5.4
Vroman effect (schematic).
See Section 9.3.3 for a discussion of the supposed role of γc in cell–surface interactions; de Palma
et al. (1972) present interesting examples of Zisman plots obtained on metallic implants before
and after blood contact.

Here, even after the total surface concentration of protein (solid line)
reaches a steady state value, the composition of the film continues to change,
as molecules of B displace those of A and, in turn, are displaced by molecules
of C. Remember also that these are equilibrium surface concentrations; the
data of Table 5.1 suggest that continuing exchange of each species may take
place.

5.6 Effects of Charged Interfaces and Ions

The discussion so far has focused upon interfaces considered to be electri-
cally neutral. That is, the deforming force on molecules results simply from
interfacial free energy and the equilibrium requirements for adhesion. If
there is a net surface charge, then a potential gradient will exist in the vicinity
of the surface. This has four major consequences:

• Uncharged molecules will suffer deformation in structure due to the

interaction of their internal dipoles (primarily associated with cova-
lent bonds) with the electrical field.
• Charged molecules and zwitterions (molecules with no net charge
but with equal amounts of negative and positive charge) will
undergo an additional set of constraints due to attraction or repul-
sion of their charge centers. These additional forces will produce
additional structural deformation.
Reactions of Biological Molecules with Biomaterial Surfaces 83

• Charged molecules and ions will also move along the electrical field
gradient, attracted to surfaces of opposite charge. This motion, called
electrophoresis, is utilized in analytical processes to separate ions
with different ratios of charge to ionic mobility. Electrophoresis can
act in vivo to change the concentration of ions as well as pH, thus
potentially altering 3° and 4° structure.
• Finally, charged molecules and ions may interact with magnetic
fields. Moving charges experience a transverse force from constant
magnetic fields; time-varying magnetic fields can produce oscilla-
tory rotation of zwitterions and local current flow through motion
of ions and molecules with net charge.

One should also remember that net (nonzero) surface potentials may arise
(1) by electrochemical equilibria; (2) from differences in dielectric constant
across the interface; (3) from external sources such as direct potential impo-
sition (as suggested by the experiments of Sawyer et al., 1965, in the reduction
of thrombogenic behavior by changing surface potential; see Section 9.3.2);
or (4) from the other member of a galvanic (corrosion) couple (see Section
4.7.2).

5.7 Final Comments

The reader will note that much of the material cited in this chapter does not
have recent publication dates. This is due to the fundamental nature of the
considerations and to the lack of broad interest in the topic of nonchemical
physiochemical events at interfaces, other than in the issues of blood coag-
ulation and hemolysis. However, recent developments in several areas,
including design, evaluation, and, in some cases, clinical application of so-
called “bioactive” materials — as well as the rising interest in cellular and
tissue engineering (see Chapter 11) — suggest that this will become a much
more vigorous research area. This is especially true because the combination
of ultrahigh speed chemical analysis and large capacity very fast computers
now permits gathering molecular configuration data and fitting it to models
of molecule–surface interaction (West et al. 1997). These developments are
certain to have a profound effect on knowledge of the fundamental mechan-
ics of biological performance of materials; once again, they emphasize the
need for biomaterials investigators to maintain a high level of alertness for
scientific and technical advances in fields seemingly far removed from their
day-to-day concerns.
84 Biological Performance of Materials: Fundamentals of Biocompatibility

References
Andrade, J.D., Interfacial phenomena and biomaterials, Medical Instrumen. 7, 110,
1973.
Baszkin, A. and Lyman, D.J., The interaction of plasma proteins with polymers. I.
Relationship between polymer surface energy and protein adsorption/desorp-
tion, J. Biomed. Mater. Res., 14, 393, 1980.
Brash, J.L., Uniyal, S. and Samak, Q., Exchange of albumin adsorbed on polymer
surfaces (1974), Trans. Am. Soc. Artif. Int. Organs, XX, 69, 1974.
dePalma, V.A. et al., Investigation of three-surface properties of several metals and
their relation to blood compatibility, J. Biomed. Mater. Res. (Symp.), 3, 37, 1972.
Sawyer, P.N. et al., Electrochemical precipitation of blood cells on metal electrodes:
an aid in the selection of vascular prostheses, Proc. Natl. Acad. Sci., 53, 294, 1965.
Smith, G.K., Systemic transport and distribution of iron and chromium from 316l
stainless steel implants, Ph.D. thesis, University of Pennsylvania, Philadelphia,
1982.
Stern, I.J. et al., Immunogenic effects of foreign materials on plasma proteins, Nature,
238, 151, 1972.
West, J.K., Latour, R., Jr. and Hench, L.L., Molecular modeling study of the adsorption
of poly-L-lysine onto silica glass, J. Biomed. Mater Res., 37, 585, 1997.
Williams, R.L. and Williams, D.F., Albumin adsorption on metal surfaces, Biomaterials,
9, 206, 1988.
Woodman, J.L., Black, J. and Jiminez, S.A., Isolation of serum protein organometallic
corrosion products from 316LSS and HS-21 in vitro and in vivo, J. Biomed. Mater.
Res., 18, 99, 1984.

Bibliography
Adamson, A.W. and Gast, A.P., Physical Chemistry of Surfaces, 6th ed., John Wiley &
Sons, New York, 1997.
Andrade, J.D. et al., Proteins at interfaces: principles relevant to protein-based de-
vices, in Proc. 2nd Intern. Symp. Bioelectron. Molecular Electron. Devices, Dec.
12–14, 1988, Fujiyoshida, Japan (unpaginated), 1988.
Bamford, C.H., Cooper, S.L. and Tsurta, T. (Eds.), The Vroman Effect, VSP, Utrecht, 1992.
Bernabeu, P. and Caprani, A., Influence of surface charge on adsorption of fibrinogen
and/or albumin on a rotating disc electrode of platinum and carbon, Bio-
materials, 11, 258, 1990.
Elwing, H., Protein absorption and ellipsometry in biomaterial research, Biomaterials,
19(4–5), 397, 1998.
Friedberg, F., Effects of metal binding on protein structure, Q. Rev. Biophys., 7(1), 1,
1974.
Gabler, R., Electrical Interactions in Molecular Biophysics, Academic Press, New York,
1978.
Gray, J.J., The interaction of proteins with solid surfaces, Curr. Opin. Struct. Biol.,
14(1), 110, 2004.
Reactions of Biological Molecules with Biomaterial Surfaces 85

Hallab, N.J. et al., Systemic metal–protein binding associated with total joint replace-
ment, J. Biomed. Mater. Res., 49, 353, 2000.
Hench, L.L. and Wilson, J., Surface-active biomaterials, Science, 226, 630, 1984.
Ivarsson, B. and Lundstrom, I., Physical characterization of protein adsorption on
metal and metaloxide surfaces, Crit. Rev. Biocompatibil., 2(1), 1, 1986.
Lomer, M.C.E., Thompson, R.P.H. and Powell, J.J., Fine and ultrafine particles of the
diet: influence on the mucosal immune response and association with Crohn’s
disease, Proc. Nutr. Soc., 61, 123, 2002.
Manly, R.S. (Ed.), Adhesion in Biological Systems, Academic Press, New York, 1970.
Morra, M. (Ed.), Water in Biomaterials Surface Science, New York, John Wiley & Sons,
2001.
Tanford, C., Physical Chemistry of Macromolecules, John Wiley & Sons, New York, 1961.
Vogler, E.A., Structure and reactivity of water at biomaterial surfaces, Adv. Colloid
Interface Sci., 74, 69, 1998.
Zangwill, A., Physics at Surfaces, Cambridge University Press, Cambridge, 1988.
6
Mechanics of Materials: Deformation and
Failure

6.1 Introduction
Mechanical integrity is a nearly universal requirement for biomaterials. All
materials must cohere or “hold together” if they are to be expected to stay
in one shape and in one location, and to perform their designed function.
The requirement may be only that they withstand the various stresses that
exist in an implant site. A more rigorous requirement exists if part of the
intended function for the implant is a mechanical one, such as a heart valve
replacement or a fracture fixation device. Then, the application may require
the preservation of a minimum value of a property, such as ability to with-
stand permanent deformation, or of a design (mean) value of another one,
such as possessing a particular spring constant. These are extrinsic behaviors
but they depend in part on intrinsic properties: in this case, on yield strength
and elastic modulus, respectively.
Unfortunately, environmental exposure of materials alters their mechanical
properties in a variety of ways. As discussed in Section 2.2, the chemical and
physical environment of the human body is different from external engineer-
ing environments and is, by comparison to many, quite severe. This chapter
will briefly consider the origin of intrinsic mechanical properties of materials.
It will also consider how materials fail in mechanical applications and how
the biological environment affects these properties and types of failure.

6.2 Mechanics of Materials

The simplest experiment that can be performed to characterize the mechan-
ical properties of a solid material is to machine a specimen with well-defined
dimensions (a “standard” specimen) and load it to failure in tension. Figure
6.1 shows the result of such a model experiment, obtained by plotting the

87
88 Biological Performance of Materials: Fundamentals of Biocompatibility

Fm
Fu x

LOAD

ELONGATION ∆L

FIGURE 6.1
Load-elongation curve.

applied load directly against the resulting elongation. The following points
characterize this load-elongation curve:

Fm: maximum load that can be sustained

ΔL: elongation to failure
Fu: load at failure

Unfortunately, these numbers are extrinsic (dependent upon the specific

dimensions of the specimen). It is common practice to transform the load-
elongation curve into a stress–strain curve. Stress, σ, is given by the force
divided by the cross-sectional area perpendicular to the direction of force
application:

F
σ= (6.1)
A

Strain, ε, is given by the ratio of the change in length, ΔL, to the original
length, Lo:

ΔL
ε= (6.2)
Lo

These conversions produce intrinsic values that are independent of spec-

imen dimensions as long as certain basic rules are obeyed in the design of
the specimen. In particular, it is necessary to assure that the specimen is
uniform in cross section in the region in which Lo, ΔL, and A are measured.
Figure 6.2 shows a stress–strain curve that might be obtained by conversion
of the model load-elongation curve of Figure 6.1. This curve is characterized
by several intrinsic parameters (Table 6.1).
Mechanics of Materials: Deformation and Failure 89

σu x

σ0.2%
σy

STRESS
M

0.002 εy εu
STRAIN
Model

x
x

Metal Ceramic Polymer

FIGURE 6.2
Stress–strain curves.

TABLE 6.1
Intrinsic Parameters from a Stress–Strain Curve
Symbol Name Units Definition
σy Yield stress MPa Stress to start permanent (plastic) deformation
εy Yield strain MPa Strain at the moment of yielding
σ0.2% 0.2% offset stress MPa Stress to produce 0.002 strain
σu Ultimate stress MPa Stress to produce fracture
εu Ultimate strain None Total strain to fracture
M Modulus GPa Ratio of stress divided by strain (slope of line
in elastic (proportional) region (= σy/εy))
E Young’s modulus GPa Modulus determined in tension
-- Work of fracture J/m3 Area under the stress–strain curve
Notes: MPa: megapascals; GPa: gigapascals; J/m3: joules/cubic meter.

Of course, other intrinsic parameters characterize materials, and these may

be determined in forms of load application other than pure tension. How-
ever, this chapter will concentrate on tensile behavior and, in particular, the
elastic modulus, E, the yield stress, σy , and the ultimate stress, σu. These
three parameters tend to dominate mechanical design. The relationship of
these parameters to mechanical design may be summarized as follows:

• Elastic modulus is the intrinsic “spring constant” of the material;

thus, it specifies the proportional deformation as a result of stress
within the limit of recoverable deformations.
90 Biological Performance of Materials: Fundamentals of Biocompatibility

• Yield stress sets the upper stress limit for the design of a body
fabricated from a plastically deformable material that under load
must not undergo permanent deformation from its original shape.
• Ultimate stress defines the stress that produces fracture and thus
sets the maximum stress, termed the strength, that the material can
withstand.

Taken together, these three parameters provide a measure of stiffness,

deformability, and strength of a material.
It should be noted that the model stress–strain curve (Figure 6.2) is only
a schematic. The relative values of these three parameters are governed to
a large degree by the nature of the atomic bonds within a material; thus,
classes of materials tend to have different general shapes of stress–strain
curves. These are shown schematically for metallic, ceramic, and polymeric
materials in the lower part of Figure 6.2.

6.3 Elastic Modulus

6.3.1 Fundamental Aspects
The elastic behavior of materials has its origin in the basic chemical bond
structure at the atomic level. That is, for deformations below the yield point,
the major effect is the elastic (recoverable) deformation of interatomic bonds.
Each of the four major materials classes — metals, ceramics, polymers, and
composites — possesses its characteristic bond structures.
The metals are characterized by the looseness of binding of their valence
electrons. Thus, solid metals are thought of as aggregates of positively
charged ions with a neutralizing negative electron cloud. The resulting cohe-
sion is high but the individual bonds lack strong directionality. In real metals,
regions of high order with almost exclusively metallic bonding (grains) are
separated by zones of disorder (grain boundaries) that contain impurities
and other forms of bonding.
Ceramic materials, on the other hand, are primarily ionically bonded. They
are made up of geometric arrays of cations and anions with strong ordering
and a resulting high directionality of bonds. Again, real (nonideal) ceramics
contain grains (ordered) and grain boundaries (disordered) in combination,
as do metals. In this case, ionic bonding is possible in the boundary regions,
but is usually weaker than within the grains because of inclusions, mis-
matches between adjacent grains, and disorder effects.
Polymers exhibit the third type of bond, the covalent bond. Although it is
not particularly strong, this bond, formed by orbital sharing of electrons
between atoms, is highly directional. Engineering polymers consist of long-
chain molecules with covalently bonded “backbones.” The chains may be
Mechanics of Materials: Deformation and Failure 91

ordered in a regular array in regions, forming crystals, or may be uniformly

amorphous; the more usual structure is, again, a combination of order and
disorder. The overall structure is stabilized by occasional interchain covalent
bonds or ionic bonds between charged side groups (cross links) and by
diffuse attraction of hydrogen, oxygen, and nitrogen atoms to –OH groups
(Van der Waals bonds).
In order of strength, these bonds may be classified as follows:

Ionic > metallic > covalent > Van der Waals

Thus, it should come as no surprise that elastic moduli can be ranked as:

Ceramic > metallic > polymeric

Composite materials do not appear in this range because they can be

particle- or fiber-reinforced ceramics, metals, or polymers and thus display
a wide range of elastic moduli, depending upon the arrangements and
relative moduli of their components. A further complication is introduced
by the nature of the bonding between the matrix and the reinforcing material.
This bonding depends on the chemistry and microstructure of the interface;
it affects yield and failure but has little effect upon elastic behavior. The
ability to adjust the modulus (and other properties) to meet the requirements
of a specific application is, of course, one of the great attractions of compos-
ites.

6.3.2 Environmental Effects

Practically speaking, the elastic moduli of metals and ceramics are unaffected
by exposure to biological environments. This is due to the great strength of
internal bonding in these materials, to the relative simplicity of their struc-
ture when compared with polymers and composites and to the typical tem-
peratures encountered in vivo. Polymers may experience profound changes
in elastic moduli in response to the internal environment. Table 6.2 summa-
rizes the principal mechanisms and their effects.

TABLE 6.2
Environmental Effects on Mechanical Properties of
Polymers
Effects on
Phenomenon Modulus (E) Yield stress (σ
σy)
Absorption Decrease (“plasticizing”) Increase
Leaching Increase (“antiplasticizing”) Decrease
Chain scission Decrease Decrease
Cross linking Increase Increase
92 Biological Performance of Materials: Fundamentals of Biocompatibility

Absorption and leaching have been discussed in Section 3.3 and Section
3.6. The principal effect of absorption of low molecular weight species is to
swell the amorphous matrix, moving the crystalline “islands” further apart
and thus weakening the already weak bonds between them. This permits
easier deformation in the same way that lubrication makes it easier for
surfaces to move over each other. However, many polymers already contain
plasticizers in the form of low-molecular-weight fragments of the basic poly-
mer, deliberately added low-molecular-weight agents, and water. Thus, the
loss of these by leaching would be expected to reverse the effect of absorption
and increase the elastic modulus. In real applications, there is competition.
However, because biomedical polymers tend to be simple (low additive)
materials due to host response considerations and tend to have high molec-
ular weight due to strength and stability considerations, the usual effect of
exposure to physiological fluids is to lower the effective elastic modulus.
Elastic moduli of highly crystalline or highly cross-linked polymers should
be less sensitive than amorphous, low-molecular-weight ones.
An illustration of plasticizing and antiplasticizing effects can be seen in
data on the elastic moduli of polymers in compression in Table 6.3. In this
study (Jacobs 1974), standard compression test cylinders were made of a
commercial poly(methyl)methacrylate (PMMA) surgical cement, a duplicate
formulation compounded in the laboratory, and a commercial medical grade
of ultrahigh molecular weight polyethylene (UHMWPE). These were tested
as fabricated (except for UHMWPE, for which the fabrication date was

TABLE 6.3
Variation of Compressive Moduli of Polymers with Environmental Exposure
PMMA PMMA UHMWPE
(raw materials) (commercial)a (commercial)b
Test condition: E1% (×
× 105 psi) E1% (×
× 105 psi) E1% (×
× 105 psi)
As fabricated 3.0 ± 0.2c 3.4 ± 0.2 NA
Postlaboratory storage 3.9 ± 0.2 3.8 ± 0.5 0.83 ± 0.06
(24°C/120 days)
Post humid storage 2.7 ± 0.3d 2.6 ± 0.2d 0.81 ± 0.05
(97%RH/37°C/120 days)
Post saline storage (0.9% 2.7 ± 0.4d 2.8 ± 0.2d 0.80 ± 0.06
NaCl/37°C/120 days)
Post implantation (rabbit, 3.2 ± 0.2d 3.1 ± 0.4 0.80 ± 0.05
subcutaneous/120 days)
Notes: E1% = tangent modulus at 1% strain; NA = not available.
a Simplex-P™ (North Hills Plastics, Ltd.).
b Zimmer-USA.
c ±95% confidence interval.
d Different from “post laboratory storage” (p < 0.05).
Source: Adapted from Jacobs, M.L., M.S. thesis, University of Pennsylvania, Philadel-
phia, 1974.
Mechanics of Materials: Deformation and Failure 93

unknown) and after 120 days’ exposure to a variety of environments, includ-

ing subcutaneous implantation in the rabbit. The slope of the stress–strain
curve at 1% strain, E1%, was used for comparison because, in common with
most other polymers, these polymers do not possess a single well-defined
elastic modulus in the elastic region of the stress–strain curve. Exposure of
both PMMA formulations to high humidity or saline solutions that dupli-
cated the ionic concentration of serum at 37°C produced a reduction of ~30%
in E1% when compared with dry, room-temperature storage. This illustrates
the plasticizing effect of absorbed water. Although it produced a similar
reduction, implantation was much less damaging. This could be interpreted
in one of two ways:

• Implantation might have prevented the loss of residual monomer

by leaching. An effect similar to leaching — loss of monomer by
evaporation — is probably responsible for the increase of E1% due
to dry storage when compared with the as-fabricated value. Residual
monomer would serve as a plasticizer; however, because it is hydro-
phobic, it might exclude the more efficient plasticizer, water.
• A cross-linking agent or an antiplasticizer might be absorbed from
serum in the animal, counteracting the plasticizing effects of water
absorption.

On the other hand, UHMWPE, with its more crystalline nature and far
higher average molecular weight (≈2 × 106 vs. 2 × 104), is unaffected by the
environmental exposures used in this experiment.*
Chain scission is the polymeric equivalent of the processes of corrosion
and dissolution of metals discussed in Section 4.1 and Section 4.2. The prin-
cipal mechanisms are intrinsic scission (no external chemical species
involved), oxidation, hydrolysis, or chemical attack. Figure 6.3 summarizes
these mechanisms and provides some generic examples.
Chain scission reduces the elastic moduli of polymers through three routes:

• The scission reaction may release a very small molecular fragment,

such as a water molecule, that can act as a plasticizer.
• Although the principal resistance to small deformations in polymers
is due to stretching and/or disruption of weak bonds, some contri-
bution is due to “tangling” of long molecules. In much the same
way that long strands of spaghetti tend to entrap each other, this
tangling forces an elongation of a portion of the strong, covalently
bonded molecules, even at modest deformations. Thus, scission of
molecules by reducing average molecular weight releases these

* In this study, the implants were sterilized chemically. If they had been irradiated, chain scis-
sion, cross-linking, and oxidation of residual free radicals would have affected the results. See
the subsequent part of this section.
94 Biological Performance of Materials: Fundamentals of Biocompatibility

INTERNAL MECHANISMS
X H Elimination
– CH = CH – + HX
C
C Depolymerization X
CH = CH2
H
H
n Random Scission X X·
C H ·CH2 or CH – CH2
e.g. PMMA ·
COOCH3 COOCH3 Depolymerization slow
CH3 unless catalyzed
C n C CH2
CH2 CH3 Accelerated by O2 , light
n
e.g. PVC
Cl H
C C n ( – CH = CH – + HCI)
H H
n
OXIDATION Free radical reaction Chain oxidation
with O2 and photo-oxidation
HYDROLYSIS
COOCH3 Amine
H20 O OH
C elimination
C
H+, OH - CH
also possible
CH2 CH2
n n
OTHER CHEMICAL REACTIONS
- -
Addition of HCl, SO2, SO3 , NO3 , etc., to unsaturated bonds
e.g. H H
H
C C + 2 HCl C H + H C
H
Cl Cl

FIGURE 6.3
Mechanisms of chain scission in polymers. (Adapted from Adams, R. and McMillan, P.W., J.
Mater. Sci., 12, 643, 1977.)

trapped molecules and permits greater strain before covalent bond

stretching can contribute significantly to the elastic modulus.
• The disorder associated with shorter chain length may reduce crys-
tallinity and thus reduce the average strength of bonding, leading
to lower moduli.

Cross linking is the reverse of scission. The formation of new bonds

between chains increases the effective molecular weight, further tangles and
traps molecules, and may reduce the effective concentration of plasticizers
by chemical combination. A common mechanism for cross-linking polymers
in the laboratory is exposure to ionizing radiation. This produces active free
radicals, as in chain scission, that link with free radicals in neighboring
chains, forming covalent cross links. Although clinical doses of x-radiation
do not produce measurable changes in the properties of polymeric implants
in patients (Eftekhar and Thurston 1975), high-dose γ-radiation in the
Mechanics of Materials: Deformation and Failure 95

laboratory is a convenient process to enable studying the mechanical conse-

quences of cross linking.
Irradiation of simple pure polymers such as polyethylene suggests a linear
increase of modulus with the number of cross links (Grobbelaar et al. 1978),
and a more pronounced effect may occur if a number of low-molecular-
weight agents that can be incorporated are present during irradiation. Fur-
thermore, unless they are removed by annealing or doping with reducing
agents such as vitamin E, residual free radicals produced by γ or electron
beam irradiation may produce progressive in vitro and in vivo property
changes by continuing reaction with polymer molecules and diffusible small
molecules, such as water or oxygen.
The environmental effects on the elasticity of composites are more difficult
to generalize. In an ideal model, the elastic modulus of a randomly oriented
composite, EC, made of materials A (matrix) and B (reinforcing or filler phase)
can be calculated from

EC = EAVfA + EBVfB (6.3)

where Vfi = volume fraction of material (phase) i. Then, any effect on the
modulus of either material is seen as a proportional effect on the modulus
of the composite. A special case would be the formation of voids in a material,
by leaching of a second phase or by aggregation of internal defects. Because
the modulus of a void is zero, the modulus of a porous material, for small
pore volumes, can be given by

E = Eo(1 – VfP) (6.4)

where Eo = elastic modulus of fully dense material. Thus, the modulus would
be expected to decrease linearly with increasing volume fraction of pores.
In real materials, the effect is somewhat greater at small void volume
fractions but becomes less pronounced for more porous materials. Equation
6.5 was derived for rigid ceramics (MacKenzie 1950) and has been shown
experimentally to describe effects in materials with Poisson ratios near 0.3:

E = Eo(1 – 1.9 VfP + 0.9 VfP2) (6.5)

There is a further problem in describing the effects of environment on the

elastic moduli of composites. Equation 6.3 is based upon an assumption that
a perfect bond exists between the phases so that each phase experiences an
equal internal strain for a given external uniform deformation of the com-
posite material. Real composites rarely display such perfect bonding, and
the bond is often the weak point for environmental attack. The consequences
of this are unpredictable but the usual effect is a reduction in modulus.
96 Biological Performance of Materials: Fundamentals of Biocompatibility

6.4 Yield Strength

6.4.1 Fundamental Aspects
The yield strength is defined by the stress necessary to produce unrecover-
able deformation in a material. Deformation at lower stresses may be linear
in the case of a simple solid or increasingly nonlinear as strain increases, as
in the case of many polymers. Recovery may be rapid at lower strains and
become slower as peak strain increases. Finally, at the yield stress, conditions
of deformation are such that a residual unrecoverable strain remains, even
after long times at zero stress.
Within crystals, unrecoverable strain is produced by migration and aggre-
gation of defects and by the slippage of material along defect planes. How-
ever, in complex materials and composites, slip, leading to unrecoverable
deformation, may occur preferentially along grain and phase boundaries.

6.4.2 Environmental Effects

At room and body temperature, the processes leading to crystalline defor-
mation or grain boundary slip in metals and ceramics are little affected by
environmental exposure because of the relatively high bonding energies.
However, the situation for polymers and polymer-based composites is dif-
ferent, as noted in the earlier discussion of elastic modulus. The effects are
summarized in Table 6.1.
Absorption and leaching produce what appear to be paradoxical effects
on yield strength. That is, a lower modulus, as results from absorption of a
plasticizer, might be expected to accompany a lower yield strength. In gen-
eral, however, the yield strength is raised. Motion along a particular grain
boundary may become easier; however, this may lead to increased load
sharing with adjacent material and, in fact, may produce modest elevations
of yield stress in inhomogeneous materials. A similar but inverse effect is
seen when plasticizers are leached from the material.
On the other hand, chain scission produces an overall reduction in molec-
ular weight, making plastic deformation more dependent upon the inter-
ruption of weak bonds and thus reducing the yield stress. Cross linking
increases the tangling effect of long molecules, thus substituting strong cova-
lent bonds for weaker bonds and raising the yield stress.
The situation in composites is more complex, and no generalizations can
be made. This is the case because the environment may affect the matrix and
the matrix-filler bond as well. The results depend upon the details of the
composite material in question and its exposure.
It is possible for materials to undergo unrecoverable deformation under
constant load at stress below the yield stress. This is the familiar creep
process. Creep rates are generally very slow for temperatures below one-half
Mechanics of Materials: Deformation and Failure 97

the melting temperature of the material. However, for temperatures above

one-half the melting temperature, or in the presence of plasticizers, creep
can be significant. Creep is possible in many biomedical polymers at room
temperature and is generally increased at body temperature.
Creep is characterized by an initial or primary creep phase in which the
creep rate diminishes rapidly. This is followed by a long secondary creep
phase with a strain rate that is essentially constant in logarithmic time. In
this secondary creep phase, the Dorn–Weertman equation can be used to
describe the creep rate:

ε· = Aσne–Q/RT (6.6)

where
ε· = creep rate
σ = stress
n = experimentally fitted parameter (≅ 5)
Q = activation energy

The activation energy (Q) is usually taken to be the activation energy for
self-diffusion, but may be considered more generally as an intrinsic activa-
tion energy for creep (Parsons and Black 1977). Thus, environmental effects
on the creep rate can be interpreted in terms of changes in the activation
requirements of the creep process.
The final or tertiary process of creep is characterized by a rapidly increasing
strain rate leading to fracture. Little is known about the mechanism of this
process or about environmental effects on it.
It should also be clear from this discussion that creep in biomedical appli-
cations is observed primarily in polymers and polymer-based composites.
In general, secondary creep rates decrease with increasing yield stress at a
given temperature, but the relationship is weak. However, they increase with
increasing temperature and with the presence of plasticizers. This latter effect
may dominate in polymer matrix composites, producing significant
increases in creep rate (Soltész 1986).

6.5 Fracture Strength

6.5.1 Fundamental Aspects
Fracture occurs when the cohesive strength of a material is exceeded. It
represents an accentuation and final stage of the processes that earlier led
to yielding — if that is possible in a particular material. However, it is
generally observed that ultimate strengths, such as the ultimate tensile stress,
98 Biological Performance of Materials: Fundamentals of Biocompatibility

F (s = F/A)

c 2c

FIGURE 6.4
The ideal Griffith crack.

are small compared with those expected, based upon cohesive energy cal-
culations.
Typical calculations of cohesive energy or theoretical maximum strength
lead to values of σu equal to E/10. This would predict a tensile strength of
12.7 GPa for Ti6Al4V, a common alloy useful in implant applications. The
actual value of σu is typically 0.9 GPa (900 MPa), that is, ≈ E/140. This is a
relatively strong material; weaker materials such as stainless steels have σu
in the range of E/250 to E/350.
Griffith (Guy 1971) was the first to explain this observation for brittle
materials (those that fail without significant unrecoverable strain) by sug-
gesting that defects exist on the surface and within the body of real materials
as seen in Figure 6.4. He calculated that the presence of an elliptical crack
in brittle materials produces a stress concentration at the “point” of crack,
as given by:

1/2
⎛ c⎞
σ m ≅ 2σ ⎜ ⎟ (6.7)
⎝ r⎠

where
σ = apparent or “macro” stress
σm = elevated stress at “point” of crack
c = 1/2 major diameter of internal elliptical crack = major width of sur-
face crack
r = radius of curvature at “point” of crack (r << c)

Thus, Griffith suggested that, although the macrostress might be well

below the true ultimate stress, the elevated local stress, σm, near a defect
might exceed the ultimate strength, and a crack would propagate. By con-
sidering the energy required to form the crack (the difference between elastic
strain energy released in the material near the newly formed crack and the
increase of interfacial surface energy due to formation of the new
Mechanics of Materials: Deformation and Failure 99

material–environment interfacial area along the crack), he also calculated the

minimum stress required to propagate the crack:

1/2
⎛ 2 γE ⎞
σ=⎜ (6.8)
⎝ πc ⎟⎠

where
γ = surface tension (material–environment)
E = elastic modulus

Fortunately, most materials undergo plastic deformation before fracture.

Thus, as stresses about a defect are increased, as predicted by Equation 6.7,
plastic deformation will take place before fracture, even if the macrostress
is below the yield stress. Orowan (Guy 1971) dealt with this problem by
replacing the term γ in Equation 6.8 with the quantity (γ + p), where p = the
work of plastic deformation at the “point” of the propagating fracture.
Because p is typically 1000 times γ in magnitude, Equation 6.6 then becomes
approximately,

1/2
⎛ Ep ⎞
σ≅⎜ ⎟ (6.9)
⎝ c ⎠

Such materials will be proportionally stronger because they will require far
higher stresses to propagate existing defects into fracture surfaces.
A special case of Equation 6.7 occurs for spherical pores, the situation
discussed previously with respect to the reduction of elastic modulus by
pores. This has been studied empirically, and the usual relationship (parallel
to Equation 6.5) derived by Ryskewitsch (Kingery 1976) is

σ ′u = σ e e( )
− nV fP
(6.10)

where σ′u is the actual fracture strength for a material with pore volume
fraction VfP , and n is an empirically fitted constant with a value between 4
and 7.
The difference between the stresses predicted by Equation 6.8 and Equa-
tion 6.9 results in the classification of materials as those that fail in a brittle
mode and those that fail in a ductile mode. Characteristic of ceramics and
of polymers at low temperatures, brittle failure occurs without significant
residual deformation and is governed by relations of the form of Equation
6.8. Materials with yield stresses well below ultimate stresses tend to be
ductile and fail in a manner governed by Equation 6.9. Because yielding
occurs during failure, they also exhibit significant unrecoverable strain.
100 Biological Performance of Materials: Fundamentals of Biocompatibility

(I) (I')

A-type

(II)

C-type
(III)

B-type

(IV)

FIGURE 6.5
Plausible crack discs in section. Note: vertical dimension exaggerated. (Adapted from Taka-
hashi, K., J. Macromol. Sci. Phys., B8(3–4), 673, 1973.)

The existence of Griffith defects has been shown repeatedly. An elegant

example is the work of Takahashi (1973) in poly(methyl)methacrylate, as
shown in Figure 6.5. The A-type cracks are those seen after modest stresses,
and B- and C-types after higher stresses, presumably exceeding the limit
imposed by Equation 6.9.
Given that Griffith defects exist two conclusions can be drawn from this
analysis:

• Equation 6.7 predicts that the magnitude of the stress concentration

near a defect will vary inversely with the minimum radius of cur-
vature, r, of the defect. Thus, a sharp crack is a more extreme stress
riser than a semicircular notch. This effect is the basis of the com-
mercial practice of drilling a hole at the advancing tip of a slowly
propagating crack, as in a cracked bridge strut, to prevent further
defect propagation by reducing σm. In implant design, this suggests
that it is desirable to avoid “sharp” features, such as edges, grooves,
and surface structure, to minimize stress concentration effects.
• As the defect major diameter (= 2c) increases, stress concentration
increases (Equation 6.7), and the stress necessary to propagate brittle
fracture (Equation 6.8) or ductile fracture (Equation 6.9) decreases.
Thus, there is a critical minimum size for a Griffith defect to con-
tribute to fracture, given that r remains constant. In practice, larger
defects decrease strength to a limit beyond which the process is
reversed because r begins to increase.

Calculations of minimum crack lengths in composites are much more com-

plex. Marom (1975) has shown that the minimum or critical defect size in
Mechanics of Materials: Deformation and Failure 101

polymer-based composites is much greater than that in the unfilled resin

and depends upon orientation of stresses with respect to the reinforcing fiber.
Attempts to improve mechanical properties of polymers by processing
may lead to paradoxical results, due to the differences between brittle and
ductile fracture strength. Fatigue processes require accumulation of microf-
ractures to produce reduction in area leading to a final single cycle failure.
Thus, cross linking a polymer, such as ultrahigh molecular weight polyeth-
ylene, can increase its ultimate and yield strengths; however, because this
also reduces ductility, it may produce dramatic decreases in fatigue strength
at high cycle numbers (N > 106) due to brittle rather than ductile crack
propagation (Sauer et al. 1996).

6.5.2 Environmental Effects

The biological environment can have significant effects upon the details of
crack propagation and, thus, upon the strength of materials. These effects
will now be briefly discussed.
Any form of chemical attack, whether corrosion, oxidation, dissolution, or
leaching, that can increase the size of pre-existing defects or produce new
defects by preferential attack clearly weakens a material. Such an attack may
take place preferentially near defects in stressed materials and is termed
“stress-enhanced attack” or “stress corrosion” in the case of metals. Such
effects are well recognized in metals and have been demonstrated in silicone
rubber (Rose et al. 1973). Crazing due to swelling can also produce Griffith
defects where none previously existed or can expand preexisting ones.
In brittle materials, the fact that γSL < γSA for all but hydrophobic materials
reduces the required propagation stress predicted by Equation 6.8. Thus, any
of the degradative phenomena discussed in Chapter 3 and Chapter 4 can be
expected to reduce the ultimate strength of biomaterials, and all classes of
materials are susceptible.
Table 6.4 summarizes behavior for a range of typical polymeric implant
materials, nonabsorbable sutures, during a 24-month experiment in rabbits
(Postlethwait 1970). The data given in the table are the ultimate tensile loads
normalized by dividing by the strength of materials retrieved after 1 week
of implantation (in the abdominal wall) to remove the effect of differing
diameters of specimens.
Absorbable materials, such as gut (natural) or polygalactic acid (synthetic)
sutures, will show more pronounced and rapid loss of strength. However,
the rate for an individual material and surgical situation is hard to predict.
The loss of strength is affected by the material composition, fabrication, and
postfabrication handling (production of surface defects), and because of
possible pH dependence of degradation by hydrolysis (Chu 1982) and/or
enzymatic attack (Salthouse et al. 1969; Lotan et al. 1995).
In addition to failure by fracture at stresses that exceed ultimate strength,
materials may fail by fatigue. Fatigue fracture — fracture at stresses below
102 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 6.4
Degradation of Tensile Strength of Sutures in Vivo
Suture Type
Multifilament Monofilament
Period of Polyamide Polyester
Implantation Silk Cotton (Nylon™) Polypropylene (Dacron™)
2 Weeks 0.87 0.98 0.98 1.05 0.92
4 Weeks 0.58 1.07 0.97 1.14 1.03
3 Months 0.20 0.67 0.88 1.03 0.91
6 Months 0.36 0.52 0.79 0.93 0.70
12 Months 0.58 0.50 0.89 0.97 0.96
24 Months Dis. 0.58 0.72 0.99 0.96
Comments Slow Separated Swollen No visible No visible
dissolution change change
Note: Data normalized by 1-week value of ultimate tensile stress.
Source: Adapted from Postlethwait, R.W., Ann. Surg., 171, 892, 1970.

ultimate after a number of cyclic deformations — is recognized as one of the

major sources of mechanical failure of implants. Fatigue failure is character-
ized by the construction of an S–N (stress vs. number of cycles) curve as
shown in Figure 6.6. The ultimate (fracture stress) decreases with cyclic
loading in air until an apparent limit, termed the “endurance limit,” is
reached. In this example, the endurance limit is reached between 106 and 107
cycles. However, in an aqueous or corrosive environment where stress cor-
rosion is possible during the high-stress part of each cycle, this endurance
limit is apparently abolished, and the fracture strength continues to decrease
with cyclic loading. The presence of defects, such as surface markings, may
lead to stress concentration and apparent reduction in fatigue strength; alter-
ation of such defects by corrosion processes may accentuate the effect of
their presence (Naidu et al. 1996).

100
Fatigue test in air
80
Stress (psi/1000)

Pre-corroded (one week)

60 Fatigue test in air

40 Fatigue test in
freshwater
Fatigue test in
20 seawater (estimated)

105 106 107

Number of cycles to failure

FIGURE 6.6
S–N curves for a Cr-V steel. (From Dumbleton, J.H. and Black, J., An Introduction to Orthopedic
Materials, Charles C Thomas, Springfield, IL, 1975.)
Mechanics of Materials: Deformation and Failure 103

100
Key:

Maximum Interfacial Shear Stress (MPa)

Dry
80 Saline
Exudate

60

40

20
1
3

101 102 103 104 105 106

Cycles to failure

FIGURE 6.7
S–N curves for a polysulfone-carbon fiber interface (group mean values; n = 6 per point). --->#
= number of unbroken specimens at this stress level (N = 2.5 × 105). (From Latour, R.A., Jr. and
Black, J., J. Biomed. Mater. Res., 27, 1281, 1993.)

In polymers and polymer-based composites, similar effects are also seen,

probably secondary to plasticizing and/or “decoupling” (secondary to bond
failure) of the reinforcing phase from the matrix. This phenomenon may be
explored by using drops of polymer matrix formed on long specimens of a
fiber-reinforcing phase (Latour and Black 1993). Figure 6.7 shows the S–N
curves for such specimens, fabricated from polysulfone resin and carbon
fiber, after equilibration in various environments (37°C, 24 h). The two solu-
tions used were physiological saline and inflammatory exudate collected
previously by implanting porous capsules of the thermoplastic matrix resin
in soft tissue in rabbits. No endurance limit was observed under dry or wet
conditions up to 2.5 × 105 cycles. The effect on shear strength of both solutions
is the same, suggesting that water is the active agent.
In polymer-based composites, a progressive reduction in modulus is also
observed as internal damage accumulates with an increasing number of load
cycles; the decrease is pronounced if the aqueous environment is such that
the fiber–matrix interface is wetted preferentially (is hydrophilic).
Ceramics display an additional fatigue problem called static fatigue, which
is a sudden failure by brittle fracture at stresses well below ultimate when
subjected to steady (noncyclic) loads. In Al2O3 (Krainess and Knapp 1978)
and in glasses (Adams and McMillan 1977), this effect is believed to be due
to the formation of weak bonds by absorbed water displacing stronger ionic
bonds.
104 Biological Performance of Materials: Fundamentals of Biocompatibility

Thus, it should be clearly understood that biological environments, as

encountered by implants, can be expected to produce a wide range of sig-
nificant changes in the mechanical properties of materials.

6.6 Final Comment

Notwithstanding the general principles discussed here, two points should
be emphasized:

• Actual or real materials have more complex structures than can be

dealt with here in this brief discussion. Therefore, it should be no
surprise when apparently paradoxical property changes are encoun-
tered as a consequence of exposure to biological environments.
• Because it is extremely difficult to predict quantitative mechanical
property changes, care should always be taken to determine the
properties of materials before environmental exposure, whether sim-
ulated or by implantation. This should be performed on actual spec-
imens, from the same batches of materials to be tested, treated
identically to the specimens to be exposed (including sterilization
and other postfabrication treatments), concurrently with initiation
of testing, because it cannot be assumed that properties will remain
unchanged over periods of months and years during laboratory
storage.

References
Adams, R. and McMillan, P.W., Review: static fatigue in glass, J. Mater. Sci., 12, 643,
1977.
Allara, D.L., Aging of polymers, Environ. Health Perspec., 11, 29, 1975.
Chu, C.C., The effect of pH on the in vitro degradation of poly(glycolide lactide)
copolymer absorbable sutures, J. Biomed. Mater. Res., 16, 117, 1982.
Dumbleton, J.H. and Black, J., An Introduction to Orthopedic Materials, Charles C
Thomas, Springfield, IL, 1975.
Eftekhar, N.S. and Thurston, C.W., Effect of irradiation on acrylic cement with special
reference to fixation of pathological fractures, J. Biomech., 8, 53, 1975.
Grobbelaar, C.J., duPlessis, T.A. and Marais, F., The radiation improvement of poly-
ethylene prostheses, J. Bone Joint Surg., 60B, 370, 1978.
Guy, A.G., Introduction to Materials Science. McGraw–Hill, New York, 1971.
Jacobs, M.L. Evaluation of three polymer resins for use in polymer-based composites
for hard tissue prostheses. M.S. thesis, University of Pennsylvania, Philadel-
phia, 1974.
Kingery, W.D., Introduction to Ceramics. 2nd ed. John Wiley & Sons, New York, 1976.
Mechanics of Materials: Deformation and Failure 105

Krainess, F.E. and Knapp, W.J., Strength of a dense alumina ceramic after aging in
vitro, J. Biomed. Mater. Res., 12, 241, 1978.
Latour, R.A., Jr. and Black, J., Development of FRP composite structural biomaterials.
Fatigue strength of the fiber/matrix interfacial bond in simulated in vivo envi-
ronments, J. Biomed. Mater. Res., 27, 1281, 1993.
Lotan, N., Azhari, R. and Sideman, S., Enzymic degradation of polymeric biomate-
rials: a review, in Encyclopedic Handbook of Biomaterials and Bioengineering, Part
A, Vol. 1, Wise, D.L. et al. (Eds.), Marcel Dekker, New York, 1995, 757.
MacKenzie, J.K., The elastic constants of a solid containing spherical holes, Proc. Phys.
Soc. (London), B63, 2, 1950.
Marom, G., Calculation of effective crack length in composite materials, Int. J. Frac.,
11, 534, 1975.
Naidu, S.H., Warner, C.P. and Laird, C., Mechanical stamping: a cause of fatigue
fracture, Clin. Orthop. Rel. Res., 328, 261, 1996.
Parsons, J.R and Black, J., On the thermodynamics of the viscous deformational
mechanism of articular cartilage, Trans. SFB, 1, 78, 1977.
Postlethwait, R.W., Long-term comparative study of nonabsorbable sutures, Ann.
Surg., 171, 892, 1970.
Rose, R.M., et al., The role of stress-enhanced reactivity in failure of orthopaedic
implants, J. Biomed. Mater. Res. Symp., 4, 401, 1973.
Salthouse, T.N., Williams, J.A. and Willigan, D.A., Relationship of cellular enzyme
activity to catgut and collagen suture absorption, Surg., Gyn. Obstet., 129, 691,
1969.
Sauer, W.L., Weaver, K.D. and Beals, N.B., Fatigue performance of ultra-high-molec-
ular-weight polyethylene: effect of gamma radiation sterilization, Biomaterials,
17, 1929, 1996.
Soltész, U., Fracture, fatigue, and aging behavior of carbon fiber reinforced plastics,
in Materials Sciences and Implant Orthopaedic Surgery. Kossowsky, R. and Koss-
ovsky, N. (Eds.), Martinus Nijhoff, Dordrecht, 1986, 355.
Takahashi, K., Cracking of poly(methyl methacrylate) caused by plane stress waves,
J. Macromol. Sci. Phys., B8(3–4), 673, 1973.

Bibliography
Anseth, K.S., Bowman, C.N. and Brannon–Peppas, L., Mechanical properties of hy-
drogels and their experimental determination, Biomaterials, 17, 1647, 1996.
Chu, C.C., Survey of clinically important wound closure biomaterials, in Biocompatible
Polymers, Metals, and Composites, M. Szycher, M. (Ed.), Technomic, Lancaster,
PA, 1983, 477.
Coury, A.J., Chemical and biochemical degradation of polymers, in Biomaterials Sci-
ence, 1st ed., Ratner, B.D., Hoffman, A.S, Schoen, F.J. and Lemons, J.E. (Eds.),
Academic Press, San Diego, 1996, 243.
Ducheyne, P. and Lemons, J.E. (Eds.), Bioceramics: material characteristics versus in
vivo behavior, Ann. N.Y. Acad. Sci., 523, 1988.
Edidin, A.A. et al., Degradation of mechanical behavior in UHMWPE after natural
and accelerated aging, Biomaterials, 21, 1451, 2000.
Ferry, J.D., Viscoelastic Properties of Polymers, 3rd ed., John Wiley & Sons, New York,
1980.
106 Biological Performance of Materials: Fundamentals of Biocompatibility

Gibson, L.J. and Ashby, M.F., Cellular Solids: Structure and Properties, 2nd ed., Cam-
bridge University Press, Cambridge, 1997.
Hayden, W., Moffatt, W.G. and Wulff, J., Mechanical Behavior, Vol. III of The Structure
and Properties of Materials, Wulff, J. (Ed.), John Wiley & Sons, New York, 1965.
Hertzberg, R.W. and Manson, J.A., Fatigue of Engineering Plastics, Plenum Press, New
York, 1980.
Hull, D. and Clyne, T.W., An Introduction to Composite Materials, 2nd ed., Cambridge
University Press, Cambridge, 1996.
Jamison, R.D. and Gilbertson, L.N. (Eds.), Composite Materials for Implant Applications
in the Human Body: Characterization and Testing STP 1178, American Society for
Testing and Materials, Philadelphia, 1993.
Jones, R.M., Mechanics of Composite Materials, 2nd ed., Taylor & Francis, New York,
1998.
Kronenthal, R.L., Biodegradable polymers in medicine and surgery, in Polymers in
Medicine and Surgery, Kronenthal, R.L., Oser, Z. and Martin, E. (Eds.), Plenum
Press, New York, 1975, 119.
Kurtz, S.M. et al., Degradation of mechanical properties of UHMWPE acetabular
liners following long-term implantation, J. Arthroplasty, 18(7), Suppl 1, 68, 2003.
Ward, I.M. and Sweeney, J., An Introduction to Mechanical Properties of Solid Polymers,
2nd ed., John Wiley & Sons, New York, 2004.
7
Friction and Wear

7.1 Introduction
The previous chapter considered the mechanical behavior of materials under
stress. The areas dealt with were those concerning the properties of singular
parts or components. When devices contain more than one component or
are able by design or chance to move against natural tissue, another class of
mechanical effects must be considered.
The general resistance to the motion of one material body over another is
termed friction. When static friction is overcome and relative motion takes
place, it is accompanied by a modification of the interface by a variety of
processes that are collectively known as wear. Introduction of surface treat-
ments or interposed materials to make relative motion easier is called, col-
lectively, lubrication. The study of these three phenomena (friction, wear,
and lubrication) is the science of tribology. This chapter will consider these
phenomena and their presence in and alterations by biological environments.

7.2 Friction
If an attempt is made to move one body over the surface of another, a
restraining force oriented to resist motion is encountered. This restraining
or frictional force, Ff, is given by

Ff = μF⊥ (7.1)

where
F⊥ = force perpendicular to interface
μ = coefficient of friction

107
108 Biological Performance of Materials: Fundamentals of Biocompatibility

The force perpendicular to the surface, F⊥, may be generated by compres-

sive or gravity forces. The coefficient of friction, μ, is a ratio or unitless
number, with values usually between 0 and 1, which describes the relation-
ship of the frictional restraining force to this perpendicular force. It is char-
acteristic of the interface, depending upon the composition and finish of the
pair of materials involved, and is affected by lubrication. Furthermore, the
coefficient is greater just before surfaces begin to move (initial conditions →
μi) than when the surfaces are in continuing or steady motion (sliding con-
ditions → μs). Table 7.1 gives some typical values of μi and μs.
Frictional behavior arises from the physical situation of the surfaces having
a relatively small area of contact due to microscopic surface roughness. The
small size of this actual contact area, perhaps as little as 1% of the geometric
interface area, leads to local yielding and bonding due to high stresses at
the points of actual contact. Thus, relative motion results only when these
bonded areas can be disrupted in shear and moved relative to one another.
This disruption produces the frictional restraining force and also clearly
leads to the wear process.
Frictional restraining forces are complex, but a number of generalizations
can be made:

• The coefficients of friction, μi and μs, are essentially independent of

F⊥ (they may be affected, however, by tangential (lateral) forces).
• For a given value of F⊥, coefficients of friction are generally inde-
pendent of stress — that is, of the apparent or geometric interfacial
surface area.

TABLE 7.1
Initial and Sliding Coefficients of Friction
Materials Combinations Lubricant μi μs
Rubber tire/concrete None (dry) 1.0 0.7
Rubber tire/concrete Water 0.7 0.5
Leather/wood None (dry) 0.5 0.4
Steel/steel None (dry) — 0.5
Steel/polyethylene None (dry) — 0.1
Steel/ice Water 0.03 0.01
Cartilage/cartilage (hip) Synovial fluid — 0.002
Ringer’s — 0.01–0.005
CoCr/CoCr (hip prosthesis)a None (dry) — 0.55
Veronate buffer — 0.22
Serum — 0.13
Synovial fluid — 0.12
Albumin (sol.) 0.11
CoCr/PE(UHMW)a Serum — 0.08
Al2O3/Al2O3b Ringer’s — 0.1–0.05
a Weightman, B. et al., J. Lubric. Tech., 94, 131, 1972.
b Dörre, E. et al., Arch. Orthop. Unfall.-Chir., 83, 269, 1975.
Friction and Wear 109

• Coefficients of friction depend upon surface texture, the material

pair, and the lubricant involved. However, in general, coefficients
are lower for a pair of unlike materials of the same roughness than
for identical materials and are lower for a given material pair in the
presence of lubricating agents than in their absence.
• Static and dynamic coefficients of friction are not closely related to
wear rates (Galante et al. 1973). In particular, low frictional coeffi-
cients do not lead necessarily to low wear rates.

7.3 Lubrication
The principle of lubrication is to provide a film or layer to separate two
surfaces during relative motion in order to reduce frictional restraining forces
and wear. Lubrication modes or processes are classified by the nature and
the magnitude of the average surface separation characteristic for each type.
It is clear from Table 7.1 that artificial material pairs do not possess coeffi-
cients of friction that closely approach those possible in natural joints,
particularly at the low velocities at which joints operate. Little can be done
about this situation as long as body fluids are depended upon for lubrication.
However, it is important from a design point of view to know the actual
coefficients of friction that may be expected.
Table 7.1 suggests the importance of using an appropriate lubricant in
laboratory determinations of friction and wear. For materials in contact with
blood, such as heart valve components, the appropriate lubricant is fresh
serum. For device components in soft tissue locations, a 50:50 mixture of
serum and normal saline approximates the intracellular exudates. For joint
replacement components, the appropriate lubricant is synovial fluid. Wood-
man and colleagues (1977) showed that the synovial tissue remaining in the
vicinity of a joint produces essentially normal synovial fluid that is available
for lubrication of the artificial joint replacement.
Differences in composition between synovial fluid and serum (Table 7.2)
suggest that dilute serum:saline solutions are generally superior to saline to
simulate synovial fluid. However, serum, whether human or animal, is a
highly variable material. In vitro simulator results, especially for wear rates,
most closely resemble those encountered in implanted joints when total
protein concentration is in the range of 20 to 30 g/L and the albumin to
globulin ratio is adjusted (by albumin addition) to a range of 1 to 1.5 (Wang
et al. 2004). However, such studies, performed on metal/polymer wear pairs,
may not be fully generalizable to other wear pairs because synovial fluid
contains one or more so-called “surfactant” species, such as phosphatidyl
choline, which avidly bind to surfaces and reduce dynamic coefficients of
friction under boundary lubrication regimes (see Section 7.3.4) (Hills and
110 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 7.2
Composition of Synovial Fluid in Comparison to Serum
Synovial Fluid Serum
Component (g/l) (g/L) Synovial/Serum
Protein (total) 18 70 0.26
Albumin 11.3 34.3 0.33
α1-Globulin 1.26 4.2 0.30
α2-Globulin 1.26 8.4 0.15
β-Globulin 1.62 11.9 0.14
γ-Globulin 3.06 11.2 0.27
Lipid (total) 2.4 7.0 0.34
Phospholipids 0.8 2.0 0.40
Urate 0.016 0.018 0.88
Glucose 0.66 0.91 0.73
Hyaluronate: 2-4 4.2 × 10–5 ~7 × 104
Albumin/globulin ratio 1.57 0.96
Sources: Proteins: Lentner, C. (Ed.), Geigy Scientific Tables, Vol. 1, Ciba–Geigy,
Basle, 1981; other: Levick, J.R., in Joint Loading, Helminen, H.J., Kiviranta, I.A.,
Säämänen, A.-M., Tammi, M., Paukkonen, K. and Jurvelin, J. (Eds.), Wright,
Bristol, 1987, 149.

Butler 1984). In addition, laboratory testing may not faithfully reproduce in

vivo conditions in which synovial fluid is constantly replaced, even when
the lubricating bath is regularly renewed.

7.3.1 Hydrodynamic
Hydrodynamic lubrication is perhaps the most usual process encountered
in human prosthetic joints in vivo; it occurs when the motion of one body
relative to the other draws a continuous film of lubricant into the contact
area. The characteristic surface separation for typical lubricants and engi-
neering finishes is between 10–3 and 10–4 cm. In this mode, all of the work
of friction is dissipated by viscous shear of the lubricant.

7.3.2 Elastohydrodynamic
Elastohydrodynamic lubrication occurs at smaller surface separations,
between 10–4 and 10–5 cm. In this case, the motion of one body of the pair is
able to transmit force through the lubricant to generate sufficient stress for
transient elastic deformation of the other body. Although this may be satis-
factory in the short term, in the long term it may lead to localized fatigue
failure of one or the other surface, with an accompanying increase in
wear rate.
Friction and Wear 111

7.3.3 Squeeze Film

Squeeze film lubrication occurs in hydrodynamic or elastohydrodynamic
conditions if the lubricant is sufficiently viscous to respond elastically (rather
than by increased flow) to temporarily increased normal loads. Thus, a
squeeze film lubricant, although highly viscous, may reduce wear in situa-
tions in which transient overloads occur. Hydrodynamic, elastrohydrody-
namic, and squeeze film conditions are collectively termed fluid film
lubrication.

7.3.4 Boundary
Boundary lubrication occurs when the lubricant coats the opposing surfaces
rather than acting as a low-shear interface. This coating acts to modify the
frictional character of the surfaces to reduce frictional restraining forces and
wear. Characteristic mean surface separations depend sensitively on the
nature of the lubricant but are usually less than 10–5 cm.

7.3.5 Mixed
Mixed lubrication occurs when a fluid lubricant operating in hydrodynamic
or elastohydrodynamic mode is able to coat the surfaces by an adhesive
process, thus providing additional protection at high loads through bonding
lubrication. Natural synovial joints in the skeletal system probably demon-
strate a combination of these lubrication modes (Wright 1969):

• Boundary lubrication during motion initiation

• Elastohydrodynamic lubrication during motion
• Squeeze film lubrication during high load events

This combination of behavior results from the structure of the joint and from
peculiarities in the nature of the lubricant, synovial fluid. As previously
noted (Table 7.2), normal synovial fluid resembles serum in inorganic species
but contains 30 to 50% of the amount of protein and lipids, a higher albumin
to globulin ratio, and significantly more hyaluronate.

7.3.6 Types of Lubricant Behavior in Response to Shear

In general, lubricants display three types of relationships between apparent
viscosity and shear rate (Dintenfass 1963) as shown in Figure 7.1. Conven-
tional lubricants have a viscosity that is independent of shear rate. Thus,
surface separation, hc, as a function of relative velocity may be determined
fairly easily, taking the lubricant properties as a constant (Hamrock and
Dowson 1981). Under these conditions the lubrication process may remain
112 Biological Performance of Materials: Fundamentals of Biocompatibility

ROUGHER

λ↓

SMOOTHER

λ↑
hc

FIGURE 7.1
Effects of counterface roughness on γ.

constant over a wide range of velocities. For simple parallel geometries, as

in concentric or parallel sliding interfaces A and B, one may then derive a
dimensionless parameter, λ:

λ = hc/(R2rmsA + R2rmsB)1/2 (7.2)

where RrmsA, RrmsB = root means square roughnesses of the opposing surfaces.
Thus, for λ < 1, one would expect predominantly boundary lubrication;
for 1< λ< 3, one would expect mixed lubrication and for λ > 3 some form
of fluid film lubrication would be expected.
If the lubricant and the relative surface velocity remain constant, the sur-
face separation (hc) remains nearly constant and the mode of lubrication is
affected primarily by surface roughness. Here one can see graphically that,
as the roughness of one of the counterfaces decreases, lubrication can tran-
sition from boundary to mixed to fluid film mode.
However, some lubricants are thixotropic; that is, they become reversibly
less viscous as shear rates increase. An everyday example of such a fluid
(although not a lubricant) is nondrip ceiling paint. This appears nearly solid
in the can but becomes quite thin as it is brushed on the wall. The inverse
of thixotropic is also possible; such a material would be called dilatant and
would become reversibly more viscous with increasing shear rate. This
material would not contribute to easy relative motion but might act to reduce
wear at high relative velocities.
Thixotropic or dilatant lubricants can produce a change in lubrication
mode as a function of relative velocity of the surfaces because hc and thus λ
now depend on actual viscosity, pressure, and relative velocity. For instance,
a material pair with a thixotropic lubricant can display hydrodynamic lubri-
cation at intermediate pressures and velocities and boundary lubrication at
high pressures and velocities (Figure 7.2).
Friction and Wear 113

Viscosity (log. η) poise

Dilatant

Thixotropic

Newtonian

Shear rate (log. D) sec-1

FIGURE 7.2
Types of lubricant behavior.

Synovial fluid, which is an ultrafiltrate of serum with the addition of a

long chain polysaccharide complex, hyaluronic acid, is a highly thixotropic
lubricant (Figure 7.3). Trauma resulting in joint effusion may reduce the
osmolarity of synovial fluid, producing a generally less viscous but still
thixotropic lubricant. However, in the presence of a persistent joint disease
such as rheumatoid arthritis, the fluid thins and tends to lose its thixotropic
property. This permits closer approach of the joint surfaces and may produce
increased wear as a contributing factor to joint degeneration.

1000
N

100
O
R
M
AL

η (poise)
PO
ST

10
-T
R
AU

DEG
M

ENER
AT

ATE
I

1.0 D
C

0.1
0.01 0.1 1.0 10 100
D (sec-1)

FIGURE 7.3
Viscosity–shear rate relationships for synovial fluid. (Adapted from Dintenfass, L., J. Bone Joint
Surg., 45A, 1241, 1963.)
114 Biological Performance of Materials: Fundamentals of Biocompatibility

7.4 Wear
7.4.1 Introduction
Wear is a more pronounced problem than frictional restraint for two reasons:

• Wear produces biologically “active” particles that can excite an

inflammatory response (see Chapter 8).
• Wear produces shape changes that can affect function.

There are several mechanisms of wear. Probably the most important mech-
anism in biomedical applications is adhesive wear. This arises from the
junction-making and -breaking process previously described. The rate of
production of wear debris, expressed as a volume, is given most generally by:

kF⊥ x
V= (7.3)
3p

where
V = volume of wear debris
k = Archard’s coefficient
F⊥ = perpendicular force
p = surface hardness
x = total sliding distance

Figure 7.4 shows the range of k values of typical engineering situations.

For the situation of a polymer on a metal in vivo, values for k should lie
between 10–5 and 10–7 with conditions described by the lower right-hand
corner of the diagram. Note that the ordinates are labeled differently. The
left ordinate refers to transfer film formation (see Section 7.4.2) and the right
refers to the production of loose particles.
It is interesting to note that, although values of μi and μs lie within a small
range (between 0 and 1), k and thus wear rates vary over many orders of
magnitude. This further reinforces the previous observation that friction
forces and wear rates have little direct relationship.

7.4.2 The Transfer Film

Figure 7.4 suggests that two major wear processes are possible for any
combination of materials: particle production as described previously, or the
formation of a transfer film. This film is produced when a hard material,
such as a metal, moves on a softer material (for instance, a polymer) and
shears off and picks up a coating of polymer, as shown in Figure 7.5. This
Friction and Wear 115

WEAR COEFFICIENT FOR PARTICLES

WEAR COEFFICIENT FOR TRANSFER
Sim 10-2
-2 ila
10 rm
Dissimilar m e
etals tals
10-4
10-4 Nonmetal ag
ainst nonm
Nonmetal ag etal
ainst metal
-6
10-6
10

CLEAN NO LUBRICANT
High LUBRICANT POOR FAIR GOOD
Vacuum in air Water Pure Mineral oil
Gasoline mineral with lubricity
Non- oil additive
wetting Molten Fatty oil
liquid glass Good
metal Wetting synthetic
liquid lubricant
metal

FIGURE 7.4
Variation of Archard’s constant, k, with wear and lubrication conditions. (Adapted from
Rabinowicz, E., Mater. Sci. Eng., 25, 23, 1976.)

film bridges across the asperities on the surface of the metal, replacing
metal–polymer contact with polymer–polymer contact; by increasing the
actual contact area (as a function of the apparent contact area), it reduces
local stresses.

POLYMER METAL

LOOSE PARTICLE
TRANSFER FILM
PRODUCTION

“big” k transfer film

particle “small” k stable film unstable film
smaller k bigger k

FIGURE 7.5
Transfer film vs. particle production.
116 Biological Performance of Materials: Fundamentals of Biocompatibility

The formation of a transfer film may lead to one of two circumstances:

• If the film is stable, wear rates may be reduced after an initial high-
wear interval during film formation.
• If the film is unstable, it may peel and result in increased wear by
abrasive or “three-body” wear.

McKellop et al. (1978) reported some interesting results involving transfer

films (Table 7.3). The material pair is F138 type steel (316L [vacuum melted])
and ultrahigh molecular weight polyethylene (UHMWPE), with serum, dis-
tilled water, or saline as lubricants. Wear rates are reported as a calculated
linear recession of the UHMWPE surface, based upon measurement of the
volume of wear debris. Thus, the wear rate does not include a creep com-
ponent. Distilled water apparently permitted the formation of a transfer film
and produced the lowest wear rate. Wear in saline was some 65 times greater
(than in distilled water) and was apparently accompanied by crevice corro-
sion of the stainless steel in the interface between substrate and transfer film
as indicated by orange discoloration. This leads to occasional release of the
film and very high wear rates — apparently case 2 cited earlier.

TABLE 7.3
Wear of UHMWPEa in Different Lubricants in Vitro
Average Wear Rate Observations
No. (μ
μm/106 cycles) (μ
μs= dynamic coefficient of
Lubricant Specimens (range) friction)
Serum (bovine) 4 0.65 (±17%) μs = 0.07–0.12 normally; μs = 0.35
during temporary high friction.
Polymer transfer onto metal
counterfaces occurred only
during high-friction phase.
Distilled water 3 0.08 (±60%) μs = 0.07–0.13 at start. A heavy
polymer transfer layer formed by
3 × 105 cycles; μs then ranged from
0.14 to 0.18. The transfer layer
remained intact for the duration
of the test.b
0.9% saline 3 5.2 (±17%) μs = 0.07–0.10 at start. Heavy,
(Ringer’s orange-colored transfer layers
solution) formed as μs increased to 0.27.
These layers occasionally broke
up and μs dropped to the initial
level.
a Against 316L (VM) stainless steel counterface at 3.45 MPa (500 psi) nominal contact stress,
106 sliding cycles @ 60 cpm (travel; 5 × 104 m [est.]).
b McKellop et al. (1978) also report transfer layer to be unstable at 6.90 Mpa.
Source: Adapted from McKellop, H. et al., J. Biomed. Mater. Res., 12, 895, 1978.
Friction and Wear 117

Serum produced an intermediate wear rate with only occasional transfer.

This suggests that, although serum does not completely reproduce synovial
fluid, active molecules present in vivo, other than surfactants, can coat metal
surfaces in metal/polymer wear combinations and, although preventing the
formation of stable transfer films, the molecules reduce wear through surface
lubrication.
Surface roughness, which plays a role in lubrication, also affects wear.
Early polishing of rough surfaces may produce elevated wear rates that
decline significantly after an initial “wearing in” period. However, a persis-
tently rough hard surface bearing against a softer counterface can be
expected to produce wear rates above those predicted by Equation 7.3 (Kurtz
et al. 2000). Finally, surface roughness may change adversely in vivo through
mechanical damage by third bodies or secondarily to intrinsic materials
property changes, such as phase transformations in ceramic surfaces (Hara-
guchi et al. 2001).

7.4.3 Other Wear Mechanisms

Three other mechanisms of wear are of concern in this chapter. The first is
abrasive wear — wear of a soft surface produced by a “plowing” of the
surface by large asperities in the harder countersurface. Although this is a
general mechanism in deliberately articulating interfaces, it may also play a
role in supposedly fixed interfaces in modular devices. In this case, such
wear is referred to as fretting and, in the case of metals, may have a corrosive
component. Abrasive wear clearly occurs in vivo, as will be seen in the later
discussion of the size of wear debris.
Corrosive wear of metals occurs secondarily to the physical removal of a
passive or protective surface layer. The exposed surface may be more sus-
ceptible to wear, perhaps being softer and/or more chemically reactive, or
wear may be accelerated by the repetition of cycles of passive film formation
and mechanical removal. Figure 7.6 suggests this possibility graphically. In
this experiment, a rod of passivated F-75 cobalt-base alloy was pressed
against a dimple in a block of UHMWPE (Jablonski et al. 1986). After the
equilibrium potential has been established (in an aerated 0.9% saline solu-
tion, reference: standard calomel electrode [S.C.E.]), a half sinusoidal stress,
with a peak value of 3.4 MPa, was applied during 40% of a 60° back-and-
forth rocking cycle, at 36 cpm (cycles per minute). These conditions replicate
those thought to exist in the human hip joint prosthesis during slow walking.
The potential (vs. S.C.E.) became more negative, indicative of the flow of an
increased corrosion current. When motion ceased, the potential became less
negative, paralleling the (presumed) re-establishment of the passive layer.
The time to 50% of the maximum potential change (t1/2max) decreased as
the peak stress and the rate of rocking were increased. Similar results have
been reported in all metal total joint replacement prostheses (Thull 1977).
118 Biological Performance of Materials: Fundamentals of Biocompatibility

Potential (vs. S.C.E.)(mv)

-200 t1/2 max

-300 rest potential

-400
-500
-600 equilibrium potential
-700 on simulator off
0 2 4 6 8 10 12 14
Time (minutes)

FIGURE 7.6
Effect of articulation on corrosive wear. (From Stone [Jablonski, J.E. et al., Trans. SFB, 9, 196,
1986], unpublished data.)

The third and final type of wear to be considered is a surface fracture or

fatigue wear. Three routes lead to this process:

• Stresses produced by asperities may not exceed yield stress but may
exceed the endurance limit for the softer of the material pair (gen-
erally a polymer in biomedical applications). This will lead to local
fatigue failure of the softer surface and local fracture, “mud-caking,”
or spalling. This process may be accelerated by diffusion and internal
reaction of chemical species, such as oxygen, with polymers such as
UHMWPE.
• Design of devices may produce a rigid support to the soft compo-
nent. If this component is too thin in comparison to the magnitude
of the contact stress and its apparent contact area, local stress eleva-
tion occurs (Bartel et al. 1985), contributing to fatigue failure, as
observed previously. Although suggesting alternative mechanisms,
Dowling et al. (1978) were some of the first workers to observe such
polymer surface failure in UHMWPE/metal hip prostheses after 8
years or more of implantation.
• Free particles, perhaps adventitial (e.g., bone or PMMA fragments),
or fragments of incomplete or shed transfer films may roll between
the moving surfaces. If these particles are relatively undeformable,
they can easily generate stresses above the ultimate tensile strength
of the surface, leading directly to surface cracking. As cracks join
up, wear debris is released. This process is usually called “three-
body” wear.

7.4.4 Evidence of Wear In Vivo

Wear is a considerable problem in vivo. Of the five heart valve poppets
studied and reported in Table 3.2, four showed signs of wear as evidenced
by strut grooves. These would be expected to have a profound effect on local
Friction and Wear 119

hemodynamics and on valve closure. All 21 hip cups recovered after periods
of 14 to 159 months and studied by Dowling et al. (1978) showed signs of
adhesive (early) or fatigue (late) wear. Of these, nine showed evidence of
the formation of a secondary socket or bearing area, with resultant theoretical
effects on joint function (Dumbleton et al. 1984). As implant designs and
materials have improved, wear rates in vivo have generally declined, but the
majority of articulating components retrieved after service in vivo show signs
of wear.
However, of more importance may be the formation of wear debris. Par-
ticulate materials usually elicit different host responses than bulk materials
do. The observed cellular response is frequently different in nature (see
Section 8.2.2) and far more vigorous than to comparable bulk materials.

7.4.5 Size of Wear Debris

Rabinowicz (1976) discussed a criterion for predicting the diameter of wear
particles produced by adhesive wear processes:

6 × 10 4 W12
d= (7.4)
p

where
d = diameter of wear particle
W12 = surface energy of adhesion between materials 1 and 2
p = hardness of wearing surface

Similar relationships for abrasive wear and for stress concentration phenom-
ena would suggest that hardness is the key parameter in determining typical
wear debris particle size. Because the elastic modulus, E, is a good estimator
of the hardness, p, one would expect particle diameter, d, to vary inversely
with E.
Although Rabinowicz suggests that Equation 7.4 correlates well with
observation (presumably in predominantly adhesive wear situations in vitro),
broad ranges of wear debris particle sizes from submicron to hundreds of
microns in principle dimension are seen in biomedical applications. Savio
et al. (1994) conducted an extensive literature survey of debris associated
with total joint replacements, comparing reports of findings at revision, at
autopsy, during in vitro simulator testing, and, finally, during attempts to
reproduce wear debris by other means for host response testing. Polymeric
particles exhibit the largest absolute size as well as the largest size range;
ceramic particles show much smaller sizes and quite small size ranges.
Metals display an intermediate size range, with great differences in shape
among different alloys and situations.
120 Biological Performance of Materials: Fundamentals of Biocompatibility

At the time of this literature study, it was not appreciated that very small
particles below the resolving power of optical microscopy (≈0.25 μm) are
present under many circumstances. Even in vitro wear and joint simulator
studies failed to detect the presence of such ultrafine particles due to the
extensive use of nuclear pore filters with a minimum pore size of 0.22 μm.
More recently, it has come to be appreciated that such small particles may
constitute a significant portion of the volume and the vast majority of the
number of any species of wear debris.
Figure 7.7 compares polymeric wear debris, generated by cobalt-base
alloy/UHMWPE wear pairs in a size range less than 2.0 μm in maximum
dimension, extracted from capsular tissue recovered during total hip and
total knee replacement revision surgery (Shanbhag et al. 1994). Note that,
although essentially all THR-associated debris observed in this study are
less than 2 μm in major dimension, only ~80% of TKR-associated debris is
this small. Similarly, the median major dimension of THR debris is < that of
TKR debris (0.45 vs. 1.10 μm). The differences in distribution may well reflect
the differences in dominate wear mechanisms in the two types of devices,
as well as differences in regional transport away from the two joints. Finally,
comparison with the size distribution observed in a common resin precursor
of polyethylene suggests that many of the very small particles (<0.25 μm)
may be released by a fatigue mechanism rather than formed by adhesion or
abrasion processes. Note that pre-1990 studies of wear debris tend to under-
estimate the prevalence of submicron size particles due to the widespread
use of 0.22-μm pore size filters.

100

THR
80
TKR C
u
D = 0.45 D = 1.10 m
60 u
l
50 a
t
40 i
v
e
%
20

0
0 0.5 1.0 1.5 2.0
Major Dimension (μm)

FIGURE 7.7
Cumulative percentile distributions of polymeric wear debris. (From Shanbhag, A.S. et al., J.
Bone Joint Surg., 76B, 60, 1994, and personal communication, A.S. Shanbhag.)
Friction and Wear 121

The finding of wear debris particles, in implant sites or in regional lymph

nodes where they were deposited by phagocytosis and transport, during
revision surgery or at autopsy, is quite variable. This suggests that several
different wear processes are taking place simultaneously and that the
mechanical and environmental details of each application influence the types
of particles produced.
Finally, it should be pointed out that no animal or clinical study to date
has shown a good correlation between wear loss of an articulating implant
component and a volumetric estimate of wear debris in the tissues of the
experimental subject. This is due to a number of factors:

• Systemic distribution of particles (previously referred to)

• Dissolution of metallic debris
• Deformational changes in polymeric components (creep, fluid
absorption, etc.)

Thus, estimates of wear from examination of implants probably provide only

upper bounds, and examination of tissues provides unrealistic lower bounds
on wear rates. Carefully conducted in vitro wear testing with appropriate
mechanical conditions and lubricants is required to measure actual wear
rates with any degree of certainty.

7.4.6 Anomalous Wear

One final comment is in order. One normally thinks of wear in terms of a
harder material wearing away a softer material. However, ample evidence
suggests that the reverse can occur. In natural tissues, it is observed that soft
tissues, such as tendons, moving over bone will rapidly form grooves while
appearing unchanged. This may reflect dynamic remodeling but may also
involve a direct wear process of the harder bone by the softer tendon. Gent
and Pulford (1979) have shown a similar situation in the wear, in vitro, of
steel and bronze alloys by a variety of elastomers. Obviously, no dynamic
remodeling is possible here. The mechanism proposed is the formation of
active radicals on the elastomer surface by mechanical cleavage, followed
by chemical attack of the metal surface by these active sites. The enhance-
ment of the effect by the exclusion of oxygen seems to support such a
mechanism. The evidence discussed in Section 4.10 for the participation of
amino acids in corrosion processes certainly suggests that one should con-
sider the possibility of such anomalous wear processes when implant mate-
rials move in contact with biological materials.
122 Biological Performance of Materials: Fundamentals of Biocompatibility

7.5 Conclusions
Frictional restraint to relative motion of device components and the associ-
ated wear, releasing particulate debris, are inherent in the nature of materials.
Careful material selection and component design can minimize but not elim-
inate these phenomena in biomedical devices fabricated from present biom-
aterials. Increased knowledge of the mechanistic details of friction and wear
phenomena in these applications can be expected to produce improved
performance. However, it may be the case that accumulation of wear debris
and the host response to such accumulation will prove to be the ultimate
limitation on the useful lifetime of articulating biomedical devices, such as
total joint replacements. With this in mind, the next section of this book
considers host response to biomaterials and their degradation products.

References
Bartel, D.L. et al., The effect of conformity and plastic thickness on contact stresses
in metal-backed plastic implants, J. Biomech. Eng., 107, 193, 1985.
Dintenfass, L., Lubrication in synovial joints: a theoretical analysis, J. Bone Joint Surg.,
45A, 1241, 1963.
Dörre, E., Beutler, H. and Geduldig, D., The properties required of bioceramics for
artificial joints (author’s transl.), Arch. Orthop. Unfall.-Chir., 83, 269, 1975.
Dowling, J.M. et al., The characteristics of acetabular cups worn in the human body,
J. Bone Joint Surg., 60B, 375, 1978.
Dumbleton, J.H. and Black, J., Some long-term complications, in Complications of Total
Hip Replacement, Ling, R.S.M. (Ed.), Churchill Livingstone, Edinburgh, 1984,
212.
Galante, J.O. and Rostoker, W., Wear in total hip prostheses. An experimental eval-
uation of candidate materials, Acta Orthop. Scand. (Suppl. 145), 1973.
Gent, A.N. and Pulford, C.T.R., Wear of metal by rubber, J. Mater. Sci., 14, 1301, 1979.
Hamrock, B.J. and Dowson, D., Ball Bearing Lubrication. The Elastohydrodynamics of
Elliptical Contacts, John Wiley & Sons, Inc., New York, 1981.
Haraguchi, K. et al., Phase transformation of a zirconia ceramic head after total hip
arthroplasty, J. Bone Joint Surg., [Br.] 83B, 996, 2001.
Hills, B.A. and Butler, B.D., Surfactants identified in synovial fluid and their ability
to act as boundary lubricants, Ann. Rheu. Dis., 43, 641, 1984.
Jablonski, J.E., Stone, J. and Black, J., The effect of articulation on the corrosion
potential of cobalt–chromium alloy in vitro, Trans. SFB, 9, 196, 1986.
Kurtz, S.M., Muhlstein, C.L. and Edidin, A.A., Surface morphology and wear mech-
anisms of four clinically relevant biomaterials after hip simulator testing, J.
Biomed. Mater. Res., 52, 447, 2000.
Lentner, C. (Ed.), Geigy Scientific Tables, Vol. 1, Ciba–Geigy, Basle, 1981.
Levick, J.R., Synovial fluid and trans-synovial flow in stationary and moving normal
joints, in Joint Loading, Helminen, H.J., Kiviranta, I.A., Säämänen, -M., Tammi,
M., Paukkonen, K. and Jurvelin, J. (Eds.), Wright, Bristol, 1987, 149.
Friction and Wear 123

McKellop, H. et al., Wear characteristics of UHMW polyethylene: a method for

accurately measuring extremely low wear rates, J. Biomed. Mater. Res., 12, 895,
1978.
Rabinowicz, E., Wear, Mater. Sci. Eng., 25, 23, 1976.
Savio, J.A., III, Overcamp, L.M. and Black, J., Size and shape of wear debris, Clin.
Mater., 15, 101, 1994.
Shanbhag, A.S. et al., Composition and morphology of wear debris in failed unce-
mented total hip replacement arthroplasty, J. Bone Joint Surg., 76B, 60, 1994.
Thull, R., The long-term stability of metallic materials for use in joint endoprostheses,
Med. Prog. Tech., 5, 103, 1977.
Wang, A., Essner, A. and Schmidig, G., The effects of lubricant composition on in
vitro wear testing of polymeric acetabular components, J. Biomed. Mater. Res.
Part B: Appl. Biomater., 68B, 45, 2004.
Weightman, B. et al., Lubrication mechanism of hip joint replacement prostheses, J.
Lubric. Tech., 94, 131, 1972.
Woodman, J.L. et al., Isolation of lubricating fraction from human synovial fluid after
total-hip implantation, Trans. ORS, 2, 13, 1977.
Wright, V. (Ed.), Lubrication and Wear in Joints, J.B. Lippincott Company, Philadelphia,
1969.

Bibliography
Affatato, S. et al., Wear behavior of cross-linked polyethylene assessed in vitro under
severe conditions, Biomaterials, 20, 3259, 2005.
Bartel, D.L., Bicknell, V.L. and Wright, T.M., The effect of conformity, thickness, and
material on stresses in ultrahigh molecular weight components for total joint
replacement, J. Bone Joint Surg., 68A, 1041, 1986.
Bayer, R.G., Engineering Design for Wear, Marcel Dekker, New York, 2004.
Bayer, R.G. and Ruff, A.W. (Eds.), Tribology. Wear Test Design and Application, STP
1247, American Society for Testing and Materials, Philadelphia, 1993.
Bowden, F.P. and Tabor, D., The Friction and Lubrication of Solids, Oxford University
Press, Oxford, 2001.
Bradgon, C.R. et al., Third-body wear of highly cross-linked polyethylene in a hip
simulator, J. Arthrop., 18(5), 553, 2003.
Davies, D.V., Synovial fluid as a lubricant, Fed. Proc., 25, 1069, 1966.
Divakar, R. and Blau, P.J., (Eds.), Wear Testing of Advanced Materials, STP 1167, Amer-
ican Society for Testing and Materials, Philadelphia, 1992.
Dumbleton, J.H., Tribology of Natural and Artificial Joints, Elsevier, Amsterdam, 1981.
Freeman, M.A.R., Swanson, S.A.V. and Heath, J.C., Study of the wear particles pro-
duced from cobalt-chromium-molybdenum-manganese total joint replacement
prostheses, Ann. Rheum. Dis. (Suppl.), 28, 29, 1969.
Good, V.D., Clarke, I.C. and Anissian, A., Water and bovine serum lubrication com-
pared in simulator PTFE/CoCr wear model, J. Biomed. Mater. Res., 33, 275, 1996.
Hall, R.M. and Unsworth, A., Friction in hip prostheses, Biomaterials, 18, 1017, 1997.
Heisel, C., Silva, M., and Schmalzried, T.P., Bearing surface options for total hip
replacement in young patients, AAOS Instruct. Course Lecs., 53, 49, 2004.
Hutchins, I.M., Tribology. Friction and Wear of Engineering Materials, CRC Press, Boca
Raton, FL, 1992.
124 Biological Performance of Materials: Fundamentals of Biocompatibility

Howell, J.R. et al., In vivo surface wear mechanisms of femoral components of ce-
mented total hip arthroplasties, J. Arthroplast., 19, 88, 2004.
Jacobs, J.J. and Craig, T.L. (Eds.), Alternate Bearing Surfaces in Total Joint Replacement,
ASTM STP 1346, ASTM, West Conshohocken, PA, 1998.
Lazennec, J.-Y. and Dietrich, M., (Eds.), Bioceramics in Joint Arthroplasty, 9th BIOLOX®
Symp. Proc., Steinkopf, Darmstadt, 2004.
Lee, L.-H. (Ed.), Polymer Wear and Its Control. ACS Symposium Series 287, American
Chemical Society, Washington, D.C., 1985.
Ludema, K.C., Friction, Wear, and Lubrication, CRC Press, Boca Raton, FL, 1996.
Mears, D.C. et al., Ferrographic analysis of wear particles in arthroplastic joints, J.
Biomed. Mater. Res., 12, 867, 1978.
McCutcheon, C.W., Lubrication of joints, in The Joints and Synovial Fluid, Vol. I.,
Sokoloff, L. (Ed.), Academic Press, New York, 1978, 437.
McKellop. H.A., Bearing surfaces in total hip replacements: state of the art and future
developments, AAOS Instruc. Course Lec., 50, 165, 2001.
Rieker, C., Oberholzer, S. and Wyss, U. (Eds.), World Tribology Forum in Arthroplasty,
Hans Huber, Bern, SW, 2001.
Schmalzried, T.P., Dorey, F.J. and McKellop, H. The multifactorial nature of polyeth-
ylene wear in vivo, J. Bone Joint Surg., 80A, 1234, 1998.
Semlitsch, M. and Willert, H.-G., Implant materials for hip endoprostheses: old proofs
and new trends, Arch. Orthop. Trauma Surg., 114, 61, 1995.
Streichler, R.M. et al., Metal-on-metal articulation for artificial hip joints: laboratory
study and clinical results, Proc. Instn. Mech. Engs., 210, 223, 1996.
Torrance, A.A., A method for calculating boundary friction and wear, Wear, 258, 924,
2005.
Walker, P.S. and Bullough, P.G., The effects of friction and wear in artificial joints,
Ortho. Clin. N. Am., 4(2), 275, 1973.
Weightman, B., Frication, lubrication, and wear, in The Scientific Basis of Joint Replace-
ment, Swanson, S.A.V. and Freeman, M.A.R. (Eds.), John Wiley & Sons, New
York, 1977, 46.
Wright, T.M. and Goodman, S.B. (Eds.), Implant Wear: The Future of Total Joint Replace-
ment, American Academy of Orthopedic Surgeons, Oakbrook, 1996.
Yust, C.S. and Bayer, R.G. (Eds.), Selection and Use of Wear Tests for Ceramics, STP 1010,
American Society for Testing and Materials, Philadelphia, 1988.
Interpart 1
Implant Materials: Properties

I1.1 Introduction
The preceding discussions concerning material response have necessarily
been generic; that is, they have largely treated biomaterials by classes related
to their predominant chemical bond type (and resulting physical properties)
rather than dealing with the response of a particular, specific composition
of material to the biological environment. This is how it should be because
the details of such response depend upon the composition (including impu-
rities) of the specific material, the methods by which it is fabricated and
finished, the device application in which it is used, the animal and/or clinical
model in which it is evaluated, etc. The professional literature on material
response is broad, so much of the information needed to make a preliminary
selection based upon material response is usually available.
However, before such a selection may be made, a more fundamental ques-
tion must be answered: does the material meet the physical requirements of
the design? Again, this is a complex problem, requiring a well-structured
design process for successful solution. This and related issues of design are
dealt with in Chapter 21. In this interpart, tables of basic materials’ properties
are provided as an introduction to the broader issue of materials selection.
The reader must be warned that most properties tabulated here are nom-
inal or typical properties. Materials specifications, whatever their source,
usually permit a range of compositions and processing conditions; in addi-
tion, the choice of starting materials (feed stocks, master blends, etc.) may
also affect final properties, particularly of polymers, fibers, and composites.
Therefore, reference should be made to original sources (as provided in the
reference and bibliography sections or elsewhere) for exact details of com-
position, fabrication, and test conditions producing the values or ranges
cited.
The properties tabulated are density and typical hardness, when reported,
and those mechanical parameters that may be obtained from a conventional
stress–strain curve. Standards, such as those produced by the ASTM Inter-
national (American Society for Testing and Materials), BSI (British Standards

125
126 Biological Performance of Materials: Fundamentals of Biocompatibility

Institute), ANSI (American National Standards Institute), and ISO (Interna-

tional Standards Organization) are cited when relevant.* However, whether
they are consensus or regulatory in nature, standards most often describe
minimum requirements; properties of actual manufactured materials fre-
quently exceed these values in general and in the case of specific production
lots. Other properties, such as fatigue endurance limit, coefficients of friction
(for material pairs), workability parameters, etc., may be of equal or greater
importance in material selection and should be sought out during the design
process.
It should be remembered that high performance requirement materials
applications, such as medical and surgical devices and implants, require a
high degree of confidence in design and performance parameters. Thus,
some form of intrinsic property verification up to and, in some cases, includ-
ing 100% nondestructive preuse “proof” testing is often required to assure
safe and effective material response in medical and surgical applications.
Finally, researchers and designers should remember that when a bioma-
terial is used in a new design and/or for a new medical indication, it should
regain its status as a candidate material. That is, although successful perfor-
mance under one set of conditions in one design provides useful knowledge,
it does not produce any guarantee of success in another situation. Therefore,
nothing in this interpart should be construed as qualifying or otherwise
recommending any material as safe and/or effective for use in any specific
medical or surgical instrument, implant, or device.

I1.2 Metals
Metallic biomaterials in common use are drawn from the stainless steels
(Table I.1), the cobalt-base “superalloys” (Table I.2), and the titanium/
titanium-base alloy system (Table I.3). Several refractory and precious met-
als that have seen limited use in biomedical applications are covered in
Table I.4.

* Where feasible and relevant, the most recent revision of the standard has been cited in foot-
notes to the tables. However, some values may have appeared in earlier versions and then been
deleted in revision.
Implant Materials: Properties 127

TABLE I.1
Stainless Steels
Material: F138 F138 F138 F745 F1314 F1314 F1586 F1586
type 2 type 2 type 2 High N2 High N2
Condition: AN HF CW An CW AN CW
Source: [1,2] [1,2] [1,2] [3] [1,4] [4] [5] [1,5]
Density(g/cm3): 7.9 7.98 7.9 — 7.98 — — —
E (tensile) (GPa): 200 200 200 — 200 — — 200
Hardness (Hv): — — 350 — 205 — — ~365
σ0.2% (MPa): 190 240 690 207 380 862 430 1000
σUTS (MPa): 490 550 860 483 690 1035 740 1100
Elong. (min.%): 40 55 12 30 35 12 35 10
Notes: AN: annealed; CW: cold worked; HF: hot forged.
Sources: [1]: BSI 3531 (Part 2, Sec.2) (Amend. 2, 1983); [2]: ASTM F138-03; [3]: ASTM F745-00;
[4]: ASTM F1314-01; [5]: ASTM F 1586-02.

TABLE I.2
Cobalt-Base Alloys
Material: Cast Wrought Wrought Wrought Wrought Wrought Wrought
CoCrMo CoCrMo CoNiCr CoNiCr CoCrMo CoNiCr CoNiCr
Mo MoWFe Mo MoWFe
Condition: AN AN AN AN CW CWA CW
Source: [1,2] [1,3] [4] [5] [1,3] [4] [5]
Density(g/cm3): 7.8 9.15 — — 9.15 — —
E (tensile) (GPa): 200 230 — — 230 — —
Hardness (Hv): 300 240 — — 450 — —
σ0.2% (MPa): 455 310 241–449 275 1000 1585 1310
σUTS (MPa): 665 860 793–1000 600 1500 1795 1172
Elong. (min.%): 8 30 50 50 9 8 12

Material: Wrought Wrought Wrought Wrought

CoCrNiMoFe CoCrNi CoCrMo CoCrMo
MoFe MoFe
grade 2 grade 2
Condition: CW CWA AN HW
Source: [6] [6] [7] [7]
Density(g/cm3): — — — —
E (tensile) (GPa): — — — —
Hardness (Hc): — — 25 28
σ0.2% (MPa): — 1240–1450 517 700
σUTS (MPa): 1515–1795 1860–2275 897 1000
Elong. (min.%): — 1–17 20 12
Notes: AN: annealed; CW: cold worked; CWA: cold worked, aged; HW: hot worked; Hv: Vickers
hardness; Hc: Rockwell C hardness.
Sources: [1]: BSI 3531 (Part 2, Sec.4–5) (Amend. 2, 1983); [2]: ASTM F75-01; [3]: ASTM F90-86-
01; [4]: ASTM F562-02; [5]: ASTM F563-00; [6]: ASTM F1058-02; [7]: ASTM F1537-00.
128 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE I.3
Titanium α–β and β Titanium-Base Alloys
Material: Ti Ti Ti3Al2.5V Ti6Al4V Ti6Al7Nb Ti6Al4V Ti5Al2.5
grade 1 grade 4 Fe
Type: — — α–ββ α–ββ α–ββ α–ββ α–ββ
Condition: An AN AN AN HF HF AN
Source: [1,2] [1,2] [3,4] [1,5] [6,7] [8] [9]
Density(g/cm3): 4.5 4.5 4.51 4.4 4.52 4.4 4.45
E (tensile) (GPa): 127 127 105–120 127 105 127 —
Hardness (Hv): — 240–280 — 310–350 400 — —
σ0.2% (MPa): 170 483 517–560 760–795 800 825–869 815
σUTS (MPa): 240 550 621–650 825–860 900 895–930 965
Elong. (min.%): 24 15 15 8 10 6–10 16

Material: Ti5Al2.5 Ti12Mo Ti13Nb Ti13Nb Ti15Mo Ti15Mo Ti35Nb

Fe 6Zr2Fe 13Zr 13Zr 3Nb0.3O2 5Ta7Zr
Type: α–ββ β β β β β β
Condition: HF AN AN QA AN AN AN
Source: [9] [10,11] [12,13] [12–14] [15] [16] [17]
Density(g/cm3): 4.45 5.0 — — — 4.94 —
E (tensile) (GPa): — 74–85 79 77 — 82 55
Hardness (Hv): — 34–35* 26* 28* — — —
σ0.2% (MPa): 900 897–1000 900 725 483–552 1020 547
σUTS (MPa): 985 931–1060 1030 860 690–724 1020 597
Elong. (min.%): 13 12–22 15 8 12–20 — 19
Notes: *: Rockwell C; AN: annealed; HF: hot forged; QA: Water quenched, aged.
Sources: [1]: BSI 3531 (Part 2, Sec.6) (Amend. 2,1983); [2]: ASTM F67-00; [3]: ASTM F2146-01;
[4]: TIMET, Inc.; [5]: ASTM F136-96; [6]: IMI Titanium Ltd, 1989; [7]: ASTM F1295-01; [8]:
ASTM F1472-93; [9]: Borowy, K.-H. and Kramer, K.-H., in Titanium Science and Technology,
Vol. 2, Luterjering, G., Zwicker, U. and Bunk, W. (Eds.), Deutsche Gesell. f. Metallkunde e.
V., Obureresel, 1985, 1381; [10]: Wang, K. et al., in Beta Titanium Alloys in the 1990s, Eylon,
D., Boyer, R.R. and Koss, D.A. (Eds.), Min. Met. and Materials Society, New York, 1993, 49;
[11]: ASTM F1813-01; [12]: U.S. Patent 5,169,597; [13]: Davidson, J.A. et al., Bio-Med. Mater.
Eng., 4(3), 231, 1994; [14]: ASTM 1713-03; [15]: ASTM F2060-01; [16]: Long, M. and Rack,
H.J., Biomaterials, 19, 1621, 1998; [17]: Ahmed, T. et al., Proc. Inst. Metal. 8th World Titanium
Conf., 1996, 742.
Implant Materials: Properties 129

TABLE I.4
Other Metals and Alloys
Material: TaP TaP Ti54.5 Ti55.8 PtP Pt10 Pt10 WP Zr-2.5
NiMA NiMA RhTC RhTC Nb α–ββ
75%
Condition: An CW AN AN AN AN CW SN CW/AN
Source: [1,2] [1,2] [3,4] [3,4] [1] [1] [1] [1] [5]
Density(g/cm ):
3
16.6 16.6 6.5 6.5 21.5 20 20 19.3 6.64
E (tensile) (GPa): 186 186 28–41 41–75 147 — — 345 99
Hardness (Hv): — — — — 38–40 90* 165* 225 —
σy (MPa): 140 345 — — — — — — 379
σUTS (MPa): 205 515 1100 1070 135–165 310 620 125–140 552
Elong. (min.%): 20–30 2 10 10 35–40 35 2 ~0 16
Notes: AN: annealed; CW: cold worked; MA: memory alloy; P: pure (elemental); SN: sintered
bar; TC: thermocouple alloy; *: Brinell hardness.
Sources: [1]: Metals Handbook, 8th ed., Vol. 1 (ASM, Metals Park, 1961); [2]: ASTM F 560-04; [3]:
ASTM F 2063-00; [4]: National Devices & Components, Inc.; [5] Wah Chang (Zircadyne 705
[contains Hf]); see ASTM B 351, F- standard pending.

I1.3 Polymers
The cautionary note previously sounded concerning the reliability of tabu-
lated materials’ properties applies especially to polymers. In the case of this
material class, four additional problems interfere with interpretation of data:

• All polymers are viscoelastic; therefore, mechanical property mea-

surements depend upon the strain rate used in evaluation. Because
viscoelastic materials generally become stiffer and less ductile as
strain rates increase, testing rates should equal or exceed those
expected to be encountered in service.
• Properties of engineering polymers are closely related to average
molecular weight and molecular weight distribution as well as to
curing conditions and time (thermosets) and fabrication tempera-
tures and postfabrication heat treatment (thermoplastics).
• Postsynthesis processing and/or sterilization, especially by 60Co γ
or electron beam irradiation, may alter final properties of some mate-
rials. This is particularly true for ultra high molecular weight poly-
ethylene, for which more than a dozen commercial grades with
varying amounts of irradiation and postirradiation heat treatment
are now available (Kurtz 2004).
130 Biological Performance of Materials: Fundamentals of Biocompatibility

• Many polymers are subject to possible oxidation during room tem-

perature storage; therefore, the chronology of production of the
material and of the device may affect final properties in clinical use.

The data are presented in two tables: Table I.5 for thermosets and Table
I.6 for thermoplastics. Both types of polymers have important roles as bio-
materials, although thermoplastics tend to be preferred due to the relatively
greater ease in fabricating them without low molecular weight leachable
components.
Finally, fatigue behavior of biomedical polymers, especially of PMMA-
type “bone-cements,” is a sufficiently controversial subject that no data are
provided here. The reader is referred to the professional literature.

TABLE I.5
Thermoset Resins
Material: EP PMMA PMMA; PEU PSU SR SR (HP)
10BaSO4
Condition: CR 24 h CR 24 h CR CR CR HV HV
Source: [1] [2,3] [3] [4] [5] [5,6] [6]
Density(g/cm3): 1.11–1.40 1.183 1.088 1.1 1.2 1.12–1.23 1.15
E (tensile) (GPa): 2.4 1.31 2.4–3.1 5.9* 3.7* <1.4* 2.4*
Hardness (Sh. A): — — — 75 88 25–75 52
σy(C) (MPa): — — 15.8 — — — —
σUCS (MPa): 100–170 70 69–125 — — — —
σUTS (MPa): 28–90 28–46 9.7–32 45 40 5.9–8.3 8.3–10.3
Elong. (min.%): 3–6 4.6 2.4–5.4 750 540 350–600 700
Notes: CR: room temperature cured; EP: epoxy; HV: heat vulcanized; PMMA: poly-
methyl methacrylate; SR: silicone rubber; HP: high performance; *: MPa.
Sources: [1]: Modern Plastics Encylopedia, McGraw–Hill, New York, 1990; note: typical
values; not specific medical grades; [2]: ASTM F 451-99a; [3]: Lautenschlager, E.P. et
al., in Functional Behavior of Orthopedic Biomaterials, Vol. II, Ducheyne, P. and Hastings,
G.W. (Eds.), CRC Press, Boca Raton, FL, 1984, 87; [4]: Boretos, J.W. and Pierce, W.S.,
J. Biomed. Mater. Res., 2, 121, 1968; [5]: Braley, S., J. Macromol. Sci.-Chem., A4(3), 529,
1970; see also ASTM F604-94 (withdrawn); [6]: Frisch, E.E., in Polymeric Materials and
Artificial Organs, ACS Symp. 256, C.G. Gebelein (Ed.), American Chemical Society,
Washington, D.C., 1984, 63.
Implant Materials: Properties 131

TABLE I.6
Thermoplastic Resins
Material: PE PE PE PE PLA PMMA PSF PEEK
(UHMW) (UHMW) (UHMW) (UHMW) (STCP)
Condition: MM EX CM HC CM CM IM IM
Source: [1] [1,2] [1–3] [3] [4] [5] [6,7] [8]
Density(g/cm3): 0.927–0.944 0.927–0.944 0.93–0.944 — — 1.186 1.23-1.25 1.28–1.32
E (tensile) (GPa): — 1.24 1.36 2.17 4–5 2.6-3.2 2.3–2.48 3-8.3*
Hardness (Sh. D): 60 60 62 66 — — — —
σy (MPa): 19–21 19–28 19–29 28 — — 65–96 70
σUCS (MPa): — — — — — 80–125 — —
σUTS (MPa): 27–35 37–47 27–40 — 50–60 50–75 106* 90–152
Elong. (min.%): 300 250–300 350 230 2–3 2–10 20–75 4.9**

Notes: IM: injection molded; MM: molded, machined; EX: extruded; CM: compression molded;
HC: high crystallinity; PE (UHMW): ultra high molecular weight polyethylene; PLA
(STCP): polylactic acid stereo copolymer; PMMA: polymethyl methacrylate; PSF: polysul-
fone; *: flexural; **: at yield.
Sources: [1]: ASTM 648-00; [2]: Roe, R.-J. et al., J. Biomed. Mater. Res., 15, 209, 1981; [3]: Depuy
(1989); [4]: Christel, P. et al., in Proceedings of the 1st International Conference on Composites in
Biomedical Engineering, November 19–20, London. Imprint of Luton, Luton, 1985, p. 11/1; note:
resorbable; properties depend upon L/D ratio; [5]: Lautenschlager, E.P. et al., in Functional
Behavior of Orthopedic Biomaterials, Vol. II, Ducheyne, P. and Hastings, G.W. (Eds.), CRC Press,
Boca Raton, FL, 1984, 87; [6]: Dunkle, S.R., in Engineered Materials Handbook, Vol. 2: Engineering
Plastics, Dostal, C.A. (Ed.), American Society for Metals Int., Metals Park, 1988, 200; [7]: ASTM
F 702-98a; [8]: Lewis (1990); see also ASTM F 2026-02.

I1.4 Ceramics
At room and body temperature, ceramic materials suitable for biomedical
applications possess negligible ductility; thus, no tensile or elongation data
are included in Table I.7. Note that strength of ceramics depends very
strongly on density (as percent of ultimate) and grain size, so data on actual
commercial formulations must be sought for engineering use. Data are
provided only for some of the most common structural ceramics; no infor-
mation is provided on resorbable or so-called bioactive ceramics due to
their variety and complexity. (See deGroot, 1983; Hench and Wilson, 1984;
and Manley, 1993.)
132 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE I.7
Ceramic Materials
Material: Al2O3 C C C ZrO2
Condition: HP LTI VT ULTI SHP
Source: [1,2] [3] [3] [3] [4,5]
Density (g/cm3): 3.98 1.7–2.2 1.4–1.6 1.5–2.2 6.1
Grain size(μm): 1.8–4 30–40* 10–40* 8–15* <0.5
E (tensile) (GPa): 400–580 18–28 24–31 14–21 200
Hardness (Hv): 2300 150–250 150–200 150–250 1300
σUFS (MPa): 550 280–560 70–210 350–700 1200
σUCS (MPa): 4500 — — — —
Notes: HP: high purity; LTI: low-temperature isotropic; SHP: sintered, hot isostatic
pressed (5% Y2O3 stabilized); ULTI: ultra low temperature isotropic; VT: vitreous
(glassy); *: angstroms.
Sources: [1]: CeramTec; [2]: ASTM F603-00; ISO 6474; [3]: various; compilation by Inter-
medics Orthopedics (1983); [4]: Christel, P. et al., J. Biomed. Mater. Res., 23, 45, 1989; [5]:
Norton Demarquest; 6ASTM F 2393-04.

I1.5 Composites
Composites, or more properly composite materials, is a term that has come
to describe a wide range of engineered or designed materials. Mechanical
properties of composite materials depend upon the properties of the phases
(matrix and strengthening, or reinforcing, phase or phases) as well as on the
nature of the phase interfaces, their volume fractions, net porosity, and the
local and global arrangement of the reinforcing phases. This complexity
dictates an economy of action here because an entire volume could be (and
has frequently been) devoted to the subject. Thus, as a brief guide, Table I.8
presents properties of some more common reinforcing phases useful in com-
posite biomaterials design, and Table I.9 presents properties of a few repre-
sentative composites that have been evaluated to some degree as
biomaterials.
Implant Materials: Properties 133

TABLE I.8
Reinforcing Phases
Material: E-glass S-glass C-glass C* C* PA PA PA
(low E) (High E) (K29)+ (K49)+ (K149)+
Condition: AN, CF AN, CF AN, CF HT, CF HT, CF CF CF CF
Source: [1,2] [1,2] [1,2] [1] [1] [1] [1] [1]
Density (g/cm 3): 2.62 2.50 2.56 1.76 1.9 1.44 1.44 1.47
Diameter(μm): 3–20 3–20 3–20 7–8 7 12 12 12
E (tensile) (GPa): 72–81 85–89 69 230 390 83 131 286
σUTS (MPa): 3450 4580 3000–5300 3300 2400 2800–3600 3600–4100 3400
Elong. (min.%): 4.9 5.7 4.8 1.4 0.6 4.0 2.8 2.0

Material: Al2O3 Al2O3– SiO2 βSiC αSiC βSiC BN B 4C

48SiO2
Condition: CF DF CF CF WH WH CF WH
Source: [1] [1] [1] [1] [1] [1] [3] [3]
Density (g/cm 3): 3.95 2.73 2.2 2.55 3.2 3.19 1.91 2.52
Diameter(μm): 20 2–3 9 10–15 0.6 0.1–0.5 7 —
E (tensile) (GPa): 379 100 69 180–200 690 400–700 90 483
σUTS (MPa): 1380 1900 3450 2500–3200 6900 3000–14000 1380 13800

Notes: AN: annealed; CF: continuous fiber; DF: discontinuous fiber; WH: whisker; HT: heat
treated; *: polyacrylonitrile (PAN) precursor; +: Kevlar™ (DuPont).
Sources: [1]: Composites, Engineered Materials Handbook, Vol. 1, ASM International, Metals Park,
1987; [2]: Lubin, G. (Ed.), Handbook of Composites, Van Nostrand, New York, 1982; [3]: Schwartz,
M.M., Composite Materials Handbook, McGraw–Hill, New York, 1984.

TABLE I.9
Composite Materials
Material: PMMA- C60SiC C60SiC CFRC CFRC EP- CFPSU PE
2C (5% por.) (7% por.) 12.5C (UHMW)-
Condition: RT ET 10C
Source: [1] [2] [3] [2] [3] [4] [5] [6]
Density (g/cm3): — 2.6 2.4 1.7 1.78 — — 0.98
E (tensile) (GPa): 5.52 100 80–90 140 40–58 14 110 1.94**
σUCS(MPa): — 1000 250–370 800 230–320 — — 14.2
σUTS (MPa): 38 220* 220–360* 800* 350–600 200* 1600* 22
Elong. (min.%): 0.7 <1 — >4 — — 1.3 150
Notes: PMMA: polymethyl methacrylate; RT: room temperature cured; ET: 70°C cured; CFRC:
carbon fiber-reinforced carbon; CFPSU: continuous fiber carbon reinforced polysul-
fone; EP: epoxy; por.: open porosity; *: flexural strength; **: estimated.
Sources: [1]: Pilliar, R.M. et al., J. Biomed. Mater. Res., 10, 893, 1976; [2]: Brückmann, H. and
Hüttinger, K.J., Biomaterials, 1, 67, 1980; [3]: Christel, P. et al., J. Biomed. Mater. Res, Appl.
Biomater., 21(A2), 191, 1987; [4]: Hastings, G.W., Composites, 7, 193, 1978; [5]: Claes, L. et al.,
in Biological and Biomechanical Performance of Biomaterials, Christel, P., Meunier, A. and Lee,
A.J.C. (Eds.), Elsevier, Amsterdam, 1986, 81; [6]: Zimmer, 1978.
134 Biological Performance of Materials: Fundamentals of Biocompatibility

References
Ahmed, T. et al., A new low-modulus, biocompatibile titanium alloy, Proc. Inst. Metal.
8th World Titanium Conf., 1996, 742.
Boretos, J.W. and Pierce, W.S., Segmented polyurethane: a polyether polymer, J.
Biomed. Mater. Res., 2, 121, 1968.
Borowy, K.-H. and Kramer, K.-H., On the properties of a new titanium alloy
(TiAl5Fe2.5) as implant material, in Titanium Science and Technology, Vol. 2,
Luterjering, G., Zwicker, U. and Bunk, W. (Eds.), Deutsche Gesell. f. Met-
allkunde e. V., Obureresel, 1985, 1381.
Braley, S., The chemistry and properties of the medical-grade silicones, J. Macromol.
Sci.-Chem., A4(3), 529, 1970.
Brückmann, H. and Hüttinger, K.J., Carbon, a promising material in endoprosthetics.
Part 1: the carbon materials and their mechanical properties, Biomaterials, 1, 67,
1980.
Christel, P. et al., Mechanical properties and short-term in-vivo evaluation of yttri-
um–oxide partially stabilized zirconia, J. Biomed. Mater. Res., 23, 45, 1989.
Christel, P. et al., Development of a carbon–carbon hip prosthesis, J. Biomed. Mater.
Res, Appl. Biomater., 21(A2), 191, 1987.
Christel, P. et al., PGA (polyglycolic acid)-fiber-reinforced-PLA (polylactic acid) as
an implant material for bone surgery, in Proceedings of the 1st International
Conference on Composites in Biomedical Engineering, November 19–20, London.
Imprint of Luton, Luton, 1985, p. 11/1.
Claes, L., Hüttner, W. and Weiss, R., Mechanical properties of carbon-fiber reinforced
polysulfone plates for internal fracture hardware, in Biological and Biomechanical
Performance of Biomaterials, Christel, P., Meunier, A. and Lee, A.J.C. (Eds.), Else-
vier, Amsterdam, 1986, 81.
Composites, Engineered Materials Handbook, Vol. 1, ASM International, Metals Park,
OH, 1987.
deGroot, K. (Ed.), Bioceramics of Calcium Phosphate, CRC Press, Boca Raton, FL, 1983.
Davidson, J.A. et al., New surface-hardened, low modulus, corrosion-resistant Ti-
13Nb-13Zr alloy for total hip arthroplasty, Bio-Med. Mater. Eng., 4(3), 231, 1994.
Depuy, Warsaw, IN, 1989.
Dunkle, S.R., Polysulfones (PSU), in Engineered Materials Handbook, Vol. 2: Engineering
Plastics, Dostal, C.A. (Ed.), American Society for Metals Int., Metals Park, 1988,
200.
Frisch, E.E., Silicones in artificial organs, in Polymeric Materials and Artificial Organs,
ACS Symp. 256, Gebelein, C.G. (Ed.), American Chemical Society, Washington,
D.C., 1984, 63.
Metals Handbook, 8th ed., Vol. 1, ASM, Metals Park, OH, 1961.
Hastings, G.W., Carbon fiber composites for orthopedic implants, Composites, 7, 193,
1978.
Hench, L.L. and Wilson, J., Surface-active biomaterials, Science, 226, 630, 1984.
Kurtz, S.M., The UHMWPE Handbook: Ultra-High Molecular Weight Polyethylene in Total
Joint Replacement, Academic Press (Elsevier), New York, 2004.
Lautenschlager, E.P., Stupp, S.I. and Keller, J.C., Structure and properties of acrylic
bone cement, in Functional Behavior of Orthopedic Biomaterials, Vol. II, Ducheyne,
P. and Hastings, G.W. (Eds.), CRC Press, Boca Raton, FL, 1984, 87.
Implant Materials: Properties 135

Lewis, G., Selection of Engineering Materials, Prentice Hall, Englewood Cliffs, NJ, 1989.
Long, M. and Rack, H.J., Titanium alloys in total joint replacement — a materials
science perspective, Biomaterials, 19, 1621, 1998.
Lubin, G. (Ed.), Handbook of Composites, Van Nostrand, New York, 1982.
Manley, M.T., Introduction, in Hydroxyapatite Coatings in Orthopedic Surgery, Geesink,
R.G.T. and Manley, M.T. (Eds.), Raven Press, New York, 1993, 1.
Modern Plastics Encylopedia, McGraw–Hill, New York, 1995.
Pilliar, R.M. et al., Carbon fiber-reinforced bone cement in orthopedic surgery, J.
Biomed. Mater. Res., 10, 893, 1976.
Roe, R.-J. et al., Effect of radiation sterilization and aging on ultrahigh molecular
weight polyethylene, J. Biomed. Mater. Res., 15, 209, 1981.
Schwartz, M.M., Composite Materials Handbook, McGraw–Hill, New York, 1984.
Wang, K., Gustavson, L. and Dumbleton, J., The characterization of Ti-12Mo-6Zr-2Fe
— a new biocompatible titanium alloy developed for surgical implants, in Beta
Titanium Alloys in the 1990s, Eylon, D., Boyer, R.R. and Koss, D.A. (Eds.), Min.
Met. and Materials Society, New York, 1993, 49.
Zimmer, Nonauthored technical material, Warsaw, IN, 1978.

Bibliography
2004 Annual Book of ASTM Standards, Vol. 13.01, Medical Devices, ASTM International,
West Conshohocken, PA, 2004.
Ashby, M., Materials Selection in Mechanical Design, 2nd ed., Butterworth–Heinemann,
London, 1999.
Ashby, M. and Johnson, K., Materials and Design: The Art and Science of Material
Selection in Product Design, Butterworth–Heinemann, London, 2002.
Black, J. and Hastings, G. (Eds.), Handbook of Biomaterial Properties, Chapman & Hall,
London, 1998.
Boretos, J.W., Concise Guide to Biomedical Polymers, Charles C Thomas, Springfield, IL,
1973.
Boutin, P. et al., The use of dense alumina–alumina ceramic combination in total hip
replacement, J. Biomed. Mater. Res., 22, 1203, 1988.
Brown, S.A. and Lemons, J.A. (Eds.), Medical Applications of Titanium and its Alloys,
STP 1272. ASTM, West Conshohocken, PA, 1992
Bush, R.B., A bibliography of monographic works on biomaterials and biocompati-
bility, J. Appl. Biomater., 4, 195, 1993.
Catledge, S.A. et al., Nanostructured ceramics for biomedical implants, J. Nanosci.
Nanotechnol., 2(3–4), 293, 2002.
Cowin, S.C., Bone Mechanics Handbook, 2nd ed., CRC Press, Boca Raton, 2002.
Disegi, J.A., Kennedy, R.L. and Pilliar, R. (Eds.), Cobalt-Base Alloys for Biomedical
Applications, STP 1365, ASTM, West Conshohocken, PA, 1999.
Donachie, M.J., Jr. (Ed.), Superalloys Source Book, ASM Int., Metals Park, 1984.
Donachie, M.J., Jr. (Ed.), Titanium: A Technical Guide, ASM Int., Metals Park, 1984.
Epinette, J.-L. and Manley, M.T. (Eds.), Fifteen Years of Clinical Experience with Hy-
droxyapatite Coatings in Joint Arthroplasty, Springer–Verlag, Paris, 2004.
Fisher, L.W., Selection of Engineering Materials and Adhesives, Taylor & Francis, Boca
Raton, FL, 2005.
136 Biological Performance of Materials: Fundamentals of Biocompatibility

Ito, A. et al., In vitro biocompatibility, mechanical properties, and corrosion resistance

of Ti-Zr-NB-TA-PD and Ti-Sn-Nb-Ta-Pd alloys, J. Biomed. Mater. Res., 29, 893,
1995.
Kingery, W.D., Introduction to Ceramics, 2nd ed., John Wiley & Sons, New York, 1976.
Kokubo T. et al., Bioactive metals: preparation and properties, J. Mater. Sci. Mater.
Med., 15(2), 99, 2004.
Lambda, N.M.K., Woodhouse, K.A. and Cooper, S.L. (Eds.), Polyurethanes in Biomed-
ical Applications. CRC Press, Boca Raton, FL, 1997.
Mishra, A.K. et al., Ti13Nb13Zr: A new low modulus, high strength, corrosion resis-
tant, near beta alloy for orthopaedic implants. In: Eylon, D., Boyer, R.R. and
Koss, D.A., (Eds.), Beta Titanium Alloys in the 1990’s. Warrendale, PA: The Min-
erals, Metals and Materials Society, 1993, 61.
Pelton, A.R., Stoeckel, D. and Duerig, T.W., Medical uses of nitinol, Mater. Sci. For.,
327–328, 63, 2000.
Portnoy, R.C. (Ed.), Medical Plastics: Degradation, Resistance, and Failure Analysis, Plas-
tics Design Library, Norwich, CT, 1998.
Ratner, B.D. and Bryant, S.J., Biomaterials: where we have been and where we are
going, Annu. Rev. Biomed. Eng., 6, 41, 2004.
Ravglioli, A. and Karjewski, A. Bioceramics and the Human Body, Kluwer Academic,
London, 1992.
Santavirta S. et al., Alternative materials to improve total hip replacement tribology,
Acta Orthop. Scand., 74(4), 380, 2003.
Semlitsch, M., Staub, F. and Weber, H., Titanium–aluminium–niobium alloy devel-
opment for high-strength surgical implants, Biomed. Technik, 30, 334, 1985.
Shackelford, J.F., Alexander, W. and Park, J.S. (Eds.), CRC Practical Handbook of Ma-
terials Selection, CRC Press, Boca Raton, FL, 1995.
Szycher, M. (Ed.), Biocompatible Polymers, Metals, and Composites, Technomic, Lancast-
er, PA, 1983.
Toni, A. et al., Ceramics in total hip arthroplasty, in Encyclopedic Handbook of Bio-
materials and Bioengineering, Part A, Vol. 1, Wise, D.L. et al. (Eds.), Marcel Dekker,
New York, 1995, 1501.
Willert, H.-G., Buchorn, G.H. and Eyerer, P. (Eds.), Ultra-High Molecular Weight Poly-
ethylene as a Biomaterial in Orthopedic Surgery, Hogrefe & Huber, Toronto, 1990.
 Part III

Host Response: Biological

Effects of Implants
8
The Inflammatory Process

8.1 Introduction
Inflammation is a nonspecific physiological response to tissue damage in
animal systems. It arises as a response to trauma, infection, intrusion of
foreign materials, or local cell death, or as an adjunct to immune or neoplastic
responses. If the initiating agent causes damage to or frank rupture of vas-
cular tissue, blood coagulation, which is related, may be superimposed on
the inflammatory response. The general inflammatory response will be con-
sidered in this chapter. Coagulation is dealt with in Chapter 9, immune,
or specific, responses are dealt with in Chapter 12, and neoplastic transfor-
mation, whether chemically or mechanically mediated, is the subject of
Chapter 13.

8.2 The Inflammatory Response

8.2.1 Introduction
The four classical clinical signs of inflammation in animals and humans are
redness (rubor), swelling (tumor), pain (dolor), and heat (calor). The magnitude
of these signs is related to the intensity and extent of the inflammatory
process. The presence of an implant does not produce additional signs or
symptoms but may alter their severity and duration.

8.2.2 Initial Events

The first events that occur in response to an inflammatory stimulus are rapid
dilation of the local capillaries and an increase in the permeability of their
endothelial cell linings. The dilation (vasodilation) arises from the activation
of a coagulation factor, factor XII (the Hageman factor), most probably by

139
140 Biological Performance of Materials: Fundamentals of Biocompatibility

contact with collagen or a foreign protein or material. Through the interme-

diate activation of a polypeptide, kallikrein, this leads to conversion of a
group of additional molecules to kinins. The kinins are strong mediators of
vasodilation and endothelial permeation.
The vasodilation leads to an increase in blood entry into the capillary beds.
Local aspect ratio changes in the capillaries, combined with loss of plasma
through the capillary walls and a tendency for the platelets and erythrocytes
to become “sticky,” lead to slower flow and sludging. This results in the first
clinical sign, redness (erythema), simply reflecting a higher local concentra-
tion of erythrocytes.
The increased permeability of the capillary endothelium allows fluid to
move into the surrounding tissue bed. Under normal conditions there is a
10 to 15 mmHg positive pressure differential between the arteriole end of a
capillary and the external tissue bed. However, the endothelium is tight and
permits only a very slow flow of water and small molecules into the sur-
rounding tissue. This fluid is normally drained away by local lymphatic
vessels maintaining a constant tissue volume. As permeability increases,
water and larger molecules, including normal plasma proteins and locally
activated kinins, move into the tissue. The increased fluid influx, if not
promptly balanced by increased lymphatic drainage, distends or swells the
tissue. The local lymphatics may be constricted or blocked due to the original
trauma, occlusion by cell fragments, or hydraulic compression. Additionally,
the presence of plasma fractions raises the local osmotic pressure and tends
to hold fluid in place. Thus, swelling (edema), the second clinical sign, is a
usual early concomitant of inflammation. In the extreme case, all inflow and
outflow of fluid may be prevented, leading to the so-called “compartment
syndrome” that, if not promptly relieved, results in cell death and tissue
necrosis.
Pain, the third clinical sign, results from at least two causes. First, edema
may activate local deep pain receptors. In patients, this is felt as a throbbing
pain as it peaks repetitively with peak systolic pressure. Second, the kinins
act directly on nerve ends to produce pain sensation. The familiar acute pain
of a bee sting is associated with the activation of a kinin called bradykinin
in bee venom.
The origin of the local heating effect, the fourth sign, is unclear. In the
general case, it may be associated with local disturbances of fluid flow in
the presence of increased cellular metabolic activity (and resulting heat pro-
duction). Although not yet shown conclusively, a group of contaminants
termed pyrogens, which are known to cause systemic fever, may be gener-
ated locally by tissue necrosis or, in the presence of infection, as a result of
activation by bacterial or viral toxins, especially endotoxin. Fine particles or
bacterial fragments left on implants that are sterilized after inadequate clean-
ing may also be pyrogenic. The presence of such pyrogens or other surface
contaminants may mask the intrinsic local host response.
These early events of inflammation are largely chemical in origin and
effect. Shortly after initiation, however, a series of cellular invasions take
The Inflammatory Process 141

place. These cells are responsible for the removal of dead tissue and the
repair of the resulting defect.

8.2.3 Cellular Invasion

The first new cells to appear at the site of injury are neutrophils. They are
the most common of the granulocytes, also termed polymorphonuclear leu-
kocytes (PMNs) due to their horseshoe-shaped, multilobed nuclei. Because
their granules are readily stainable by neutral dyes, they can be distinguished
from related cell types, basophils and acidophils.*
At the earliest stages of the inflammatory response, the neutrophils become
sticky. At first they stick momentarily to the capillary endothelium and then
are released. More of them stick and remain for longer periods. Then, neu-
trophils begin to penetrate between the endothelial cells and move into the
surrounding damaged tissue. Neutrophil emigration (diapedisis) begins
minutes to hours after insult and may continue for as long as 24 hours. The
time course varies with the nature and severity of the insult. Although
following an initial but transient episode of increased vascular permeability,
this emigration is accompanied by longer sustained endothelial permeability.
This second phase of permeability parallels the maturation of edema and
erythema and subsides as neutrophil emigration draws to a halt.
The primary role of the neutrophil is phagocytosis. This process, the
engulfing and degradation or digestion of fragments of tissue or material,
is common to this and several other cell types. What distinguishes each cell
type in regard to its phagocytic function is the nature and amount of the
degradative enzymes (lysozymes) that each has available. Lysozymes are
stored as cytoplasmic granules, perhaps 200 per neutrophil that are readily
seen in the electron microscope. Before encountering foreign particles,
phagocytic cells are inactive; after such contact, they become activated. Acti-
vation is characterized by a change in metabolic activity, especially an
increase in oxygen consumption (the “respiratory burst”), a change in cell
shape, and an internal degranulation as lysozymes are released into vacu-
oles, termed phagosomes, or more particularly, primary lysozomes.
Phagocytosis proceeds from an initial contact between the cell membrane
and a foreign material (inorganic or organic). Neutrophils are specialized to
find and phagocytize bacteria; how they find the foreign material particles,
such as splinters or wear debris, is poorly understood, but there appear to
be several mechanisms. The neutrophil may be attracted by a particular
chemical composition (chemotaxis), by local pH differences, or by electro-
chemical factors associated with the foreign particle and its surroundings.
In addition, a group of native molecules, opsonins, coat foreign surfaces
(opsonization), thus rendering them chemoattractive. Opsonins are small
molecules excreted by a variety of cells and include the activated form of

* So called because of their granule stainability by, respectively, basic and acidic histological
dyes.
142 Biological Performance of Materials: Fundamentals of Biocompatibility

complement factor five (C5a). When in contact with and recognizing a for-
eign material through antigen (opsonin)-membrane receptor binding, a dim-
ple develops in the cell wall (invagination) of the phagocyte. The particle is
drawn in and the dimple closes, often breaking off to form a vacuole, termed
a secondary lysozome, which leaves the particle inside the cell and sur-
rounded by an everted unit membrane. This vacuole merges with primary
lysosomes, whose contents are then released into the vacuole and attack
foreign or denatured proteins. This process of degranulation, as previously
noted, is accompanied by a respiratory burst involving production of hydro-
gen peroxide and superoxide anions, which act nonspecifically to digest
nonproteinaceous foreign materials that are not attacked by lysozymes. The
vacuole may clear and be absorbed, or its contents, presumably altered to
be less objectionable, may be released outside the cell.
The complement system, of which factor C5 is an element, is a complex
series of glycoproteins and protein inhibitors present in the circulating and
interstitial fluids of mammals. It constitutes a powerful, noncellular part of
the biological defense system against foreign materials. It can be activated
by contact with antigen–antibody complexes (see Chapter 12), bacterial pro-
teins and polysaccharides, endotoxins, and many polymers, of natural or
synthetic origin. Once activation is initiated, it proceeds along two pathways
through a series of enzymatic cascade amplifying steps, much like the blood
coagulation cascade (see Chapter 9). The end products are activated com-
plement factors, which can mediate membrane damage directly, or molecular
fragments that can influence inflammatory and immune processes (Ander-
son and Miller 1984; Frank 1979).*
Accompanying the neutrophils is a far smaller number of another type of
leukocyte, the eosinophil. These are similar in structure to the neutrophils
and, although they perform a similar phagocytic function, they can also
phagocytize antigen–antibody complexes (see Section 12.2.1). Their name
stems from the observation that, unlike the neutrophils, their granules can
be stained by eosin, an acidic dye. They contain a different distribution of
enzymes and are usually only present in significant numbers in association
with immune responses.
The neutrophils and eosinophils serve as the first active line of defense
against foreign material in tissue. Leukocytes have a life span of only hours
in blood and a few days in tissue. They are end-state cells and cannot
replicate by division (mitosis). Although capable of oxidative phosphoryla-
tion, they obtain their energy primarily from the anaerobic metabolism of
stored glycogen, so they can persist in areas with highly disturbed metabolic
states. After fulfilling their function, they die rapidly. Normally, they consti-
tute an “emergency squad” whose duties are later supplanted by another
cell, the monocyte. Thus, the presence of extensive numbers of live neutro-
phils in tissue may be interpreted as evidence of continued inflammatory
challenge.

* The complement system is discussed at greater length in Section 12.2.2.

The Inflammatory Process 143

hydrolases

Exocytosis
nucleus

go
Endocytosis

lgi
bo
die
s
1° lysosomes
Phagocytosis rough
and endoplastic
Pinocytosis reticulum

2° lysosomes autophagic
vacuole
phagolysosome
residual
body

FIGURE 8.1
Phagocytic cell functions.

The monocyte is the largest of the freely circulating leukocytes, extending

to perhaps 15 μm in diameter. It is distinguished from the neutrophil by its
larger size and its single, centrally located large nucleus. Once in tissue, it
becomes a macrophage or mononuclear phagocyte (MNP) (Figure 8.1). In
addition to circulating monocytes, a resident population of monocytes in
tissue can also rapidly become macrophages and migrate to the site of injury.
From whatever source, macrophages arrive at the site of inflammation after
the neutrophilic invasion has begun to subside and concentrate in apprecia-
bly smaller numbers. Their role is similar to that of the neutrophil in that
they can actively phagocytize materials and digest them. They also possess
specialized membrane sites that mediate specific reactions. When activated
by a suitable challenge, the macrophage possesses a number of functional
capabilities, as seen in Figure 8.1.
The activated macrophage is not an end state cell; it has an active aerobic
glycolysis cycle and can undergo a number of transformations. In addition
to releasing lysozymes, both internally and externally, related to attempts to
digest foreign materials, activated macrophages synthesize and release a
wide range of biochemical factors which can mediate the activity of many
other cells including lymphocytes, fibroblasts, osteoblasts, osteoclasts, and
foreign body giant cells (Ziats et al. 1988).
Macrophages may multiply by mitosis (Figure 8.2). By fusion, the macro-
phage is the progenitor of the next major cell type seen, the multinuclear
foreign body giant cell (FBGC). This cell is up to 80 μm in diameter, is found
primarily in foreign body or implant sites, and can be more aggressive and
active than the neutrophil, eosinophil, or macrophage. Because of its larger
size, it can phagocytize still larger particles. It has a relatively short life span
144 Biological Performance of Materials: Fundamentals of Biocompatibility

BONE MARROW BLOOD

lysosomes

PROMONOCYTE MONOCYTE MACROPHAGE

MITOSIS TISSUE MONOCYTE

IMPLANT SITE

FOREIGN BODY
GIANT CELL ACTIVATED MACROPHAGES

FIGURE 8.2
Development of the foreign body giant cell. Key: I = insult; CP = cellular proliferation; R =
reorganization; C = chronic host response.

that is on the order of days. Thus, the presence of multinuclear giant cells
long after implantation would suggest the presence of a chronic FB reaction.
It is possible that macrophage fusion to form multinuclear giant cells is
specifically stimulated by foreign bodies (Mariano and Spector 1974; Cham-
bers 1977).
The presence and activity of phagocytic cells is particularly related to the
presence of “small” particles. The relationship between size and stimulus is
not understood, except in a general way. Maximum stimulus seems to occur
when the average particle size is in the 0.1- to 1.0-μm range. Particles bigger
than 50 μm do not excite a reaction greater than bulk materials, unless they
possess a dimension in the size range previously indicated (for instance, long
slender fibers). Large particles or even bulk implants may still activate
phagocytic cells, resulting in external release of lysozymes and oxidative
products. This process is often termed frustrated phagocytosis and may
prevent the phagocytes from performing their normal functions. Smaller
particles, especially less than 0.1 μm (<100 nm) in major dimension (variously
termed microparticles or nanoparticles), tend to be phagocytozed in clumps
or clusters in a manner termed pinocytosis by which immiscible liquids are
taken up by phagocytic cells.
Although the intrinsic chemical activity* of phagocytozable particles does
not seem to be primarily responsible for their cellular stimulatory effect, the

* Care must be taken in in vitro studies to avoid drawing erroneous conclusions due to particle
surface contamination by endotoxin (Brooks et al. 2002) or other foreign materials that may affect
behavior of phagocytes.
The Inflammatory Process 145

cellular response depends indirectly upon ease of degradation. If the particle

is toxic to the ingesting cell and/or is expelled unchanged, the result may
be massive accumulations of dying and dead neutrophils and macrophages.
The resulting “pus” or cellular debris accumulation (caseation) resembles
that which accompanies massive bacterial infection. Unsuccessful or success-
ful phagocytosis is a mechanism for moving foreign particulate material
away from the implant sites. Cells and cell fragments pass into the local
lymphatic drainage system and may accumulate in local and regional lymph
nodes. Small particles may accumulate in such nodes, as well as in remote
organs such as liver and spleen (Urban et al. 2000), or be cleared into the
lungs via the lymphatics and exhaled or swallowed (Styles and Wilson 1976).
In numbers and in temporal displacement of cell types, the progression
described here reflects the increased threat that the intrusive agent repre-
sents. In parallel with this progression, an invasion of blood-borne lympho-
cytes occurs. These cells, which normally pass directly through the
endothelial lining and are taken up by the lymphatic processes, will pass
through the lymphatic nodes and eventually return to arterial circulation.
Their function in inflammation is not well understood, and they are associ-
ated primarily with immunologic response (see Chapter 12). Their numbers
do increase in areas of inflammation, and they are found associated with
neutrophils and foreign particles in regional lymph nodes. It is possible that
they respond to foreign proteins or to proteins denatured by foreign contact
and have some ability to phagocytize particles. They may participate in
fibroplasia, the final stage of inflammation.

8.2.4 Remodeling
Successful response to an inflammatory challenge will result in a locally
decreased tissue mass. Dead cells have been phagocytized and removed by
neutrophils and macrophages. Cells will be produced in situ by mitosis of
the cell types present or by maturation of more primitive precursor or stem
cells. This newly forming tissue is termed granulation tissue because of the
pebbly specular appearance it has when seen growing at a free surface. The
“pebbles” are vascular “buds” — capillary and arteriole loops that grow
rapidly out from stable tissue into the disturbed area. They “burrow”
through, bringing the benefits of improved circulation and stimulating cel-
lular activity.
In addition, large amounts of muccopolysaccharides and collagen will be
synthesized. The resulting tissue expansion is termed fibroplasia. These new
tissue elements, forming the familiar scar, create a scaffold for cellular recon-
struction and remodeling of the damaged area. The primary mediators of
fibroplasia are fibroblasts, although in vitro studies suggest that a wide
variety of cells can be stimulated to produce collagen and, to a lesser degree,
muccopolysaccharides.
146 Biological Performance of Materials: Fundamentals of Biocompatibility

The remodeling of granulation tissue proceeds differently in different tis-

sues. In skin, the reformation of tissue may be nearly complete, except for
a possible absence of hair follicles and a small residual collageneous scar
representing the collapsed remnant of the capsule that formed around the
injury site. Bone has the ability to remodel completely to such a degree that
it is said to regenerate. At the other extreme, articular (hyaline) cartilage
never remodels but is repaired or replaced by formation of a loose tissue
called fibrocartilage, which is different from normal cartilage in composition
and structure and eventually deteriorates under repeated mechanical stress.
Remodeling is also the primary mechanism involved in tissue adaptation
(see Chapter 10).

8.2.5 Capsule Formation

The maturation of the scar tissue (in soft-tissue sites) marks the end of the
process termed inflammation. The continuing presence of an implant pre-
vents the collapse of the capsule that forms around the injured tissue and
its maturation into a scar. The degree of scarring or capsule formation seen
at later times depends upon the degree of original insult, the amount of
subsequent cell death, and the location of the site.
Grading the host response at the implant site by a method such as that
described in Appendix 1 and Appendix 2 in Chapter 18 actually evaluates
the sum of two responses: the intrinsic inflammatory response to trauma
and the host response to the implant. The response to trauma is relatively
brief, passing through phases of insult, cellular proliferation, and reorgani-
zation in 1 to 2 weeks, as seen in the lower left of Figure 8.3. The host response
to an implant is similar in its early phases, but then enters a chronic phase.
The time required to enter the chronic phase or, in other words, achieve
steady state, varies with the degree of host response elicited. Thus, although
it may be safe to evaluate a low-reactivity implant after 3 to 4 weeks, one
may not gain a true picture of the host response to a high-reactivity implant
unless studies are extended to 6 or even 8 weeks after implantation.
Explorations of an implant site 4 or more weeks after implantation will
usually reveal a relatively acellular fibrous capsule. This capsule is main-
tained by the continuing presence of the implant. In the absence of an implant
(or after its removal or resorption), the capsule may contract or collapse into
a residual scar or be completely remodeled, as described earlier. Spindle-
shaped fibroblasts are usually associated with the capsule, and small num-
bers of macrophages may be found in the vicinity. The presence of neutro-
phils suggests the possibility of a continuing inflammatory challenge,
including infection. Detection of multinuclear (foreign body) giant cells
(FBGCs) suggests the production of small particles by corrosion, depolymer-
ization, dissolution or wear, and a continuing general tissue response. Obser-
vation of large numbers of lymphocytes suggests the possibility of a specific
immune response (see Chapter 12).
The Inflammatory Process 147

CP Chronic
Response
Trauma + “High Reactivity” Implant
I

CP Chronic
I

CP Trauma + “Low Reactivity” Implant

I R

Time (weeks) ~2 4 6

Key: I = insult C = chronic host response

R = reorganization CP = cellular proliferation

FIGURE 8.3
Variations of local host response with implant “reactivity.” Key: C = chemical mediation; M =
mechanical mediation; E = electrical mediation; “?” = may be a mediating factor or may progress
to resolution; ⊕ = enhances effect present without implant.

The thickness of the fibrous capsule is related to several factors. Materials

that are chemically active, such as metals that corrode freely or polymers with
leachable constituents, will mediate formation of a capsule whose thickness
is directly proportional to the rate (and thus local concentration) of release of
these small molecules. In addition to concentration, the chemical nature of the
released material is important. It may be cytotoxic, inhibitory, or neutral. The
combination of these factors may lead to specific capsular morphology at a
histological and an ultrastructural level, which is characteristic of the particular
material composition of the implant. These differences can only be fully appre-
ciated when the responses to otherwise identical implants of different compo-
sitions are studied (McNamara and Williams 1982).
Mechanical factors are also important in mediating capsule formation.
Capsule thickness is presumed to increase with increased relative motion
between implant and tissue. In extreme cases, a fluid-filled bursa mimicking
a synovial capsule may form; these bursae may be painful. The shape of the
implant also affects the fibrous capsule thickness. The capsule will be thicker
over edges and sharp changes in surface features. Thus, the capsule around
a rectangular slab of reactive material will be dog-bone or club shaped. For
this reason, the phenomenon is called clubbing (Wood et al. 1970).* Electrical
currents, such as those emanating from an implanted stimulating electrode,
also produce capsules whose thicknesses are related to current density.
Because electrodes can also mediate changes in local pH and pO2 as well as
releasing corrosion products,** effects due to direct electrical (faradic) and
indirect electrochemical (electrodic) stimulation may easily be confused.
* See Section 18.2.1 for a more complete discussion of this point.
** See Section 4.7.2 and Section 4.8.
148 Biological Performance of Materials: Fundamentals of Biocompatibility

In addition to depending upon tissue type, these general responses are

species dependent and, in humans, may have age dependence. Furthermore,
identification of intrinsic tissue response to an implant may be complicated
by the presence of an overlying acute or chronic infection. Section 8.3 will
take up the relationship between implants and infection as a special case of
the inflammatory response to implant materials.

8.2.6 Resolution
When injury occurs to tissue, the overall aim of the process initiated is to
produce a return to the status quo ante, a re-establishment of homeostasis.
The presence of an implant in an operative site must, of necessity, prevent
attainment of the original condition, but a steady state is possible in many
cases. Reaching this condition is termed resolution. Figure 8.4 shows an
overall schematic of the progression of the local host response from insult
through resolution in the presence of an implant. The nature of the physical
signals that mediate the local host response is still somewhat unclear; the
symbols C, M, and E are used here to reflect the most likely situation. In
particular circumstances, it is clear that these signals can interact in syner-
gistic or antagonistic ways (Chesmel et al. 1995).

INCISION IMPLANT

M C
C,
,E

M
C,M

M ,E
C, Particulate Soluble
Debris Products
C,
,E M C
C,M ? ?

Acute Chronic Immune Neoplastic

response response response response

Granuloma
?

RESOLUTION
Extrusion Resorption Integration Incapsulation

Key: C = chemical mediation M = mechanical mediation

= enhances effect present E = electrical mediation
without implant
? = may be a mediating factor or may progress to resolution

FIGURE 8.4
Schematic of development of local host response.
The Inflammatory Process 149

Four types of response are noted: acute, chronic, immune, and neoplastic.
The first two usually reach resolution, the third may lead to a granuloma (a
benign tumorous condition due to continued tissue elaboration without
progressive remodeling), and the fourth does not resolve and always repre-
sents a failure of the implant. If resolution — that is, an achievement of a
final state after which no more progressive biological changes occur — is a
possibility, a distinction may be made among four possible outcomes:

• Extrusion: if the implant is in contact with epithelial tissue, the local

host response will be the formation of a pocket or pouch continuous
with the adjacent epithelial membrane. The process is termed mar-
supialization due to the similarity of the resulting structure to a
kangaroo’s pouch. In the case of the external epithelium (skin), this
results in the effective externalization or extrusion of the implant
from the host.
• Resorption: if the implant is resorbable, the implant site eventually
resolves to a collapsed scar or, in the case of bone, may completely
disappear.
• Integration: in a very limited number of cases, such as the implan-
tation of pure titanium (Albrektsson et al. 1983) or tantalum (Aron-
son et al. 1977) in bone, a close, possibly adhesive, approximation
of nearly normal host tissue to the implant is possible without an
intervening capsule, although inflammatory cells may persist in
small numbers. This special case of resolution is termed osseointe-
gration (sometimes osteointegration).
• Incapsulization (the most usual response as previously described):
if the implant is placed in a location where bone may form, such as
within a medullary space, and integration does not take place, the
capsule may become mineralized, in which case the structure is
called a sequestrum.

In addition, it is possible with the passage of time for granulomas to resolve

if the challenge presented by the implant changes or subsides.
Unlike the chronic granuloma and neoplastic transformation responses,
whether each of these resolution outcomes represents success or failure of
the implant depends on the circumstances — that is, upon the desired con-
sequences of the insertion of the implant. This is the basic idea of biocom-
patibility, as discussed earlier: biological performance in a specific
application that is judged suitable to that situation.
150 Biological Performance of Materials: Fundamentals of Biocompatibility

8.3 Infection
8.3.1 Types
A common phrase in research and clinical situations is “infected implant.”
This is a misnomer because the tissue surrounding the implant is infected
rather than the implant. Infection is specifically the invasion and multipli-
cation of microorganisms in tissue. However, the misuse of the term serves
as a reminder of special problems that arise when bacterial infection is
associated with an implant in animal experimentation or in human clinical
practice. Immediate postoperative implant site infections were a consider-
able clinical problem in the 1960s and 1970s; incidence rates approached 5
to 10% in some series. Now, with a greater appreciation of the problem and
better operating room techniques, rates are well below 1% for most perma-
nent devices.
Three types of infection are associated with implants. The first, the super-
ficial immediate infection, is due to the growth of organisms on the skin (or
near it) in association with an implant. Examples of this include suture
infections and growth of microorganisms under burn dressings. These infec-
tions can usually be traced to bacteria residing normally on human skin,
such as Staphylococcus aureus or Staphylococcus epidermidis, or to airborne
bacteria that are trapped and cultured in the moist conditions at the super-
ficial site.
The second and third types are, collectively, deep infections. The second,
the deep immediate infection, is the usual type of (now) low-frequency
infection seen immediately after surgery. The bacteria responsible are usually
skin-dwelling types, generally staphylococci, carried into the implant site
during the surgical procedure. Less commonly, they are airborne types, and,
even less frequently, bacteria already present, but relatively inactive, in the
operative site. This last source reflects the physical disruption of tissue dur-
ing surgery, resulting in the release of material in cysts or sequestered sites
into areas where bacteria experience better growth conditions. Infections of
this type, although extremely infrequent, may arise in patients with a pre-
vious history of infection in the operative site, such as a septic joint, or a
systemic chronic infection. The common external origin of deep immediate
infections is underlined by the well-known positive correlation between rate
of infection and length of the surgical procedure.
Standard operating room environments contain 50 to 500 cfu/m3 of bac-
teria (Antti–Poika et al. 1990). A colony-forming unit (cfu) is the minimum
number of bacteria required to grow a cell cluster or colony on a suitable
solid culture medium. Special precautions, including air-tight gowns and
hoods for surgeons, body exhaust systems for operating room personnel,
and control of air supply to cause laminar flow away from the patient, can
reduce the level of airborne bacteria to as low as 1 to 5 cfu/m3. Risk of clinical
infection for similar operation duration is subsequently reduced.
The Inflammatory Process 151

The last of the three types, the late infection, has the most cryptic origin.
This infection, of the nonsuperficial or deep type, may occur months to years
after surgery in sites with no prior history of infection. Its cause is generally
believed to be the transport and seeding of blood-borne bacteria from an
established infection, such as a tooth-root abscess or a urinary tract infection,
at a remote site. However, late infections that occur more than 1 month but
less than 3 months after surgery are sometimes distinguished as delayed
infections and ascribed to slow development of intraoperative bacterial con-
tamination. Delayed and late infections represent a major problem in many
procedures, such as total joint replacement or heart valve implantation. This
is the case not because they are a frequent occurrence but because of the
great difficulty encountered in treatment of the infection once it is established
on and around the implant.
Whether infections become established in a site containing an implant has
a great deal to do with the nature of the implant–tissue interface. Gristina
(Gristina and Costerton 1984; Gristina 1987) has suggested that infection is
a competition or “race for the surface” between bacteria and host cells,
which, through encapzulization and resolution, produce the desired tissue
integration. If the bacteria arrive first (before encapulization) or later, due to
transient damage to the surrounding tissue, they can establish residence on
the surface. Some bacteria, such as Staphylococcus aureus and Staphylococcus
epidermidis, are said to be slime formers, in that they elaborate a muccoplysac-
chride film, termed glycocalyx, that protects them from outside influences
such as phagocytic cells and diffusion of antibiotics.
Positive culture from carefully acquired specimens is proof of infection in
an implant site in animals or patients. However, failure to obtain such pos-
itive results, even with multiple cultures, is not proof of the absence of
infection because the culture conditions may have been inappropriate or
each inoculum too small (containing less than 1 cfu). Even in the absence of
clinical signs of frank infection, low-level or occult chronic infection may be
present and unrecognized. In some cases, occult infection may be verified
by the detection of bacterial DNA through the use of molecular multiplica-
tion techniques; however, such apparently positive results may be due to
specimen contamination and should be interpreted with care.

8.3.2 Use of Antibiotic-Impregnated Implants

The advent of modern bacteriostatic and bactericidal antibiotics has sug-
gested the possibility of utilizing them to suppress or prevent infection
associated with implants. An obvious approach would be to design a diffu-
sion-controlled delivery system that could be made part of the clinical
implant or implanted as an adjunctive device.
A common form of late infection is that associated with the use of poly
(methyl) methacrylate cements in total joint replacement; thus, a number of
investigators have proposed that appropriate antibiotics could be mixed with
152 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 8.1
Influence of Antibiotic in PMMA Cement on Immediate and Late
Infection in the Rat
Results
Groupa Number of Legs Antibiotic (wt. %) Infected (%) Sterile (%)
A 60 —b 42 (70) 18 (30)
42 1.25b,c 2 (5) 40 (95)
B 64 —b 11 (17) 53 (83)
12 1.25b,c 1 (8) 11 (92)
a In group A, the injection (2.5 × 107 Staphylococcus aureus) was given 1/2 hour
after closure of leg wound; in group B, it was delayed for 6 weeks. Plugs of
cement and adjacent bone were cultured after 2 weeks.
b Palacos™ (Sulzer Bros., Bad Homberg, Germany).
c Gentamicin added (0.5 g/40 g PMMA).
Source: Adapted from Elson, R.A. et al., J. Bone Joint Surg., 59B, 452, 1977.

the cement at insertion. Release by dissolution or diffusion could then be

depended upon to maintain an inhibitory but decreasing antibiotic concen-
tration in the vicinity of the joint replacement components and reduce the
incidence of late infection. This technique has been used clinically for some
time but can be criticized on three grounds:*

• It is difficult to be able to select the appropriate antibiotic because a

range of bacteria with different sensitivities produces the late infec-
tion.
• The patient may develop an antibiotic sensitivity that would require
the removal of the device to alleviate it.
• The antibiotic may not be able to penetrate all of the infectible tissue
around the implant effectively.

Elson et al. (1977) performed an experiment that sheds some light on these
concepts. Defects were created in rat tibiae; plugs of cement with and without
antibiotic impregnation were inserted and allowed to cure in place before a
bacterial inoculum was injected systematically within 1/2 hour of surgery
(group A) or after a 6-week delay (group B) (Table 8.1).
The antibiotic-impregnated cement was highly effective (χ2 = 48.55, p <
0.001) in reducing immediate postoperative infection, but not to a statistically
significantly extent by 6 weeks (χ2 = 0.185, n.s.). The latter comparison is
complicated by the low incidence of infection in the animals receiving unim-
pregnated cement and the small size of the antibiotic-impregnated cement
group at 6 weeks. Thus, the conclusion seems to be that the technique is
effective if infection by a sensitive organism occurs immediately after the
operation, but the effect on late infections is doubtful. This latter point is no

* PMMA bone cements with particular antibiotics added by the manufacturer are now widely
available.
The Inflammatory Process 153

surprise because the diffusion release of antibiotic by cement must fall rap-
idly with time.*
Bone cement cured in place seems especially to permit local infection, but
all implanted materials increase the risk of infection to some degree (Petty
et al. 1985). The efficacy of antibiotic impregnation of bone cement in reduc-
ing or preventing immediate infection was verified in a more severe canine
model using a direct infusion of Staphylococcus aureus into a bony site just
before cement insertion (Petty et al. 1988). This principle has been extended
to the use of precured antibiotic-impregnated PMMA “beads” (balls) (Blaha
et al. 1990) or preforms in the shape of the removed device (Younger et al.
1997) as temporary implants to treat deep-seated infections. However, anti-
biotic-impregnated cement has clearly been seen to be a failure in preventing
late infection in more realistic models such as a clinically functional total
knee replacement in the rabbit subjected to a systemic bacteremia 6 to 8
weeks after implantation (Blomgren 1981).
This last point attracts attention to the question of associations between
implants and infection. Bacterial growth in the presence of implants is tena-
cious and difficult to deal with. Some bacteria undergo a transformation of
behavior and increase their activity and pathogenicity. Conventional treat-
ments for bacterial infections that may be successful in nonsurgical cases
often have reduced efficacy or fail in the presence of implants.
The difficulty of treatment often leads to the removal of the implant and
the need to delay reimplantation until after the infection is successfully
suppressed. This is possible in some cases as total hip replacement, but not
in others as mitral (heart) valve replacement. Thus, “infection” constitutes a
major source of implant “failure,” despite the inability of an implant to
become infected. A final concern is that the chronic use of antibiotics in
treatment of patients and in decontamination of hospital facilities has led to
the emergence of ever more resistant and virulent strains of microorganisms.
Thus, although the incidence of implant infection can be expected to continue
to fall, the problems raised by each case may well become more difficult,
especially as the practice of combining antibiotics with implants to form
combination products increases in popularity. More benign approaches may
be introduction of molecular species that may interfere with bacterial adhe-
sion and formation and/or properties of the glycocalyx (Trampuz et al. 2003).

8.3.3 Geometric Factors

Geometric factors are extremely important to the consideration of interac-
tions between implants and infecting microorganisms. Initiation (seeding or
colonization) and propagation of infection constitute a competitive process

* The result of studies such as Elson et al. (1977) has been the growing contemporary use of anti-
biotic-containing PMMA formulations in primary joint replacement procedures, even in the
absence of known (proven) infection. The long-term consequences of such practice are still
unknown.
154 Biological Performance of Materials: Fundamentals of Biocompatibility

pitting the ability of the invading bacteria to replicate against the ability of
the host tissues and support cells to annihilate. This competition defines the
size of the cfu for each microorganism. Therefore, the exposure of the bacteria
to tissue and, especially, vascular processes is a critical factor in successful
host defense against infection.
As has been shown, tissue immediately adjacent to an implant is relatively
acellular at maturity. Bacteria that can grow within the capsular membrane
will encounter little active opposition until they break out into the surround-
ing tissues. In addition to general and specific immune responses, the imme-
diate response to infection is the inflammatory response. The cells that
mediate the response are blood borne. Thus, the reduction of the solid angle
of surrounding tissue near an implant takes on extreme importance; it
reduces the accessibility of the infected tissue near an implant to neutrophils
and macrophages.
The existence of a “dead space,” a volume filled with cell-free fluid rather
than tissue, represents a special hazard of the geometric type. Design and
implantation practices should try to avoid such features because this fluid
can act as an in vivo culture medium for bacteria. Even in the absence of an
acellular membrane and/or a dead space, as previously noted, some bacteria
can produce their geometric defense by forming a glycocalyx about them on
the surface of the implant.
Although porous bodies represent no particular structural hazard when
viewed from the aspect of carcinogenesis (see Chapter 13), they might be
expected to present a geometric hazard in the initiation and/or support of
infection. Kiechel et al. (1977) have studied the influence of porosity in
poly(methyl) methacrylate implants in rabbits on the propagation of an
aerobic infection by Staphylococcus aureus and its treatment by penicillin.
Using a muscle site (corresponding to an immediate deep infection), they
concluded that there was no effect on the inoculum size needed to initiate
infection or on the response of established infection to penicillin. Whether
the outcome would have been different in the case of a late infection or an
infection due to an anaerobic bacterium cannot be determined from this
experiment.
A somewhat more sophisticated repetition of this study (Merritt et al. 1979)
reached a somewhat different set of conclusions. Using a variety of porous
and dense implant materials, but the same microorganism, and comparing
the effects of immediate (at implantation) injection of an inoculum to an
inoculation after a 1-month delay, this later study concluded that porosity
had an effect on infection rates. In the immediate injection (acute) group, the
infection rate for porous materials was higher, and in the 1-month delay
(chronic) group, the infection rate for dense materials was higher. The
authors suggested that these findings support the hypothesis that bacteria
can evade host defense mechanisms if they enter the pores of an implant
before tissue invasion; but once the implant is invaded by host tissue, the
bacteria are more exposed to cellular attack.
The Inflammatory Process 155

However, these conclusions are drawn from a comparison of porous poly-

ethylene and dense alumina ceramic and devolve from relatively small
experimental groups. Furthermore, as the authors advance it, the hypothesis
can explain the acute results but not the chronic results. Thus, the question
is still open to debate. Using a slightly different model, Cordero et al. (1994)
reproduced this result for immediate (intraoperative) infection and further
were able to distinguish CoCrMo implant sites as more easily infectable
(requiring fewer cfus) by Staphylococcus aureus than sites with pure titanium
implants of identical configuration, probably reflecting differences in corro-
sion products. A study of the outcome of porous percutaneous carbon
implants in rabbits for up to 4 years (Krouskop et al. 1988) further supports
the view that implant porosity, if not immediately colonized by bacteria,
represents no long-term hazard for infection.
The clinical practice of removing the “infected implant” is based most
directly on geometric arguments and on their physiological consequences.
Two other types of arguments can be proposed: one based upon chemically
mediated interactions between the implant and the bacterial infection and
the other based upon chemical effects on the cell populations in the host that
act against infection. Both of these are controversial today.
It is thought that if implants can provide metals valuable to metabolism
(see Section 14.1) through the corrosion of metallic parts, then they can also
be shown to participate in bacterial metabolism. Thus, the view is that
metallic implantation is similar in effect to putting fertilizer on a garden.
However, one metal — iron — occupies a special position. Iron levels in the
body are evidently subject to active control and change rapidly in response
to infection (see Chapter 14). It has been suggested that the interference of
endogenous iron sources with this control system plays a role in the prop-
agation of certain infections (Weinberg 1974).

8.4 Effects of Implant Degradation Products

8.4.1 Effects on Phagocytosis
One comes then to a consideration of the effects of implants, primarily
metals, on the cellular elements of the host defense system. There are many
possible mechanisms for such effects but this chapter will focus on one:
alteration of phagocytic function. Phagocytic cells play key roles in inflam-
matory response, as previously discussed. Several studies that explore var-
ious aspects of reduction of phagocytic function through a variety of animal
models will be considered.
Graham et al. (1975) were interested in determining whether the inhalation
of metallic particles from industrial pollution would reduce the number
or the activity of lung macrophages. They used cultured rabbit alveolar
156 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 8.2
Effects of Trace Metals (as Ions) on Alveolar Macrophages
VO–3 Cr+3 Mn+3 Ni+2 Ni+2 Ni+2
Ion concentration (M) (ppm): 7 × 10–5 3 × 10–3 2 × 10–3 5 × 10–4 8 × 10–4 1 × 10–3
6.9 156 110 29 47 59
Phagocytic index (%) 100 70 75 50 20 13
Cell survival (%) 79 84 69 92 85 76
Phagocytic efficiency (%) 79 59 52 46 17 10
Notes: The phagocytic index is the percentage of living cells that contain one or more
latex spheres after a 1-hour exposure at a ratio of 120 spheres/cell. Phagocytic
efficiency = phagocytic index × cell survival.
Source: Original data from Graham, J.A. et al., Infect. Immunol., 11, 1278, 1975.

macrophages, which were challenged by a 1-day incubation in the presence

of metallic salts of various types, and examined the ability of the survivors
to phagocytize 1.1-μm-diameter latex spheres. For the metals of interest in
implant applications, they obtained the data shown in Table 8.2.
Each of the metal ions in this table exhibited cytotoxic effects at the tested
concentrations. All but vanadate also produced a reduction in phagocytotic
behavior, generally independent of the cytotoxic effect. In the case of nickel,
in which three concentrations were employed, the sign of both effects were
the same, but the reduction of phagocytosis was more extreme. Note that
the true in vivo effect would depend upon the product of the cell survival
index and the phagocytic index, here called the phagocytic efficiency. Thus,
for nickel, an increase of concentration by a factor of ~2 (from 29 to 58 ppm)
produces a depression of macrophage effect of 78%.
The authors point out that the ion concentrations used were one to two
orders of magnitude below those found in human lung tissue obtained in
highly polluted industrial environments. However, this comment is based
upon comparison to average tissue concentrations; in Chapter 14, it will be
seen that concentrations immediately adjacent to corroding implants may
be two to three orders of magnitude higher than average values. Thus, the
concentrations used by Graham et al. are relevant. They also report a study
showing a positive correlation between metal aerosol inhalation in mice and
susceptibility to infection, as well as studies in human populations showing
positive correlations between metal dust inhalation and respiratory infection
rate.
Rae (1975) has pursued this issue further. He incubated mouse macroph-
ages with finely divided implant alloys and pure metals and examined the
activity of two enzyme systems:

• Lactic dehydrogenase (LDH) is an enzyme normally contained

wholly within cells. Thus, its detection is a measure of the leakiness
of, or damage to, cell membranes.
The Inflammatory Process 157

• Glucose-6-phosphate dehydrogenase (G6PD) is an enzyme required

for synthetic processes and must be present for phagocytosis to take
place. Thus, its reduction signals a reduction in the phagocytotic
ability of cells.

Cobalt and nickel particles were found to be cytotoxic. At the concentra-

tions obtained in this study, no cytotoxic effect was observed for chromium
or molybdenum. Nickel, cobalt, and cobalt–chromium alloy particles were
found to reduce the phagocytic ability of the surviving macrophages signif-
icantly (as measured by G6PD activity).
The mechanisms of these two effects, cell death and reduction of phago-
cytotic ability, are unclear. In particular, they interacted in Rae’s experiment,
and an effect of particle size was displayed; smaller particles were at the
same time more frequently phagocytocized and more toxic.
A study by Ward et al. (1975) suggests that a wide variety of effects exist
that are not closely related. This view supports the results of the previous
two studies. The authors studied a wide variety of metallic salts and their
effects on neutrophil chemotaxis and on the incorporation of amino acids
into neutrophils and fibroblasts. Rabbit neutrophils, HeLa (a human cancer
cell line), and human gingival fibroblasts were used. An Escherichia coli
filtrate was used as a neutrophil challenge to study chemotaxis by neutro-
phils. Incorporation inhibition (a reduction in phagocytic efficiency) was
shown for a variety of metal salts; however, the antichemotaxis results are
of more interest here. An extract of the results of these studies is given in
Table 8.3.
The authors discuss these chemical agents primarily as anti-inflammatory
agents. It is interesting to note that these data do not support either of the
two previous studies. This disagreement stems from several origins:

• Chemotaxis is only one factor in phagocytosis.

• Rae’s study associated a small effect (reduction in phagocytic effi-
ciency) of chromium and molybdenum with low solubility of metal-
lic particles; this study bypassed the problem by using soluble salts.
• Graham’s study utilized a passive challenge to the macrophage
rather than one of biological origin.

TABLE 8.3
Chemotaxis Inhibition by Metallic Ions
Cr+3 Mn+3 Fe+2,+3 W+4 Mo+4 Co+2
Ion concentration for 50% M: 3 × 10–4 5 × 10–3 10–3 10-3 10–3 10–3
inhibition of chemotaxis ppm: 16 27 56 18 96 59
Source: Data extracted from Ward, P.A. et al. J. Reticuloendothel. Soc., 18, 313, 1975.
158 Biological Performance of Materials: Fundamentals of Biocompatibility

Using a bacterially derived chemoattractant with human neutrophils,

Remes and Williams (1990) concluded that Cr+3 (2.5 ppm) (56% inhibition),
Co+2 (≥20 ppm) (100%), and Ni+2 (≥30 ppm) (50%) all actively interfered with
chemotaxis. The greater apparent sensitivity to chromium and cobalt than
seen in Graham’s study may reflect differences in experimental technique,
challenge agent, and/or cell source and underline the difficulties inherent
in such.

8.4.2 Effects of Phagocytosis

Section 8.2.3 considered the process of phagocytosis as carried forward by
macrophages and foreign body giant cells. This is a normal process existing
in all tissues, primarily for the removal of dead cells, cellular debris, and
damaged tissue matrix, as well as to aid in defense against infection. In the
majority of these situations, phagocytosis, digestion, and/or transport and
excretion are successful. However, as previously noted, excessive amounts
of foreign material or debris too large for phagocytosis or cytoxic materials
result in abnormal cellular behavior.
Willert (Willert and Semlitsch 1977) was the first to suggest that abnormal
response to small particles might play a role in the modification of interfaces
between implants and surrounding tissue, particularly in the case of joint
replacement components secured to bone with poly (methyl) methacrylate
“bone cement.” Patients with such devices often show a localized or pro-
gressive loss of bone adjacent to the bone–bone cement interface. This inter-
face often contains wear debris with associated cells in a fibrous membrane,
perhaps 100 μm in thickness; in some cases this may resemble the synovial
membrane found in normal articular joints.
Willert suggested that although small amounts of indigestible debris could
be stored locally or transported away through the lymphatic drainage, large
amounts overwhelmed the normal process and produced a histiocytic gran-
ulation tissue with accompanying fibrosis, due to attempts to encapsulate
and isolate the reaction. He suggested that these tissues interfere with motion
of the joint and invade the interface between the implant and bone, produc-
ing progressive tissue loss through necrosis and attempts at remodeling.
This phenomenon — previously called cement disease or polyethylene
disease, but now most generally small particle disease — is currently under-
stood as shown in Figure 8.5. Macrophages and tissue monocytes ingest
foreign particles and, in the process of being activated, release a variety of
organic molecules collectively called cytokines (Masui et al. 2005). Cytokines
are naturally occurring molecules used by cells to communicate with each
other, particularly to regulate stages in processes in which many cell types
are involved. For instance, the invasion of capillaries during formation of
granulation tissue is stimulated by a group of cytokines called AGFs or
angiogenic growth factors that are released by a variety of other cells.
The Inflammatory Process 159

BIOMATERIAL
DEBRIS
CYTOKINES:
ACTIVATED
IL-1, IL-6,TNFα, PGE2
MACROPHAGES

MACROPHAGES,
TISSUE MONOCYTES

OSTEOCLASTS
MACROPHAGES
(ACTIVATED?) OSTEOBLASTS

IMPLANT SITE
Bone Bone Inhibition of
EFFECTS:
Resorption Resorption Bone Formation

FIGURE 8.5
Small particle disease. (Adapted from Willert, H.-G. and Semlitsch, M., J. Biomed. Mater. Res.,
11, 157, 1977; Howie, D.W. et al., Orthop. Clin. N. Am., 24(4), 571, 1993.)

Although more than 40 human cytokines are known, only a few have been
so far implicated in small particle disease. Principal among these are several
interleukins (IL-1 and IL-6), tumor necrosis factor (TNF), and prostaglandin
E2 (PGE2). Each cytokine may have numerous effects on cell activity, depend-
ing upon the exact nature of the cytokine (many cytokines have variants,
such as IL-1α, IL-1β, etc.), the concentration in tissue, the presence of other
cytokines, and the nature of the target cell. Thus, research in this area, which
is extremely active in the biomaterials field as well as the general medical
and biological community, is remarkably complex and, to a great degree,
opaque to nonspecialists.
The primary effect in the presence of small particles appears to be the
production of one or more cytokines that stimulate large phagocytic cells
called osteoclasts to resorb (remove) bone (Teitelbaum 2000). Osteoclasts are
present on bone surfaces throughout the skeleton of mammals and the
amount of bone present is due, in part, to the balance between bone resorp-
tion by osteoclasts and bone formation by osteoblasts. Increase in osteoclastic
activity stimulated by cytokines released by phagocytes, without an accom-
panying increase in osteoblastic activity, results in overall bone loss, with
subsequent loss of integrity of the implant–bone interface, loosening, pain,
and the eventual need for surgical intervention. The actual process is far
more complex: macrophages (which may not have been previously acti-
vated) may resorb bone directly; osteoblastic activity may be suppressed
and cells at the bone–implant interface also produce and release cytokines,
160 Biological Performance of Materials: Fundamentals of Biocompatibility

affecting each other’s actions. Finally, because tissue necrosis may occur,
depending on the nature and composition of the foreign material, the normal
events of the inflammatory process (Section 8.2) may also occur, superim-
posed on the increased osteoclastic activity and its consequences.
As a result of these complex and inter-related effects, small particle disease
is still poorly understood and efforts to prevent or treat it in animal models
and patients are still in their very early stages (Goodman et al. 2005).

8.5 A Final Comment

In the first edition of this work, I emphasized that the effects of apparent
chemical interactions between implant and bacteria and implant and host
cells were at that time highly speculative and based almost completely on
in vitro studies. Similar mistakes were being made in the study of the clinical
consequences of particulate phagocytosis (small particle disease). In vitro
studies lack the humeral, metabolic, catabolic, and excretory influences seen
in whole-animal experiments. Most animal models are also incomplete, in
that they fail to reproduce key aspects of actual clinical implant composition
and structure and rely, in large part, on single or repeated (divided) doses,
rather than on continual infusion or release. I had further pointed out that
many critical experiments remained to be done before any of these supposed
effects could be established or disproved.
Unfortunately, the situation remains much the same today, if not more
confused due to the growing awareness of the role of physical and electrical
factors mediating cell behavior. These chemical, physical, and electrical fac-
tors may act separately or in concert in synergistic or antagonistic manners.
Until studies use appropriate animal models and forms of implant materials,
as well as control chemical, physical, and electrical aspects of the biomaterials
under study in single experiments, much smoke and heat will be produced
but little light shed on key issues of local host response to implants in
patients.

References
Albrektsson, T. et al., The interface zone of inorganic implants in vivo: titanium
implants in bone, Ann. Biomed. Eng., 11, 1, 1983.
Anderson, J.M. and Miller, K.M., Biomaterial biocompatibility and the macrophage,
Biomaterials, 5, 5, 1984.
Antti–Poika, I. et al., Hip arthroplasty infection, Acta Orthop. Scand., 61, 163, 1990.
Aronson, A.S., Hansson, L.I. and Selvik, G., Roentgen stereophotogrammetry for
determination of bone growth. Comparison with the tetracycline method, Acta
Radiol. Diagn. (Stockh.), 18, 87, 1977.
The Inflammatory Process 161

Blaha, J.D. et al., The use of Septopal (polymethylmethacrylate beads with gentam-
icin) in the treatment of chronic osteomyelitis, Instruct. Course Lect., 39, 509,
1990.
Blomgren, G., Hematogenous infection of total joint replacement, Acta Orthop. Scand.,
52(Suppl. 187), 1, 1981.
Brooks, R.A., Wimhurst, J.A. and Rushton, N., Endotoxin contamination of particles
produces misleading inflammatory cytokine responses from macrophages in
vitro, J. Bone Joint Surg., 84B, 295, 2002.
Chambers, T.J., Fusion of macrophages following simultaneous attempted phagocy-
tosis of glutaraldehyde-fixed red blood cells, J. Pathol., 122, 71, 1977.
Chesmel, K.D. et al., Cellular response to chemical and morphological aspects of
biomaterial surfaces: II. The biosynthetic and migratory response of bone cell
populations, J. Biomed. Mater. Res., 29, 1101, 1995.
Cordero, J., Munuera, L. and Folgueira, M.D., Influence of metal implants on infec-
tion, J. Bone Joint Surg., 76B, 717, 1994.
Elson, R.A. et al., Bacterial infection and acrylic cement in the rat, J. Bone Joint Surg.,
59B, 452, 1977.
Frank, M.M., The complement system in host defense and inflammation, Rev. Infect.
Dis., 1, 483, 1979.
Goodman, S.B. et al., Pharmacologic modulation of periprosthetic osteolysis, Clin.
Orthop. Rel. Res., 430, 39, 2005.
Graham, J.A. et al., Effect of trace metals on phagocytosis by alveolar macrophages,
Infect. Immunol., 11, 1278, 1975.
Gristina, A.G., Adhesive colonization of biomaterials and antibiotic resistance, Sci-
ence, 237, 1588, 1987.
Gristina, A.G. and Costerton, J.W., Bacterial adherence and the glycocalyx and their
role in musculoskeletal infection, Ortho. Clin. N. Am., 15, 517, 1984.
Howie, D.W. et al., The response to particulate debris, Orthop. Clin. N. Am., 24(4),
571, 1993.
Kiechel, S.F. et al., The role of implant porosity on the development of infection, Surg.
Gynecol. Obstet., 144, 58, 1977.
Krouskop, T.A. et al., Bacterial challenge study of a porous carbon percutaneous
implant, Biomaterials, 9, 398, 1988.
Mariano, M. and Spector, W.G., The formation and properties of macrophage
polykaryons (inflammatory giant cells), J. Pathol., 113, 1, 1974.
Masui, T., Expression of inflammatory cytokines, RANKL and OPG induced by
titanium, cobalt–chromium and polyethylene particles, Biomaterials, 26, 1695,
2005.
McNamara, A. and Williams, D.F., Scanning electron microscopy of the metal-tissue
interface. I. Fixation methods and interpretation of results, Biomaterials, 3, 160,
1982.
Merritt, K., Shafer, J.W. and Brown, S.A., Implant site infection rates with porous and
dense materials, J. Biomed. Mater. Res., 13, 101, 1979.
Petty, W. et al., The influence of skeletal implants on incidence of infection, J. Bone
Joint Surg., 67A, 1236, 1985.
Petty, W., Spanier, S. and Shuster, J.J., Prevention of infection after total joint replace-
ment, J. Bone Joint Surg., 70A, 536, 1988.
Rae, T., A study on the effects of particulate metals of orthopedic interest on murine
macrophages in vitro, J. Bone Joint Surg., 57B, 444, 1975.
162 Biological Performance of Materials: Fundamentals of Biocompatibility

Remes, A. and Williams, D.F., Chemotaxis and the inhibition of chemotaxis of human
neutrophils in response to metal ions, J. Mater. Sci.: Mater. Med., 1, 26, 1990.
Styles, J.A. and Wilson, J., Comparison between in vitro toxicity of two novel fibrous
mineral dusts and their tissue reactions in vivo, Ann. Occup. Hyg., 19, 63, 1976.
Teitelbaum, S.L., Bone resorption by osteoclasts, Science, 289, 1504, 2000.
Trampuz, A. et al., Molecular and antibiofilm approaches to prosthetic joint infection,
Clin. Orthop. Rel. Res., 414, 69, 2003.
Urban, R.M. et al., Dissemination of wear particles to the liver, spleen, and abdominal
lymph nodes of patients with hip or knee replacement, J. Bone Joint Surg., 82(4),
457, 2000.
Ward, P.A., Goldschmidt, P. and Greene, N.D., Suppressive effects of metal salts on
leukocyte and fibroblastic function, J. Reticuloendothel. Soc., 18, 313, 1975.
Weinberg, E.D., Iron and susceptibility to infectious disease, Science, 184, 952, 1974.
Willert, H.-G. and Semlitsch, M., Reactions of the articular capsule to wear products
of artificial joint prostheses, J. Biomed. Mater. Res., 11, 157, 1977.
Wood, N.K., Kaminski, E.J. and Oglesby, R.J., The significance of implant shape in
experimental testing of biological materials: disc vs. rod, J. Biomed. Mater. Res.,
4, 1, 1970.
Younger, A.S. et al., The outcome of two-stage arthroplasty using a custom-made
interval spacer to treat the infected hip, J. Arthoplasty, 12(6), 615, 1997.
Ziats, N.P., Miller, K.M. and Anderson, J.M., In vitro and in vivo interactions of cells
with biomaterials, Biomaterials, 9, 5, 1988.

Bibliography
Aderem, A., How to eat something bigger than your head, Cell, 110, 5, 2002.
Bauer, T.W., Particles and peri-implant bone resorption, Clin. Orthop. Rel. Res., 405,
138, 2002.
Berry C.L. (Ed.), The Pathology of Devices, Springer–Verlag, Berlin, 1994.
Bisno, A.L. and Waldvogel, F.A., Infections Associated with Indwelling Medical Devices,
American Society of Microbiologists, Washington, D.C., 1989.
Brandwood, A. et al., Phagocytosis of carbon particles by macrophages in vitro,
Biomatials, 13, 646, 1992.
Cameron, H.U. (Ed.), Bone Implant Interface, Mosby-Year Book, St. Louis, 1994.
Chang, C.C. and Merritt, K., Infection at the site of implanted materials with and
without preadhered bacteria, J. Orthop. Res., 12, 526, 1994.
Coleman, D.L., King, R.N. and Andrade, J.D., The foreign body reaction: a chronic
inflammatory response, J. Biomed. Mater. Res., 8, 199, 1974.
Costerton, J.W. et al., Microbial biofilms, Annu. Rev. Microbiol., 49, 711, 1995.
Craig, M.R. et al., A novel total knee arthroplasty infection model in rabbits, J. Orthop.
Res., 23(5), 1100, 2005.
Duffield, J.S., The inflammatory macrophage: a story of Jekyll and Hyde, Clin. Sci.,
104, 27, 2003.
Delle Valle, C.J., Zuckerman, J.D. and Di Cesare, P.E., Periprosthetic sepsis, Clin.
Orthop. Rel. Res., 420, 26, 2004.
Ellingsen, J.F. and Lyngstadaas, S.P., Bio-implant Interface: Improving Biomaterials and
Tissue Reactions, CRC Press, Boca Raton, 2003.
The Inflammatory Process 163

Esterhai, J.L., Jr., Gristina, A.G. and Poss, R. (Eds.), Musculoskeletal Infections, American
Academy of Orthopedic Surgeons, Park Ridge, IL, 1992, 153–300.
Gatti, A.M., Biocompatibility of micro- and nanoparticles in the colon, Biomaterials,
25, 385, 2004.
Greco, R.S. (Ed.), Implantation Biology. The Host Response and Biomedical Devices, CRC
Press, Boca Raton, FL, 1994.
Heggeness, M.H. et al., Late infection of spinal instrumentation by hematogenous
seeding, Spine, 18, 492, 1993.
Hunt, J.A., Remes, A. and Williams, D.F., Stimulation of neutrophil movement by
metal ions, J. Biomed. Mater. Res., 26, 819, 1992.
Ingham, E. and Fisher, J., Biological reactions to wear debris in total joint replacement,
Proc. Inst. Mech. Eng., Part H, 214, 21, 2000.
Inghram, J.H. et al., The influence of molecular weight, crosslinking, and counterface
roughness on TNF-alpha production by macrophages in response to ultra high
molecular weight polyethylene particles, Biomaterials, 25, 3511, 2004.
Ito, A. et al., In-vitro analysis of metallic particles, colloidal nanoparticles, and ions
in wear-corosion products of SUS317L stainless steel, Mater. Sci. Eng., C17, 161,
2001.
Kawaguchi, H. et al., Phagocytosis of latex particles by leucocytes. I. Dependence of
phagocytosis on the size and surface potential of particles, Biomaterials, 7, 61,
1986.
Leibovich, S.J. and Ross, R., The role of the macrophage in wound repair, Am. J.
Pathol., 78, 71, 1975.
Maloney, W.J. et al., Isolation and characterization of wear particles generated in
patients who have had failure of a hip arthroplasty without cement, J. Bone
Joint Surg., 77A, 1301, 1995.
Merritt, K., Role of medical materials, both in implant and surface applications, in
immune response and in resistance to infection, Biomaterials, 5, 47, 1984.
Nagase, M., Host reactions to particulate biomaterials, in Encyclopedic Handbook of
Biomaterials and Bioengineering, Part A: Materials, Vol. 1, Wise, D.L. et al. (Eds.),
Marcel Dekker, New York, 1995, 269.
Peters, K. et al., Effects of nanoscaled particles on endothelial cell function in vitro:
studies on viability, proliferation, and inflammation, J. Mater. Sci.: Mater. Med.,
15, 321, 2004.
Rae, T., Cell biochemistry in relation to the inflammatory response to foreign mate-
rials, in Fundamental Aspects of Biocompatibility, Vol. 1, Williams, D.F. (Ed.), CRC
Press, Boca Raton, FL, 1981, 159.
Rae, T., Localized tissue infection and the influence of foreign bodies, in Fundamental
Aspects of Biocompatibility, Vol. 2, Williams, D.F. (Ed.), CRC Press, Boca Raton,
FL, 1981, 139.
Revell, P.A., Pathology of Bone, Springer–Verlag, Berlin, 1986, 217–223.
Rimondini, L., Fini, M. and Giardino, R., The microbial infection of biomaterials: A
challenge for clinicians and researchers, J. Appl. Biomater. Biomech., 3(1), 1, 2005.
Sanzén, L. and Linder, L., Infection adjacent to titanium and bone cement implants:
an experimental study in rabbits, Biomaterials, 16, 1273, 1995.
Schoen, F.J., Interventional and Surgical Cardiovascular Pathology: Clinical Correlations
and Basic Principles, W.B. Saunders, Philadelphia, 1989.
Sethi, R.K. et al., Macrophage response to cross-linked and conventional UHMWPE,
Biomaterials, 24, 2561, 2003.
164 Biological Performance of Materials: Fundamentals of Biocompatibility

Spector, W.G. and Wynne, K.M., Proliferation of macrophages in inflammation,

Agents Actions, 6, 123, 1976.
Sugarman, B. and Young, E.J. (Eds.), Infections Associated with Prosthetic Devices, CRC
Press, Boca Raton, FL, 1984.
Trowbridge, H.O. and Emling, R.C., Inflammation: A Review of the Process, 5th ed.,
Quintessence Pub., Carol Stream, IL, 1997.
Vogler, E.A., Water and the acute biological response to surfaces, J. Biomater. Sci.
Polymer Edn., 10(10), 1015, 1999.
Williams, D.F., Tissue–biomaterial interactions, J. Mater. Sci., 22, 3421, 1987.
Yamamoto, A. et al., Cytoxicity evaluation of ceramic particles of different sizes and
shapes, J. Biomed. Mater. Res., 68A, 244, 2004.
9
Coagulation and Hemolysis

9.1 Introduction
In the previous chapter, inflammation was presented as a nonspecific
response to tissue damage. This chapter will consider responses to two more
specific events involving the circulatory system and its tributaries:

• Damage to blood carrying vessels and/or contact with foreign mate-

rials leading to coagulation
• Damage to the tissue of blood, leading most generally to cellular
destruction or hemolysis

9.2 The Coagulation Cascade

9.2.1 Intrinsic Pathway
The coagulation cascade may be thought of as a biological amplifier that
permits an initiating event to be magnified into a process sufficiently wide-
spread to produce local homeostasis, prevents further loss of blood (if the
vascular process has been breached), and permits repair of the damaged
tissue. The overall process is referred to as coagulation or thrombosis and
the resulting hemostatic plug is termed a thrombus. If damage to the wall
of a blood vessel (endothelium) is the initiation factor, then coagulation
proceeds by an intrinsic pathway.
In this case, the first two events closely parallel those of inflammation. In
fact, they are essentially those of the inflammatory process. Initially, the
smooth muscle in the vessel wall dilates and then constricts, probably stim-
ulated by activated factor XII (also called Hageman factor) (see Figure 9.1).
Although this is formally the first step in the intrinsic pathway, it is preceded
by surface contact (and denaturation) of other molecules, including kinino-
gen and prekallikrein. Endothelial permeation also increases but is masked

165
166 Biological Performance of Materials: Fundamentals of Biocompatibility

Surface contact
XII active XII
Tissue
active XI Thromboplastin
XI

IX active IX Extrinsic
Pathway
Intrinsic
Pathway
Ca++ VII
Ca++ VIII
Surface contact
Platelets

active X X

Ca++ V
Surface contact
Platelets
Fibrinogen

Prothrombin Thrombin
Soluble fibrin
XIII active XIII Ca++
Ca++ Fibrin

FIGURE 9.1
The coagulation cascade. (Adapted from Salzman, E.W., in The Chemistry of Biosurfaces, Vol. II.
Hair, M.L., Ed., Marcel Dekker, New York, 1972, 489.)

by the release of serum and blood-borne cells if the vessel wall is ruptured.
In addition, the endothelial lining becomes sticky.
The combination of the presence of activated factor XII and the sticky
quality of the endothelial lining triggers the first step unique to coagulation,
the adhesion and aggregation of platelets. Adhered platelets rapidly lyse
(undergo membrane rupture) through mechanical and biochemical paths,
releasing adenosine diphosphate (ADP), serotonin, and epinephrine. ADP
encourages further platelet adhesion, and the latter two agents cause vaso-
constriction. The combination of these agents rapidly produces a platelet
“plug” that serves to staunch further blood loss from the area.
Activation of factor XII and platelet adhesion trigger a complex chain of
events leading to transformation of inactive fibrinogen into an active mole-
cule, fibrin. This transformation, accompanied by a shrinking or retraction
of the platelet plug, produces a mature clot, a mesh of polymerized fibrin
strands trapping leukocytes, erythrocytes, and platelet fragments. Within
minutes to hours this clot is invaded by the same series of cells seen in pure
inflammation and then later is perfused by numerous new capillaries. Even-
tually, the clot is removed and replaced by fibrous scar and remodeled tissue.

9.2.2 Extrinsic Pathway

The intrinsic pathway to coagulation, as previously described, depends upon
interaction of normal blood components (macromolecules, cells, platelets,
Coagulation and Hemolysis 167

etc.) after alteration by an initiation event, usually surface contact. An alter-

nate initial coagulation pathway, the extrinsic pathway, involves release of
materials from cells external to the vascular processes. The released material
is termed tissue thromboplastin. This is a protein–phospholipid complex
derived from normal cell contents (otherwise termed factor III) that, with
factor VII and Ca++, activates factor X (Figure 9.1). The intrinsic and extrinsic
coagulation pathways merge in this common event leading to the terminal
event of fibrin production and clot formation.

9.2.3 Implant-Induced Coagulation*

The events of the intrinsic pathway are triggered, as previously stated, by
contact between blood elements and a surface. The normal endothelial lining
of blood vessels is not able, fortunately, to induce this response. Damage to
the endothelium may expose collagen, its major structural molecule. Nor-
mally, collagen is rendered nonthrombotic (not able to induce coagulation)
by a covering of a family of macromolecules, glycopolysaccharides. This
coating is electropositive, which attracts a layer of neutralizing negative ions
and thus repels the negatively charged erythrocytes and platelets. Exposed
collagen is electronegative, and thus highly thrombotic. This property is
utilized in surgery when powdered crystalline collagen may be used as a
topical hemostatic agent on highly vascular tissues in which electrocauter-
ization and/or ligation are not possible — for example, the liver.
The intrinsic pathway may also be triggered by contact with a foreign
body** (such as an implant) and the resulting events are quite similar. In this
case, however, the implant persists after the initial insult. This difference
from the case of isolated vessel wall damage has several possible conse-
quences (Figure 9.2):

• Blood-borne proteins retaining native structure may coat the surface

and prevent activation of factor XII or adhesion of platelets. Such a
surface would have a high degree of hemocompatibility.
• Even if platelet adhesion occurs, it may not be accompanied by
sufficiently rapid lysis to promote further progression of the cascade.
Such a surface would not actively form a thrombus. However, it
would deplete the circulating blood of platelets and thus might
reduce the coagulatibility of the host. Such an effect is common in
chronic (repeated) hemodialysis, or intraoperatively when an ex vivo
blood oxygenator is used. Brash (Skarja et al. 1997) has developed

* See Banerjee et al. (1997) for a review of the interaction between blood and biomaterials.
** We speak here of contact with the surface of a “foreign body”; however, the surface of an
implant so rapidly becomes coated with serum proteins and coagulation factors that it is better
to think of the foreign body contact effect as mediated by surface adhered and denatured organic
molecules. See Vroman (1971) and Chapter 5, particularly Section 5.5.
168 Biological Performance of Materials: Fundamentals of Biocompatibility

No Thrombus

c
eni
m bog
nth
r o ma
BLOOD ELEMENTS No o - inti
ud
pse
or Thrombogenic
+ Adherent
Th
rom and adherent Thrombus
an bo
dn ge
nic
on
ad
he
ren
t

Shed
Thrombus

FIGURE 9.2
Interaction of implants with blood elements.

an in vitro test for quantification of platelet adhesion under flow

conditions (see Section 17.4.3).
• A thrombus may form but be rapidly removed by blood dynamic
forces. Such an implant will “shed” emboli and cause damage by
infarction at a remote site. This effect is utilized in the Kusserow test
for whole blood compatibility of materials (see Section 18.2.3).
• Remodeling will not remove the obstruction but will tend to encap-
sulate it, as in the case of the fibrous encapsulation found around
implants in soft tissue. If the implant surface is structured (such as
a felt or velour) to encourage cellular trapping, the surface exposed
to the blood may come to function as the endothelial wall of a normal
vessel and is termed “pseudointima” (see Section 10.3.2). It differs,
of course, from the normal intima because it has no smooth muscle
component and thus no ability to dilate or constrict.

These effects have been summarized by Baier (1972) as seen in Figure 9.3.
In this figure, Baier also indicates possible points and methods of interven-
tion that may reduce the thrombogenic potential of implant surfaces.
In the ordinary (implant-free) progression of the intrinsic pathway, the
fibrin clot gradually isolates flowing blood from the site of possible surface
contact insult. However, on a foreign (implant) surface, it is possible for
fibrin to initiate coagulation by itself due to its altered configuration after
surface binding (Lindon et al. 1986). Implants may also initiate coagulation
by binding (and possible activation) of molecules such as complement C3*
that are not normally involved in the intrinsic pathway (Hayashi 1990; Her-
zlinger et al. 1981).

* The proteins involved in the complement system are usually associated with blood plasma,
where they are present in abundance. However, due to the close association between comple-
ment and immune response, this topic is discussed in Section 12.2.2.
 Coagulation and Hemolysis

FIGURE 9.3
Coagulation events, timing, and intervention strategies. (From Baier, R.E., Bull. N.Y. Acad. Med., 48, 257, 1972. With permission.)
169
170 Biological Performance of Materials: Fundamentals of Biocompatibility

9.3 Approaches to Thromboresistant Materials Development

9.3.1 General Considerations
In the same article from which Figure 9.3 is drawn (Baier 1972), the various
historical approaches taken to design new surfaces or to render surfaces of
older materials less thrombogenic were summarized (Figure 9.4). The
approaches shown by the solid-line arrows pointing to the left are those that
mimic properties that the natural intact endothelium is known to have. Those
shown by the dotted lines pointing to the right are other proposals not based
so directly upon known natural properties, but rather on theoretical
approaches. All have met with mixed success and failure because, with the
limited exceptions of knitted polyester grafts; felted polyurethane; poly (tet-
rafluoro) ethylene and related materials that can sustain “pseudointimal”
linings; certain bulk polymers; and graphites, most materials evoke unac-
ceptable host responses for chronic exposure to blood. Even satisfactory
materials tend to thrombose in prosthetic devices with blood conduit internal
diameter less than 6 mm.
Although Figure 9.4 is now more than 30 years old, it is interesting that
no really new approaches to solution of the so-called “blood contact” prob-
lem have been developed. The only addition that could be made to this
schematic, other than subtopics under each of the eight main headings,
would be the addition of viable biomaterials as a fifth mimicking approach

FIGURE 9.4
Approaches to producing thromboresistant surfaces. (From Baier, R.E., Bull. N.Y. Acad. Med.,
48, 257, 1972. With permission.)
Coagulation and Hemolysis 171

(right side). This is an extension of the “natural” materials approach, which

is specifically the modification of permanent implantable materials with
natural molecules or a surface-dwelling cell population. In the newer
approach, a composite of resorbable materials and living cells (smooth mus-
cle cells, endothelial cells, etc.) is made and implanted with the hope that it
will eventually remodel into native (host) tissue (Weinberg and Bell 1986).
The lack of new approaches to producing noncoagulating biomaterial
surfaces reflects the complexity of the natural system and the difficulty of
performing reproducible blood contact experiments with generality of
results, whether in vitro or in animal models. This natural complexity has
aroused considerable interest in finding simplifying or dominant factors in
surface properties that affect thrombogenesis. Several of the directions of
biomaterials efforts indicated in Figure 9.4 reflect such a search and are of
historical and practical interest. Three of these will be considered: negative
surface charge, critical surface tension, and “natural” surfaces. Each has
yielded important results and contributed to a better understanding of
thrombogenesis, but has failed to produce the final answer in suppressing
host response to materials in contact with blood.

9.3.2 Negative Surface Charge

The recognition that erythrocytes and platelets have net negative surface
charges has suggested to many investigators that electrostatic repulsion
could be utilized to keep them away from implant surfaces, thus suppressing
thrombogenesis. Sawyer and Srinivasan (1972) were the most active initiators
and proponents of this idea. A series of early experiments suggested that a
relative positive potential on a biomaterial surface exposed to blood pro-
motes thrombogenesis and negative potentials tend to suppress thrombo-
genesis proportionally to the potential depression below local neutral. These
conclusions arise from a complex series of experiments. Perhaps the most
convincing data are given in Figure 9.5 (Sawyer and Srinivasan 1972), sum-
marizing the results of implantation of Gott rings in the canine vena cava
(see Section 18.2.3) and of a related design (Edwards) made from various
materials in the canine aorta.
Under biophysiological conditions, metals with a negative electromotive
potential form a positive interfacial potential with blood (due to the attrac-
tion of counter ions, as previously noted) and vice versa. Here, one can see
the rapid and complete occlusion (coagulation sufficient to prevent blood
flow) for silver and platinum, which have positive interface potentials in
vivo, and the increasing patency (proportion of devices permitting flow) at
increasing times for metals, such as iron (in stainless steel), aluminum, and
magnesium, that have negative interface potentials.
The approach is, however, quite limited. The use of rigid metallic surfaces
for vascular prostheses might prove nonthrombogenic by surface interaction
but might produce a pseudoextrinsic thrombogenesis through mechanical
172 Biological Performance of Materials: Fundamentals of Biocompatibility

FIGURE 9.5
Relationship between interface potential of vascular implants and patency. (From Sawyer, P.N.
and Kaplitt, M.J., Eds., Vascular Grafts, Appleton–Crofts, New York, 1978.)

damage to blood-borne cells.* Furthermore, control of interfacial potential

might require an active power source, thus making the implant far more
complex and less reliable than the knitted prostheses discussed in Section
10.3.2.**

* However, such considerations might well play a role in selection of materials for uncoated
metallic vascular stents.
** A personal note: as an experimental physicist, I have always been fascinated by Sawyer’s
results in this area of research. While I was a predoctoral student in the late 1960s, I conceived
an experiment, based on Milliken’s classic oil drop experiment for determination of the charge
of a single electron, to study the approach of platelets to charged surfaces in serum-free saline
solution. I had the pleasure of discussing this idea with Phil Sawyer at the First World Bio-
materials Congress (Baden, Austria, 1980). Although he thought it was an interesting idea, nei-
ther he nor anyone else to his knowledge had ever done this study (to my knowledge it has not
subsequently been done).
Coagulation and Hemolysis 173

9.3.3 Critical Surface Tension

The critical surface tension approach was proposed some time ago and has
been strongly advanced by Baier (1972) as a significant but not necessarily
dominant solution to the problem of surface thrombogenesis. The basic idea
is to utilize the Young–Dupree equation (Section 5.4; Equation 5.4):

γSL = γPS + γPLcosθ (9.1)

If surface tensions combine so that θ = 180°, no adhesion of the particle,

p, may occur. Then, of necessity, cosine θ must equal –1. This condition can
be achieved if:

γSL = γC = γPS – γPL (9.2)

(γC = critical surface tension).

Because γC is determined by the biological system, the trick is to find a
surface with a suitable surface tension so that Equation 9.2 is satisfied; then
no adhesion would occur for a particular molecular or cellular species and
one or more of the surface contact steps in Figure 9.1 could be prevented.
This can be done at two points in the coagulation cascade:

• The surface may have the appropriate critical surface tension to

prevent adhesion of factor XII.
• The surface may have the appropriate critical surface tension to
prevent adhesion of platelets in the intrinsic or the final (common)
pathway.

This turns out to be far more difficult in practice than in principle. For instance,
a surface might be selected that does not permit adherence of factor XII.
However, other molecular species will probably adhere, γSL may change, and
adhesion of factor XII may then become possible. Measurements of surface
tension of nonbiological surfaces, such as silicon, exposed in vitro or ex vivo to
sera or whole blood suggest complex time-dependent changes in γSL.
Notwithstanding these practical problems, a theoretical range of the
material–blood interfacial critical surface tension, γC, should suppress throm-
bogenesis (Figure 9.6). Because most blood contact materials are polymeric, a
fruitful embodiment of this approach has been chemical surface modification
of implants. Studies of a wide variety of as-synthesized and surface-modified
materials show relative improved thromboresistance in this range (20 to 30
dyn/cm*) of surface tension. It is interesting to note that 30 dyn/cm is the so-
called “Berg limit” (Vogler 1998); the essential balance point between hydro-
phobic and hydrophobic interfacial forces at which the solid/liquid phase
boundary should exert a minimum deforming effect on proteins.

* Older units; preserved for historical purposes in Figure 9.6. 1 dyn/cm = 10–3 N.m–1.
174 Biological Performance of Materials: Fundamentals of Biocompatibility

FIGURE 9.6
Proposed relationship of critical surface tension (γc) to biological response. (From Baier, R.E.,
Bull. N.Y. Acad. Med., 48, 257, 1972. With permission.)

9.3.4 “Natural” Surfaces

The thromboresistance of the undamaged internal surfaces of blood vessels
has attracted the surgeon and the bioengineer for a long time. Direct trans-
plants (allografts) and implants of animal material (xenografts) are limited
in utility by host rejection of the implant through an immune response (see
Section 12.1) and by biological degradation of the foreign material. Various
methods of processing have been tried to reduce the immune response and
to improve resistance to degradation. A number of processes in use involve
cleaning the tissue, removing cellular debris, and crosslinking the collagen
component (“tanning”) by a variety of agents such as glutaraldehyde, form-
aldehyde, etc. Although human material has been used, the pig is a favorite
donor. Kiraly and Nosé (1974) have summarized some of the early applica-
tions of such materials.
These materials seem to have considerable degrees of thromboresistance.
However, they are not incorporated into the body in the same way that
knitted grafts are (Nosé et al. 1977). Thus, a pseudointima does not form
and the surfaces exposed to blood are slow to mature. Although cells are
found on their surfaces, the same low surface tensions that suppress throm-
bogenesis seem to retard cellular adhesion. Additionally, these processed
materials are relatively impervious to diffusion of fluids and tend to develop
late calcification, similar to that which occurs naturally in arteriosclerosis.
Coagulation and Hemolysis 175

However, the approach remains interesting, particularly as how calcification

can be inhibited becomes known (Levy et al. 1995), and bears considerable
promise for the future, perhaps utilizing cultured (grown in vitro) tissue. The
“natural” approach is an obvious precursor to the more recent ideas involv-
ing viable biomaterials for blood-contacting surfaces (see earlier discussion).

9.3.5 An Overview of Thromboresistant Materials Development

It is difficult to consider in further detail the efforts at bulk and surface
modification that have been made to render implant surfaces friendlier to
blood. In general, the experiments with bulk materials are straightforward,
but their interpretation depends strongly on the validity of the blood expo-
sure test used to evaluate them. Chapter 17 and Chapter 18 will return to
this point.
I think that the experiments in surface modification and coating remain
open to broad, general criticism. In the first place, there is rarely very good
characterization of the bulk materials used. One may ask what the actual
(rather than calculated) structural and energetic properties of the surfaces
of these materials are and how these properties affect the resulting surface
exposed to cells. Furthermore, there is rarely good evidence that the surface
treatments or coatings are homogeneous or even cover the surface com-
pletely. The “catalytic” nature and inherent amplification of coagulation
processes suggest that the presence of occasional high-energy defects may
be highly effective as foci for initiation of thrombogenesis and may be far
more important than the anti- or nonthrombogenic properties of the balance
of the surface.
As pointed out previously, it can be generally assumed that implant sur-
faces become rapidly protein coated after insertion and that the proteins are
denatured in some degree. I have suggested (Section 5.4) that if a normally
free protein adheres to a biomaterial surface, by definition it must be dena-
tured. Whether this is mild and reversible (3 or 4°), moderate (2°), or severe
and essentially irreversible (1°) depends upon the nature of the protein–sur-
face interaction forces.* Thus, the central issues concerning adhesion of pro-
teins to surfaces are whether they are uniformly distributed or associated
with surface defects and whether the type of denaturation that occurs will
evoke a specific cellular response — in this case, the initiation of coagulation.
Finally, most studies neglect to measure the important physiochemical
properties of the actual material/surface/protein complex during and/or
after blood contact and instead rely on secondary determinations, frequently
in simplified systems with only one or two proteins present at low concen-
tration or on theoretical considerations. It seems most likely that, in the
absence of surface-bound molecules for which target cells (platelets, eryth-
rocytes, etc.) have specific surface receptors, cells involved in coagulation
are affected by the following properties of biomaterials:
* See Chapter 5 for a more complete discussion.
176 Biological Performance of Materials: Fundamentals of Biocompatibility

• Potential gradients and associated ion fluxes near surfaces

• Features of surface geometry with dimensions between 0.1 and 5 μm
• Actual surface/solution interfacial free energy
• Stresses at impact (governed by the preceding three points as well
as surface hardness and conditions of the experiment or clinical
exposure)
• Presence of specific proteins sufficiently denatured to evoke a bio-
logical response

9.4 Hemolysis
9.4.1 General Description
Foreign materials may trigger thrombosis by contact with blood in the
absence of motion at the blood–surface interface. However, motion may
trigger thrombus formation or may cause damage to blood cells in the
absence of thrombosis. Such initially nonthrombotic damage resulting in cell
death and release of cell contents is termed hemolysis. Although hemolysis
can be detected by a reduction in red cell (erythrocyte) count or by the
presence of cell “ghosts” (fragments of empty cell membranes), it is most
usually followed by measuring the level of serum hemoglobin.
Presence of hemoglobin in blood serum is a direct result of erythrocyte
lysis. The normal concentration of serum hemoglobin (in humans) of <1.1
g/l represents in toto no more than 0.4% of hemoglobin in the blood. Serum
hemoglobin is normally bound to a carrier molecule, haptoglobin; however,
the maximum capacity of this fraction in normal serum is 1.4 g/l. Hemolysis
rates above the release due to cell death in normal turnover may overpower
the ability of haptoglobin to bind and thus “detoxify” otherwise free hemo-
globin. Modest increases lead to increases in excretion of hemoglobin and
may cause anemia if sustained for long periods. Greater elevations (25 g/l
and higher) will cause systemic clinical symptoms including cyanosis and
hematuria; still higher levels may lead to kidney failure and toxemia.
Hemolysis may occur in freely flowing blood in the absence of foreign
surfaces. Turbulent flow and shear stresses above 150 to 300 Pa will cause
direct lysis. If stagnation points exist, as in some device designs, this lysis
may lead directly to thrombus formation in the apparent absence of surface
contact. Of course, in this case, the initiating surface is damaged cell mem-
brane exposed by cell lysis.
Coagulation and Hemolysis 177

Reservoir
Ω

r 10 cm.
diameter
Blood Test Material
(both sides)

FIGURE 9.7
Schematic of apparatus for study of shear-induced hemolysis. (Adapted from Lampert, R.H.
and Williams, M.C., J. Biomed. Mater. Res., 6, 499, 1972.)

9.4.2 Experimental Relation to Flow Velocity

However, flow effects can be seen in the contact between blood and foreign
surfaces at much lower ranges of shear stress. A large number of studies
have demonstrated this effect. One of the best is still that of Lampert and
Williams (1972). These investigators studied hemolysis rates of various mate-
rials compared to a standard material, aluminum, at a variety of flow rates.
Their apparatus consisted of a fixed disc with an opposed rotating disc at a
fixed separation distance, both on a common axis (see Figure 9.7 for a
schematic representation). The interdisc separation (H) and the speed of
relative rotation (Ω) could be varied.
The Reynolds numbers (Re) for this apparatus were calculated as shown
below. With respect to separation (H) at the edge of the disc (max r):

ρΩH 2
( )
Re H =
η
(9.3)

where
ρ = density of blood
Ω = angular velocity
η = viscosity of blood

Re(H) ≤ 1.5; laminar flow to Re(H) = 102.

With respect to radius (r) at maximum separation (max H):

ρΩr 2
( )
Re H =
η
(9.4)

Re(r) ≤ 4.4 × 104, laminar flow to Re(r) = 105.

These conditions yield shear rates, G ≤ 11,200 sec–1, (r ≤ 5 cm), and shear
stresses that do not exceed 44.8 Pa. In this low shear regime, Lampert and
178 Biological Performance of Materials: Fundamentals of Biocompatibility

Williams (1972) defined a relative plasma hemoglobin concentration

increase, ΔC, with respect to aluminum:

ΔC =
( ( ))
⎡⎣ ΔChb ⎤⎦ m, t secs
(9.5)
( )
⎡⎣ ΔChb ⎤⎦ A1, 30

For a number of materials tested, the following relationship was found:

ΔC = Atβ (9.6)

For instance, for heat-cured poly (methyl) methacrylate (Plexiglass™):

ΔC = 0.047t0.84 (9.7)

In general, A was found to be apparatus dependent for this experiment

and is given by the following relationship:

A = 0.020β–5.6 (9.8)

On the other hand, β was found to be a constant characteristic of the test

material and to have a single unique value for each material composition
tested. For a group of five polymers (both test surfaces made from the same
material), β correlated to the critical interfacial tension, γC, by a linear nega-
tive relationship of the form:

β = D – E γC (9.9)

where D and E are experimentally derived constants.

This can be understood by the following argument. A high value of γC
(with respect to a low value) favors rapid early protein absorption from the
plasma. This increases the adhesion of platelets (increases the negative work
of adhesion), leading to rapid maturation of a fibrin layer. This mature fibrin
layer discourages further platelet and cell adhesion, thus lowering hemolysis
rates with respect to those of the surface with lower γC. Thus, although high
γC leads to rapid thrombus formation, it may reduce free cell hemolysis.
These experiments suggest that the ideal blood-compatible surface may not
be the least reactive one.
These results were obtained with a constant shear stress. Performing exper-
iments at varying values of Ω, Lampert and Williams (1972) formed the ratio,
R, to examine shear stress effects:
Coagulation and Hemolysis 179

R=
(
ΔC m, Ω ) t = 100 sec (9.10)
( (
ΔC m, 640 rpm ))
The results of this analysis are equivocal due to uncertainties in the true
fluid dynamics of the system. They are consistent with a linear rise of R at
low stress (below 40 Pa) and a nonlinear increase at higher apparent stresses.
These results are interpreted as reflecting a constant boundary layer effect
(material effect) and a superimposed bulk shear effect (turbulence effect) at
higher rotational speeds. Thus, this experiment neatly shows the merging
of the two flow regimes.
Although extremely enlightening, this series of studies may be criticized
on three grounds:

• The blood is exposed to an air–liquid interface and to materials other

than the test materials. Thus, hemolysis may be associated with
increased cell fragility and with contact with nontest surfaces. These
effects are contained within the constant A; however, they prevent
the derivation of an absolute hemolysis rate for a material.
• The results obtained are specific for the surfaces actually used and
not representative to the general material classes from which they
are selected. Specimen-specific features, such as type and degree of
roughness, clearly affect the outcome (Monroe et al. 1981).
• The experiment is conducted in vitro and under conditions in which
significant hemolysis rates will occur in brief periods; thus, it is very
difficult to relate these results to those that might be obtained in vivo
at more realistic low hemolysis rates.

9.5 Final Comments

It should be clear from the brief remarks in this chapter that understanding
and controlling the host response of materials exposed to blood represents
one of the great unsolved problems of biomaterials science. Valve and vas-
cular replacements, left ventricular assist devices, and total artificial heart
replacements, as well as dialyzers and oxygenators used for shorter periods,
have become very sophisticated in design and control. However, their clinical
utility continues to be severely limited by unwanted interactions between
their materials of manufacture, the degradation products of those materials,
and circulating blood proteins and cells, especially those affected by shear
forces and foreign surface contact. Thus, it is difficult to overemphasize the
need for progress in developing materials that can function well in the
cardiovascular system for long periods of time.
180 Biological Performance of Materials: Fundamentals of Biocompatibility

The problem is complicated by several factors:

• The phenomena of coagulation and hemolysis are complex and

multifactorial.
• There seems to be a broad range of host response, particularly in
temporal variations in response by individuals, unlike the less spe-
cific response to inflammation, with only quite modest responses
tolerable chronically.
• Partly because of the first two points, as well as the rapidity of
development of the coagulation cascade, adequate experimental
models and techniques for research in this field are acutely lacking.

Fortunately, the search for improved “blood compatibility” as an attribute

of biomaterials continues to be vigorous. Harker et al. (1993) provide an
excellent overview of this difficult problem. Part III of this book will return
to the specific issue of deficiencies in testing and evaluation.

References
Baier, R.E., The role of surface energy in thrombogenesis, Bull. N.Y. Acad. Med., 48,
257, 1972.
Banerjee, R. et al., Hematological aspects of biocompatibility — review article, J.
Biomater. Appl., 12, 57, 1997.
Harker, L.A., Ratner, B.D. and Didisheim, P. (Eds.), Cardiovascular Biomaterials and
Biocompatibility. Cardio. Pathol., 2(3) (Suppl), 1993, 1.
Hayashi, K., In vivo thrombus formation induced by complement activation on poly-
mer surfaces, J. Biomed. Mater. Res., 24, 1385, 1990.
Herzlinger, G.A. et al., Quantitative measurement of C3 activation at polymer sur-
faces, Blood, 57, 764, 1981.
Kiraly, R.J. and Nosé, Y., Natural tissue as a biomaterial, Biomat. Med. Dev. Art. Org.,
2(3), 207, 1974.
Lampert, R.H. and Williams, M.C., Effect of surface materials on shear-induced
hemolysis, J. Biomed. Mater. Res., 6, 499, 1972.
Levy, R.J. et al., Calcification of valved aortic allografts in rats: effects of age, crosslink-
ing, and inhibitors, J. Biomed. Mater. Res., 29, 217, 1995.
Lindon, J.N. et al., Does the conformation of adsorbed fibrinogen dictate platelet
interactions with artificial surfaces?, Blood, 68, 355, 1986.
Monroe, J.M. et al., Surface roughness and edge geometries in hemolysis with rotating
disk flow, J. Biomed. Mater. Res., 15, 923, 1981.
Nosé, Y. et al., Surface characteristics of cardiac prostheses in vivo, J. Biomed. Mater.
Res. Symp., 8, 85, 1977.
Salzman, E.W., Surface effects in hemostasis and thrombosis, in The Chemistry of
Biosurfaces, Vol. II. Hair, M.L. (Ed.), Marcel Dekker, New York, 1972, 489.
Sawyer, P.N. and Kaplitt, M.J. (Eds.), Vascular Grafts, Appleton–Crofts, New York,
1978.
Coagulation and Hemolysis 181

Sawyer, P.N. and Srinivasan, S., The role of electrochemical surface properties in
thrombosis at vascular interfaces: cumulative experience of studies in animals
and man, Bull. N.Y. Acad. Med., 48, 235, 1972.
Skarja, G.A. et al., A cone-and-plate device for the investigation of platelet biomaterial
interactions, J. Biomed. Mater. Res., 34, 427, 1997.
Vogler, E.A., Structure and reactivity of water at biomaterial surfaces, Adv. Colloid
Interface Sci., 74, 69, 1998.
Vroman, L., Summation: protein at the interface, Fed. Proc., 30, 1703, 1971.
Weinberg, C.B. and Bell, E., A blood vessel model constructed from collagen and
cultured vascular cells, Science, 231, 397, 1986.

Bibliography
Department of Health and Human Services, Guidelines for Blood–Material Interactions,
NIH Publication 85-2185, Public Health Service, National Institutes of Health,
U.S. Government Printing Office, Washington, D.C., 1986.
Bamford, C.H. (Ed.), The Vroman Effect, Coronet, Philadelphia, 1992.
Barbanel, J.C. et al., Blood Flow in Artificial Organs and Cardiovascular Prostheses, Clar-
endon Press, Oxford, 1989.
Basmadjian, D. et al., Coagulation on biomaterials in flowing blood: some theoretical
considerations, Biomaterials, 18, 1511, 1972.
Bodnar, E. and Frater, R., Replacement Cardiac Valves, McGraw–Hill, New York, 1991.
Bruck, S.D., Blood Compatible Synthetic Polymers, Charles C Thomas, Springfield, IL,
1974.
Ellis, J.T. et al., Prosthesis-induced hemolysis: mechanism and quantification of shear
stress, J. Heart Valve Dis., 7(4), 376, 1998.
Gott, V.L. and Furuse, A., Antithrombogenic surfaces, classification and in vivo eval-
uation, Fed. Proc., 30, 1679, 1971.
Hanson, S.R., Blood–material interactions, in Black, J. and Hastings, G. (Eds.), Hand-
book of Biomaterial Properties, Chapman & Hall, London, 1998, 545.
Hastings, G., Cardiovascular Biomaterials, Springer–Verlag, Berlin, 1991.
Hughes–Jones, N.C., Lecture Notes on Haematology, 7th ed., Blackwell Scientific, Lon-
don, 2003.
Kambic, H.E., Kantrowitz, A. and Sung, P. (Eds.), Vascular Graft Update, Safety and
Performance. ASTM STP 898. American Society for Testing and Materials, Phil-
adelphia, 1986.
Lefrak, E.A. and Starr, A., Cardiac Valve Prostheses. Appleton–Century–Crofts, New
York, 1979.
Leonard, E.F., Turitto, V.T. and Vroman, L. (Ed.), Blood in Contact with Natural and
Artificial Surfaces. Ann. N.Y. Acad. Sci., 516, 1988.
Magnani, A. and Barbucci, R., Hemocompatible materials, surface and interface
aspects, in Encyclopedic Handbook of Biomaterials and Bioengineering, Part B, Ap-
plications, Vol. 2. Wise, D.L. et al. (Eds.), Marcel Dekker, New York, 1995, 1101.
Merrill, E.W., Properties of materials affecting the behavior of blood at their surfaces,
Ann. N.Y. Acad. Sci., 283, 6, 1977.
Sawyer, P.N. et al., Physical chemistry of the vascular interface, in Vascular Grafts,
Sawyer P.N. and Kaplitt M.J. (Eds.), Appleton–Crofts, New York, 1978, 53.
182 Biological Performance of Materials: Fundamentals of Biocompatibility

Smith, J.P. and Sawyer, P.N. (Eds.), Modern Vascular Grafts, McGraw–Hill, New York,
1986.
Schoen, F.J., Interventional and Surgical Cardiovascular Pathology, W.B. Saunders, Phil-
adelphia, 1989.
Thubrikar, M. et al., Study of surface charge of the intima and artificial materials in
relation to thrombogenicity, J. Biomech., 13, 663, 1980.
Van Kampen, C.L. et al., Effect of implant surface chemistry upon arterial thrombosis,
J. Biomed. Mater. Res., 13, 517, 1979.
Vroman, L., Blood, American Museum Science Books, B26, Doubleday, New York,
1968.*
Wagner, W.R. et al., Blood biocompatibility analysis in the setting of ventricular assist
devices, J. Biomater. Sci. Polym. Ed., 11(11), 1239, 2000.

* Blood, by Leo Vroman, is a marvelous, amusing, and witty account of blood biochemistry and
surface interactions by one of the leading blood physiologists of the 20th century. Although one
will enjoy it as recreational reading, one cannot avoid learning a great deal about blood. In addi-
tion to being a researcher, Leo Vroman is an author of poetry and children’s books. Unfortu-
nately, most have only been published in Dutch, his native language. Love, Greatly Enlarged (1991,
Cross-Cultural Communications) is a major poem in English and, given the author, is about
many things, not the least of which is blood.
10
Adaptation

10.1 Introduction
So far in Part II, acute host responses to singular events have been considered.
The insertion of an implant may evoke inflammation, with an acute course
and a longer chronic phase. Interruption of a blood vessel by injury or
insertion of an implant may trigger acute coagulation and, perhaps, chronic
hemolysis. Chapter 13 will take up the subject of neoplastic transformation:
abnormal tissue development and elaboration as a result of chemical or
foreign-body challenge.
Between these two types of events — acute response and abnormal devel-
opment —is another class of tissue response; that is, the presence of an
implant, perhaps due to the implant’s chemical, physical, or electrical prop-
erties, affects the organization and elaboration of tissue elements in the
vicinity. These events must be considered because of the well known ability
of many of the tissues to remodel adaptively, in response to physical induc-
tion factors, to reflect changes in demand and function. From the phrase
“adaptively remodel,” I shall purloin the term “adaptation” to describe such
events, especially as influenced by implants. The recognition and manage-
ment of adaptation is an important aspect of the design, development, and
use of interactive (type 2) biomaterials.

10.2 Tissue Growth Strategies

10.2.1 General Principles
Goss (1978, p. 2) discusses the strategy of growth of tissue in terms of three
patterns previously recognized by Bizzozero (1894):

• Expanding tissues grow by mitosis to increase cell number (for

instance, the liver).

183
184 Biological Performance of Materials: Fundamentals of Biocompatibility

• Static tissues retain essentially constant cell number but grow by

individual hypertrophy (for instance, muscle).
• Renewing tissues retain essentially constant cell number by replac-
ing losses from differentiation of proliferating stem cells (for
instance, skin).

In the rapidly developing immature individual, all tissues expand by

mitosis. As maturation proceeds, the cells of some tissues lose their mitotic
ability and the tissues become static or renewing. It is also possible for tissues
to continue to display a combination of two of these patterns, varying their
response to the challenge.
The question of control of these processes is still to a certain degree unre-
solved. Because all of these processes tend towards limits, a variety of neg-
ative-feedback control systems have been proposed (see Section 10.3.6 for
one example). However, a response is evoked in each cell type by death of
a portion of the tissue or by surgical removal. Less significant challenges,
such as blunt trauma or change in mechanical functional requirements, may
also evoke responses. These observations suggest the existence of a wide
variety of control processes regulating the quantity and, to some degree, the
type of tissue in any location in a mammalian body. It is altogether reasonable
to assume that the response of any such active biological control system
facing a challenge may be affected by the presence of an implant.

10.2.2 Fracture Healing

The process of fracture healing — the restoration of the integrity and conti-
nuity of mineralized tissue after mechanical injury — includes an interesting
example of natural adaptation in the absence of implants. Successful fracture
healing can be considered most generally to entail four stages:

1. Hematoma: from injury to formation of a “soft” callus

2. Soft callus: from initiation of soft callus formation until its conden-
sation into a “hard” callus
3. Hard callus: from initiation of hard callus until normal stiffness is
restored
4. Remodeling: from restoration of normal stiffness to full restoration
of normal structure

Stages 1 and 2 (see Figure 10.1) represent the acute or healing phase leading
to the formation of a natural splint or callus. This is a weak provisional tissue
consisting of fibrocartilage and fibrous bone with little internal organization.
However, when 50 to 75% of the normal stiffness is reached, the healing
fracture undergoes a very dramatic and fairly rapid transformation. The soft
Adaptation 185

1 2 3 4 (end)

Stiffness (100%)
Strength (100%)

Injury Time

FIGURE 10.1
Stages of fracture healing.

callus rapidly condenses and shrinks to form a more organized hard callus,
then is slowly resorbed through a progressive remodeling process leading
to full restoration of normal structure (including reopening of the medullary
canal in long bones). These latter two stages appear to be mediated by
mechanical requirements through Wolff’s law (see Section 10.3.4) and lead
to an efficient structure with normal properties. Thus, stages 3 and 4 of
fracture healing can be thought of as a chronic phase and appear to represent
adaptive remodeling of a functionally healed bone (end of stage 2).
Note that normal stiffness is obtained first and, in fact, generally exceeded,
due to the combined presence of callus and healing cortical bone in the
defect; normal strength is obtained only towards the end of the remodeling
phase.
In any particular bone, fracture healing, with a certain degree of injury,
can be expected to progress to completion in an average time characteristic
of the initial condition and, to some degree, the type of treatment (internal
fixation, external splinting, etc.). If there is a significant delay but eventual
complete healing and remodeling, the condition is termed delayed union.
However, it is possible for the process to be interrupted at any stage. If
this occurs in stages 1 and 2, before any degree of structural integrity is
obtained, and is permanent, the result is a nonunion, sometimes called a
“pseudarthrosis.”* Nonunions are frequently sites of active tissue turn-
over but rarely heal without additional intervention, validating the dis-
tinction made here between nonadaptive stages (1 and 2) and the adaptive
ones (3 and 4).

* Black (1987) includes a more detailed discussion of these diagnostic distinctions.

186 Biological Performance of Materials: Fundamentals of Biocompatibility

10.3 Examples of Adaptation in Implant Applications

10.3.1 Introduction
Implant applications offer a wide variety of possible examples of adaptation.
Consideration here will be restricted to the following examples:

• Growth of the “neointima” in arterial prostheses

• Attachment of tendon prostheses to soft tissue
• Hard tissue remodeling in the presence of implants
• Bone response to electrified implants

Many other examples of adaptation to the presence of implants can be found.

In fact, it may be useful to consider that, when resolution to a stable situation
is possible (see Section 8.2.6), accommodation of the local host tissues to
implants consists most generally of two phases: an acute or healing phase
and a chronic or adaptive phase.
In all of the adaptive processes, the control mechanism invoked is phe-
nomenologically described by a unifying principle. In hard tissue, this prin-
ciple is termed Wolff’s law. Julius Wolff, a 19th century German anatomist,
suggested in 1892 that bone (and, by inference, other load-bearing tissues)
remodels in an attempt to maintain a constant (optimal) pressure or local
stress. Although he did not make such a specific statement, it is usual to say
that Wolff’s law is: “The form being given, tissue adapts to best fulfill its
mechanical function.”
Thus, as load increases over a period of time, bone mass would be expected
to increase to maintain a constant level of stress. Conversely, a reduction in
load should produce a loss of tissue. In a more sophisticated interpretation,
realignment of principal stresses should be reflected by a modification of
tissue structure. Observations in a variety of static and renewing tissues
suggest a considerable pragmatic generality for Wolff’s law.

10.3.2 Vascular Adaptation

The growth of a new tissue layer, or neointima, on the internal surfaces of
knitted arterial prostheses is an interesting example of adaptation. Here, the
prosthesis is introduced to serve as a framework to support host tissue that
will serve as a long-term nonthrombogenic blood contact surface. The body
of the prosthesis, which is most commonly made of polyester or similar
polymeric fiber, provides the mechanical resistance to internal pressures that
was once provided by the original vessel.
Upon insertion of the prosthesis, the features of the clotting cascade
described in Section 9.2 take place. In fact, the surgeon usually preclots
the prosthesis using the patient’s blood in an effort to provide a smooth,
Adaptation 187

defect-free surface and to reduce blood loss. The fibrin surface, inside and
outside, matures and achieves stable dimensions within 24 hours. At this
point, healing is clearly initiated. Within a few weeks, the outside (tissue
side) of the prosthesis is covered by granulation tissue, and a capsule of
fibrous tissue matures with increasing organization. The resolution or heal-
ing response of the internal surface facing flowing blood is a different matter.
Schoen (1989) distinguishes between two forms of vascular resolution:
pseudointimal formation, the mere coating of the implant’s surface with
proteins and cells other than endothelial cells, and neointimal formation, the
formation of an endothelial lined surface, usually overlying a layer of smooth
muscle cells. The choice of outcome is apparently governed by the nature of
the biomaterial surface and, as will be seen, can be considered as an
example of adaptation of the natural healing process to the properties of the
prosthesis.
During ideal or optimal healing, blood vessels penetrate to the lumen of
the prosthesis and “tufts” of tissue spread along the inner surface, merging
to form a smooth neointimal surface that functions like the natural arterial
lining. Some investigators (e.g., Annis et al. 1978) suggest that this process
of “through-growth” is not essential to the formation of the neointima and
that longitudinal growth from the ends inward is possible in impermeable
vascular prostheses. In a pig, this process is completed within a month, but
it takes up to a year in a human patient.
However, this healthy adaptive vascular maturation has been shown
(Wesolowski et al. 1968) to be critically dependent upon the prosthesis poros-
ity. Figure 10.2a displays the relationship developed for a variety of porous
prosthetic materials tested as arterial grafts in the pig. The porosity is given
in units of liters/min/cm2 of water expressed from the lumen through the
wall of an unclotted non-blood-exposed graft with a pressure differential of
120 mmHg. The calcification index is the product of the average calcification
(on a subjective scale of 1+ to 4+) of all animals implanted with a given
material and the percentage of all specimens of the same material that dis-
play calcification. Here, high porosity above a value of 1 favors maturation
of the neointima; lower porosity leads to failure of microvascular ingrowth,
focal necrosis of the neointima, and possible calcification of the resulting
pseudointima, in many respects mirroring the events of arteriosclerosis in
natural tissue.
Attempts have been made to moderate this process by “seeding” the lumen
wall with autologous cells (Kahn and Burkel 1973), with cultured endothelial
cells (Mansfield et al. 1975), and by various pretreatments; however, the
relationship found by Wesolowski et al. (1968) appears to be well founded.
The generality of this relationship is suggested by Figure 10.2b. This dis-
plays the relationship of a “net acceptability index” and the previously
discussed physical water porosity. It represents the results of the evaluation
of a variety of knit and velour polyester arterial prostheses (Sawyer et al.
1979). The net acceptability index is a transformation of the “net evaluation
index” used by the authors so that the low scores are satisfactory, as in Figure
188 Biological Performance of Materials: Fundamentals of Biocompatibility

Porosity 2

0 100 200 300

Calcification Index
(a)

4
Porosity

0 1 2 3
Net Acceptability Index
(4 - Net Evaluation Index)
(b)

FIGURE 10.2
(a) Relationship between physical water porosity of arterial (woven) prostheses and calcification
in the pig. (b) Relationship between physical water porosity and net acceptability index of
arterial grafts (dashed lines indicate range). (Figure 10.2a adapted from Wesolowski, S.A. et al.,
Ann. N.Y. Acad. Sci., 146(1), 325, 1968; Figure 10.2b adapted from Sawyer, P.N. et al., J. Biomed.
Mater. Res., 13, 937, 1979.)

10.2a. The net evaluation index is a combined score based upon in vivo
performance and postimplantation evaluation. It is clear that the same result
is obtained as that found by Wesolowski et al. (1968); high porosity favors
good in vivo performance. This is clearly an example of a structural attribute
of an implant strongly affecting tissue adaptation after surgery. Unlike the
case of Wolff’s law adaptation in bone, where the amount of tissue is affected
Adaptation 189

by the implant attribute, here the type of tissue is affected by the (nonchem-
ical) implant attribute.

10.3.3 Prosthetic Replacement of Tendons

The second area of adaptation to be considered is in the area of tendon
prostheses. Here, the natural system provides a junction between the soft
tissue (tendon) and the hard tissue (bone) by a series of collagenous fibers
(Sharpey’s fibers) that pass into the bone. There has been considerable inter-
est in reproducing such a system by inserting a porous material into the
bone or by providing a porous bridge between the end of the natural tendon
insertion and a tendon remnant or the muscle. In either case, a situation is
desired in which the natural tissue will grow into the prosthesis and provide
mechanical strength. The topic of bony ingrowth will be discussed later.
Homsy et al. (1972) performed an interesting study of this subject using a
porous graphite-reinforced poly (tetrafluoro) ethylene (Proplast™, Vitek,
Inc., Houston, TX).* Blocks of this material were inserted in the rabbit cal-
caneal tendon, replacing segments of the natural structure, and sutured in
place to be load bearing immediately. At periods of up to 26 weeks, the
rabbits were sacrificed and the strength of the bond between natural and
prosthetic material was tested in tension. The results are shown in Figure
10.3. These results are reported to be the same for tendon/prosthesis and
muscle (gluteal)/prosthesis interfaces and to reach a plateau value of rupture
strength shortly after 8 weeks — about 5 weeks later than for collagenous

4
Unit Rupture Stress

3
(KgF/cm2)

0 2 4 6 8 10 12 14 16 26
Time (weeks)

FIGURE 10.3
Prosthesis/tissue anastamosis strength. (From Homsy, C.A. et al., Clin. Orthop. Rel. Res., 89, 220,
1972.)

* I discussed this study in the first edition of this work (published in 1981). Since then, the sad story
of the misapplication of this material to load-bearing applications in bone, such as in the attempted
replacement of the human temperomandibular joint, has become well known. However, in prepar-
ing this edition, I elected again to retain this study for two reasons: (1) it is, for its time, a well done
piece of work; and (2) it is an object lesson concerning the limits of animal studies and the reality
of the need to define biological performance in relationship to specific applications.
190 Biological Performance of Materials: Fundamentals of Biocompatibility

20

Rupture Strength
Normal
16 range

(KgF)
12
8
4

0 2 4 6 8 10 12
Time (weeks)

FIGURE 10.4
Breaking strength of pseudotendon. (Adapted from Jenkins, D.H.R. et al., J. Bone Joint Surg.,
59B, 53, 1977.)

ingrowth in bony sites. Here the gradual development of strength, as con-

trasted with the rapid (within 3 weeks) complete penetration of unorganized
collagen that the authors would have predicted from earlier studies suggests
that an organized structure is forming under control of the axial mechanical
load exerted by the muscle.
This picture is substantiated by studies of a different system in this appli-
cation. Here, the system is a complete tendon prosthesis that gradually
disintegrates under use. Studies with a braided carbon ligament in rabbits
(Forster et al. 1978) and sheep (Jenkins et al. 1977) showed that, as the
prosthesis frayed and fatigued, it served as a scaffold for the growth of
fibrous tissue that took over the function of the prosthesis. Figure 10.4 shows
the breaking strength of the rabbit calcaneal tendon prostheses from this
study. The authors ascribe the rapid development of strength to a stimulation
effect of the carbon, but one might equally suggest that the gradual transfer
of stress to the ingrowing tissue evoked this adaptive response. The devel-
opment of maximum strength at 6 weeks (somewhat earlier than in the study
of Homsy et al. in 1972) suggests that the disintegration of the prosthesis
served to transfer stress more rapidly than was the case with a more durable
material.
This experiment has been carried to a natural conclusion by incorporating
the carbon fibers in a matrix that is degradable in the internal environment
(Alexander et al. 1979). The matrix used here was polylactic acid, and the
implant was inserted as a replacement for the canine patellar tendon. A
gradual dynamic replacement of prosthesis with organized fibrous tissue
was reported as the prosthesis matrix disintegrated, with reasonable matu-
rity and linear tissue arrangement after 2 months of implantation. However,
in neither of these experiments did the resulting tissue closely come to
resemble normal tendon. Thus, it should be best termed neotendon, in par-
allel with Schoen’s (1989) terms for healing vascular endothelial tissue. In
more extreme cases when the new tissue more resembles a fibrous scar or
capsule, it is more appropriately termed pseudotendon and its long-term
durability under mechanical loads must surely be questioned. Balduini at
Adaptation 191

al. (1986) have summarized further clinical developments that attempted to

build on these observations.

10.3.4 Adaptive Remodeling of Bone near Implants

10.3.4.1 Stress Shielding
Wolff’s law suggests that the introduction of a load-bearing implant coupled
to a previously load-bearing natural structure should result in some atrophy
or tissue loss. Because the assumption of a portion of the tensile load or
bending moment by the implant must necessarily reduce the local stress in
the adjacent tissue, this phenomenon has come to be called “stress shielding”
(Huiskes 1988). The most commonly cited example is that of the progressive
loss of bone material in the proximal medial femoral cortex (calcar) observed
in animals and humans after total replacement of the hip joint. However,
observation of the remodeling response is complicated by a simultaneous
presence of an osteolytic response mediated by the presence of wear debris
released from the articulating interface (Amstutz et al. 1992).
A more easily examined example is an unfortunate concomitant of the
use of metallic internal fracture fixation devices. These devices are usually
far more rigid than the bone to which they are attached, even if the bone
is intact (as in experimental situations). Although they provide excellent
support and maintain reduction and fixation during healing, a considerable
amount of osteoporosis,* or loss of bone mass without external change of
shape, occurs in the bone under the plate or adjacent to the rod. A study
of experimental fractures in rabbits (Brown and Mayor 1978) seems to
reinforce this point. Fractures were produced in the tibias of rabbits and
were internally fixed with rods made of a variety of metals and polymers.
The rods were a constant diameter to provide a range of stiffness relative
to the intact bone from 10 to 0.03×. The animals were sacrificed and studied
at 9 and 16 weeks after fracture and fixation. The results were confused by
an anomalous response to one metal alloy (Ti6Al4V); however, at 16 weeks,
fractures fixed with rods that were less stiff than the intact bone were
significantly stronger and tougher in torsion than those fixed with rods
that were stiffer than the intact bone. Additionally, in the weaker, less tough
bones (fixed with the stiffer rods), a greater degree of osteoporosis was
seen histologically.
This experiment is difficult to interpret solely in terms of an adaptive
response to implants because the changes in bone are presumably caused by
two factors: a healing response and an adaptive response. Moyen et al. (1978)
conducted an experiment in dogs in which no fracture was involved and
obtained somewhat similar results. In this experiment, metallic plates of two

* The term osteoporosis is more generally applied to metabolic rather than adaptive loss of bone.
(See NIH Consensus Statement, Vol. 17 (1)), Osteoporosis Prevention, Diagnosis and Therapy,
Washington, D.C., 2000.) However, the structural and mechanical consequences are essentially
the same as in adaptive remodeling, so the same term is used here.
192 Biological Performance of Materials: Fundamentals of Biocompatibility

different stiffnesses varying by a factor of approximately 5 were attached to

the midshaft of the femur in the dog. The bone mass under the plates was
compared with the control (unplated) side after 6 and 9 months implantation.
Interestingly, a small constant porosity of 1 to 3% was seen. However, after 6
months, the bone mass under the rigid plate had decreased 26.4% vs. only
16.4% under the more flexible plate. The decreases in bone mass continued to
develop more slowly up to 9 months and showed modest reversals in another
group implanted for 6 months and studied 3 months after plate removal. The
authors ascribed the failure to observe the greater porosity seen in other studies
(such as the one by Brown and Mayor in 1978) to the absence of the fracture-
healing process accompanying the remodeling (adaptive) process.
Perren et al. (1988) further criticized the possible adaptive role in the
production of porosity near fracture fixation devices and suggested that the
physical presence of the device interferes with revascularization after frac-
ture. He showed an inverse relationship between the area of plate–bone
contact and cortical porosity in a sheep tibial model. One may, in turn,
criticize this work because reducing the plate–bone contact area probably
reduces the coupling between plate and bone, thus rendering the plate
effectively less stiff — that is, less able to remove load (and thus reduce local
stress) from the bone.
Another study by Bradley et al. (1979) combined some of the aspects of
the studies of Brown and Mayor (1978) and Moyen et al. (1978). In this case,
a variety of fracture fixation plates, with stiffnesses between 4 and 40% of
the bone to be fixed, were used in a 16-week study of the healing of femoral
midshaft osteotomies in dogs. A definite relationship was seen between plate
rigidity and strength of the bony material and the femoral midshaft as a
structure. This is shown in Figure 10.5.
Two comments should be made about this study. In the first place, the
parameter examined was strength rather than rigidity, as in the case of the
previous two studies. Although rigidity may parallel strength, the correlation
is not exact, even in materials that are simpler in microstructure than bone.
Furthermore, restoration of rigidity is probably more important in the early
phases of fracture healing because it confers functional ability before adap-
tation is complete. In the second place, porosity was not studied, as in the
work of Moyen et al. (1978). Porosity was suggested by Moyen to be a
concomitant of trauma rather than simply adaptation (my term) and by
Perren et al. (1988) as secondary to interference with revascularization; it
may severely affect material and structural strength due to stress concentra-
tion effects, but have a modest volume-fraction effect upon modulus and
thus upon material and structural bending rigidity (see Section 6.3.2). It is
impossible to isolate this effect in this study.
One should not assume from these studies that large changes in stress are
necessary to modify bone growth — that is, to produce adaptive changes in
bone. Modest changes in stress, such as those that might be produced by
simple, soft polymeric caps of bone shafts after segmental excision, have
been shown to produce profound adaptive changes (Lusskin et al. 1972).
Adaptation 193

1.4

1.2
beneath plate

Material Strength
opposite plate

| σ/σ control |
1.0
Normalized
0.8

0.6

0.4

0.2

0 2 4 6 8 10 12 14 16 18
Plate Flexural Rigidity (N - m2)

1.4
beneath plate
1.2 opposite plate
Structural Strength

1.0
| s/s control |
Normalized

0.8

0.6

0.4

0.2

0 2 4 6 8 10 12 14 16 18

Plate Flexural Rigidity (N - m2)

FIGURE 10.5
Changes in bone material and structural strength with fixation plate rigidity. (Adapted from
Bradley, G.W. et al., J. Bone Joint Surg., 61A, 866, 1979.)

10.3.4.2 Ingrowth into Porous Biomaterials

It is an easy step from the earlier discussion of the ingrowth associated with
the prosthetic replacement of tendons to consideration of the more general
problem of ingrowth into porous bodies. One of the responses to implants
is the formation of a fibrous capsule, so it is no surprise that tissue will
invade the internal spaces of an implant with an open, connected pore
structure.
Extensive studies of this phenomenon have been conducted. Ingrowth
occurs into porous implants fabricated from a wide variety of metals,
polymers, and ceramics. The nature of the ingrowing tissue in the presence
of sufficient interfacial mechanical stability* depends upon the size of the
* The issue of the role of interfacial shear (producing the so-called “micromotion”) on tissue
ingrowth remains sufficiently confused that an analytical discussion of this point is not possible.
However, see Brunski (1988) and Prendergast et al. (1997) for fuller discussions of this topic.
194 Biological Performance of Materials: Fundamentals of Biocompatibility

pore (Friedenberg and Lawrence 1959) or, more properly, on the minimum
size of the interconnections between pores (Klawitter and Weinstein 1974).
Soft tissue elements will be found in interconnects as small as 1 to 5 μm;
at some minimum interconnect diameter between 50 and 100 μm, miner-
alized tissue will be found and organized osteonal bone will grow into
interconnects as small as 250 μm. Maximum interfacial shear strengths
develop between 8 and 16 weeks after implantation, depending upon
anatomical locations, species of animal, and type of tissue ingrowth. Veloc-
ity of ingrowth appears to increase with pore sizes above 50 μm and to
reach a peak near pore sizes of 400 and 500 μm, as determined in a single
pore model (Howe et al. 1974).
The ability of tissue to mineralize and organize as interconnect size
increases is another clear example of adaptation. A study of Proplast™ (Vitek,
Inc., Houston, TX) by Spector et al. (1979) confirmed this finding and dem-
onstrated that it is not a false conclusion based upon comparison of studies
with different materials and/or test conditions. If implanted directly, this
material exhibits a pore size near 76 μm with an interconnect size of 50 μm.
In a canine cortical bone site, only fibrous ingrowth was observed for periods
of up to 20 weeks. However, if the material was “teased” before implantation
to increase the size of interconnects, a variable degree of bony ingrowth
occurred.
There appears to be a difference of opinion over the interpretation of the
findings in this report (Homsy 1979; Spector 1979). This finding, combined
with earlier reports (Klawitter and Hulbert 1971), suggests a practical lower
interconnect limit of 100 μm for bone ingrowth. The mechanism of control
of ingrowth and the manner in which pore interconnect size controls min-
eralization are unknown. Although Wolff’s law arguments can be invoked
to explain tissue remodeling near the bone–implant interface, it is presumed
that tissue more than one pore diameter deep within the implant porosity
will be essentially load free if the modulus of the implant exceeds that of
bone to any significant degree. However, tissue maturation internal to porous
implants appears to have little dependence on implant material modulus,
but may be related more to electrical, chemical, and morphological effects
on the pericellular environment.

10.3.4.3 Adhesion
Tissue is not inherently “sticky.” Cells adhere to each other and to their
extracellular matrices through the interaction of a variety of specific and
nonspecific adhesion molecules and specific cell surface receptors for por-
tions of these molecules (see Section 11.3.3), in addition to more diffuse Van
der Waals bonding mechanisms. However, attempts to cause implants pur-
posefully to adhere to tissue have aroused considerable interest.
When such adhesion is produced by the mere close (molecular-scale)
approximation of tissue and implant, without an intervening fibrous capsule
or other elements of an inflammatory response, it is termed most generally
Adaptation 195

tissue integration; in the case of bone, the more specific term is osseointe-
gration (Albrektsson and Hansson 1986). In the latter case, the apparent
adhesion is produced by cellular binding to proteins adsorbed to the implant
surface. Originally thought to be a property of pure titanium alone, such
tissue integration has now been shown for a variety of metallic implant
surfaces (Linder 1989), including tantalum (Black 1994). In fact, Linder (1989)
has suggested that, in general, “osseointegration is a response of bone to a
tolerable implant material inserted under tolerable conditions” without spec-
ifying the meaning of “tolerable” in either case.
When tissue adhesion to an implant is accompanied by a chemical alter-
ation of the implant surface, a true bonding process with a continuous
gradation of structure and composition across the tissue–implant interface
may occur. Although there is no generally accepted term for this condition,
interactive biomaterials that produce it have been termed surface active
(Hench and Wilson 1984) in recognition of the necessity of chemical reaction
with the local host environment prior to bond formation.* A number of
ceramic and glassy materials have been produced that develop such bonds
to bone and soft tissue (Figure 10.6).
In the case of integration or of bonding, the implant becomes mechanically
coupled to the adjacent tissue. In the case of hard tissue, this results in a
strain incompatibility due to the differences in moduli between bone and
the implant. Several adaptive changes are possible. The most common situ-
ations are:

SiO2

Inert
Resorbable
Cervital® (Leitz)
A/W Glass Ceramic Adhesive 45S5 Bioglass® (U. of Florida)
(Kyoto U.)
Bone bonding boundary

CaO(MgO) + P2 O5 Na2 O(K2O)

Compositions where soft tissue adhesion occurs

FIGURE 10.6
Ceramic and glassy tissue-bonding materials. (From Hench, L.L. and Wilson, J., Science, 226,
630, 1984. With permission: L.L. Hench.)

* The term bioactive has also been used for such materials (see Section 1.4). However, this is an
apparent misnomer because it appears that the necessary surface modification is a consequence
of exposure to the physiological environment, rather than to life processes.
196 Biological Performance of Materials: Fundamentals of Biocompatibility

• The implant is placed in cancellous bone and is of very significantly

higher modulus than the surrounding tissue. The frequently
observed response in this case is the formation of a bony “plate”
much resembling a subchondral plate in an articular joint and a
relative rarefaction of the cancellous trabeculae behind the plate.
This results in introducing a relatively compliant tissue zone adja-
cent to the implant and “shielding” the bone further away from the
mechanical consequences of the stiff implant.
• The implant is placed in cortical bone and is of significantly higher
modulus than surrounding tissue. In this case, bone near the implant
becomes porous, much as in the stress-shielding examples previ-
ously discussed (Section 10.3.4.1). However, this porosity may
increase and proceed to a remodeled condition resembling cancel-
lous bone. Thus, this process is termed cancellization.

In either case, the tissue structure changes are a result of the change in
mechanical conditions near the newly formed interface and thus are true
examples of adaptive remodeling.

10.3.5 Bone Response to Electrified Implants

It has been frequently suggested that the mechanism of Wolff’s law in bone
is electrically controlled. Bone and other tissues produce potentials when
deformed; these potentials are called piezoelectric potentials or, more gener-
ally, strain-related potentials. A variety of other potential sources also exists.
Bassett (1971) proposed a generalized, closed-loop control system (as
shown in Figure 10.7) to relate these signals to hard tissue remodeling. There
is a conceptual error in his scheme because, presumably, the structural
response (to adjust stress) results in a change in the osseous transducer (see
added dashed line), rather than in a modification of the extrinsic force as
proposed. Nevertheless, this general idea, first proposed some years before
this early review was published, has been a motivating factor in the inves-
tigation of the effects of electrical phenomena on the modification of bony
growth and remodeling. Numerous studies have shown a correspondence
between endogenous electrical phenomena and growth, repair, and remod-
eling processes; however, no critical experiments that show that these signals
are necessary and sufficient to serve as stimuli for the observed processes
have yet been performed.
Although it is not clear that electrical control of bone growth, remodeling,
and repair is an example of adaptive response to implants, it is worth noting
some general conclusions from research in this area for completeness of this
discussion (Black 1987). A physical electrode may be considered as a special
case of metallic implantation: one in which the electrode, instead of being
allowed to find its appropriate mixed corrosion potential (see Section 4.6),
is maintained under a controlled relative potential or current condition. The
Adaptation 197

EXTRINSIC
FORCE

STRUCTURAL (Correct effect) OSSEOUS

RESPONSE TO TRANSDUCER
RESIST FORCE

PROPORTIONAL
BIOLOGICAL ELECTRICAL
TRANSDUCER CONTROL SIGNAL

FIGURE 10.7
Negative feedback control system proposed as a basis for Wolff’s law. (Adapted from Bassett,
C.A.L., in Biochemistry and Physiology of Bone, Vol. 2, 2nd ed., Bourne, G.H., Ed., Academic Press,
New York, 1971, 1.)

responses to implantation of such a physical electrode in or near hard tissue

are summarized as:

• The presence of a metallic cathode, with a negative potential, stim-

ulates the conduct of cells involved in bone formation.
• In particular, bony repair in sites of trauma is accelerated, and cells
in medullary sites may be induced to form bone in the absence of
bony trauma or to maintain bone formed in response to transient
trauma.
• In all cases, apparently a narrow stimulatory “window” is defined
by limits on current and electrode potential.
• Finally, monophasic negative pulses of frequencies from 10 to 750
Hz, with duty cycles between 50 and 5%, provide stimulation that
approaches but does not exceed that of direct uninterrupted current.

Although the relationship of this line of research to the more general problem
of adaptive growth is unclear, it does represent one of the more systematic
attempts to elucidate the phenomena involved.

10.4 A Final Comment on Adaptation

It now seems clear that one of the original goals of implantation — that is,
to produce minimal tissue (host) response — is an outmoded view that may
198 Biological Performance of Materials: Fundamentals of Biocompatibility

limit further development of implant materials and devices. It is proper to

ask that adverse host response be kept within bounds acceptable to the
application. This is the necessary condition of biocompatibility. It is essential
to rule out a priori certain classes of response, such as neoplastic transforma-
tion, as unacceptable. Now it appears equally reasonable to pursue active
tissue response in the form of adaptation so that natural tissue can take over
the role of the implant as completely as possible.
One of the ways of viewing the research literature on local host response
is to note that it has three somewhat unconnected parts:

• The majority of the studies reported emphasize the chemical com-

position of the implant and its degradation products. The local host
response is seen as a physiological response to soluble chemical
species.
• A smaller body of work, primarily oriented to orthopaedic surgery,
emphasizes the stiffness of the implant in relation to surrounding
tissues, strains imposed by function, and, especially recently, the
surface texture and configuration of the implant surface. The local
host response is seen as a structural embodiment of consequences
of Wolff’s law.
• A still smaller set of studies report the cellular and tissue response
to imposed electrical fields and currents, especially in the context of
biomaterials, by implanted electrodes. The local host response is
seen, primarily, as mediated by information content of the electrical
effects and secondarily by local electrochemically induced pericel-
lular environmental changes.

Figure 10.8 illustrates the general schematic form of the proposition: physical
factors induce local host response through various mechanisms.
However, it may well be the case that these effects are not independent,
but interdependent and perhaps, in some cases, synergistic. Figure 10.9
summarizes the local host effects discussed in Chapter 8, Chapter 9, Chapter
12, and Chapter 13 and their relationships to the degree or intensity (“dose”)

PHYSICAL INDUCTION
FACTORS
CHEMICAL

ELECTRICAL LOCAL HOST

IMPLANT RESPONSE
MECHANICAL

FIGURE 10.8
Physical induction factors in local host response.
Adaptation 199

IMPLANT TISSUE

LOCAL HOST RESPONSE

PHYSICAL INDUCTION FACTORS
CHEMICAL ELECTRICAL MECHANICAL
LOW LOW STRESS “MOTION”

ACUTE, MILD GROWTH and ATROPHY ADHESION (?)

INFLAMMATION HEALING CANCELLIZATION INGROWTH
FB NEOPLASIA (?) HEALING
REMODELING ADAPTIVE PHAGOCYTOSIS
REMODELING CLASIS
ADHESION (?) HYPERPLASIA
HYPERSENSITIVITY SUB-CHONDRAL ENCAPSULIZATION
CHEM. NEOPLASIA CONDENSATION (?)
CHRONIC, SEVERE NECROSIS NECROSIS CHRONIC
INFLAMMATION RESORPTION FRACTURE INFLAMMATION

HIGH HIGH

(?): Possible effect

FIGURE 10.9
Adaptive responses to physical induction factors. (Adapted from Black, J., Orthopaedic Bio-
materials in Research and Practice, Churchill–Livingstone, New York, 1988, 288.)

of physical induction factors. Here I distinguish between interfacial stress

and resulting shear displacement (motion): these are, in fact, coupled because
the latter depends upon the former and on the surface structure of the
implant. A “rough” implant permits far smaller interfacial displacements
than a “smooth” nonadhesive one under the same stress. Normal host
response, the quiescent state after full, acceptable resolution has occurred
(see Section 8.2.6), is most generally characterized by low chemical activity,
intermediate (near charge neutral) electrical activity, and intermediate inter-
facial stress, accompanied by an appropriate physical configuration of the
implant surface. Particular cases, such as direct tissue adhesion, require other
combinations of levels of these factors.
Much work remains to be done until the principles of adaptation are fully
understood and harnessed to the solution of patient problems.

References
Albrektsson, T. and Hansson, H.-A., An ultrastructural characterization of the inter-
face between bone and sputtered titanium or stainless steel surfaces, Bio-
materials, 7, 201, 1986.
200 Biological Performance of Materials: Fundamentals of Biocompatibility

Alexander, H. et al., Ligament and tendon replacement with resorbable polymer-

filamentous carbon tissue scaffolds, Trans. Orthop. Res. Soc., 4, 27, 1979.
Amstutz, H.C. et al., Mechanism and clinical significance of wear debris-induced
osteolysis. Clin. Orthop. Rel. Res., 276, 7, 1992.
Annis, D. et al., An elastomeric vascular prosthesis, Trans. Am. Soc. Artif. Intern.
Organs, XXIV, 209, 1978.
Balduini, F.C., Clemow, A.J.T. and Lehman, R.C., Synthetic Ligaments: Scaffolds, Stents,
and Prostheses, Slack, Thorofare, NJ, 1986.
Bassett, C.A.L. Byophysical principles affecting bone structure, in Biochemistry and
Physiology of Bone, Vol. 2, 2nd ed., Bourne, G.H. (Ed.), Academic Press, New
York, 1971, 1.
Bizzozero, G., Brit. Med. J., 1, 728, 1894 (cited in Goss, 1978).
Black, J., Electrical Stimulation: Its Role in Growth, Repair, and Remodeling of the Muscu-
loskeletal System, Praeger, New York, 1987.
Black, J., Orthopaedic Biomaterials in Research and Practice, Churchill–Livingstone, New
York, 1988, 288.
Black, J., Biological performance of tantalum: a review, Clin. Mater., 16(3), 167, 1994.
Bradley, G.W. et al., Effects of flexural rigidity of plates on bone healing, J. Bone Joint
Surg., 61A, 866, 1979.
Brown, S.A. and Mayor, M.B., The biocompatibility of materials for internal fixation
of fractures, J. Biomed. Mater. Res., 12, 67, 1978.
Brunski, J.B., The influence of force, motion, and related quantities on the response
of bone to implants, in Non-Cemented Total Hip Arthroplasty, Fitzgerald, R.H.,
Jr. (Ed.), Raven, New York, 1988, 7.
Forster, I.W. et al., Biological reaction to carbon fiber implants: the formation and
structure of a carbon-induced “neotendon,” Clin. Orthop. Rel. Res., 131, 299,
1978.
Friedenberg, Z.B. and Lawrence, R., Bone growth in polyvinyl sponge, Surg. Gynecol.
Obstet., 109, 291, 1959.
Goss, R.J., The Physiology of Growth, Academic Press, New York, 1978.
Hench, L.L. and Wilson, J., Surface active biomaterials, Science, 226, 630, 1984.
Homsy, C.A., Comments on “characteristics of tissue growth into Proplast and porous
polyethylene implants in bone,” J. Biomed. Mater. Res., 13, 987, 1979.
Homsy, C.A. et al., Porous implant systems for prosthesis stabilization, Clin. Orthop.
Rel. Res., 89, 220, 1972.
Howe, D.F., Svare, C.W. and Tock, R.W., Some effects of pore diameter on single-pore
bony ingression patterns in Teflon, J. Biomed. Mater. Res., 8, 399, 1974.
Huiskes, R., Stress patterns, failure modes, and bone remodeling, in Non-Cemented
Total Hip Arthroplasty, Fitzgerald, R.H., Jr. (Ed.), Raven, New York, 1988, 283.
Jenkins, D.H.R. et al., Induction of tendon and ligament formation by carbon im-
plants, J. Bone Joint Surg., 59B, 53, 1977.
Kahn, R.H. and Burkel, W.E., Propagation of pseudointimal linings of vascular pros-
theses, In Vitro, 8, 451, 1973.
Klawitter, J.J. and Hulbert, S.F., Application of porous ceramics for the attachment
of load-bearing internal orthopedic applications, J. Biomed. Mater. Res. Symp.,
2, 161, 1971.
Klawitter, J.J. and Weinstein, A.M., The status of porous materials to obtain direct
skeletal attachment by tissue ingrowth, Acta Orthop. Belgica, 40, 755. 1974.
Linder, L., Osseintegration of metallic implants. I. Light microcopy in the rabbit, Acta
Orthop. Scand., 60, 129, 1989.
Adaptation 201

Lusskin, R. et al., Bone contouring under silicone polymer implants, Clin. Orthop.
Rel. Res., 83, 300, 1972.
Mansfield, P.B., Wechezak, A.R. and Sauvage, L.R., Preventing thrombus on artificial
vascular surfaces: true endothelial cell linings, Trans. Am. Soc. Artif. Intern.
Organs, XXI, 264, 1975.
Moyen, B.J.-L. et al., Effects on intact femora of dogs of the application and removal
of metal plates, J. Bone Joint Surg., 60A, 940, 1978.
Perren, S.M. et al., Early temporary porosis of bone induced by internal fixation
implants. A reaction to necrosis, not stress protection? Clin. Orthop. Rel. Res.,
232, 139, 1988.
Prendergast, P.J., Husikes, R. and Soballe, K., ESB research award 1996. Biophysical
stimuli on cells during tissue differentiation at implant interfaces, J. Biomech.,
30, 539, 1997.
Sawyer, P.N. et al., In vitro and in vivo evaluations of Dacron velour and knit pros-
theses, J. Biomed. Mater. Res., 13, 937, 1979.
Schoen, F.J., Interventional and Surgical Cardiovascular Pathology: Clinical Correlations
and Basic Principles, W.B. Saunders, Philadelphia, 1989, 36.
Spector, M., Reply to comments on “characteristics of tissue growth into Proplast
and porous polyethylene implants in bone,” J. Biomed. Mater. Res., 13, 991, 1979.
Spector, M., Harmon, S.L. and Kreutner, A., Characteristics of tissue growth into
proplast and porous polyethylene implants in bone, J. Biomed. Mater. Res., 13,
677, 1979.
Wesolowski, S.A. et al., Arterial prosthetic materials, Ann. N.Y. Acad. Sci., 146(1), 325,
1968.
Wolff, J., Das Gesetz der Transformation der Knochen, A. Hirschwald, Berlin, 1892.

Bibliography
Brighton, C.T., Black, J. and Pollack, S.R. (Eds.), Electrical Properties of Bone and Car-
tilage: Experimental Effects and Clinical Applications, Grune & Stratton, New York,
1979.
Burke, J.F. et al., Successful use of a physiologically acceptable artificial skin in the
treatment of extensive burn injury, Ann. Surg., 194, 413, 1981.
Chehroudi, B. and Brunette, D.M., in Encyclopedic Handbook of Biomaterials and Bioengi-
neering, Part A, Vol. 1, Wise, D.L. et al. (Eds.), Marcel Dekker, New York, 1995,
813.
Dadsetan, M. et al., Surface chemistry mediates adhesive structure, cytoskeletal or-
ganization, and fusion of macrophages, J. Biomed. Mater. Res., 71A, 439, 2004.
Draenert, K.D. et al., Strain adaptive remodeling in total joint replacement, Clin.
Orthop. Rel. Res., 430, 12, 2005.
Fitzgerald, R.H., Jr. (Ed.), Non-Cemented Total Hip Arthroplasty, Raven, New York, 1998.
Giavaresi, G. et al., Mechanical and histomorphometric evaluations of titanium im-
plants with different surface treatments inserted in sheep cortical bone, Bio-
materials, 24, 1583, 2003.
Homsy, C.A., Implant stabilization. Chemical and biomechanical considerations, Or-
thop. Clin. N. Am., 4, 295, 1973.
Huiskes, R. et al., A biomechanical regulatory model for periprosthetic fibrous-tissue
differentiation, J. Mater. Sci. Mater. Med., 8, 785, 1997.
202 Biological Performance of Materials: Fundamentals of Biocompatibility

Lane, J.M. (Ed.), Fracture Healing, Churchill Livingstone, New York, 1987.
Lossdörfer, S. et al., Microrough implant surface topographies increase osteogenesis
by reducing osteoclast formation and activity, J. Biomed. Mater. Res., 70A, 361,
2004.
Rubin, C.T. and Hausman, M.R., The cellular basis of Wolff’s law. Transduction of
physical stimuli to skeletal adaptation, Rhem. Clin. N. Am., 14, 503, 1988.
Smith, I.O., Baumann, M.J. and McCabe, L.R., Electrostatic interactions as a predictor
for osteoblast attachment to biomaterials, J. Biomed. Mater. Res., 70A, 436, 2004.
Søballe, K. et al., Tissue ingrowth into titanium and hydroxyapatite-coated implants
during stable and unstable mechanical conditions, J. Orthop. Res., 10, 285, 1992.
Thompson, D’A.W. On Growth and Form, Bonner, J.T. (Ed.), Cambridge University
Press, Cambridge, 1977.
Vogel, S., Life’s Devices, Princeton, NJ, Princeton University Press, 1988.
Wainwright, S.A. et al., Mechanical Design in Organisms, John Wiley & Sons, New
York, 1976, 348.
Woo, S. L.-Y. et al., Less rigid internal fixation plates: historical perspectives and new
concepts, J. Orthop. Res., 1, 431, 1984.
11
In Vitro Tissue Growth and Replantation

11.1 General Considerations

In 1992, I defined the field of biomaterials (Section 1.4) as having four historic
phases and, in turn, identified four classes or types of biomaterials:

• Type 1: inert
• Type 2: interactive
• Type 3: viable ([bio]hybrid)
• Type 4: replant

The current increasing interest in tissue engineering (TE) represents a shift

in focus from types 1 and 2 to types 3 and 4 biomaterials. The earlier
biomaterials, especially type 2, are not passé; on the contrary, they will
continue to be represented in devices used in the bulk of conventional clinical
applications, will be incorporated in some phase 3 devices, and will serve
in devices intended as bridges to replantation* of phase 4 “devices” (organs).
In some applications, due to problems of cost, supply, etc., types 1 and 2 will
continue to be the biomaterials of choice indefinitely. Research and devel-
opment of type 2 materials also continues very actively: in fact, the use of
such materials to produce adaptive affects in the host tissues (see Chapter
10) can be considered, in many cases, as in vivo precursors of tissue engi-
neering. Nevertheless, tissue engineering is a rapidly growing area of bio-
materials science and engineering (BSE). This chapter will begin to define
the field, explore its early progress, and foresee its near future.

* The surgical literature uses the term “replantation” to describe reattachment of severed body
parts (fingers, etc.). If the severed part is viable and circulation is restored, such procedures can
be highly successful. I retain the term (Section 1.4 and here) because, in each case, the inserted or
reattached portion is mature, autologous tissue.

203
204 Biological Performance of Materials: Fundamentals of Biocompatibility

11.2 What Is Tissue Engineering?

The clearest early definition of tissue engineering of which I am aware is:
“Tissue engineering is the application of principles and methods of engi-
neering and life sciences toward fundamental understanding of structure-
function relationships in normal and pathological mammalian tissues and
the development of biological substitutes to restore, maintain, or improve
tissue function” (Skalak et al. 1988). After providing this definition, Skalak
went on to remark that

The basic point of the above definition is that tissue engineering involves
the use of living cells.... The definition is intended to encompass proce-
dures in which the replacements may consist of cells in suspension, cells
implanted on a scaffold such as collagen and in cases in which the
replacement consists entirely of cells and their extracellular products.

Skalak was correct and prescient: the use of engineering principles and
techniques to elaborate and incorporate living cells into constructs for ther-
apeutic use distinguishes tissue engineering from more conventional biom-
aterials studies and from the concerns of other fields of physical and
biological science and medicine. However, his definition has not been gen-
erally accepted and there are many interpretations of the scope and breadth
of tissue engineering today.
The best global discussion of this term is provided by Vacanti and col-
leagues (2000):

In essence, new functional living tissue is fabricated using living cells

which are usually associated in one way or another to a matrix or scaf-
folding which can be natural, man-made, or a composite of both. The
living cells can migrate into the implant after implantation or can be
associated with the matrix in cell culture before implantation. Concep-
tually, the field (tissue engineering) differs from the field of cell trans-
plantation insofar as organized three-dimensional tissue is desired and
designed.

However, Vacanti et al. conflate types 2, 3, and 4 biomaterials as they have

been previously defined (Section 1.4):

• Type 2 consists of materials engineered in vitro to produce a desired

host response in vivo, such as ingrowth into a porous surfaced med-
ullary stem.
• Type 3 consists of composites of cells and matrix (natural, man-
made, or a combination) produced in vitro for implantation, such as
synthetic vascular grafts incorporating one or several cell types in a
resorbable matrix.
In Vitro Tissue Growth and Replantation 205

• Type 4 consists of cells, but more generally tissue and, eventually,

organs, grown and/or modified in vitro for replantation. Carticel™
(autologous hyaline cartilage cells cultured and replanted) is a prim-
itive type 4 biomaterial because, although it is fully viable, it lacks
the (differentiated) three-dimensional aspect referred to by Vacanti.

I will not attempt a final definition of “tissue engineering” here: as is the

case for any emerging field, the definition will mature as the field does.
However, any widely adopted, successful definition (one that can clearly
delimit TE and distinguish it from other pre-existing efforts) must contain
four elements:

• TE must involve engineering — that is, the utilization of fundamen-

tal physical, chemical, and electromagnetic laws and principles as
well as proven design and development methodology to produce
practical solutions to clinical problems. Thus, merely renaming
another research field “tissue engineering” will not contribute to
progress. A sad example of this is the parallel emerging field of
“genetic engineering,” which apparently involves neither an engi-
neering design component nor any appeal to basic physical laws or
principles.
• TE must involve engineering manipulation of live cells or tissue, not
merely the preparation of interactive (type 2) biomaterials. Thus,
there may be type 3 and 4 tissue-engineered materials, but type 1
or 2 materials cannot be reasonably said to be tissue engineered
because they affect, attract, or incorporate viable cells and tissues
only after being placed in vivo.
• The organic component of a tissue-engineered product must include
viable cells (or, at the very least, functional genes). The use of pro-
cessed natural products as type 1 and 2 biomaterials in medical
devices (such as gluteraldehyde-treated porcine tissue) is very well
established and, although it has considerable clinical utility, it cannot
be fairly said to constitute TE. Such products are probably better
grouped with other “biologics”: nonviable agents or materials of
essentially natural (rather than synthetic) origin.
• The production of a tissue-engineered product must also involve
some in vitro manipulation of viable tissue, cells, or organs. If this
is not the case, it will be very difficult to distinguish, as Vacanti et
al. (2000) try to do, between tissue engineering and transplantation
of cells, tissue, and organs; the latter is a purely medical, not engi-
neering, procedure.

Tissue engineering has a considerable overlap with the emerging field of

regenerative medicine. It is as yet unclear where one can draw the
line between the two; in fact, interdisciplinary academic programs already
206 Biological Performance of Materials: Fundamentals of Biocompatibility

combine the two. In general, one can distinguish tissue engineering as pri-
marily comprising efforts to replace damaged or absent tissue, using engi-
neering techniques as outlined earlier. Regenerative medicine currently
embodies more traditional therapeutic approaches, utilizing stem cells and
genetic transfection, to restore inadequate or lost function in situ, as in
treatment of diabetes, heart disease, spinal cord injury, and Parkinson’s
disease (National Resource Council 2002).

11.3 The Cell–Receptor Paradigm

11.3.1 Early Ideas
The advent of the “unit cell” concept in materials science had a revolutionary
effect on understanding of materials properties and their dependence on
structure. Thus, it was natural, as physical scientists and engineers moved
into interdisciplinary biological research, for them to seek a similar unifying,
simplifying paradigm. In early BSE studies, tissues were regarded as con-
tinuous, largely homogenous materials possessing only limited anisotropy
of structure and properties. Because most early characterization studies were
performed on dead tissue, cells were viewed as imperfections in structure,
rather like defects or grain boundaries in polycrystalline materials, and were
largely overlooked. It was understood that cellular function defined the
living state, but that was not a concern for engineers and the general, usually
unspoken, nonvitalist assumption was that living and dead materials had
the same physical properties. This has turned out to be untrue in the general
case (Black 1984).
As BSE became more sophisticated, the biological cell came to be seen as
an equivalent, in tissues, of the unit cell in engineering materials. However,
much of the bulk of tissues is made up of water and of various molecules,
collectively termed extracellular matrix. This matrix, as well as the small
partial volume contributed by the cell walls of dead cells, is, in fact, being
characterized when investigations of structure–property relations are per-
formed on tissues. Only in particular hypercellular structures, such as the
epiphyseal growth plates of long bones or in the contents of the circulatory
system, do the physical structure and properties of biological cells dominate
in mechanical property determinations.
The cell wall, more generally termed the plasma membrane, was originally
described as a phospholipid bilayer, with a hydrophobic core, possessing
general permeability and numerous pores, some passive but others with
active ion-pumping mechanisms. The cell content was regarded as an amor-
phous gel with specialized organelles, such as the nucleus and mitochondria,
simply floating in essentially random locations. This simple model has
proven inadequate to describe and explain the details of cell morphology
In Vitro Tissue Growth and Replantation 207

and function and, in particular, the association of cells with each other and
with their extracellular matrix, in all of the specialized profusion found in
mammalian as well as nonmammalian tissues. For instance, such a simple
model predicts that, through surface tension arguments, all cells should be
spherical in solution and sessile on surfaces. This is clearly not the case.
Fortunately, in parallel with the development of BSE, mammalian cell
biology and physiology was making great strides. A new unifying paradigm
has emerged: receptor–ligand binding. Tissues are now seen as associations
of cells and matrix, connected by specific and nonspecific receptor–ligand
binding and receptors crossing the plasma membrane, providing bidirec-
tional linkages capable of transducing information among the interior of the
cell, its nearest neighbors, and its environment. Furthermore, most cells are
now recognized to have well defined internal structures; active and passive
molecular scaffolding links many of the organelles. Thus, the recep-
tor–ligand-binding model in normal tissues provides the key to understand-
ing the elaboration of those tissues and therefore the association between
cells and synthetic matrices of types 1 through 3 biomaterials.
Tissues, as always, remain artifacts of the lives of cells. However, through
the receptor–ligand paradigm, how cells make tissue matrices and, in turn,
how these matrices influence the conduct of their lives can be understood.

11.3.2 The Membrane Receptor

In its most general form, a cellular receptor is a complex of two molecular
chains traversing the cell membrane. The receptors are grouped together
into types, such as integrins, selectins, cadherins, etc., based upon similarity
of structure and of function. Within each type, the individual molecular
chains are grouped together into families, depending upon general features
of structure — primarily, the possession of a common subunit (see following).
The integrin type is perhaps of primary importance to TE studies because
these receptors play key roles in cell–cell and cell–matrix association. Inte-
grins are heterodimers made up of one each from subunits or chains termed
“ α” and “β” (see Figure 11.1). There are large numbers of α- and β-chains
and thus very large numbers of possible different receptor structures and
functions. The α- and β-chains are distinguished by the use of numerical
subscripts: for example, the α6β1 is a common receptor specialized for
binding to the matrix protein laminin (Wei et al. 1997). However, all integrins
share the following characteristics:

• The two chains, as combined, reside in the cell membrane so that

they have three principal domains, or subsections:
• An intracellular domain that can link with intracellular molecules
and, in particular, can bind to the internal molecular skeleton of
the cell
• An intramembrane domain
208 Biological Performance of Materials: Fundamentals of Biocompatibility

Ligand Binding Site

β
α
EXTRACELLULAR
Cell
Membrane
INTRACELLULAR

FIGURE 11.1
Schematic arrangement of the integrin receptor.

• An extracellular domain that forms the binding site for extracel-

lular ligands, such as free molecules (cytokines, peptides, etc.) or
portions of the extracellular matrix
• The receptors are free to move along the surface of the membrane
and, under certain conditions, such as phagocytosis, can be inter-
nalized, recycled, and reappear on the cell surface.
• Intracellular and extracellular binding is reversible and, as such, the
bound state represents an equilibrium:


Receptor + Ligand 
kb
 
 Receptor-ligand complex
k d

Thus, the action of a receptor depends upon its population concentration

on the cell membrane, the concentration of the ligand, and the energetics of
binding and dissociation, as represented by kb and kd, respectively. Further-
more, divalent ions, such as Ca++, are also frequently involved in forming
or stabilizing receptor–ligand complexes. In addition, the population con-
centration on the cell membrane depends upon equilibrium between the
assembled receptor and the intracellular concentrations of the respective α-
and β-chains. Thus, receptor activity can be affected in many ways and
receptor–ligand association has the potential to exert well differentiated
effects on the cell.

11.3.3 Receptor–Matrix Interactions

The primary interest here is in the consequences of binding of receptors to
molecules in the extracellular matrix. Again, although this complex material
holds many molecules and the appearance and concentration of the spectrum
vary from tissue to tissue, certain motifs recur. Ligand binding by receptors
depends on recognizing short sequences of three or more amino acids. One
such sequence is (-arginine-glycine-aspartamine-) (abbreviated: RGD). The
In Vitro Tissue Growth and Replantation 209

Ligand

Binding
Dissociation

β
Receptor α

Cell
Membrane

Consequences
of extracellular
coupling with ligand binding: intracellular
membrane- signaling
associated
molecules
transmembrane
transport

FIGURE 11.2
Possible consequences of extracellular receptor–ligand binding.

α5β1 integrin binds to the RGD sequence, which is found in the very common
extracellular matrix molecule fibronectin. Thus, the formation of α5β1RGD
complexes plays a strong role in association between cells and the extracel-
lular matrix (MacDonald 1989). This observation has been utilized in the
fabrication of interactive substrates that will bind cells through recep-
tor–ligand association rather than through nonspecific chemical affinity
(Massia and Hubbell 1990).
Once receptor–ligand binding takes place, several consequences can occur
(Lauffenburger and Lindermann 1993) (Figure 11.2):

• The extracellular receptor–ligand complex formation may change

the nature and/or affinity of the intracellular domains for intracel-
lular species. Such intracellular binding, in turn, may alter the behav-
ior of the extracellular domain of the receptor.
• The complex may be internalized, as in phagocytosis, thus playing
a role in transmembrane transport of the ligand.
• Changes in the extracellular domain may “transduce” information,
via the intracellular molecular skeleton (not shown in Figure 11.2
for clarity) to organelles such as the nucleus. In the short term, this
may result in modification of cellular behavior (modulation), but in
the longer term, it may produce activation of DNA, resulting in
expression of new RNA and thus a possibly irreversible change in
cell function (differentiation).
210 Biological Performance of Materials: Fundamentals of Biocompatibility

The emerging understanding of the consequences of the receptor–ligand

paradigm is profound. In regard to BSE in general and TE in particular, the
following conclusions can be drawn:

• Cells are in constant communication with their external environ-

ment; therefore, maintenance and/or alteration of the pericellular
conditions are important considerations in understanding host
response to biomaterials.
• In their native habitat, cells depend upon signals received (trans-
duced) from their extracellular matrix in order to develop normally
and to maintain their appropriate phenotype. Thus, the design of
synthetic matrices for type 3 biomaterials must focus not merely on
avoiding cytotoxic or inhibitory responses but also on providing
appropriate information so as to encourage cells to maintain their
normal habit and function or stimulate specific behavior.
• The appropriate consideration of biomaterial–(local) host interac-
tions must now move from the so-called “tissue–material” interface
to the cell–material or, more properly, the receptor–ligand interaction
realm.

11.4 Matrices and Cell Sources

11.4.1 Cells*
Whether consisting entirely of cells and their extracellular products, like
more traditional tissue grafts and organ transplants, or merely incorporating
living cells, tissue replacements come in three types: auto-, allo-, and xeno,
reflecting the original cell source (respectively, autobiopsy, allobiopsy, and
xenobiopsy). The issue of xenohybrids (type 3) or xenoreplants** (type 4)
will not be considered here because they represent special cases of xenografts
and, except in the minds of animal rights advocates, do not raise any of the
objections that will be discussed later. However, the success in transplanting
human immune systems to small rodents such as mice suggests that, in the
future, the availability of “pseudo” xenotissue and organ sources may help
to resolve some of my present concerns. It has even been proposed to use
genetically transformed larger animals such as pigs to grow immunologi-
cally matched xeno-organs, which I would term xenoreplants (Fox 1997).
Of the two remaining tissue types, the autoreplant, or true replant, as
employed by Peterson (Brittberg et al. 1994) (see Section 11.5) in attempting
* Portions of this subsection, as well as Section 11.5, were published previously in different form
(Black 1997).
** Strictly speaking, such replants could only be into animals, such as seeing eye dogs, unless
transgenic hosts were used.
In Vitro Tissue Growth and Replantation 211

to repair articular cartilage, appears to raise the lesser concern. The avail-
ability of true cultured autografts would certainly definitively dispel some
of the generic problems with transplants: each replant would exactly match
the donor genetically (especially immunologically), it would introduce no
additional infectious agents (if carefully handled in vitro and during replan-
tation), and the supply would always match the demand. The downside
would be high cost, significant delay (which in the case of culturing a full
organ such as the kidney might be several years), and availability of support
technology. The commercial and legal issues are fairly straightforward: the
cost would reflect the processing (growth and handling) of tissue removed
from the ultimate user and then returned, as well as some amortization of
research and development costs.
However, the alloreplant seems to pose more problems. On the face of it,
the use of cells from a single, possibly fetal or newborn, human donor and
the elaboration of very large quantities of tissue appear attractive. This offers
the theoretical prospects of lower costs (through economy of scale), better
quality control, and a more manageable commercial manufacturing process.
Early efforts in this area feature, in some cases, a self-contained cassette or
cartridge: a support enclosure in which the cells can be maintained, nour-
ished, and kept sterile until delivered to the sterile field within the operating
room for implantation.
However, several areas of concern arise:

• The possible use of tissue from a single source — not for 15 to 25

recipients as is commonly the case for organ transplantation, but for
perhaps as many as 25,000 recipients — raises the prospect of real
clinical catastrophe if something goes wrong, such as inadvertent
transmission of an undetected (possible previously unrecognized)
slow virus. Unfortunately, one can only test for those viri and bac-
teria that have already been detected and even such tests have a
significant rate of false negative findings.
• The use of an allobiopsy rather than autobiopsy cell source probably
enables the future identification of the replant by karotyping, espe-
cially in the case of long lived tissue such as articular cartilage or
cultured functional organs. This has potential implications for legal
proceedings related to mal-outcome of the clinical procedure.
• In the U.S., the Uniform Anatomical Gift Act (1968) and the National
Organ Transplant Act (1984) essentially forbid the sale or other com-
mercial use of dead or live human body parts. The motivation for
these laws is very clear and straightforward: they were intended to
make human material available for teaching, research, and thera-
peutic purposes without creating a market in such materials. The
concern about market forces stems from three bases:
212 Biological Performance of Materials: Fundamentals of Biocompatibility

• A recognition that the human body is special; emboding the

moral sense that human beings are ends in themselves and
should not be used as means to achieve other purposes
• A desire to prevent the poor and powerless from selling parts of
their body, perhaps under duress
• A more general need to prevent the creation of an analog of the
autobody chop shop that converts stolen cars into far more valu-
able subassemblies and parts
In the case of postmortuum organ donation, the organ is donated
by the next of kin, frequently carrying out the desires of the
donor, and the charges to the recipient cover only “harvesting,”
testing, transport, and implantation (although donor funeral ex-
penses are also sometimes provided for). However, what about
the donation of cells or portions of organs, perhaps by surviving
donors, that could benefit thousands of recipients and are incor-
porated into “products” with significant value added by manu-
facturers? What would be sold then? Who should benefit finan-
cially? Can it be done legally and ethically?
• The need to obtain allogenic cells with pluripotentiality has turned
interest first to human adult and then to fetal and embryonic stem
cells. The former can be obtained with little or no injury to the donor,
but harvesting the latter invariably results in death of the fetus or
embryo. This latter situation has produced a wide range of national
responses related, in large part, to local attitudes towards early-life
issues such as abortion. In some countries, such cell sources may not
be used and, in others, there is wide latitude, under ethical super-
vision. In the U.S., a hybrid approach is followed, permitting fetal
stem cells to be used but limiting their sources narrowly to previ-
ously* established cell lines when public funding is involved.

Thus, one of the barriers to successful TE appears to be a need to under-

stand and reach agreement concerning implications of choosing among var-
ious cell sources. As I have suggested previously, some of these objections
may be overcome, potentially, by the use of transgenic xeno- (animal)
sources. In the near term, however, the issues raised by current practices
must be addressed and workable answers provided. Similar issues, particu-
larly of ownership and its transfer, have already arisen in the commercial
preparation of biological products from human cell sources. The litigation
record, although instructive, does not resolve these questions definitively.
Furthermore, even if it did, one would still be constrained to consider the
ethical aspects of the questions. I will return to these points again in Section
11.5.

* Before August 9, 2001 (speech, 11/11/01, President G.W. Bush).

In Vitro Tissue Growth and Replantation 213

11.4.2 Matrices
There are also less ethically challenging but more pragmatic problems with
the choice of matrix material, in the case of design and fabrication of a type
3 biomaterial.
Matrices may be natural in origin, such as collagen, keratin, or chitosan.
Natural matrices are, to a lesser or greater degree, inherently “biodegrad-
able,” depending upon the amount of molecular structure change during
processing, because mammalian systems possess a wide repertoire of deg-
radative enzymes for self- and foreign organic molecules. A further attraction
of natural sources is that a matrix familiar to the selected cells can be chosen;
thus, type II collagen can be combined with chondrocytes in an attempt to
fabricate a tissue-engineered hyaline cartilage. Collagen is perhaps the most
widely used of such natural materials (Silver and Pins 1992; Bell 1995),
although concerns still remain about immune response to some types (Lynn
et al. 2004).
Matrices may also be prepared by deliberately modifying natural materials
through physical, chemical, and enzymatic treatment. This is a practice of
great antiquity, used in attempts to process natural structures, such as dem-
ineralized bone or tendon segments, to reduce their antigenic properties and
improve their physical handling and postimplantation behavior. Many such
materials derived from allo- and xeno- sources are in clinical use as type 2
biomaterials.
However, both of these approaches, although still popular, seem too crude
today. A far more promising approach is the direct synthesis of polymers
with specific structures and properties, such as including receptor recogni-
tion sequences, as the previously noted (-RGD-). Capello (1992) and his
coworkers have been pioneers in this approach, using recombinant tech-
niques to select DNA sequences that code for specific molecular arrange-
ments and then transfecting these pseudogenes into bacteria such as
Escherichia coli. They then induce them to synthesize large quantities of the
desired material, which can then be spun into fibers, coated onto surfaces,
etc. Such materials can be designed to be nonresorbable, partially resorbable,
or fully resorbable, depending upon the demands of the application. One of
the first achievements of this technique was the synthesis of spider silk
modified to include receptor recognition domains (Anderson et al. 1994).

11.4.3 Combining Cells and Matrices

The issue of combining cells with matrices will not, in the long run, be a
problem as the technology of TE moves increasingly towards type 4 or
replant materials. In the meantime, considerable concern and argument will
involve which combinations of cell and matrix sources are the most appro-
priate. Although it is hard to generalize, I suggest that, at this moment, the
possible combinations can be grouped into three classes, in order of decreas-
ing preference (Figure 11.3):
214 Biological Performance of Materials: Fundamentals of Biocompatibility

CELL SOURCE

XENOBIOPSY
AUTOBIOPSY

ALLOBIOPSY
MATRICES

NATURAL 2

MAN MADE -
1
RESORBABLE

MAN MADE -
3
NONRESORBABLE

FIGURE 11.3
Possible combinations of matrices and cell sources (numbers identify classes; see text).

• Class 1: the most desirable from pragmatic and ethical points of view
appears to be the use of autobiopsy cells with partially or fully
resorbable, synthetic matrices.
• Class 2: this can be extended, with care and consideration, to include
natural matrices such as collagen and chitosan, perhaps with some
chemical modification, and allobiopsy cell sources, particularly
primitive stem cells. This class is the one now enjoying the most
attention from industrial developers.
• Class 3: finally, and likely to be least satisfactory in the long run, is
the use of xenobiopsy cell sources and the permanent incorporation
of nonresorbable matrix elements. In some cases, such as in early
studies using encapsulated xenobiotic cells for experimental human
therapy, these approaches are unavoidable. It may continue to prove
economically desirable in certain external applications, such as burn
treatment. However, I hope that this approach will fall into disuse,
due to permanent concerns about the possibility of unsought gene
transfer and presence of undetected infectious agents.

11.5 Thinking Twice about Tissue Engineering

The dividing line between traditional biomaterials (in particular, the search
for the Philosopher’s Stone), the truly inert (type 1) biomaterial, and the field
of tissue engineering was crossed on October 2, 1967 when Dr. Christiaan
Barnard transplanted a live human heart into a patient with heart failure
In Vitro Tissue Growth and Replantation 215

(Barnard 1967). Although his first patient succumbed to a variety of clinical

complications 18 days later, the possibility of routine transplantation of major
functioning organs drew worldwide attention. An immediate impact felt in
the U.S. was a radical curtailment of funds for the multicontractor federal
program directed towards the development of an implantable (permanent)
artificial heart. It is worth noting that this research and development program
never recovered from this setback; today, successor devices such as active
left ventricular assist devices (LVADs) are seen only as “bridges” to trans-
plantation: as devices to sustain a patient’s life until a suitable donor organ
becomes available. In the same way, blood and peritoneal dialysis, which
were originally viewed as miraculous but last-stage interventions in kidney
disease and failure, are now widely employed for prolonged periods for
patients awaiting definitive treatment by allotransplantation.
The success of organ transplantation from cadavers or, in the case of
kidneys, from related or unrelated live donors cannot be denied. Perhaps as
many as 340,000* U.S. patients have received transplants of living organs,
including kidneys (the most commonly transplanted), heart, lung, liver, pan-
creas, and intestinal segments. Many patients receive multiple organ trans-
plants, such as ex-governor of Pennsylvania Robert Casey, who received a
combined heart and liver transplant in 1993 (Alexander and Baker 1993).
Casey’s situation illustrates some popular concerns about organ donation
and transplantation:

• Organs for transplantation are rare in comparison to need and wait-

ing times may be long. In 1995, it was estimated that 38,000 U.S.
patients were on waiting lists as possible organ recipients; a new
name was added very 30 minutes and at least 8 patients per day
died while awaiting transplants. Nevertheless, a match was fortu-
itously found for Casey within 1 day of the decision to perform his
transplant. Today, some 88,500 people await transplants.
• The donor involved was young and relatively disenfranchised: a 34
year old murder victim. This had been the situation in Dr. Barnard’s
initial heart transplantation; the donor was a young man killed in a
vehicular accident (Barnard 1967).
• The recipient was older and, to a considerable degree, privileged.
This issue was also highlighted by Barnard’s first transplant: in a
country still in the coils of apartheid, the donor was black and the
recipient was a retired white civil servant. This issue of status priv-
ilege was also raised when the legendary baseball star Mickey Man-
tle received an unsuccessful liver transplant (Meyerson 1995).

Without question, organ transplantation is difficult and expensive. It is

extremely difficult to screen donors for transmissible disease and accurately
match donor organs to recipients so that the best possible blood and tissue
* http://www.optn.org/latestData/rptData.asp.
216 Biological Performance of Materials: Fundamentals of Biocompatibility

compatibility, as well as appropriate size, can be obtained. It must be done

in a very short time span because organs generally become available on short
notice and deteriorate rapidly with time, even when the donor is maintained
on life support. Thus, statewide and regional networks and a national net-
work, the United Network for Organ Sharing (UNOS),* have sprung up in
the U.S. in response to the National Organ Transplant Act (1986) to facilitate
optimal usage of organs. As a result, it is now possible for organs and tissues
from a single donor to be given to many geographically widely separated
recipients. Extensive publicity drives have been conducted to encourage
individuals, especially people obtaining driver’s licenses, to register as pro-
spective organ donors. However, the fundamental problems remain: demand
far exceeds supply and costs of the procedures involved tend pragmatically
to reduce access by less well off and disenfranchised patients; as a result,
nagging questions of social justice remain.
It is fair to state that these moral, ethical, and pragmatic considerations
have also played strong roles in the development of TE so far. Early efforts
in TE, although surrounded by high enthusiasm and attracting large quan-
tities of venture capital, were relatively noncontroversial. These included,
primarily in animal experimental studies and, in some cases, early human
clinical trials, the encapsulation of Langerhans’ islet cells for the treatment
of diabetes (Scharp et al. 1994); the separation, culturing, and reinfusion of
specific lymphocyte subpopulations for cancer therapy (Chen 2000); and the
fabrication of a variety of hybrid skin replacements for use in burn treatment
(Yannas and Burke 1980; Yannas 1997). In each case, the goal was transient
therapy rather than definitive replacement, the conditions addressed were
severe and life threatening, and an easy distinction could be maintained
between “implant” and recipient.
However, in 1994, another phase began. Dr. Peterson and coworkers from
the University of Göteborg, Sweden (Brittberg et al. 1994), reported early
clinical trials of growth and replantation of articular chondrocytes in a pro-
cedure to replace full thickness focal defects in the tibial plateau. Autologous
cells were removed from non-load-bearing areas of articular cartilage, sep-
arated from their matrix, cultured in vitro for 14 to 21 days, and then
replanted in articular cartilage defects in the tibial plateaus of each donor;
autologous periosteal membrane was used to retain the replant. In this
application, the therapy is expected to be definitive (in the same way that
implanting a metal and plastic total joint “replacement” is), the disease
treated is not life threatening, and the replant is expected to become fully
integrated with the patient’s tissues.
This technique was introduced into the U.S. on an experimental basis and
received FDA approval for some medical indications in August 1997. The
“product,” as commercialized by Genzyme Tissue Repair (GTR) (a division
of Genzyme Inc.) and now distributed by U.S. Biosciences, is called Carticel™
(autologous replant chrondrocytes). The technique of Brittberg et al. (1994)

* http://www.unos.org.
In Vitro Tissue Growth and Replantation 217

is generally followed, with a 3- to 6-week interval between autobiopsy and

replantation. The culture process currently costs (nominally) $19,750 of an
estimated average total treatment cost, including surgeon’s fees, of $40,000.*
The general technique, now termed articular cartilage implantation or ACI,
has been essentially duplicated by several other companies and is in use in
a number of countries.
This development requires taking a second look at the entire concept of
TE. Before going too far down this path, it is worthwhile to think twice —
to consider the implications and ask whether some things should not be
done even when they appear to be possible. Section 11.4 discussed some
of the trade-offs necessary between benefit and risk in the selection of
matrices and cell sources for type 3 and 4 biomaterials. It is still too early
to do more than muse about which of the risks will emerge as true clinical
and legal problems and, as a result, which of the conflicting viewpoints
among TE researchers and developers will come to dominate clinical prac-
tice in replantation.
I would like to note a more fundamental concern about TE. This concern
was raised a half a century ago by Jack Vance (1956) in a novel entitled To
Live Forever. Vance imagined a city society, Clarges, on a remote world in the
last stages of societal decay. On one hand, its citizens possess an obsession
about immortality, but, on the other, they submit to an agreement for State
limitation of life span through the use of public assassins, with postponement
of “termination” based upon a continual measurement of one’s individual
contribution to the public good. However, as one would suspect, in Clarges
some people are more equal than others and society, in Vance’s account, has
evolved into five social classes: Brood, Wedge, Third, Verge, and Amaranth.
It is the privileged Amaranth, the social and economic elite, who have solved
the problem of immortality. Those few judged to have achieved the most
and contributed the most to Clarges’ society are admitted into the Amaranth
class and undergo the following procedure:

...Five cells were extracted from [the] body. After such modification of
genes as might be desired, they were immersed in a solution of nutrients,
hormones and various special stimulants, where they rapidly evolved
through the stages of embryo, infant, child and adolescent.... When in-
vested with the prototype’s memory-bank, they became the identity of
the original: full-fledged surrogates (Vance 1965).

Amaranths zealously keep their memory-bank recordings up to date and

their surrogates are carefully guarded against the day that the original (the
prototype) might have a fatal accident, develop an incurable disease, or be
irrecoverably injured or killed by violence.
Vance’s novel centers on a problematic situation: one of the surrogates
escapes and tries to lead a life independently of its prototype. Issues con-
cerning the meaning of self, the value of life, involuntary servitude, and
* http://www.firstdatabank.com; Dr. T. Minas (personal communication).
218 Biological Performance of Materials: Fundamentals of Biocompatibility

manipulation of the fundamental elements of human existence are raised.

In this prescient novel, the logical, although perhaps impossible, conse-
quence of the promise of Dolly, the first successful report of cloning a sheep
by nuclear transfer (Campbell et al. 1996), can be recognized. A dark reflec-
tion of the “baby factories” of an earlier visionary novel, Brave New World
(Huxley 1932), in which all fetal development occurs in vitro (rather than in
utero) and chemical manipulation is employed to fit each individual to his
or her predestined place in society can also be seen. Rather than rejecting
such disturbing visions out of hand as incredible, one should be aware of
the widespread current efforts in therapeutic gene implantation and of anec-
dotal rumors of conception of babies to serve as marrow donors for older
siblings.
The question that ought to be raised by Vance, Huxley, and visionaries is
essentially, “When is enough, enough?” That is to say, where is the line
between legitimate therapeutic intervention and a morbid preoccupation
with life enhancement at all costs, and who shall judge its location? This is
not a problem unique to the fields of BSE and TE. It can be recognized in
the use of life support for the terminally ill, in radical interventional surgery,
including multiple organ transplants in the elderly, in fetal (in utero) surgery,
and so forth.
In his work, Enough: Staying Human in an Engineered Age (2003), McKibben
wrote:

…[T]hese new technologies show us that human meaning dangles by a

far thinner thread than we thought. What if the ending to our story has
already been written, our compass set? What if we have been pro-
grammed, or at least must suspect each time we choose a path that we
have been nudged in that direction by our engineered cells. Who then
are we?

One can properly object that my suggesting connections between TE and,

on one hand, the fictional visions of Huxley and Vance and, on the other,
biotechnological research into gene transplantation and mammalian cloning
is extreme and far fetched. In principle, I agree. However, this is a situation
in which one can only see the extreme ends of a spectrum: at one end are
research and medical procedures, which are simple and raise no ethical
issues; at the other are dark dreams of clearly morally objectionable acts.
The reason for creating this spectrum, or linkage, is to raise the questions:
how will the boundary between the acceptable and the unacceptable be
recognized? What are the moral mileposts that should warn one to slow
down and, perhaps, to stop when, in McKibben’s view, enough has been
achieved?
Consider an illustrative example of such a spectrum, one relating to kidney
transplantation. The implantation of cadaveric kidneys can be placed at one
end; at the other, assume Vance’s hypothesis but suggest only that the cloned
human surrogate be maintained as an “organ bank” for the prototype.
In Vitro Tissue Growth and Replantation 219

Clearly, the former situation is found to be acceptable and the latter rejected
as unthinkable.*
Now consider some possible intermediate points on this spectrum of pos-
sible organ sources, in no particular order:

• Live organ transplantation:

• From a parent
• From a child
• From a sibling
• From a sibling expressly conceived for the purpose
• From a late term aborted fetus
• From a transgenic anima
• From a paid donor
• From an executed criminal
• From a condemned criminal
• From a criminal seeking reduction or remission of sentence
• Implantation of artificial constructs (hybrid artificial organs) with
cells from one or more of such sources or from a cloned surrogate

The potential use of each of these organ or cell sources raises interesting,
different, and, in some cases, morally complex problems. Their rank order
of moral acceptability is not obvious and a marker or threshold of unaccept-
ability is not clear.
Traditionally, such questions have been referred to philosophers, ethicists,
and religious leaders while scientists, engineers, and physicians have gone
on with their studies and clinical treatment. I have no simple answers to the
questions raised here. However, I am sufficiently disturbed by long-term
implications of TE to suggest that engineers and scientists need to open a
broad and deep dialogue with a cross-section of interested parties rather
than simply going ahead with a narrow focus on possible technological
solutions to human problems. The road to hell is, as Ambrose Bierce com-
mented, paved with good intentions (1906).
Finally, I want to suggest a simple functional test for engineers and scien-
tists in these (or, for that matter, other related) fields of investigation. Even
if one can personally resolve the issues raised, as one must, it is worthwhile
considering this question: will this work ultimately benefit refugees in
Rwanda** or in the other parts of the world mired in desperate, apparently
chronic, poverty and deprivation?

* If one has any doubt of this statement, simply replace “kidney” with “heart” or “pancreas” in
the first sentence of this paragraph.
** As this is written (in 2005), more than three-fourths of a million refugees still are unable or
afraid to return home, despite the best efforts of international organizations, more than a decade
after the massive intertribal massacres that occurred in Rwanda.
220 Biological Performance of Materials: Fundamentals of Biocompatibility

I have come to call this question the “Rwanda test.” I offer it to readers,
experienced professionals, and students alike — not with the suggestion that
a negative response is a necessary “show stopper,” but rather with the hope
that, among all the good and promising questions that can be asked, increas-
ingly often those that have the potential to bring the greatest good to the
most with the least risk of catastrophe will be selected. To fail to apply such
a test to future investigations runs the risk of losing oneself and one’s society.

11.6 Some Final Comments

Tissue engineering, with its hints of Frankenstein’s monster and the seduc-
tion of dealing with the very substance of life — living tissue — is enor-
mously alluring to beginning and long-time investigators in BSE, as well as
in many other more traditional fields. Beyond the ethical concerns raised in
the previous section, I want to close this discussion with a few practical
concerns.
Without question, TE will continue to grow and flourish, although many
of the early extravagant visions of the pioneers may be long delayed or prove
unattainable. In the meantime, it is important not to lose sight of these points:

• Like the larger field of BSE, TE has three primary aspects: it is a

materials science, a biological science, and a clinical science. Thus,
TE cannot be expected to be successful unless it is moved forward
by integrated multidisciplinary teams of physical and biological sci-
entists, engineers, and clinicians and physicians. It would be easy
for newer TE workers (such as materials specialists beginning to
work in TE) to become second-rate biochemists, biophysicists, cell
biologists, etc. This is a bad idea; it is far better to leave these more
fundamental fields to their traditional practitioners, harvest their
intellectual product, and focus on being broad integrators and inno-
vators, as workers in BSE have always been.
• Interdisciplinary or innovative studies are no excuse for mediocrity.
TE should not be allowed to become a refuge for workers who cannot
compete successfully elsewhere in BSE or in its foundation fields of
biological and physical sciences and engineering. On the contrary,
difficult interdisciplinary fields such as TE have little room for the
otherwise blameless but merely average investigator.
• Nothing is really new: many challenging problems faced in TE have
been investigated in other contexts and, in many cases, practical
solutions already exist. The emergence of online publications, new
journals, and fledgling scientific societies are welcome as signs of
intellectual vigor; however, abandoning the traditional literature and
In Vitro Tissue Growth and Replantation 221

forums of biological and physical science and engineering and, par-

ticularly, of BSE will only handicap students and advanced workers
alike in TE.
• TE products will continue to need a supporting superstructure of
types 1 and 2 biomaterials for their fabrication, evaluation, preser-
vation, handling, implantation, and clinical monitoring. Care must
be taken that older, successful biomaterials technologies are not
neglected in the haste to move ahead. Otherwise, future workers
will repeat old mistakes and be faced with continually “reinventing
the wheel.”
• Clinical success of TE products will require a much more sophisti-
cated understanding of the physical property differences between
normal and diseased tissues than is currently available. I have long
argued this point in the broader field of BSE, saying that, in partic-
ular, tissue mechanics is an essential prerequisite to and should be
a component of any research effort directed towards support or
replacement of natural tissues or organs with structural attributes.
With the emergence of TE, I hope for an accompanying revival of
tissue mechanics and morphometrics, with an increased focus on
changes associated with medical and surgical procedures, disease,
age, and interaction with implants.
• In any newer aspect of research such as TE, traditional methods and
procedures continue to have a place. Vision is no excuse for aban-
donment of rigorous process: good science will always require the
sequence of conception of new ideas; hypothesis formulation; careful
observation; replication of experiments; critical, skeptical statistical
analysis of outcomes; and testing and retesting of earlier hypotheses.
Good engineering practice requires clear statements of problems;
examination, development, and perfection of alternate approaches;
and exhaustive testing* and verification of numerical, process, and
hardware solutions. With regard to outcomes in science or engineer-
ing, if something seems too good to be true, it probably is, whether
it is a laboratory result or a vacation travel offer.

Nevertheless, I have significant enthusiasm for TE; I only hope that the field
and its workers are able to fulfill its great promise in responsible and ethical
ways.
* Many years ago, to my amusement and dismay, I heard a podium presentation about an
implant study in which two dogs had been used to examine local host response to a wide variety
of candidate biomaterials. When I questioned the presenter, I suffered a lapsus lingua and asked
why two dogs were used (rather than why only two were used!). The answer, after much reflec-
tion, was, “Well, I suppose I could have used just one.” Since then, at each meeting I have
attended, I have mentally bestowed the “one dog” award to the presentation containing the
greatest amount of data gained from the smallest number of test subjects, animal or human.
Regretfully, there has been a continuing need to make this award at TE as well as BSE meetings.
(See also Chapter 18, Appendix 1.)
222 Biological Performance of Materials: Fundamentals of Biocompatibility

References
Alexander, K.L. and Baker, S., Governor Casey’s timely transplant, Bus. Week, June
28, 1993.
Anderson, J.P. et al., Structural evolution of genetically engineered silk like protein
polymers, in Silk Polymers: Materials Science and Biotechnology (ACS Symposium
Series #544). Kaplan, D., Adams, W.W., Farmer, B. and Viney, C. (Eds.), Amer-
ican Chemical Society, Washington, D.C., 1994, p. 137.
Barnard, C.N., The operation. A human cardiac transplant: an interim report of a
successful operation performed at Groote Schuur Hospital, Cape Town, S. Afr.
Med. J., 41(48), 1271, 1967.
Bell, E., Strategy for the selection of scaffolds for tissue engineering, Tissue Eng., 1(2),
163, 1995.
Black, J., Tissue properties: relationship of in vitro studies to in vivo behavior, in
Natural and Living Biomaterials, Hastings, G.W. and Ducheyne, P. (Eds.), CRC
Press, Boca Raton, FL, 1984, 5.
Black, J., Thinking twice about tissue engineering, IEEE Eng. Biol. Med., 16(4), 102,
1997.
Brittberg, M. et al., Treatment of deep cartilage defects in the knee with autologous
chondrocyte transplantation, New Engl. J. Med., 331(14), 889, 1994.
Campbell, K.H. et al., Implications of cloning, Nature, 6573, 383, 1996.
Capello, J., Genetic production of synthetic protein polymers, Mater. Res. Soc. Bull.,
17, 48, 1992.
Chen, U., Lymphoid cells, in Lanza, R.P., Langer, R. and Chick, W.L. (Eds.), Principles
of Tissue Engineering, 2nd ed., Elsevier Science, New York, 2000, 611.
Fox, M., Sheep-cloners PPL to breed for transplants, Reuters News Service, 3/24/97,
1997.
Huxley, A.L., Brave New World, Doubleday, Doran & Co., Garden City, NJ, 1932.
Lauffenburger, D.A. and Linderman, J.J., Receptors: Models for Binding, Trafficking, and
Signaling, Oxford University Press, New York, 1993.
Lynn, A.K. et al., Antigenicity and immunogenicity of collagen, J. Biomed. Mater. Res.,
Part B: Appl. Biomater., 71B, 343, 2004.
MacDonald, J.A., Receptors for extracellular matrix components, J. Physiol., 257, L331,
1989.
Massia, S.P. and Hubbell, J.A., Covalent surface immobilization of Arg-Gly-Asp- and
Tyr-Ile-Gly-Ser-Arg-containing peptides to obtain well-defined cell-adhesive
substrates, Anal. Biochem., 187(2), 292, 1990.
McKibben, B., Enough: Staying Human in an Engineered Age, Henry Holt, New York,
2003, 65.
Meyerson, A.R., Final stats: Mantle’s last medical bills, NY Times, 8/20/95.
National Organ Transplant Act, Public Law No: 98-507, 1984.
National Research Council, Stem Cells and the Future of Regenerative Medicine, National
Academy Press: Washington, D.C., 2002, 4.
Scharp, D.W. et al., Protection of encapsulated human islets implanted without im-
munosuppression in patients with type I or type II diabetes and in nondiabetic
control subjects, Diabetes, 43, 1167, 1994.
In Vitro Tissue Growth and Replantation 223

Silver, F.H. and Pins, G., Cell growth on collagen: a review of tissue engineering
using scaffolds containing extracellular matrix, J. Long-Term Eff. Med. Impl., 2(1),
67, 1992.
Skalak, R. et al., Preface, in Skalak, R. and Fox, C.F. (Eds), Tissue Engineering, Alan R.
Liss, Inc., New York, 1988, 1.
Uniform Anatomical Gift Act (1968, as amended 1987), National Conference of Com-
missioners on Uniform State Laws, C. 1:12A-8 (subsequently adopted by indi-
vidual states (U.S.)).
Vacanti, J.P. and Vacanti, C.A., History and scope of tissue engineering, in Lanza,
R.P., Langer, R. and Chick, W.L. (Eds.), Principles of Tissue Engineering, 2nd ed.,
Elsevier Science, New York, 2000, 3.
Vance, J., To Live Forever, Ballantine Books, New York, 1965.
Wei, J. et al., Integrin signaling in leukocytes: lessons from the α6β1 integrin, J.
Leukocyte Biol., 61, 397, 1997.
Yannas, I.V. and Burke, J.F., Design of an artificial skin. I. Basic design principles, J.
Biomed. Mater. Res., 14, 65, 1980.
Yannas, I.V., In vivo synthesis of tissues and organs, in Lanza, R.P., Langer, R. and
Chick, W.L. (Eds.), Principles of Tissue Engineering, 1st ed., Elsevier Science, New
York, 1997, 167.

Bibliography
Atala, A. and Lanza, R.P., Methods of Tissue Engineering, Academic Press, New York,
2001.
Beauchamp, T.L. and Childress, J.F., Principles of Biomedical Ethics, 5th ed., Oxford
University Press, New York, 2001.
Bierce, A., The Cynic’s Word Book, 1906. (Reprinted as: The Devil’s Dictionary, Castle
Books: New York, 1967).
Brodt, P. (Ed.), Cell Adhesion and Invasion in Cancer Metastasis, Chapman & Hall,
London, 1996.
Germain, L. et al., Engineering human tissues for in vivo applications, Ann. N.Y. Acad.
Sci., 961, 268, 2002.
Gold, E.R., Body Parts: Property Rights and the Ownership of Human Biological Materials,
Georgetown University Press, Washington, D.C., 1996.
Fox, R.C. and Swazey, J.P., Spare Parts. Organ Replacement in American Society, Oxford
University Press (Acadia Institute), New York, 1992.
Hardie, D.G., Biochemical Messengers: Hormones, Neurotransmitters, and Growth Factors,
Chapman & Hall, London, 1991.
Katz, B.Z. and Yamada, K.M., Integrins in morphogenesis and signaling, Biochimie,
79(8), 467, 1997.
Langer, R. and Vacanti, J.P., Tissue engineering, Science, 260, 920, 1993.
Lanza, R.P., Langer, R. and Chick, W.L. (Eds.), Principles of Tissue Engineering, 2nd
ed., Elsevier Science, New York, 2000.
Lysaght, M.J. and Hazelhurst, A.L., Tissue engineering: The end of the beginning,
Tissue Eng., 10(1/2), 309, 2004.
Palsson, B., Hubbell, J.A., Plonsey, R. and Bronzino, J.D. (Eds.), Tissue Engineering,
CRC Press, Boca Raton, 2003.
224 Biological Performance of Materials: Fundamentals of Biocompatibility

Ratner, B.D. and Bryant, S.J., Biomaterials: Where we have been and where we are
going, Annu. Rev. Biomed. Eng., 6, 41, 2004.
Schwartz, M.A. et al., Integrins: emerging paradigms of signal transduction, Annu.
Rev. Cell Dev. Biol., 11, 549, 1995.
Shelly, M., Frankenstein, 1818 (reprinted by J.M. Dent, Everyman’s Library, London,
1994.)
Skalak, R. and Fox, C.F. (Eds.), Tissue Engineering, proceedings for a workshop held
at Granlibakken, Lake Tahoe, California, February 26–29, 1988, Alan R. Liss,
Inc., New York, 1988.
Younger, S.J. et al., (Eds.), Organ Transplantation, University of Wisconsin Press, Mad-
ison, 1996.
12
Allergic Foreign Body Response

12.1 Specific vs. Nonspecific Response

In earlier consideration of the inflammatory response (Chapter 8), the actions
of neutrophils and macrophages in response to a foreign material are
described as a nonspecific defense mechanism. That is, their response is
universal in nature and only slightly affected by the structure and chemical
composition of the foreign material. This chapter will discuss a second type
of response to foreign materials, the specific or immune response.
The aspects of specific response to foreign or nonself materials are grouped
together and collectively ascribed to a system of cells and mediating agents,
collectively termed the immune system. Immunity is usually understood as
the property of being secure or nonsusceptible to the adverse effects of a
particular bacterium or foreign material. Conversely, allergy is the property
of being especially sensitive (or hypersensitive) to such agents. Figure 12.1
shows the overall system and its general features in mammals.
The immune system is configured or adapts to distinguish between self
(things that are part of the natural, intact physiological system) and nonself
(all other things). It normally ignores all aspects of self; however, if it mistakes
self for nonself, adverse reactions, collectively termed autoimmunity, may
occur. Introduction of foreign tissue, also recognized as nonself, produces
an inflammatory response termed rejection. Resistance may be conferred by
genetic inheritance of a “memory” for certain nonself materials such as
proteins in bacterial cell walls, thus producing a natural resistance to certain
infections. Perhaps the most important aspect of the system is its ability to
adapt by developing a specific memory for particular foreign materials. This
may be produced by deliberate exposure to a partial or attenuated organism
or material under nonpathogenic conditions (vaccination) or by prior expo-
sure under sensitizing conditions (high dosage, physical stress, presence of
an adjuvant material, etc.). The result of this specific memory may be desir-
able, as in affording acquired (vs. natural) resistance to infection, or unde-
sirable, as in producing a form of adverse reaction to antigens or implants,
termed hypersensitivity.

225
226 Biological Performance of Materials: Fundamentals of Biocompatibility

Desirable Consequences of Immunity

Natural Resistance Acquired Resistance

LF lantation
NONSE on/reimp
in a tion Reinfecti
a c c tion
V Implanta
Infection
Gra g
ft in
E
NS
S PO
RE
UNE SPECIFIC
M MEMORY
IM
E
TIV
SELF
DAP
A

Autoimmunity Rejection Hypersensitivity

Undesirable Consequences of Immunity

FIGURE 12.1
The immune system. (Adapted from Playfair, J.H., Immunology at a Glance, 3rd ed., Blackwell
Scientific Publications, Oxford, 1979.)

12.2 Mechanisms of Immune Response

12.2.1 Antigen–Antibody Complex Formation
Specific or immune responses depend upon the exact details of the chemical
composition and conformation or structure (see Section 5.2) of the foreign
material. This class of response is directed primarily towards recognition of
foreign proteins such as toxins, viri, and bacterial cell wall components. The
response to these foreign materials takes place through two mechanisms:
humoral and cellular (also termed cell-mediated) response. In each case, the
foreign body is referred to as an antigen and the body responds by producing
an antibody. Antibody is a generic name for a macromolecular complex formed
by the association of large immunoglobulins present in serum into a Y-shaped
molecule with a molecular weight exceeding 1 × 106. The stem of the Y is
essentially the same for all antibodies, permitting it to bind to cell surfaces;
regions in the arms are variable in structure, producing the specificity, or ability
to bind with a particular antigen, characteristic of an antibody.
The humoral mechanism is based upon the production of freely circulating
antibodies. These antibodies are designed to unite with the foreign material
and “denature” or neutralize it. That is, the antibody–antigen complex lacks
the undesirable or destructive activity of the free antigen. The complex may
accumulate locally in tissue, be carried to lymph nodes by phagocytes, or
Allergic Foreign Body Response 227

be more easily catabolized than the free antigen. Circulating antibody pro-
duction is mediated by a class of lymphocytes called B-cells that arise from
primitive mesenchymal cells in the bone marrow. Antibody production is
initiated by introduction of an antigen consisting in part or whole of a foreign
(nonself) material. Whether biological, organic, or inorganic in constitution,
such material may be able to act as an antigen; it may combine with native
proteins, especially complement fragments of C3 and C5 (see Section 12.2.2),
to form an antigen; or it may undergo metabolic processing by a phagocytic
cell, usually a macrophage, to become an antigen.
Once antibodies are produced and released into the blood stream, they
may persist for long periods of time. This is the principle of immunization
to bacterial and viral infection: administering a small provocative dose of a
specific antigen or one that produces antibodies specific to the infectious
organism of interest. Then, when a later challenge is encountered, immediate
antibody–antigen complexing occurs without the delay necessary for new
antibody production. The antibody production response is quite variable,
depending upon the initiating agent, its concentration, the general state of
health of the immune system, and the presence of sensitizing agents such
as corticosteroids. Finally, a subpopulation of B-lymphocytes, memory B-
cells, are then capable of rapidly synthesizing additional amounts of the
specific antibody (which it “remembers”) upon stimulus by a later exposure
to the same foreign body.
Cell-mediated response depends upon the action of another class of lym-
phocytes, the T-cells. These cells also arise in bone marrow, but pass through
the thymus gland where they undergo a conversion that improves their
ability to differentiate. They then collect in lymph nodes and associated
tissue. Their activity continues to be affected by a hormone secreted by
epithelial cells in the thymus gland. The T-cells cannot be distinguished from
B-cells morphologically until challenged by an antigen. As in the case of
humoral response, a foreign material or one of its degradation products may
be able to act as an antigen; it may combine with native proteins, especially
complement molecule C5 (which may be activated to C5a [see Shepard et
al. 1984]), to form an antigen or it may undergo metabolic processing by a
phagocytic cell, usually a macrophage, to become an antigen.
T-cells may then be distinguished by morphological changes that include
the appearance of many polyribosomes and scant rough endoplasmic retic-
ulum. They produce and store a different class of antibodies, primarily
bound to the cell membrane surface. These antibodies are highly specific
and cannot survive with appreciable activity outside or separate from the
T-cell. T-cell antibodies act against intracellular infections, cancer, and foreign
materials of nonbiological origin. However, T-cells must be present and
aggregate in the region of antigen concentration to be effective because they
must externalize and directly present the antibody to the antigen.
In addition to direct neutralization of undesired effects of foreign materials,
the formation of antibody–antigen complexes acts to enhance inflammatory
response. This can happen directly through a nonspecific phagocytic
228 Biological Performance of Materials: Fundamentals of Biocompatibility

response because the complexes can grow to microscopic size by accretion

of additional antibody and antigen molecules, or it can occur indirectly
through complement activation (Chenoweth 1986; Tang et al. 1998).
Because particulate foreign materials such as wear debris from a joint
replacement (Urban et al. 1994, 2004) may be distributed widely in the host
and lymphocytes and circulating antibodies are also widespread, local
response to antibody–antigen complexing may occur anywhere. This results
in the wide variety of physiological effects popularly termed “allergic”:
swelling of the membranes in the respiratory system (hay fever, asthma,
etc.), rash or reddening (arthus, etc.), hives (swelling related to local kinin
activation), and so forth.

12.2.2 Complement System Activation

The formation of antigen–antibody complexes will interfere with the chem-
ical function of the antigen. However, if the antigen is a surface feature,
perhaps a receptor, on the surface of an invading bacteria, then the mere
formation of such a complex may not be enough to prevent the bacteria from
multiplying, producing toxins, etc. Thus, a humoral system of immune
defense exists that has been termed the complement system because it sup-
ports and extends the immune system. The complement system consists of
approximately 40 proteins of various molecular weights found in serum and
interstitial fluid. When activated, its main function is to produce a product
(the membrane attack complex [MAC]) that renders bacterial cell wall mem-
branes porous, leading to cell death. Thus, the MAC is directly bactericidal.
The principal molecular elements of the complement system are labeled
C1 though C9 and, in a motif similar to the coagulation cascade (see Figure
9.1), they participate in a cleavage and amplification process leading to the
formation of MACs. There are two primary pathways: the classical pathway
is initiated by the formation of antigen–antibody complexes and thus can
become active as a concomitant of humoral or cellular immune responses;
an alternate pathway can be triggered in the absence of a specific immune
response by certain foreign molecules, such as repeating sugars or proteins
(Figure 12.2).
The role of complement activation in immune response to implants is
controversial, but has been recognized by some researchers (Chenoweth
1987; Tegnander et al. 1994). It is clearly implicated in the case of T- or B-
cell activation. However, the types of materials that may be able to activate
the alternate pathway are still not well recognized. Complement activation
has several consequences:

• The MAC can attack native cells and foreign cells (as in transplants),
as well as bacteria, causing undesirable cellular necrosis.
Allergic Foreign Body Response 229

Antigen-Antibody
Complex
C1, C4, C2 C3 convertase
C3 C3a, C3b
Classical
Pathway

C5 convertase*
C5 C5a, C5b
C5 convertase*

Membrane
C6, C7, C8, C9 attack
complex
Surface
Contact
D, B, C3 C3 convertase
C3 C3a, C3b
Alternative
Pathway

*different structures; same enzymatic activity

FIGURE 12.2
Classical and alternative pathways for complement activation. (Adapted from Elgert, K.D.,
Immunology. Understanding the Immune System, Wiley–Liss, New York, 1996.)

• In sufficient concentration, several of the activated intermediates,

such as C5a, are capable of producing an inflammatory process (see
Section 8.2).
• Some intermediates, such as C3b, inhibit the growth of antigen–anti-
body complexes.
• Some intermediates, such as C3b and C5a, serve as opsonins (see
Section 8.2.3) and encourage phagocytosis.

The consequences of complement activation are not necessarily adverse.

Although cellular necrosis and inflammation are unwanted effects, reduction
in antigen–antibody complex size and improvement of phagocytosis through
more efficient opsonization may be beneficial in certain settings. The role of
complement activation in local (and systemic) host response to implants is
still largely unexplored and further investigations may yield new insights
in the future.
230 Biological Performance of Materials: Fundamentals of Biocompatibility

12.3 Classes of Hypersensitivity Reactions

Allergic responses are more properly categorized collectively as hypersen-
sitivity reactions (Merritt 1998). These may be separated into immediate
hypersensitivity, which is mediated by direct antibody–antigen combination,
and delayed hypersensitivity, which is cell mediated. Immediate hypersen-
sitivity is further subdivided into three types:

• Anaphylactic or immediate shock response

• Cytolytic/cytotoxic reactions
• Toxic complex syndrome

Delayed hypersensitivity responses are usually called type 4 reactions. If a

challenge evokes an immediate reaction (type 1, 2, or 3) and then a further,
delayed reaction, the net reaction is termed “mixed” and referred to as type
5. (Note: these responses are frequently designated by Roman numerals.)
The nature of supposed responses to implants (to be discussed later) sug-
gests that they are, collectively, examples of delayed hypersensitivity, or type
4 reactions. The type 4 reaction is clinically similar to a chronic inflammatory
response, except that lymphocytes as well as neutrophils are seen in foci of
antigen–antibody complex accumulation and additional concomitant symp-
toms, as discussed in Section 12.4, may occur.

12.4 Hypersensitivity Reactions Associated with Implants

12.4.1 Polymers
Although humoral and cell-mediated immune responses are usually directed
toward materials of natural origin, it is of interest to inquire concerning the
response to implants. As of now, few responses to polymers used as biom-
aterials are recognized, except for the special case of implants made from
processed natural tissue. The use of such materials, including processed
allografts (tissue from human donors) and xenografts (tissue from other
species, such as porcine heart valves, etc.) is finally limited by the ability to
denature or chemically mask the foreign materials so as to suppress their
antigenic activity (Bajpai 1983). Some natural materials used in medicine,
but fortunately not as implants (such as latex rubber), are widely recognized
to evoke responses (Emans 1992).
Attempts to produce immune responses to bulk polymers (Maurin 1995;
Stern et al. 1972) have produced very mild effects. They are most suggestive
of immune response to native proteins denatured by adsorption (see Section
Allergic Foreign Body Response 231

5.5) sufficiently so that they are no longer recognized as self and are able to
serve as antigens, rather than direct antibody production or T-cell activation
by polymer molecules. Clinically, cell-mediated responses have been
reported to poly methyl methacrylate “bone cement” (Haddad et al. 1995;
Clementi et al. 1980) and to silicone elastomers (Kossovsky et al. 1987).
This latter topic has attracted a wide interest in recent years. Silica (SiO2)
has long been recognized as an antigenic adjuvant; that is, when adminis-
tered in conjunction with an antigen, in animal models, it produces a height-
ened immune response. Some animal studies suggest that silicone gel may
be an adjuvant (Nicholson et al. 1996; Naim et al. 1995). Thus, scattered
reports of apparent autoimmune diseases in patients with silicone elastomer
breast augmentation devices raised the question of whether a causal con-
nection might be possible. Kossovsky has been the primary proponent of
this view (Kossovsky and Freiman 1994). However, despite continued animal
studies and extended human epidemiological studies, no association has
been demonstrated between the use of silicone implants and a wide range
of connective tissue disorders, many of which are known or suspected to be
autoimmunity diseases (Gabriel et al. 1994).*
Despite these isolated reports implicating polymers, concerns about
immune or allergic responses to biomaterials have centered primarily on
metallic materials on the skin and as implants.**

12.4.2 Metals
12.4.2.1 General
The action of metals on the immune system is a bit of a puzzle. It is generally
recognized that they must be dissolved to be active but the low molecular
weight and simplicity of structure of the resulting ions argue against their
being capable of directly inducing humoral or cell-mediated response. It is
thought that they combine with organic molecules such as albumin to form
complexes termed haptens, which possess antigenic qualities. It has also
been suggested that increases in concentrations of naturally occurring
metal–carrier protein complexes, such as Fe–transferrin or thioneins, can
render them effective haptens. Details of metal–protein complexing affecting
molecular shape changes are complex (Friedberg 1974).
However, Yang and Merritt (1994, 1996) have been able to demonstrate
the presence of antibodies to albumin–metal complexes in patients with well
functioning cobalt-base alloy joint replacement components and to produce
specific monoclonal antibodies to these complexes in a rabbit model. In the
presence of excessive wear, these findings may possibly be clinically signif-
icant. Nakamura et al. (1997) reported a case of autoimmune hemolytic

* The reader may wish to consult Angel (1997) for an account of the practical consequences of
this controversy.
** Note: immune responses to ceramic biomaterials have apparently not been reported.
232 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 12.1
Incidence of Metal Sensitivity
Patients Presenting with
Skin Problems (%) Normals (%)
Allergen Male Female Total Male Female Total
Nickel 3.1 12.9 9.6 1.5 8.9 4.2
Chromium 12.7 7.1 9.3 2.0 1.5 1.7
Cobalt 4.7 5.3 6.0 nr nr nr
Note: nr: not reported.
Source: Adapted from Hildebrand, H.F. et al., in Biocompatibility of
Co–Cr–Ni Alloys, Hildebrand, H.F. and Champy, M. (Eds.), Plenum
Press, New York, 1988, 201.

anemia associated with abnormal wear of a cobalt-base femoral head by

screw impingement.
Metal sensitivity among the general population is not rare; sensitivity to
nickel is the most common, followed by cobalt and chromium. The incidence
of sensitivity is estimated to be between 1 and 5%, with as much as 10% of
the population sensitive to at least one of these metals (Table 12.1). Incidence
rates are different for men and women, reflecting differences in home and
workplace exposure, and are higher for individuals involved in certain
industries, including mining, metal refining, electroplating, printing, etc.
In addition to acting as antigens through hapten formation, some metals
of interest for use in implants have been shown to affect directly the response
of the host immune system to other antigens. Chromium (Shrivastava et al.
2002) and nickel have been shown to suppress antibody production; the roles
of cobalt and manganese are anomalous. Iron, chromium, and nickel also
have been shown in vitro to bind with T-cell surface antigens (Bravo et al.
1990); this binding may change the specificity of the previously formed
antigens.

12.4.2.2 Dermatitis
Only two aspects of immune response to metals will be discussed. The first
of these is dermatitis as a direct response to metallic contact. This is important
for a number of reasons. It may be a direct and unacceptable side effect of
the use of external metallic support devices such as braces or dentures. It
has come to be recognized as a general indicator of systemic challenge in
the mechanism of hypersensitivity of metals.
A classic account of metallic dermatitis is that of Fisher (1986), who rec-
ognizes sensitization by nickel, chromium, cobalt, gold, and platinum among
the metals of implant interest (the last two have extensive dental applica-
tions). With the exception of nickel, these do not provoke an initial sensitivity
in the solid state due to their low solubility. However, in a previously sen-
sitized individual, all can evoke skin inflammations (that is, dermatitis)
of various types and degrees of severity. Fisher suggests that little
Allergic Foreign Body Response 233

TABLE 12.2
Sensitivity Thresholds for Metals
Clinical Challenge Threshold Conc.
Sensitivity Compound (mean, wt%)
Co CoCl2 0.27
Co, Ni CoCl2 0.25
Co, Ni, Cr CoCl2 0.51
Co, Cr CoCl2 0.30
Cr K2CrO4 0.21
Cr, Co K2CrO4 0.08
Cr, Co, Ni K2CrO4 0.14
Cr K2Cr2O7 0.22
Cr, Co K2Cr2O7 0.22
Source: Adapted from Wahlberg, J.E., Berufsder-
matosen, 21, 22, 1973.

cross-sensitivity occurs — that is, metal “A” causing a response in a patient

sensitized to metal “B” — although nickel sensitivity is often accompanied
by cobalt sensitivity.* Chromium appears to cause a response primarily when
present as a chromate [Cr+6]; the other metals are active in divalent and
trivalent ionic forms. Fisher’s chapter is especially valuable for its catalogs
of metal-bearing articles and environmental (exposure) settings. It also pro-
vides excellent clinical descriptions of the various dermatitides and of tech-
niques for skin testing for sensitivity.
Wahlberg (1973) tested a number of patients with identified clinical sensi-
tivity to metals and determined threshold concentrations for topical appli-
cation of metal salts. Of interest is his finding that the carrier used in the
patch test affects the threshold in many cases. That is, if the carrier aids
penetration of the metallic ions, a lower threshold is found. Table 12.2 pre-
sents these data for water-based solutions only. At variance with Fisher’s
conclusion, a subtle pattern of cross-sensitivity, at least with respect to the
minimum dose required for response, can be seen. These data should be
regarded with some care because individual patients showed a response to
concentrations one to two orders of magnitude below mean threshold values
for these experimental groups.
Wahlberg also looked for relationships between clinical severity of
responses and threshold values for particular patients and found poor cor-
relations except in the case of cobalt, where the correlation coefficient r =
0.61. He concluded that “…renewed contact with cobalt plays a greater role
in recurrence [of allergenic response] than [with] allergens such as chromium
and nickel, where other factors, such as infections, heat, moisture, cold,
stress, etc.…contribute to the recurrence.”
At a more subtle level, Hallab et al. (2004) found a positive correlation
between chromium and cobalt serum concentrations in nonsymptomatic

* Note that this latter finding may be due simply to the close association of these two metals in
alloys, etc. and not to a true cross-sensitivity.
234 Biological Performance of Materials: Fundamentals of Biocompatibility

patients with high release rate (metal-on-metal) hip replacements and the
responses of their lymphocytes to dissolved cobalt or nickel challenge in vitro.
It should be noted here that clinical practice is to use 1 to 5 wt% solutions
for skin patch testing (McGillis et al. 1989) for metallic allergenic response.
Comparing these concentrations with the threshold data in Table 12.1 lends
real weight to the concern that patch testing in the presence of a sensitizing
agent or condition may contribute to a later immune response in a previously
insensitive individual.
Many clinical reports detail local as well as remote skin response to metallic
devices. One that is reported completely (Brendlinger and Tarsitano 1970)
will suffice as an example. The patient, a 25-year-old woman, appeared with
generalized eruptions on her trunk, arms, and legs. Treatment with topical
corticosteroids provided some relief. She sought treatment for dermatitis on
her ring finger 8 months later. Despite treatment, she continued to experience
a spread and increase in intensity of symptoms. A rash appeared on her feet.
She began to experience pain and soreness in her mouth 2 months later —
that is, 11 months after her first symptoms appeared. On examination, she
was found to have a cobalt-chromium partial denture that she had acquired
several months before her initial skin problems and had worn intermittently
thereafter. Replacement of the metal denture with an acrylic one resulted in
prompt remission of symptoms. Reinsertion of the metal denture resulted
in a return of symptoms within 24 hours. At that time, skin testing showed
that she was sensitive to chromium as well as nickel in pure solid metallic
form.
Sensitivity to chromium, cobalt, and nickel is now well recognized in
dental applications in which contact between the device and the oral mucosa
occurs. Hildebrand et al. (1988) report a compilation of 149 cases that fulfill
three strict criteria:

• Presence of one or more clinical features suggestive of immune

response, such as eczema, redness, ulceration, etc.
• Healing (resolution of the clinical features) after removal of the
device
• Positive skin (epicutaneous) response to a metallic component of the
device

Of particular interest in this report is the observation that only 28 patients

(~20%) reported a prior history of symptoms referable to an established
metal sensitivity. Thus, it is probable that the use of stainless-steel and cobalt-
base alloys in dental applications can result in sensitizing previously non-
sensitive patients.
It should not be concluded that skin or mucosal contact by the implant is
necessary to produce dermatitis. A report (Cramers and Lucht 1977) docu-
ments the cases of three patients with 316L stainless steel and screw implants
who developed dermatitis 3 to 3.5 months after surgery. Two patients were
Allergic Foreign Body Response 235

found to be sensitive to chromium and cobalt by patch test and one was
sensitive to nickel. On surgical exploration, no infection could be cultured
in any case, and all immune response symptoms resolved promptly after
the devices were removed.

12.4.2.3 Implant Site Inflammation

The second aspect of immune response to metals to be considered is the
direct implant site inflammation. Hicks (1958) was probably the first to report
this effect. Initially, it was thought to be a simple inflammatory response
associated with the relatively high corrosion rates of alloys in use in the
1940s and 1950s. In the 1970s, interest in the problems of loosening of total
joint replacements, especially of the hip, reawakened interest in this problem,
especially in relation to possible immune response to metals.
An initial study of the possible relationship between sensitivity to metal
and problems with total joint replacement conducted by Evans et al. (1974)
aroused considerable interest. These researchers reported the results shown
in Table 12.3. They also found an apparently greater effect associated with
metal-on-metal devices. Several of their conclusions are:

1. Evidence is presented which suggests that after replacement, bone necrosis

and consequent loosening of the prosthesis may be due to the development
of sensitivity to the metals used.
4. Examination of this material [tissue of joints from sensitive patients] showed
necrosis of bone and soft tissue following obliterative changes in the vas-
cular supply.
7. We conclude that prostheses in which metal articulates with polyethylene
should be preferred; that any patient in whom loosening or fragmentation
occurs should be patch tested and that if sensitivity is found the implant
should be removed.

This study suggests a linkage between loosening, perhaps secondary to

inflammation, and sensitivity, especially for metal-on-metal devices.
Although they have overall lower wear rates, such devices would be

TABLE 12.3
Relationship between THR Loosening
and Metal Sensitivitya
Total Sensitive Insensitive
Patients (#) (#, %) (#, %)
Loose 14 9b(64) 5 (36)
Not loose 24 — 24 (100)
a By skin test.
b 11 loose prostheses.
Source: Adapted from Evans, E.M. et al., J. Bone
Joint Surg., 56B, 626, 1974.
236 Biological Performance of Materials: Fundamentals of Biocompatibility

expected to shed larger amounts of metallic products than metal-on-polymer

ones and appear to be involved more often in the supposed linkage than do
metal-on-polymer devices. The study was poorly controlled, as was a smaller
one later by Elves et al. (1975), who reported the following results:

Sensitivity to chromium, cobalt, nickel, molybdenum, vanadium and

titanium was studied by patch tests in 50 patients who had received total
joint replacements. Nineteen (38%) were sensitive to one or more metals,
primarily cobalt and nickel. In 23 patients, nontraumatic failure of the
prosthesis had occurred and 15 of these failures were sensitive to metal.
Out of 27 patients with no evidence of prosthesis loosening, four were
sensitive to nickel and cobalt or nickel alone. Dermatological reactions
occurred in 13 patients after surgery; however, only eight of these showed
evidence of metal sensitivity.

However, several questions remain unanswered. One of particular interest

is whether sensitization occurs from an implant or if it is a pre-existing
condition. No accurate incidence rates for metal sensitivity in the general
population are available. Large-scale studies, yielding the rates quoted in
the previous section, have been done only on populations that appear at
dermatological clinics — that is, patients with active skin problems — with
small control groups. A report published with that of Elves et al. (Benson et
al. 1975) favored the presensitization position. Groups of patients awaiting
total hip joint replacement were compared with those who had received the
devices already and were, in some cases, already symptomatic (device loos-
ening). Although “high” rates were found in both groups, no differential
was seen.
A subsequent study of 212 patients awaiting total hip replacement (Deut-
man et al. 1977) produced more definite data. These patients could be divided
into four groups, as shown in Table 12.4. This study showed a modest

TABLE 12.4
Metal Sensitivitya and Associated Complications
Total Sensitivea
Patient Classification Number Preoperative Postoperative
I: no previous bone operation 173 10 14b
II: previous metal implant 17 2 2
III: loose THR (to be revised) 16 2 2
IV: normal THR (contralateral) 6 — —
Note: THR = total hip replacement.
a Sensitive to at least one: Co+2, Ni+2, CrO4–2 (by skin test).
b From retest, 6 months postoperatively of 66/168 patients with no preoperative
sensitivity.
Source: Adapted from Deutman, R. et al., J. Bone Joint Surg., 59A, 862, 1977.
Allergic Foreign Body Response 237

possibility of association between sensitivity and device loosening and a

small possibility of sensitization or activation of previous sensitization by
metallic implantation.
Another clinical report (Brown et al. 1977) casts further doubt on an easy
interpretation. This study reported a group of 20 American patients with 23
hip implants of the metal-on-metal type. At least one implant was loose in
each patient. No patients were found to be sensitive to cobalt, nickel, or
chromium. This difference in incidence rates of metal sensitivity has been
ascribed to different environmental exposure between Brown’s American
patients and Elves’ and Benson’s British patients.
The questions of the relationship between immune sensitivity and device
loosening and of possible sensitization by implanted device components
have remained of interest to the clinical and research communities. Carlsson
et al. (1980) examined a group of 112 patients before and 134 patients after
metal-on-polymer hip replacement and concluded that little or no relation-
ship existed between previous sensitivity and loosening. They also felt that
it was doubtful that devices could induce sensitivity. Waterman and Schrik
(1985) studied 85 patients before and after metal-on-polymer hip replace-
ment and, although finding definite evidence of postoperative sensitization
to Cr+6, Co+2, Ni+2, and methyl methacrylate, also concluded that no relation-
ship was present between these findings and device loosening.
These latter studies suggest that probably more than one mechanism is
involved in device loosening. The results of Brown et al. (1977) would be
explained by some, such as Willert and Semlitsch (1977), as being secondary
to accumulations of wear debris and response to that accumulation. In small
amounts, wear particles are encapsulated or removed to regional lymph
nodes. When this system is overwhelmed, a foreign body response with
accumulation of giant cells may invade the tissues surrounding the joint,
causing resorption of soft and hard tissue. In such cases, device loosening
might be secondary to cellular attack of the interface between implant and
tissue (Harris 1995) or, as it were, it might be a biological analog of crevice
corrosion. However, multiple processes may occur in the same implant site
(Santavirta et al. 1990).
There has been continuing interest in the question of a possible relationship
between immune sensitivity to metal implants and bone damage in the
absence of infection (aseptic loosening) or wear debris. Leynadier and Lang-
lais (1988) reviewed the studies cited here, as well as others, and summarized
the outcome of 300 patients. In these studies, they found an overall 7.4%
incidence of sensitivity (to Ni, Cr, and/or Co in some valence state) in
patients with a good clinical outcome (N = 163) vs. a 46% incidence of
sensitivity in patients with aseptic loosening (N = 137). They further noted
an unexpectedly high rate of sensitivity to cobalt (30%) that was more than
ten times that expected from their control population. On the basis of this
review, they concluded that, except in exceptional cases, loosening was more
likely to promote development of sensitivity rather than vice versa.
238 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 12.5
Relationship of Metal Sensitivitya to Loosening in Internal Fixation of
Fractures
Postoperative Sensitive Sensitive to (#)b
Complications Number M (#) F (#) Total (%) Ni Cr Co
None 208 — 8 8 (3.9) 7 3 —
Delayed union 230 10 14 24 (10.4) 21 4 3
Infected 267 14 13 27 (10.1) 26 2 4
a By skin test.
b Some sensitive to more than one metal; no cobalt in alloy used.
Source: Hierholzer, personal communication, 1990.

This view is contradicted by Hierholzer (1990) in a study of patients with

nickel-bearing (steel) fracture fixation implants. He found that, although
tissue nickel concentrations around fracture fixation hardware were as much
as 100 times greater in infected than noninfected sites (Hierholzer et al. 1984),
a positive correlation could be found between incidence of all prosthetic
loosening (whether associated with delayed union or infection) and metal
sensitivity. His results are summarized in Table 12.5.
A major criticism of this entire line of clinical investigation of possible
immune responses to implants is the relative crudity of the skin test. Merritt
and Brown (1980) have adapted a test termed the leukocyte migration inhi-
bition factor (MIF) test to determine sensitivity to metallic ions more accu-
rately. This test is an in vitro measure of the ability of metal ions (incorporated
into haptens) to inhibit migration of human leucocytes towards a chemotactic
attractant. Inhibition of migration is taken as a measure of production of a
leukocyte migration inhibition factor by T-lymphocytes presumably acti-
vated by the specific metal-bearing hapten involved. In a later report, Merritt
and Brown (1985) reviewed a study of 283 patients who underwent routine
or cause-related device removal. Their data (summarized in Table 12.6) sug-
gest a far higher incidence of metal sensitivity in patients with metallic
devices than in any other study previously cited. In a parallel test of 629
patients coming to surgery (without prior history of metal implantation),
they found, again by the MIF test, that 25% were sensitive to at least one
metal among nickel, cobalt, and chromium.
Using a somewhat different approach, Wooley et al. (1997) demonstrated
an association between sensitivity to poly (methyl) methacrylate or cobalt-
base alloy particles and clinical diagnosis in patients receiving primary total
joint replacement or presenting for revision for component loosening and
pain, suggesting as expected a role for other health factors in determining
immune response to foreign materials.
More recently, using MIF testing, Merritt and Rodrigo (1996) screened a
carefully selected group of 22 patients, without metallic implants, who were
coming to primary total joint replacement surgery. All were insensitive to
Allergic Foreign Body Response 239

TABLE 12.6
Metal Sensitivitya at Device Removal
Alloy Number Insensitive (%) Sensitive (%) Reacting (%)b
Stainless steel 187 43 37 20
Cobalt base 55 26 36 38
a Sensitive to one or more of Ni+2, Co+2, Cr+6 (by MIF test).
b “Nonmigrators” (all chemotaxis suppressed); reverted to sensitive on retest
30 to 60 days after implant removal.
Source: Adapted from Merritt, K. and Brown, S.A., in Corrosion and Degradation
of Implant Materials: Second Symposium. ASTM STP 859, Fraker, A.C. and Griffin,
C.D. (Eds.), American Society of Testing and Materials, Philadelphia, 1985, 195.

TABLE 12.7
Metal Sensitivity Associated with Primary Total Joint
Replacement
Postimplantationa
Element Insensitive (%) Sensitive (%)b Reacting (%)c
Titanium 17 (77.3) 4 (18.2) 1 (4.5)
Cobalt 19 (86.4) 2 (9.1) 1 (4.5)
Chromium 17 (77.3) 4 (18.2) 1 (4.5)
Nickel 20 (91.0) 1 (4.5) 1 (4.5)
a 3 to 12 months postoperative.
b Seven patients showed sensitivity to one or more element.
c One patient.
Source: Adapted from Merritt, K. and Rodrigo, J.J., Clin. Orthop. Rel.
Res., 326, 71, 1996.

the metals for which they were screened (see Table 12.7) but 3 to 12 months
later, seven (31.8%) had developed sensitivities to one or more metals and
one developed a total nonmigration response characterized as a “severe
response.” See Hallab et al. (2001) for a full review of this complex topic.

12.4.2.4 Summary
It appears possible to draw the following conclusions concerning specific
immune sensitivity to metallic biomaterials:

• A significant level of immune sensitivity to metals is present in the

general population.
• Groups of patients who have had nickel-, chromium-, and cobalt-
bearing implants display higher than expected incidences of
sensitivity.
240 Biological Performance of Materials: Fundamentals of Biocompatibility

• Specific instances of symptoms consistent with a type 4 delayed

hypersensitivity reaction being specifically related to the presence
of a metallic implant have been documented.*
• Sensitivity to metal and sensitivity to other clinical symptoms, espe-
cially device loosening, are generally correlated; however, which is
cause and which is effect is unclear at this time.

12.5 Final Comment

The lack of basic knowledge and reliable statistics on clinical outcomes
associated with hypersensitivity responses to biomaterials has had severe
consequences, most notably in the debate over possible immune or autoim-
mune responses associated with the use of silicone gel-filled breast implant
prostheses. Although immune responses associated with implants may well
have a low incidence, the consequences for individual patients remain
severe. In the present medical treatment environment, in which the possi-
bility of such effects is usually ignored, some patients no doubt experience
a high level of frustration and prolonged periods of inappropriate therapy.
These possibly include patients with suspected infections that prove impos-
sible to culture from clinical specimens and who receive the working diag-
nosis of “sterile abscess” without appropriate subsequent testing for delayed
hypersensitivity to implanted materials (Hallab et al. 2000).
The entire issue of the role of specific (i.e., immune) vs. general response
to implants should remain an exciting one, particularly as biomaterials tech-
nology advances. For example, newer total joint replacements with hard-on-
hard (metal/metal or ceramic/ceramic) bearings are known to yield very
large numbers of submicron (1 to 100 nm)-sized metal oxide particles during
locomotion. Recent concerns have been raised (Gatti et al. 2004; Lomer et al.
2002) as to whether such particles may be antigens or adjuvants. Basic and
clinical research results can be expected to continue to shed more light on
these as well as more traditional concerns in the future. However, it would
be fair to say that study of effects of biomaterials on the human immune
system is one of the most neglected areas of host response research.

* This story may still have additional chapters: current interest in the use of metal/metal
articulations in total hip replacements is producing a number of clinical failures of fixation of
components to bone with adjacent soft tissues characterized as showing “…necrosis, lym-
phocyte infiltration, elevated mast cell counts and tissue fibrosis” (Lintner et al. 2005.) This is
interpreted as a type 3 toxic or arthus response to cobalt-bearing particles (Lintner, personal
communication).
Allergic Foreign Body Response 241

References
Angel, M., Science on Trial, W.W. Norton and Co., New York, 1997.
Bajpai, P.K., Antigenicity of glutaraldehyde-stabilized biological materials, in Bio-
materials in Reconstructive Surgery, Rubin, L.R. (Ed.), C.V. Mosby, St. Louis, 1983,
243.
Benson, M.K.D. et al., Metal sensitivity in patients with joint replacement arthroplas-
ties, Br. Med. J., 4, 374, 1975.
Bravo, I. et al., Differential effects of eight metal ions on lymphocyte differentiation
antigens in vitro, J. Biomed. Mater. Res., 24, 1059, 1990.
Brendlinger, D.L. and Tarsitano, J.J., Generalized dermatitis due to sensitivity to a
chrome cobalt removable partial denture, J. Am. Dent. Assoc., 81, 392, 1970.
Brown, G.C. et al., Sensitivity to metal as a possible cause of sterile loosening after
cobalt–chromium total hip-replacement arthroplasty, J. Bone Joint Surg., 59A,
164, 1977.
Carlsson, Å.S. et al., Metal sensitivity in patients with metal-to-plastic total hip
arthroplasties, Acta Orthop. Scand., 51, 57, 1980.
Chenoweth, D.E., Complement activation produced by biomaterials, Trans. Am. Soc.
Artif. Intern. Organs, 32, 226, 1986.
Chenoweth, D.E., Complement activation in extracorporeal circuits, Ann. N.Y. Acad.
Sci., 516, 306, 1987.
Clementi, D. et al., Clinical investigations of tolerance to materials and acrylic cement
in patients with hip prostheses, Ital. J. Orthop. Traumatol., 6, 97, 1980.
Cramers, M. and Lucht, U., Metal sensitivity in patients treated for tibial fractures
with plates of stainless steel, Acta Orthop. Scand., 48, 245, 1977.
Deutman, R. et al., Metal sensitivity before and after total hip arthroplasty, J. Bone
Joint Surg., 59A, 862, 1977.
Elgert, K.D., Immunology. Understanding the Immune System, Wiley–Liss, New York,
1996.
Elves, M.W. et al., Incidence of metal sensitivity in patients with total joint replace-
ments, Brit. Med. J., 4(Nov. 15), 376, 1975.
Emans, J.B., Current concepts review: allergy to latex in patients with myelodysplasia,
J. Bone Joint Surg., 74A, 1103, 1992.
Evans, E.M. et al., Metal sensitivity as a cause of bone necrosis and loosening of the
prosthesis in total joint replacement, J. Bone Joint Surg., 56B, 626, 1974.
Fisher, A.A., Dermatitis and discolorations from metals, in, Contact Dermatitis, 3rd
ed., Fisher, A.A. (Ed.), Lea & Febiger, Philadelphia, 1986, 710.
Friedberg, F., Effects of metal binding on protein structure, Q. Rev. Biophys., 7(1), 1,
1974.
Gabriel, S.E. et al., Risk of connective-tissue diseases and other disorders after breast
implantation, N. Engl. J. Med., 330(24), 1697, 1994.
Gatti, A.M., Biocompatibility of micro- and nanoparticles in the colon. Part II., Bio-
materials, 25, 2004.
Haddad, F.S. et al., Cement hypersensitivity: a cause of aseptic loosening? J. Bone
Joint Surg., 77B, 329, 1995.
Hallab, N.J. et al., Immune responses correlate with serum-metal in metal-on-metal
hip arthroplasty, J. Arthrop. (Suppl 3), 19, 88, 2004.
242 Biological Performance of Materials: Fundamentals of Biocompatibility

Hallab, N. et al., Metal sensitivity in patients with orthopaedic implants, J. Bone Joint
Surg., 83A, 428, 2001.
Hallab, N. et al., Hypersensitivity to metallic biomaterials: a review of leukocyte
migration inhibition assays, Biomaterials, 21, 1301, 2000.
Harris, W.H., The problem is osteolysis, Clin. Orthop. Rel. Res., 311, 46, 1995.
Hicks, J.H., Pathological effects from surgical metal, in Modern Trends in Surgical
Materials, Gillis, L. (Ed.), Butterworth, London, 1958, 29.
Hierholzer, S., Local tissue reactions and sensibilitization in the presence of stainless
steel implants, personal communication, 1990.
Hierholzer, S. et al., Increased corrosion of stainless steel implants in infected plated
fractures, Arch. Orthop. Trauma Surg., 102, 198, 1984.
Hildebrand, H.F. et al., Nickel, chromium, cobalt dental alloys, and allergic reactions:
an overview, in Biocompatibility of Co-Cr-Ni Alloys, Hildebrand, H.F. and
Champy, M. (Eds.), Plenum Press, New York, 1988, 201.
Kossovsky, N. and Freiman, C.J., Silicone breast implant pathology. Clinical data and
immunologic consequences, Arch. Pathol. Lab. Med., 118, 686, 1994.
Kossovsky, N. et al., The bioreactivity of silicone, CRC Crit. Rev. Biocompat., 3, 53, 1987.
Leynadier, F. and Langlais, F., Total hip arthroplasties and allergy to metals, in
Biocompatibility of Co–Cr–Ni Alloys, Hildebrand, H.F. and Champy, M. (Eds.),
Plenum Press, New York, 1988, 193.
Lintner, F. et al., Are multinucleated giant cells indicative of cobalt-related tissue
damage after metal-on-metal THR? (Ger.; author’s trans), Osteologie, 14, 117,
2005.
Lomer, M.C.E. et al., Fine and ultrafine particles of the diet: influence on the muscosal
immune response and association with Crohn’s disease, Proc. Nutr. Soc., 61,
123, 2002.
Maurin, N., Delayed in vitro immune response to long-term intraperitoneal polymer
implant in mice, J. Biomed. Mater. Res., 29, 1493, 1995.
McGillis, S.T. et al., Patch testing, Clin. Rev. Allergy, 7, 441, 1989.
Merritt, K., Immune response, in Handbook of Biomaterial Properties, Black, J. and
Hastings, G. (Eds.), Chapman & Hall, London, 1998, 513.
Merritt, K. and Brown. S.A., Tissue reaction and metal sensitivity. An animal study,
Acta Orthop. Scand., 51, 403, 1980.
Merritt, K. and Brown, S.A., Biological effects of corrosion products from metals, in
Corrosion and Degradation of Implant Materials: Second Symposium. ASTM STP
859, Fraker, A.C. and Griffin, C.D. (Eds.), American Society of Testing and
Materials, Philadelphia, 1985, 195.
Merritt, K. and Rodrigo, J.J., Immune response to synthetic materials, Clin. Orthop.
Rel. Res., 326, 71, 1996.
Naim, J.O. et al., The effect of molecular weight and gel preparation on humoral
adjuvancy of silicone oils and silicone gels, Immunol. Invest., 24(3), 537, 1995.
Nakamura, S. et al., Autoantibodies to red cells associated with metallosis — a case
report, Acta Orthop. Scand., 68, 495, 1997.
Nicholson, J.J., III et al., Silicone gel and octamethylcyclotetrasiloxane (D4) enhances
antibody production to bovine serum albumin in mice, J. Biomed. Mater. Res.,
31, 345, 1996.
Playfair, J.H., Immunology at a Glance, 3rd ed., Blackwell Scientific Publications, Ox-
ford, 1979.
Santavirta, S. et al., Aggressive granulomatous lesions associated with hip arthro-
plasty, J. Bone Joint Surg., 72A, 252, 1990.
Allergic Foreign Body Response 243

Shepard, A.D. et al., Complement activation by synthetic vascular prostheses, J. Vasc.

Surg., 1(6), 829, 1984.
Shrivastava, R. et al., Mini review: effects of chromium on the immune system, FEMS
Immunol. Med. Microbiol., 34, 1, 2002.
Stern, I.J. et al., Immunogenic effects of foreign materials on plasma proteins, Nature,
238, 151, 1972.
Tang, L. et al., Complement activation and inflammation triggered by model biom-
aterial surfaces, J. Biomed. Mater. Res., 41, 333, 1998.
Tegnander, A. et al., Activation of the complement system and adverse effects of
biodegradable pins of polylactic acid (Biofix®)) in osteochondritis dissecans,
Acta Orthop. Scand., 65, 472, 1994.
Urban, R.M. et al., Accumulation in liver and spleen of metal particles generated at
nonbearing surfaces in hip arthroplasty, J. Arthrop., 19(8 Suppl 3), 94, 2004.
Urban, R.M. et al., Migration of corrosion products from modular hip prostheses.
Particle microanalysis and histopathological findings, J. Bone Joint Surg., 76A,
1345, 1994.
Wahlberg, J.E., Thresholds of sensitivity in metal contact allergy. 1. Isolated and
simultaneous allergy to chromium, cobalt, mercury, and/or nickel, Berufsder-
matosen, 21, 22, 1973.
Waterman, A.H. and Schrik, J.J., Allergy in hip arthroplasty, Contact Derm., 13, 294,
1985.
Willert, H.-G. and Semlitsch, M., Reactions of the articular capsule to wear products
of artificial joint prostheses, J. Biomed. Mater. Res., 11, 157, 1977.
Wooley, P.H. et al., Cellular immune responses to orthopedic implant materials fol-
lowing cemented total joint replacement, J. Orthop. Res., 15, 874, 1997.
Yang, J. and Merritt, K., Detection of antibodies against corrosion products in patients
after Co–Cr total joint replacements, J. Biomed. Mater. Res., 28, 1249, 1994.
Yang, J. and Merritt, K., Production of monoclonal antibodies to study corrosion
products of Co–Cr biomaterials, J. Biomed. Mater. Res., 31, 71, 1996.

Bibliography
Abbas, A.K. et al., Cellular and Molecular Immunology, 5th ed., W.B. Saunders, Phila-
delphia, 2003.
DeLustro, F. et al., Immune responses to allogenic and xenogenic implants of collagen
and collagen derivatives, Clin. Orthop. Rel. Res., 260, 263, 1990.
Ellingsen, J.E., A study on the mechanism of protein adsorption to TiO2, Biomaterials,
12, 593, 1991.
Gabriel, S.E., Soft tissue responses to silicones, in Handbook of Biomaterial Properties,
Black, J. and Hastings, G. (Eds.), Chapman & Hall, London, 1998, 556.
Guyton, A.C., Resistance of the body to infection: II. Immunity and allergy, in Textbook
of Medical Physiology, 6th ed., W.B. Saunders, Philadelphia, 1991.
Hennekens, C.H. et al., Self-reported breast implants and connective-tissue diseases
in female health professionals, JAMA, 275(8), 616, 1996.
Hildebrand, H.F., Veron, C. and Martin, P., Nickel, chromium, cobalt dental alloys,
and allergic reactions: an overview, Biomaterials, 10, 545, 1989.
Klippel, J.H. (Ed.), Primer on the Rheumatic Diseases, 12th ed., Arthritis Foundation,
Atlanta, 2001.
244 Biological Performance of Materials: Fundamentals of Biocompatibility

Kossovsky, N. and Freiman, C.J., Immunology of silicone breast implants, J. Appl.

Biomater., 8(3), 237, 1994.
Kossovsky, N. et al., Self-reported signs and symptoms in breast implant patients
with novel antibodies to silicone surface associated antigens [anti-SSAA(x)], J.
Appl. Biomater., 6, 153, 1995.
Kumar, P. et al., Metal hypersensitivity in total joint replacement, Orthopedics, 6, 1455,
1983.
Merritt, K. and Brown, S.A., Hypersensitivity to metallic biomaterials, in Systemic
Aspects of Biocompatibility, Vol. II, Williams, D.F. Ed., CRC Press, Boca Raton,
FL, 1981, 33.
Merritt, K., Role of medical materials, both in implant and surface applications, in
immune response and in resistance to infection, Biomaterials, 5, 47, 1984.
Niki, Y. et al., Screening or symptomatic metal sensitivity: a prospective study of 92
patients undergoing total knee arthroplasty, Biomaterials, 26, 1019, 2005.
Playfair, J.H.L. and Chain, B.M., Immunology at a Glance, 7th ed., Blackwell Scientific
Publications, Oxford, 2000.
Rostocker, G. et al., Dermatitis due to orthopedic implants, J. Bone Joint Surg., 69A,
1408, 1987.
Wang, J.Y. et al., Prosthetic metals impair immune response and cytokine release in
vivo and in vitro, J. Orthop. Res., 15, 688, 1997.
Weir, D.M. and Stewart, J., Immunology, 8th ed. Churchill–Livingstone, Edinburgh,
1997.
Wooley, P.H. et al., The immune response to implant materials in humans, Clin.
Orthop. Rel. Res., 326, 63, 1996.
13
Chemical and Foreign-Body Carcinogenesis

13.1 Definitions
This chapter will consider the role of implants in carcinogenesis. Some def-
initions are needed to make the arguments clear and unambiguous:

Benign Possessing controlled self-limiting growth without invasiveness

or ability to metastasize.
Cancer Perhaps the best definition is that of Roe (1966): “Cancer is a
disease of multicellular organisms which is characterized by the
seemingly uncontrolled multiplication and spread within the or-
ganism of apparently abnormal forms of the organism’s own
cells.” Roe further states that the three key characteristics of cancer
are cellular multiplication, autonomy, and invasiveness.
Carcinogen An agent capable of causing cancer.
Carcinogenesis The production of cancer (more properly but less com-
monly termed cancerogenesis).
Carcinogenic/tumorigenic Used interchangeably to connote agents ca-
pable of causing cancer.
Carcinoma A malignant neoplasm arising from cells of epithelial origin.
Leukemia or lymphoma A malignant neoplastic transformation of cells
of the circulatory system.
Malignant Possessing uncontrolled growth, invasiveness, and ability to
metastasize.
Metastasis A neoplasm arising by “seeding” from a primary malignant
neoplasm to a remote site. Also called a secondary neoplasm.
Mutagenesis The production of inheritable (genotypic) changes in cells.
Neoplasm A tissue mass arising from an abnormal, uncoordinated pro-
liferation of cells.
Primary neoplasm A locally arising neoplasm.

245
246 Biological Performance of Materials: Fundamentals of Biocompatibility

Sarcoma A malignant neoplasm arising from cells of connective tissue.

Tumor Literally, a swelling; used to refer to neoplasms.

With these terms in mind, first, chemically induced and promoted carcino-
genesis as it relates to implants will be examined. Initially, all neoplasms
associated with implants in experimental animals and in patients were
thought to be chemical in origin. It is now recognized that these tumors can
arise, at least in animals, from chemical and nonchemical origins. The
nonchemical or so-called “solid-state” mechanism will be taken up in the
later parts of this chapter.

13.2 Chemical Carcinogenesis

13.2.1 Introduction
Chemical carcinogens have many different forms, affect a variety of cell
types, and produce a variety of neoplasms. The nature of their action leads
to neoplastic transformation being possible near implants by direct solution
or diffusion; at a distance by transport and concentration; or in the absence
of implants by contact, ingestion, or inhalation. It is not within the scope of
this book to discuss the details of these processes. However, it should be
recognized that there is more than one transformation effect. Neoplastic
growth may be initiated by alteration of metabolic processes, by alteration
of replication processes (by growth stimulation or by reduction of contact
inhibition), or by mutagenesis. Although all carcinogens are now thought to
be mutagens, not all mutagenic agents are carcinogenic. A mutation may be
lethal (to a cell), prevent cellular replication, or simply not affect metabolic
or growth processes sufficiently to produce malignant behavior.

13.2.2 What “Everybody Knows” about Cancer

Before proceeding to a discussion of classes and types of chemical carcinogens,
it would be a good idea to discuss some popular misunderstandings about
carcinogenesis in general and chemical carcinogenesis in particular. “Every-
body knows” at least three things about cancer that probably are not so:

• “Cancer is increasing.” Figure 13.1 presents a summary of cancer

death rates in the U.S. for the period of 1930 to 2000, adjusted for
age (American Cancer Society, 2004). The adjustment for age is nec-
essary because, as life expectancy and thus age at death increase, a
fixed annual incidence rate of mortality from one source will pro-
duce an increase in actual deaths from that source. Today, most death
Chemical and Foreign-Body Carcinogenesis 247

80

Rate per 100,000 female population

60

40
Uterus* Lung
Breast

20
Stomach
Ovary Colon
and Rectum

Pancreas
0
1930 1940 1950 1960 1970 1980 1990 2000

80

Lung
Rate per 100,000 male population

60

Stomach
Colon and
40
Rectum

Prostate
20
Pancreas

Liver
0
1930 1940 1950 1960 1970 1980 1990 2000

FIGURE 13.1
Cancer death rates, by site, U.S., 1930–2000. Rates adjusted to 1970 standard U.S. population;
*cervix and endometrium combined. (Adapted from American Cancer Society, 2004 Cancer Facts
and Figures, American Cancer Society, Atlanta, 2004.)
248 Biological Performance of Materials: Fundamentals of Biocompatibility

rates due to particular types of cancer are decreasing. Furthermore,

excluding lung cancer, which is related primarily to tobacco smoking
and secondarily to environmental effects, death rates due to cancer
have been constant or have decreased since 1945 for men and since
1930 for women. Note that, although the rate of death due to prostate
cancer appeared to be increasing in the 1980s, this was likely due to
better, earlier diagnosis rather than to an actual increase in incidence.
Today, although nearly 30% of Americans alive will develop cancer
and 25% will die from cancer, the overall 5-year survival rate, despite
an aging population, has increased to 63% compared with 33% in
the 1960s and 25% in the 1930s. This increased survival rate is due
in part to earlier detection, permitting a longer normal course until
death and earlier medical intervention, and to improved therapies,
especially for some specific cancers.
• “Everything causes cancer.” In fact, the contrary seems to be true.
Chemical carcinogenesis seems to be the exception rather than the
rule. The statistical estimates are as follow. Approximately 1.5 to 2
million compounds and substances are now individually chemically
identifiable. An exhaustive literature search concerning the 6000
most likely candidates uncovered evidence that only 17%, or approx-
imately 1000, were possible carcinogens. A survey by the National
Institute of Occupational Safety and Health (Christensen 1972) of
2415 suspected carcinogens produced evidence of 1905 reported as
carcinogenic effects but only 1000 thought to be carcinogenic in
animals. In the most recent compilation available, only 58 substances
or groups of related substances or occupational exposures are listed
as known to cause cancer in humans; 188 additional ones can rea-
sonably be anticipated to be carcinogenic in humans (U.S. Depart-
ment of Health and Human Services 2004).
• “Toxic materials cause cancer.” The classic study is that of Innes et
al. (1969) in which 120 pesticides and toxic industrial chemicals were
selected for evaluation. These materials were fed to two strains of
mice in the maximum tolerable doses. The animals were sacrificed
after a standard period and evaluated for tumor incidence. The
results were that 11 compounds (including five insecticides) were
significantly carcinogenic; 20 compounds were equivocal — that is,
did not show significant elevation of cancer incidence rates in this
study; and 89 compounds produced no elevation of cancer incidence
rates.
Thus, in this study of compounds specifically selected for their tox-
icity and given in maximum possible doses, fewer than 10% proved
to be carcinogenic. Furthermore, these findings have come under
increasing criticism as being too pessimistic. It has been suggested
that testing toxic potential carcinogens at high dosages may
Chemical and Foreign-Body Carcinogenesis 249

artificially accentuate their activity by inducing increased rates of

cell division (Ames and Gold 1990).

13.2.3 Types of Carcinogens

In the classification of materials that are carcinogens, three types of agents
are recognized:

• The complete carcinogen that produces neoplastic transformation

by itself
• The procarcinogen, or carcinogen precursor, that is not a carcinogen
but is converted to one by metabolic processes in the body of the
test animal or man
• The cocarcinogen, which is a weak carcinogen or has no inherent
carcinogenic activity but increases the activity of complete carcino-
gens or procarcinogens when it appears in their company

The exact roles and functions of pro- and cocarcinogens remain unclear. It
was suggested early that neoplastic transformation is a two-step process
(Friedewald and Rous 1944):

1. Initiation produces the primary cellular transformation. The cells

enter a latent period and do not ordinarily develop into a tumor.
2. Promotion is characterized by the development of previously trans-
formed cells into an active visible tumor.

However, this process is now viewed as having at least three steps, with
specific conditions necessary during the latent period if subsequent expres-
sion (development of a tumor) is to occur.
Thus, a complete carcinogen is one that is an initiator as well as a promoter,
and a cocarcinogen may be promoter or initiator but probably not both.
Similarly, a procarcinogen may not be a complete carcinogen after metabolic
conversion but its action may depend upon the presence of other initiators
and promoters (Berenblum 1969). It is clear that potential chemical carcino-
gens must be considered in their roles as complete or incomplete agents as
well as possible promoters of previously initiated processes of neoplastic
transformation.

13.2.4 Chemical Carcinogens

An excellent and still useful review of these three types of agents among
organic compounds is that of Weisburger and Williams (1975). Figure 13.2
and Figure 13.3 and Table 13.1 are drawn from this study. Figure 13.2 lists
250 Biological Performance of Materials: Fundamentals of Biocompatibility

O CO
β-Propiolactone
CH2 CH2

O O
1,2,3,4-Diepoxybutane
CH2 CH CH CH2

NH
Ethyleneimine
CH2 CH2

O SO2
Propane sulfone
CH2 CH2 CH2
Dimethyl sulfate CH3OSO2OCH3

Methyl methanesulfonate CH3SO2OCH3

CI CH2CH2
Bis(2-chloroethyl) sulfide
(mustard gas or yperite) S
CI CH2CH2

CI CH2CH2
Nitrogen mustard (HN2)
N CH3
CI CH2CH2

Bis(chloromethyl) ether CICH2OCH2CI

Benzyl chloride C6H6CH2CI

Dimethylcarbamyl chloride (CH3)2NCOCI

FIGURE 13.2
Typical direct-acting chemical carcinogens. (From Weisburger, J.H. and Williams, G.M., in Can-
cer: A Comprehensive Treatise, Vol. 1, Becker, F.F. (Ed.), Plenum Press, New York, 1975, 185. With
permission.)

some of the typical stronger, pure organic carcinogens with their chemical
structures. Table 13.1 lists some of the better known procarcinogens. The
details of metabolic conversion are still unclear for many of these agents.
Table 13.1 suggests the form of the converted carcinogen and Figure 13.3
provides examples of possible intermediates and structures. Of particular
interest is the inclusion of vinyl halide or acetate in the procarcinogen list
and a variety of epoxides in the activated list in Table 13.1. Polyvinyl chloride
(PVC) and polyvinyl acetate (PVA) have some popularity in medical appli-
cations and epoxide conversion is possible for many polymeric implant
materials.
Chemical and Foreign-Body Carcinogenesis 251

Procarcinogen --> (proximate carcinogen) --> Ultimate Carcinogen

O
BENZO (a) ANTHRACENE 5,6 - EPOXIDE
(BENZO (a) PYRENE WITH
ADDITIONAL RING)
O - ESTER

NHCOCH3 N-R

OH
N-2-FLUORENYLACETAMIDE
ACTIVE ESTER
NCOCH3
(SULFATE, ACETATE)
R= -H or -COCH3
N-HYDROXY
DERIVATIVE

+
CCI4 CCl3

CARBON TETRACHLORIDE O

H2C - CHCI H2C - CHCI

VINYL CHLORIDE EPOXIDE

FIGURE 13.3
Typical procarcinogen activation reactions. (From Weisburger, J.H. and Williams, G.M., in Can-
cer: A Comprehensive Treatise, Vol. 1, Becker, F.F. (Ed.), Plenum Press, New York, 1975, 185. With
permission.)

Testing for possible agents, especially of the pro- or co- type, is quite
difficult due to the necessity of following the products through the various
steps of the metabolic chain. Also, many of the small animals used for these
tests have significant and not inconsiderable rates of spontaneous neoplastic
transformation. Although many agents that produce neoplastic transforma-
tion in test animals have not been definitely shown to be carcinogenic in
man, it is essentially correct to assume that all human carcinogens also
produce neoplastic transformation in animals. All known (specifically iden-
tified or associated with occupational exposure) chemical carcinogens in
humans have been shown to have carcinogenic activity in at least one test
animal species. However, the neoplasms may vary widely in location, dose
dependency, malignancy, etc. between species.
The vast majority of recognized chemical carcinogens are organic com-
pounds. Ceramic-body induction of carcinogenesis through a chemical route
252 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 13.1
Principal Procarcinogens and Key Derived Active Metabolites
Procarcinogen Proximate or Actual or
Ultimate Proposed Carcinogen
Polycyclic aromatic hydrocarbons Epoxide; radical ion?
Aflatoxin Epoxide
Arylamine or amide; azo dyes N-Hydroxylamino-O-esters; radical
ion (?); epoxide (special case:
cutaneous cancers)
Nitro aryl or heterocyclic N-Hydroxylamino-O-esters
compounds
3-Hydroxyxanthine, related purines O-Esters
Safrole 1′-Hydroxy-O-Esters
Urethane, alkylcarbamates Active esters
Pyrrolizidine alkaloids Pyrrolic esters
Alkynitrosamines or -amides, Alkyl carbonium ion
alkyhydrazines or -triazenes
Halogenated hydrocarbons Haloalkyl carbonium ions
Vinyl halide or acetate Epoxide?
Source: Adapted from Weisburger, J.H. and Williams, G.M., in Cancer: A Com-
prehensive Treatise, Vol. 1., Becker, F.F. (Ed.), Plenum Press, New York, 1975, 185.

has not been reliably identified in animals at this time. This is probably due
to the low solubility of ceramics used in implants and the paucity of testing.
The role of metals as possible chemical carcinogens will be examined in the
next section.

13.2.5 Metals as Chemical Carcinogens

The status of metals as carcinogens is less clear. One of the difficulties in
determining this is the problem of distinguishing between chemical and
foreign-body (FB) action (see Section 13.3). Mechanical implantation or inha-
lation of metal dust may proceed to neoplastic transformation by a chemical
route after corrosion, by an FB route by the presence of the residual metal,
or perhaps due to aggregation of corrosion products at the implant or at a
remote site. Furthermore, unlike most organic molecules, metals can display
a wide range of electronic valences.
Metals may be placed in a classification system as given in the earlier
discussion. They may be directly (or completely) carcinogenic (pure action)
or they may potentiate other agents and their compounds. Reaction products
or organometallic complexes may be carcinogenic, thus classing the original
form as a procarcinogen. Potentiation (classing metals as cocarcinogens) is
a very broad, nonspecific activity because many forms of neoplasms tend to
concentrate metallic ions and complexes. It is difficult to distinguish cause
and effect here: the concentration of metals may be causal or merely the
consequence of the higher level of metabolic activity of the neoplastic cells.
Chemical and Foreign-Body Carcinogenesis 253

Summarizing a broad range of early animal studies, Sunderman (1971)

has made a strong case for carcinogenic roles for chromium, cobalt, iron,
nickel, titanium, and for some metals not found in implant alloys. Environ-
mental and industrial workplace studies support the presumed carcinoge-
nicity of chromium, cobalt, nickel, and, perhaps, iron.
Furst (1978) has extensively reviewed the status of metals as carcinogenic
agents. This review is noteworthy because the author had previously (Furst
and Haro 1969) proposed strict criteria that a material should meet before it
could be considered carcinogenic:

Tumors must appear both at the site and at a distance from the point of
application; more than one route [of application] must be effective; more
than one species must respond; the growth should be transplantable;
and, if malignant, invasion and/or metastasis must be noted. Most im-
portant, all histological slides must be evaluated by a pathologist knowl-
edgeable in animal tumors.

These criteria, although more than 35 years old, are still applicable and very
relevant today. Of importance are Furst’s conclusions drawn from then avail-
able in vitro and animal studies subject to the preceding criteria:

• Metals for which pure metal and compounds are carcinogenic: Ni

• Metals for which pure metal is carcinogenic but no carcinogenic
compound is known: Co
• Metals for which pure metal is not a carcinogen but that have car-
cinogenic compounds (given in parentheses): Cr (CrO4–2), Fe (dext-
ran, dextrin), Ti (titanocene?), Mn (MnCl2?) (“?” = there is some
doubt)

Only metals of interest in implant applications are included here. Furst (1978)
lists several others, including cadmium, lead, and beryllium, that fall into
one of these categories. However, it is fair to state that Furst regards metallic
carcinogenesis as a well-established, real effect.
The complexity of the problem presented by potentially carcinogenic
metallic implants is shown in a study by Gaechter et al. (1977). These inves-
tigators implanted polished rods of seven common implant alloys, including
common stainless steel and cobalt- and titanium-base implant alloys, in rats
and followed them for 2 years. Each of these alloys contained at least one
element recognized by Furst as carcinogenic. Neoplasms of a wide variety
of types were found, but no statistical elevation above the control (nonim-
planted) group incidence rates was seen. This study was possibly suggested
by an earlier one by Heath et al. (1971) in which wear-produced particles
from a Co–Cr metal-on-metal total joint replacement were shown to be
carcinogenic in rat muscle 4 to 15 months after implantation.
There are two possible arguments to explain these conflicting findings. In
the first place, the rods used by Gaechter et al. (1977) may have released
254 Biological Performance of Materials: Fundamentals of Biocompatibility

metal at a slower rate than seen in the works referenced by Sunderman (1971)
or in the study of Heath et al. (1971). Thus, dilution may have prevented
direct chemical carcinogenesis or indirect (pro-) carcinogenesis by maintain-
ing pool concentrations below critical levels. This possibility is supported
by a later, much larger and somewhat longer rat implant study using rods
and powder (Memoli et al. 1986), which demonstrated a small but significant
increase in incidence of sarcomas and lymphomas in animals with implants
containing cobalt, chromium, or nickel.
The possibility that dilution may reduce the risk of neoplastic transforma-
tion leads directly to the question of whether a “threshold” of effect exists.
That is, is there a concentration of a carcinogenic agent below which it looses
its effectiveness? This is a matter of considerable importance in the implant
field because corrosion rates of successful alloys are relatively quite low. A
high-corrosion-rate alloy would probably be rejected for implant applica-
tions because of an acute tissue response. These low corrosion rates result
in modest serum and tissue concentration increases, except in instances of
local concentration, as will be discussed in Chapter 15.
A great deal of attention has been paid to this possibility of threshold levels
by legislators and administrators concerned with food purity and workplace
safety. The common view is that no threshold exists; that is, the transforming
effect is like a molecular “trigger” and reduced concentration simply reduces
the likelihood that the critical event will take place. Therefore, given random
chance enhanced by continued exposure, any concentration of a carcinogen
can eventually evoke a neoplastic response. This view was the precipitating
factor in the adoption in 1958 of the now famous Delaney Amendment to
the Pure Food, Drug, and Cosmetic Act. This amendment imposed a zero
(!) permissible level of carcinogenic agents as deliberate food additives,
stating that “…no additive shall be deemed to be safe if it is found to induce
cancer when ingested by man or animal, or if it is found, after tests which
are appropriate for the evaluation of the safety of food additives, to induce
cancer in man or animals…”.* Contrast this with the discussion in Chapter
1 on value judgments inherent in definitions and the carefully enunciated
position of Furst (1978) on the carcinogenic status of metals and their com-
pounds.
It is clear that, if one were to apply the Delaney criteria to medical devices,
none of the metals listed by Furst (1978) could be judged satisfactory, even
for short-term implantation. However, the wide utility of probably carcino-
genic food substances, such as certain dyes and saccharin, has resulted in a
case-by-case relaxation of the Delaney criteria for deliberate food additives.
These decisions have been made by balancing risk against benefit with,
admittedly, a portion of political judgment added in some cases. In 1996, the
issue of residual materials, such as pesticides, was dealt with by the enact-
ment of the Food Quality Protection Act.** Administered by the U.S.

* Cited in Federal Register 42(192), Tuesday Oct. 4, 1977, Part VI, page 54166.
** PL 104-170.
Chemical and Foreign-Body Carcinogenesis 255

Environmental Protection Agency, this law establishes a system of scientific

review to establish (and review on a 10-year cycle) tolerable limits for the
presence of such materials in food stuffs.
The same careful judgments will probably need to be made concerning
metallic implants. To do so, it is necessary to know the dose–response rela-
tionship for carcinogenesis accurately in the various animal models, how to
project this to low-dose–long-response-time conditions (in which animal
experiments become prohibitively expensive), and, most importantly, how
to translate the animal projections to rate predictions in humans. Very little
of the required information is now available.
The difficulty of doing such studies was illustrated by Bouchard et al.
(1996) in a large-scale, essentially lifetime study also performed in rats. In
this study, nearly cylindrical metal implants were fabricated from Ti6Al4V
(F138) or CoCrMo (F-75) in solid form and the latter in a fully porous form
with approximately 20 times the total surface area of the nonporous implants.
Groups of approximately 100 rats, with equal numbers of males and females,
had an implant placed on the lateral surface of one femur; an additional
group received a quantity of 50- to 80-μm F-75 microspheres implanted
subcutaneously. Animals were autopsied as they died of apparently natural
causes during the experiment and all remaining animals were sacrificed 24
months after implantation. No overall differences in tumor incidence or
death rates were found among the groups; however, 55 implant site tumors
of various types were observed (Table 13.2).
The data in Table 13.2 show a very significant association between implant
fixation (p < 0.001) with no apparent association between local dose (presumed
to be proportional to implant surface area) of potential carcinogens released
by CoCrMo alloys, with the possible expectation of a higher incidence in the
highest exposure group (IV) in which the implants might be considered to be
fixed due to individual encapsulation. Here, only one complication (implant
fixation) was studied closely; others known to affect cancer incidence and
prevalence include environmental, dietary, and hereditary factors.

TABLE 13.2
Implant-Associated Tumors
Implant Fixation Statusa
Group No. Tumors I II III IV
Solid Ti6Al4V 23 22 0 0 1
Solid CoCrMo 14 12 0 0 2
Porous CoCrMo 3 0 0 3 0
CoCrMo microspheres 15 n/a
a Key: I — loose in soft tissue; II — loose, but in contact with bone;
III — fixed to bone (ingrown); IV — unclassified; n/a — not
applicable (all microspheres encapsulated in soft fibrous tissue).
Source: Adapted from Bouchard, P.R. et al., J. Biomed. Mater. Res., 32,
37, 1996.
256 Biological Performance of Materials: Fundamentals of Biocompatibility

Moreover, one disappointing factor is emerging. It appears that linear

projections of response rates to low dose rates, even when the dose–response
curve is linear, provide underestimates of the effect. One reason for this is
discussed briefly in the following section.

13.2.6 The Latent Period

A second argument that may shed light on the results of Gaechter et al.
(1977) is based on the issue of latency. It is common in animal and human
neoplastic transformation for a period of time to pass between exposure
(initiation) and manifestation of neoplastic transformation. This waiting or
latent period differs from species to species and is different for each agent.
In humans, latency periods are typically 15 to 20 years and may be as long
as 40 years (Schottenfeld and Haas 1979). Furthermore, no simple way to
“scale” the effect — that is, to predict the latent period for an agent in one
species from that observed in another species — is known. Thus, one may
argue that the latent period in Gaechter’s experiment exceeded the test
period, despite the fact that 2 years is more than half the life span of most
laboratory rats.
The latency argument is particularly important in the generalization of
conclusions such as those of Sunderman (1971) to expectations of implant
site tumor incidences in patients. The vast majority of implants have been
in patients for only 15 or fewer years because of the advanced age of the
average implant patient and the relatively recent advances in total joint
replacement. For instance, Table 13.3 suggests that patients 65 years or
older at time of surgery have less than a 50% chance of outliving a 20-
year latency period, although some 2.5% will live to age 100 or more
(Arias 2004). Thus, the appearance of metal carcinogenesis in humans may
be awaiting the passage of an unelapsed latency period in the younger
patients who have received implants in large numbers in the last decade
(however, see Section 13.4). The differences in life expectancy at any age
between men and women depend upon a number of factors, including
occupational exposure, recreational pursuits, etc., and may not be related
merely to gender difference.
Excellent reviews of metal-associated chemical neoplastic transformation
have been provided by Sky–Peck (1986), Vahey et al. (1995), Rock (1998),
and Adams et al. (2003); the latter three reviews are directed primarily
towards human clinical experience.
Chemical and Foreign-Body Carcinogenesis 257

TABLE 13.3
Life Expectancya by Age in the U.S.
Age (years) Male Female
At birth 74.5 79.9
5 70.2 75.4
10 65.3 70.5
15 60.3 65.5
20 55.6 60.7
30 46.3 51.0
40 37.0 41.4
50 28.3 32.2
60 20.2 23.5
65 16.6 19.5
70 13.2 15.8
75 10.3 12.2
80 7.8 9.4
85 5.7 6.9
90 4.2 5.0
95 3.2 3.7
100 2.5 2.8
a Mean; all races; alive in 2002.
Source: Arias, E., United States Life Tables,
2002, National Vital Statistic Reports, 53(6),
National Center for Health Statistics, U.S.
Government Printing Office, Washington,
D.C., November 10, 2004.

13.3 Foreign Body Carcinogenesis

13.3.1 Early Observations of Foreign Body Carcinogenesis
So far the various classes of chemical carcinogens have been considered and
the available information concerning their metabolism and ability to produce
neoplastic transformation has been summarized. Across all the classes of
pro-, co-, and complete chemical cocarcinogens, it may be stated that the risk
of neoplastic transformation increases at least linearly with the concentration
and period of exposure.
Studies of chemical carcinogenesis show interesting differences in action
depending upon the manner and form of administration of the agent. Such
differences led early investigators to study the influence of the physical form
of the carcinogenic agent on its ability to induce transformation. A startling
finding was that many agents not previously thought to be carcinogens
produced dramatic neoplasm incidence rates in rodents when implanted in
a solid form rather than injected or fed in soluble or dispersed form. This
258 Biological Performance of Materials: Fundamentals of Biocompatibility

effect was called foreign body (FB) carcinogenesis and is known more
recently as solid state* carcinogenesis.
Among the early investigators of FB carcinogenesis were E. and B.S.
Oppenheimer who, in conjunction with a number of co-investigators, pub-
lished a long series of papers in the 1940s and 1950s (see Oppenheimer et
al. 1955). Their studies and those of other investigators of the period estab-
lished the following points:

• Solid materials without chemical carcinogenic activity can induce a

variety of neoplasms in several small rodent species.
• Induction activity generally increases with the size of the implant.
• Induction activity varies inversely as the inflammatory response;
that is, in the long run, well-tolerated materials are more effective
FB carcinogens.
• Porosity with an average diameter above 0.22 μm (the smallest size
studied) reduces the risk of transformation.

13.3.2 Mechanisms of Foreign Body Carcinogenesis

An excellent contemporary summary of these early investigations, primarily
using plastic films as challenge agents, is that of Alexander and Horning
(1959), who proposed that “the most likely process [of neoplasm induction]
would appear to be that the film alters the normal environment of the
neighboring cells in such a way as to favor the induction (or selection) of
discontinuous variations leading to malignancy.” That is, the survival of
viable products of normally occurring cell damage or mutation are somehow
favored by the presence of the solid body and protected from physiological
processes until they are ready to enter the rapid growth phase characteristic
of malignancy.
Two considerations appear to favor this argument. The first is geometric
and was discussed briefly in Chapter 8 when the role of implants in infection
was examined. A cell is normally surrounded by a volume of tissue sub-
tending a 4π solid angle. As an implant is approached, the solid angle
decreases to 2π. Furthermore, if the implant is invaginated with a surface
roughness that has a characteristic dimension on the order of cell sizes (2 to
20 μm), the solid angle might even be less than 2π for some selected cells.
The results would be less access to microvasculature, poorer diffusional
supply, and reduced cell contact inhibition. The flaw in this general line of
argument is, of course, the observation that materials with distributed poros-
ity of cellular dimensions are less carcinogenic in rodents than smooth non-
porous materials. Perhaps one of these geometric factors is dominant or

* I dislike this phrase due to its confusion with semiconducting materials and thus the implica-
tion of electronic causality.
Chemical and Foreign-Body Carcinogenesis 259

perhaps the improved diffusion and cellular activity associated with

microporous surfaces offsets the other aspects of the near-surface geometry.
The second consideration that favors the argument proposed by Alexander
and Horning (1959) is that chemical and electrical conditions near an
implant–tissue interface are different from those at a distance; this has been
discussed extensively in previous chapters. The question remains, however,
of which field and concentration effects might favor protection of deviant
cells. Russian investigators (cited by Bischoff and Bryson 1964) felt that
piezoelectric materials with sharp points and asperities were more tumori-
genic than the same materials in smooth or colloidal form. This suggested
a role for very high gradient electric fields in FB carcinogenesis. A study by
Andrews et al. (1979) attempted to investigate this question by subcutaneous
implantation of plates of polystyrene resin in mice. The resin was implanted
in neutral condition or as poled (polarized) electrets of various strengths.
Tumors associated with the control plates were evenly distributed on both
sides, and those associated with the electrets were found predominantly on
the electronegative side.* The investigators felt that the trend was to higher
incidence rates and shorter latency periods for electrets with higher fields.
A more complete review (Bischoff and Bryson 1964) expanded this critique.
The authors posed and answered three questions:

• Question 1: “Is the concept of nonspecific (rather than by a specific

chemical agent) solid state carcinogenesis justified?”
Answer: After considerable criticism of experimental method and
test subject, Bischoff and Bryson conclude that “on the basis of
the responses to rather stable, unrelated substances, there is a
type of nonspecific carcinogenesis in rodents that is dependent
upon a minimum surface requirement.”
• Question 2: “Does solid state carcinogenesis occur in humans?”
Answer: For humans, the authors note the low reported incidence
of neoplasms associated with implants, natural deposits (choles-
terol plaques, gall stones, etc.), and chronic low-level inflamma-
tory processes (leading to acellular fibrotic tissue, which has FB
attributes). They admit an exception to this basic pattern in the
observation of carcinoma associated with silicosis and asbestosis.
(Note, however, that in 1964 the very high correlation between
a specific type of asbestos [chrysotile] and mesothelioma was not
yet known.) They also recognize the problem of the latency pe-
riod and of the relatively smaller implants (with respect to body
weight) used clinically as compared with those in the experi-
ments of Oppenheimer and others. Thus, they concluded that
“the incidence of sarcoma [in humans] arising from subcutane-
ous FB granuloma is minimal.”
* However, this difference, as well as all other conclusions reported for this study, was not sta-
tistically significant.
260 Biological Performance of Materials: Fundamentals of Biocompatibility

• Question 3: “Is the subcutaneous site in rodents valid for testing for
carcinogenic hazards?”
Answer: This question reflects their observation of a “widespread
disenchantment” with subcutaneous studies in rodents. Their
arguments are rather vague, but essentially arrive at the point of
view that, through nonspecific irritation, FB carcinogenesis and
chemical carcinogenesis may occur together and, without ade-
quate controls, cannot be easily distinguished in the subcutane-
ous site. They suggest, however, that the determining factor is
the difference between response to irritation in the chemical case
and transformation after noninflammatory isolation of the for-
eign body in the FB case. In this latter comment, they presage
the more modern studies of FB carcinogenesis.

13.3.3 Additional Studies of Foreign Body Carcinogenesis

Further studies by Ott and by Brand and their students have focused upon
the mechanisms of neoplastic transformation in rodents in order to distin-
guish FB from chemical carcinogenesis. Brand (1975) has more fully sum-
marized his research and theoretical ideas. He made extensive use of related
species of mice that will accept tissue transplants without immune response,
but in which cells and their daughters can be identified, by species, through
an examination of chromosomes (karyotyping). I will paraphrase two sec-
tions of his paper dealing with the mechanism of the tumorigenic process
and hypotheses concerning initiation and promulgation of the process.
Brand reached the following conclusions, which were documented by
thorough and ingenious research:

• The most probable target cell in FB carcinogenesis is the pericyte, a

small cell type associated with microvasculature.
• After implantation, the transformation needed to produce a preneo-
plastic parent cell, with all the genetic information for later expres-
sion in the active neoplasm, occurs quite rapidly. In Brand’s mice
populations, this occurs within 4 to 8 weeks after implantation.
• Although the transformation occurs “near” the FB–tissue interface,
actual close contact with the FB is not required.
• Transformation is quite uncommon; thus, neoplasms appear to
develop from single parent cells representing one in the several
million affected by the presence of the implant.
• Although neoplasm production will occur in a capsule after FB
removal, a significant period of implantation after the initial trans-
formation event is required.
• A latent period always occurs between transformation and neoplas-
tic expression. However, this period is characterized more by
Chemical and Foreign-Body Carcinogenesis 261

inactivity of microphages than of the transformed parent cell that

may be cloning (undergoing mitosis without inheritable change) at
a slow rate. In the presence of active macrophages, as in moderate
to severe chronic inflammation, later neoplastic expression is sup-
pressed.
• When the latent period is over and rapid malignant growth begins,
all daughter cells, even if transplanted, appear to act in synchrony.

Boone and coworkers’ (1979) set of in vitro tissue culture experiments have
partially verified Brand’s in vivo studies. These researchers studied the effects
of attachment of mouse fibroblasts to polycarbonate plates in an in vitro
tissue culture system. Cells implanted after in vitro exposure produced trans-
plantable, undifferentiated sarcomas. Notwithstanding a decrease in latent
period with increased time in tissue culture, the authors concluded, as had
Brand, that the smooth surfaces of the plates acted as an FB carcinogen, for
at least initiation, independently of chemical composition.
Brand (1975) cited six proposed mechanistic origins of FB carcinogenesis
(with accompanying criticisms):

• Chemical activity of components of the FB. Brand suggested a mod-

erating or modifying role for chemical agents; however, he felt that
the nonchemical mechanism for FB carcinogenesis was well estab-
lished.
• Physiochemical surface properties of the FB. Again, Brand recog-
nized a possible role for interface physical effects but suggested that
they are overpowered in importance by other physical factors, such
as porosity.
• Interruption of cellular contact or communication. Brand felt that
this is an open question, but indicates that it would be expected to
play a more important role in neoplasm expression and maturation
than in induction.
• Tissue anoxia and insufficient exchange of metabolites. Brand
rejected this hypothesis based upon comparison of cell distances
from vascular processes in normal and neoplastic tissue and upon
induction studies with vascularized microporous surfaces.
• Virus (as an unseen contaminant of FBs). Although some viri are
now recognized as mammalian carcinogens, the evidence is still
weak for viri playing a role in FB transformation; Brand evidently
did not favor it.
• Disturbance of cellular growth regulation. Brand clearly favored this
mechanism, based on the heritability of neoplastic behavior in the
growing cell population. He suggested a wide variety of possible
aberrations in growth control and communication processes in cells.
262 Biological Performance of Materials: Fundamentals of Biocompatibility

Thus, his view is that the nonspecific surface effect, whatever its
origins, acts in a mutagenic fashion on cell populations.

This discussion of FB carcinogenesis has focused on nonchemical neoplas-

tic transformation effects produced by materials external to cells. However,
solid materials in a form that can penetrate cells can also produce FB trans-
formation. The best known example is crysotile asbestos, which was recog-
nized as a human carcinogen only because it produced a relatively rare lung
tumor (mesothelioma) (U.S. Department of Health and Human Services,
2004). Studies of asbestos and other fibers in animal models has led to the
Stanton hypothesis: mesothelioma can be induced by fibers less than 0.5 to
1 μm in diameter and more than 8 μm in length,* regardless of fiber com-
position (Lipkin 1980). Lipkin has also shown that in vitro fiber cytotoxicity
correlates well with these dimensions rather than with fiber composition.
Thus, slender stiff fibers, such as mineral whiskers, that are apparently able
to penetrate cells and produce direct mechanical damage, presumably to the
nucleus, appear to be undesirable components of biomaterials.
Perhaps the best place to end this discussion is with a quotation from
Brand (1975, p. 487):

Despite the rarity of FB tumors in man, it would be irresponsible to look

at the situation with complacency. Several measures are at our disposal
which would minimize the probability of FB tumors in man…. These
include (1) a more restrictive approach to artificial implantations, espe-
cially the exclusion of medically unnecessary cosmetic procedures, unless
they are indicated for psychiatric reasons; (2) smallest possible size of
implants; (3) reexamination of implant carriers at frequent intervals; (4)
a centralized registry for gathering information on general complications
as well as instances of neoplasia; (5) continued research (a) on implant
materials regarding their suitability for specific surgical purposes and
(b) on etiological questions concerning this type of neoplasia.

This passage certainly contains food for thought for the bioengineer.

13.4 Nonspecific Carcinogenesis

A final form of neoplastic stimulation is also recognized. Neoplasms can
arise in response to chronic irritation (leading to chronic inflammation).
Chemicals (as well as foreign bodies), infection, and mechanical trauma have
all been recognized as leading to this type of neoplastic transformation. This
was possibly a feature in the large, lifetime rodent implantation study

* That is, with an aspect ratio (L/D) > 8 to 16. However, other factors, such as fiber stiffness, may
play a role.
Chemical and Foreign-Body Carcinogenesis 263

discussed previously (Section 13.2.5; Bouchard et al. 1996). It is characterized

by an infidelity of replication (producing a daughter cell not identical to its
parent). The formation of keloids (hyperplastic, expansive scars) is a non-
malignant example of this effect. The occasional, apparently spontaneous,
malignant transformation of benign lesions such as fibrous histiocytomas is
a somewhat more ominous example (Heselson et al. 1983).

13.5 Evidence for Implant Carcinogenesis in Humans

In light of the long use of metallic implants in clinical orthopaedics and other
surgical specialties, it is fair to ask whether any evidence indicates chemical
or FB carcinogenesis associated with implants. The number of reports of
tumors at implant sites in animals is mounting. Sinibaldi et al. (1976) reported
sarcomas in animals (seven dogs and one cat) occurring 6 months to 4 years
after implantation of stainless steel devices. Of interest is the fact that five
of the eight cases accompanied the use of the Jonas telescoping splint. With
its integral deep “crevice” between parts, this device would be expected to
display an abnormally high rate of corrosion (see Chapter 4). Harrison et al.
(1976) had earlier reported two cases in dogs, one 6 years and one 12 years
after stainless steel implantation during clinical treatment of fractures. An
additional case of sarcoma occurring 12 years after Jonas splint implantation
in a dog has been reported recently (Madewell et al. 1977). A survey by
Stevenson et al. (1982), including these cases, reported 35 fracture-associated
sarcomas in animals with an average of 5.8 years between injury (and inter-
nal fixation) and diagnosis.
Finally, these tumors in animals are apparently associated with implant
site infection. Such infection may be expected to produce elevated rates of
corrosion and thus elevated concentrations of metal-bearing species near
implants, due to the resulting local acidosis. However, a large-scale case-
control study of possible association between tumors and the use of metallic
fracture fixation devices in 222 dogs with tumors (Li et al. 1993) failed to
show a significant relationship between the incidence of bone and soft-tissue
sarcomas and the use of implants predominantly fabricated from stainless
steel.
The early human orthopaedic literature reports only three cases of fracture-
related implant site tumors: Delgado (1958), sarcoma after fracture of tibia,
internally fixed; Dube and Fisher (1972), hemangioendothelioma after frac-
ture of tibia, internally fixed; and McDougall (1956), sarcoma (Ewing’s type)
in fractured humerus, after internal fixation. The last case is the best known
and occurred more than 30 years after the initial injury and metallic implan-
tation. Additional cases continue to be reported, typically occurring more
than 5 years after implantation. Although fracture fixation hardware should
be removed routinely within 2 years after implantation, if the patient’s health
264 Biological Performance of Materials: Fundamentals of Biocompatibility

permits, less than half is currently removed. Therefore, although such cases
are expected to continue to be rare, some concern remains, especially for
younger individuals.
More than two dozen cases of tumors associated with partial or total joint
replacements in humans have now been published;* the average postoper-
ative period before diagnosis is 7 years. The early reports have been dis-
cussed previously (Black 1988); more recent reports are reviewed by Rock
(1998) and Adams et al. (2003). These tumors fall into two general groups:

• Tumors of various etiologies occurring in fairly short periods after

implantation
• Primarily malignant fibrous histiocytomas occurring 10 to 15 years
after implantation

To date, all of these tumors have been associated with stainless-steel or

cobalt-base alloy devices. The origin of the former group is somewhat
obscure; however, the latter group may reflect direct chemical carcinogenesis
(associated with elevated tissue concentrations of metals near the implant;
see Chapter 15) or possibly malignant transformation of the previously
benign implant capsule.
Orthopaedic devices are placed in soft and hard connective tissues, which
are not especially sensitive to primary neoplastic transformation in humans
(Black 1984, 1985). Until recently, no evidence had suggested remote site
tumors possibly resulting from concentrations of suspected carcinogens,
such as chromates, because no large epidemiological study had been done
to detect their presence or absence. However, due in part to my suggestions
(Black 1984), at least three epidemiological studies have now been com-
pleted, the first by Gillespie et al. (1988).
Gillespie and colleagues identified 1358 patients who had received total
hip replacements in New Zealand between 1967 and 1978. They investigated
the health status of over 1000 of these patients who could be followed for
more than 10 years postimplantation and found a highly significant 70%
elevated incidence of tumors of lymphatic and hemopoietic origin (Table
13.4). In addition, they observed a significant suppression of soft tissue
(colon, bowel, and breast) tumor incidence up to 10 years postimplantation
followed by an apparent but nonsignificant increase in incidence. Even 10-
year follow-up may be insufficient for expression of primary (chemical)
tumors at the low doses encountered at remote sites in implant-bearing
human patients. These results may simply reflect an effect of corrosion
products producing a chronic immune system stress — that is, playing the
part of indirect promoters of soft tissue neoplasias produced by other causes.

* The word “published” is operative here: during a career of more than three decades of lectur-
ing on this topic to clinical audiences, almost without fail, after the formal program, I would be
approached by a surgeon who knew of someone who had such a case — but as yet unpublished!
Chemical and Foreign-Body Carcinogenesis 265

TABLE 13.4
Risks of Cancer in Patients with Total Hip Replacement
Study
Feature Gillespie et al. (1988)a Visuri et al. (1991) Nyrén et al. (1995)b
Total patients 1358 433 39,154
Site New Zealand Finland Sweden
Total THRs unk. (>1358) 511 46,547
Total patient years 14,286 5729 327,922
Average duration, 10.5 9.6 8.4
years
Type of THR Many; mostly McKee–Farrar Many; no
McKee–Farrar McKee–Farrar
Increased cancer Lymphoma/leukemia Lymphoma/ Bone (2.08);
riskc (1.38) (total period: leukemia (3.01d); melanoma (1.44d);
1.68d) intra-abdominal connective tissue
(2.78) (1.42); kidney
(1.33d); prostate
(1.18d)
Decreased cancer Colon/rectum (0.52d); Respiratory (0.86); Gastric (0.74);
riskc bronchus/lung (0.54); female lymphoma (0.89)
breast (0.30d) reproductive
system (0.56)

Overall riskc Decreased (0.77d) Unchanged (1.02) Increased (1.05d)

Comment Patients followed > 10
years; SIR = 1.60e
Note: McKee–Farrar is a metal/metal articulating THR device.
a SIRs for 5- to 10-year follow-up.
b SIRs for 5- to 9-year follow-up.
c Standard incidence ratio (SIR).
d Significant change in SIR(p < 0.05).
e Significant change in SIR(p < 0.01).
Source: Adapted from Black, J., Clin. Orthop. Rel. Res., 329S, S244, 1996.

Longer term follow-ups and larger, better defined study groups will be
required to explore these preliminary results.
Since this study, two additional ones have been done to explore the same
question (Table 13.4). The work by Visuri et al. (1991) appears to support the
earlier study, but that of Nyrén et al. (1995) appears to contradict it. However,
when viewed in light of the apparent prevalence of metal-on-metal devices,
which have been shown to produce perhaps as much as a 10- to 15-fold
elevation in circulating serum chromium concentration (Jacobs et al. 1996),
one might well conclude that a positive relationship exists between metal
release and the incidence of lymphoma and leukemia in patients. This con-
clusion should be viewed with some caution because it has been known for
some time that patients with rheumatoid arthritis are at greater risk than the
overall population for developing lymphoma and leukemia (Isomäki et al.
1978). Studies such as those reported in Table 13.5 are based upon compar-
isons to overall population statistics; in total hip replacement recipient
266 Biological Performance of Materials: Fundamentals of Biocompatibility

populations, individuals with rheumatoid arthritis are probably over-repre-

sented in comparison to the overall population in the country of the study.
In reviewing this growing body of conflicting information, Rock (1998)
concludes that “[although] the incidence of primary tumors in close prox-
imity to implants appears consistent with that expected in the general pop-
ulation…the frequency of occurrence and associated individual and group
risks of systemic and remote site malignancy remains unresolved.”
The question of the occurrence of FB tumors in patients also remains a
mystery. Brand (1983) reviewed 43 tumors occurring in humans at implant
sites, at up to 53 years after implantation; 25% occurred within 15 years and
50% within 25 years of implantation. In light of the large upsurge in medical
and cosmetic implant use in the 1950s and 1960s, he would have expected
to find orders of magnitude for more cases if causation in humans was the
same as for mice. Thus, although continuing to express concern about the
possibilities for FB carcinogenesis in humans, Brand concluded that little
clinical evidence for its occurrence existed.
Similarly, Berkel and colleagues (1992) studied a group of 11,676 women
with silicone elastomer breast implants, at an average of 10.2 years after
implantation, and concluded that they experienced 52% fewer primary
breast tumors than expected. Thus, although longer term data still are
required, it is beginning to appear likely that the Oppenheimer effect is a
consequence of the relatively primitive immune system of rodents in com-
parison to that in humans.
Perhaps the best overview is provided by a study published by the Inter-
national Agency for Research on Cancer (IARC 1999). The IARC has classi-
fied agents suspected of being carcinogens in a well-defined system based
upon balancing animal- and clinically-derived evidence. This system estab-
lishes five groups of agents and exposures as carcinogenic, probably or

TABLE 13.5
IARC Classification of Evidence for Human Carcinogenicity Risk
Degree of Evidence for Carcinogenicity
Group Classification in
Humans In Humans In Animals
1: Carcinogenic < Sufficient but Sufficient
2a: Probably carcinogenic Limited and Sufficient
Or inadequate and Sufficient
2b: Possibly carcinogenic Limited and < Sufficient
Or inadequate and Sufficient
3: Not classifiable Inadequate and Limited or inadequate
Or inadequate and Sufficienta
4: Not carcinogenic Suggested lack ??
Inadequate but Suggested lack
a However, mechanism of causation inoperative in humans.
Source: IARC, Evaluation Carcinogenic Risks Hum., Vol. 74, WHO, IARC,
Lyon, France, 1999.
Chemical and Foreign-Body Carcinogenesis 267

possibly carcinogenic, not classifiable, or not carcinogenic (Table 13.5). Based

upon an exhaustive study of animal data and human clinical results, the
IARC concluded that the following overall evaluation could be made con-
cerning the carcinogenic risk of clinical implants:

• Group 1(carcinogenic): none

• Group 2A (probably carcinogenic): none
• Group 2B (possibly carcinogenic): smooth metallic and polymeric
films, solid bodies of metallic cobalt, nickel, and a nickel alloy (66
to 67% Ni, 13 to 16% Cr, 7% Fe)
• Group 3 (not classifiable): orthopaedic implants of complex compo-
sition, cardiac pacemakers, silicone breast implants, metallic chro-
mium, titanium, Co-, Cr-, and Ti-based alloys, stainless steels and
depleted uranium, as well as dental materials and solid ceramic
bodies
• Group 4 (noncarcinogenic): none

It is of interest that, despite the extensive clinical use of biomaterials as

human implants, the IARC could not conclude that any are actually carci-
nogenic (group 1) or totally lacking in risk (group 4).
Two additional ways of approaching this issue parallel Furst’s (1978)
analysis of metal carcinogenesis in animals. The U.S. Department of Health
and Human Services is required by statute* to publish a list of agents (A)
known to be human carcinogens or (B) reasonably anticipated to be human
carcinogens. These correspond to the IARC groups 1 and 2A, and 2B and
3, respectively (see Table 13.5). The most recent of these lists (U.S. Depart-
ment of Health and Human Services 2004) contains the following agents,
which could be elements of metallic implants or be released by implant
degradation:

• Known (A): chromium hexavalent compounds, nickel compounds

and metallic nickel
• Reasonably anticipated (B): cobalt sulfate

Additionally, the state of California adopted Proposition 65, the Safe Drink-
ing Water and Toxic Enforcement Act of 1986, by ballot initiative. This act
requires the periodic publication of a list of chemicals known (to the state)
to cause cancer or reproductive toxicity. This list draws from many sources,
including IARC analyses; because it is precautionary, it is somewhat conser-
vative. It also provides for labeling requirements for any product known to
contain such chemicals, unless the release (dose) rate can reasonably be
expected to be below a previously determined acceptable level.

* Section 301(b)(4) of the Public Health Service Act as amended by Section 262, PL, 95-622.
268 Biological Performance of Materials: Fundamentals of Biocompatibility

The most recent version of the Proposition 65 list (12/31/04)* contains the
following carcinogenic agents, which could be elements of metallic implants
or be released by implant degradation:

• Chromium (hexavalent compounds)

• Cobalt (metal powder; [II] oxide, sulfate heptahydrate)
• Nickel (metallic, acetate, carbonate, carbonyl, oxide, refinery dust,
subsulfide), nickelocene

It remains clear that the traditional problem of projecting animal experi-

ence to human occupational and clinical situations applies in the consider-
ation of carcinogenesis as a possible consequence of biomaterial
implantation. In particular, it should be noted that, despite Brand’s com-
ments in his 1983 study, it is only now that significant human populations
with implants in place for more than 15 to 20 years are coming into existence.
Survivors of these large modern cohorts are now beginning to pass the
average latency period for low-concentration chemical carcinogenesis. There
are no animal experiments to guide one accurately in what expectations
should be. Perhaps Brand’s previous remarks might be paraphrased:
“despite the rarity of [implant-associated] tumors in [hu]man[s]…” — and
more careful attention should be given in the future to the challenging and
difficult issues of possible chemical and FB carcinogenesis by clinical
implants.

References
Adams, J.E. et al., Prosthetic implant associated sarcomas: a case report emphasizing
surface evaluation and spectroscopic trace meta analysis, Ann. Diagn. Pathol.,
7(1), 35, 2003.
Alexander, P. and Horning, E.S., Observations on the Oppenheimer method of in-
ducing tumors by subcutaneous implantation of plastic films, in Carcinogenesis:
Mechanisms of Action, Ciba Foundation Symposium, Wolstenholme, G.E.W. and
O’Connor, M. (Eds.), Little, Brown, Boston, 1959, 12.
American Cancer Society, 2004 Cancer Facts and Figures, American Cancer Society,
Atlanta, 2004.
Ames, B.N. and Gold, L.S., Too many rodent carcinogens: mitogenesis increases
mutagenesis, Science, 249, 970, 1990.
Andrews, E.J. et al., Surface charge in foreign body carcinogenesis, J. Biomed. Mater.
Res., 13, 173, 1979.
Arias, E., United States Life Tables, 2002, National Vital Statistic Reports, 53(6),
National Center for Health Statistics, U.S. Government Printing Office, Wash-
ington, D.C., November 10, 2004.

* http://www.oehha.ca.gov/prop65/prop65_list/Newlist.html.
Chemical and Foreign-Body Carcinogenesis 269

Berenblum, I., A re-evaluation of the concept of co-carcinogenesis, Prog. Exp. Tumor

Res., 11, 21, 1969.
Berkel, H., Birdsell, D.C. and Jenkins, H., Breast augmentation: a risk factor for breast
cancer? New Engl. J. Med., 326(25), 1649, 1992.
Bischoff, F. and Bryson, G., Carcinogenesis through solid state surfaces, Prog. Exp.
Tumor Res., 5, 85, 1964.
Black, J., Systemic effects of biomaterials, Biomaterials, 5, 11, 1984.
Black, J., Metallic ion release and its relationship to oncogenesis, in Fitzgerald, R.H.,
Jr. (Ed.), The Hip, Vol. 13, C.V. Mosby, St. Louis, 1985, 199.
Black, J., Orthopedic Biomaterials in Research and Practice, Churchill Livingstone, New
York, 1988, 292.
Black, J., Metal on metal bearings: a practical alternative to metal on polyethylene
bearings? Clin. Orthop. Rel. Res., 329S, S244, 1996.
Boone, C.W. et al. “Spontaneous” neoplastic transformation in vitro: a form of foreign
body (smooth surface) tumorigenesis, Science, 204, 177, 1979.
Bouchard, P.R et al., Carcinogenicity of CoCrMo (F-75) implants in the rat, J. Biomed.
Mater. Res., 32, 37, 1996.
Brand, K.G., Foreign body induced sarcomas, in Cancer: A Comprehensive Treatise, Vol.
1, Becker, F.F. (Ed.), Plenum, New York, 485, 1975.
Brand, K.G., Human foreign-body carcinogenesis in the light of animal experiments
and assessment of cancer risk at implant sites, in Biomaterials in Reconstructive
Surgery, Rubin, L.R. (Ed.), C. V. Mosby, St. Louis, 36, 1983.
Christensen, H.E. and Fairchild, E.J. (Eds.), Suspected Carcinogens, 2nd ed., CDC,
National Institute for Occupational Safety and Health, Cincinnati, OH, 1972.
Delgado, E.R., Sarcoma following surgically treated fractured tibia, Clin. Orthop., 12,
315, 1958.
Dube, V.E. and Fisher, D.E., Hemangioendothelioma of the leg following metallic
fixation of the tibia, Cancer, 30, 1260, 1972.
Friedewald, W.F. and Rous, P., The initiating and promoting elements in tumor
formation, J. Exp. Med., 80, 101, 1944.
Furst, A. and Haro, R.T., A survey of metal carcinogenesis, Progr. Exp. Tumor Res., 12,
102, 1969.
Furst, A., An overview of metal carcinogenesis, Adv. Exp. Med. Biol., 91, 1, 1978.
Gaechter, A. et al., Metal carcinogenesis, J. Bone Joint Surg., 59A, 622, 1977.
Gillespie, W.J. et al., The incidence of cancer following total hip replacement, J. Bone
Joint Surg., 70B, 539, 1988.
Harrison, J.W. et al., Osteosarcoma associated with metallic implants, Clin. Orthop.
Rel. Res., 116, 253, 1976.
Heath, J.C. et al., Carcinogenic properties of wear particles from prostheses made in
cobalt–chromium alloy, Lancet (March 20), 564, 1971.
Heselson, N.G. et al., Two malignant fibrous histiocytomas in bone infracts, J. Bone
Joint Surg., 65A, 1166,1983.
IARC, Surgical implants and other foreign bodies, IARC Monogr. Evaluation Carcino-
genic Risks Hum., Vol. 74, WHO, IARC, Lyon, France, 1999.
Innes, J.R.M. et al., Bioassay of pesticides and industrial chemicals for tumorgenicity
in mice: a preliminary note, J. Natl. Cancer Inst., 42, 1101, 1969.
Isomäki, H.A. et al., Excess risk of lymphomas, leukemia, and myeloma in patients
with rheumatoid arthritis, J. Chron. Dis., 31, 691, 1978.
Jacobs, J.J. et al., Cobalt and chromium concentrations in patients with metal-on-
metal total hip replacements, Clin. Orthop. Rel. Res., 329S, S256, 1996.
270 Biological Performance of Materials: Fundamentals of Biocompatibility

Li, X.Q. et al., Relationship between metallic implants and cancer: a case-control
study in a canine population, Vet. Clin. Orthop. Trauma, 6, 70, 1993.
Lipkin, L.E., Cellular effects of asbestos and other fibers: correlations with in vivo
induction of pleural sarcoma, Environ. Health Persp., 34, 91, 1980.
Madewell, B.R. et al., Osteogenic sarcoma at the site of a chronic nonunion and
internal fixation device in a dog, J. Am. Vet. Med. Assoc., 171, 187, 1977.
McDougall, A., Malignant tumor at site of bone plating, J. Bone Joint Surg., 38B, 709,
1956.
Memoli, V.A. et al., Malignant neoplasms associated with orthopedic implant mate-
rials in rats, J. Orthop. Res., 4, 346, 1986.
Nyrén, O. et al., Cancer risk after hip replacement with metal implants: a population-
based cohort study in Sweden, J. Natl. Cancer Inst., 87(1), 28, 1995.
Oppenheimer, B.S. et al., Further studies of polymers as carcinogenic agents in ani-
mals, Cancer Res., 15, 333, 1955.
Rock, M., Cancer, in Handbook of Biomaterial Properties, Black, J. and Hastings, G. (Eds.),
Chapman & Hall, London, 1998, 529.
Schottenfeld, D. and Haas, J.F., Carcinogens in the workplace, CA, 29, 144, 1979.
Sinibaldi, K. et al., Tumors associated with metallic implants in animals, Clin. Orthop.
Rel. Res., 118, 257, 1976.
Sky–Peck, H.H., Trace metals and neoplasia, Clin. Physiol. Biochem., 4, 99, 1986.
Stevenson, S. et al., Fracture-associated sarcoma in the dog, J. Am. Vet. Med. Assoc.,
180, 1189, 1982.
Sunderman, F.W., Jr., Metal carcinogenesis in experimental animals, Food Cosmet.
Toxicol., 9, 105, 1971.
U.S. Department of Health and Human Services, Report on Carcinogens, 11th ed., U.S.
Department of Health and Human Services, Public Health Service, National
Toxicology Program, December 2004.
Vahey, J.W. et al., Carcinogenicity and metal implants, Am. J. Orthop., 24(4), 319, 1995.
Visuri, T. and Koskenvuo, M., Cancer risk after Mckee-Farrar total hip replacement,
Orthopedics, 14(2), 137, 1991.
Weisburger, J.H. and Williams, G.M., Metabolism of chemical carcinogens, in Cancer:
A Comprehensive Treatise, Vol. 1, Becker, F.F. (Ed.), Plenum Press, New York,
1975, 185.

Bibliography
Ambrose, E.J. and Roe, F.J.C., The Biology of Cancer, D. Van Nostrand, New York, 1966.
Berenblum, I., Carcinogenesis as a Biological Problem, North–Holland Pub. Co., Am-
sterdam, 1974.
Becker, F.F., Cancer: A Comprehensive Treatise, Vol. 1: Etiology: Chemical and Physical
Carcinogenesis, Vol. 4: Biology of Tumors: Surfaces, Immunology, and Comparative
Pathology, Plenum Press, New York, 1975
Haag, M. and Adler, C.P., Malignant fibrous histiocytoma in association with hip
replacement, J. Bone Joint Surg., 71B, 701, 1989.
Kolstad, K. and Högstorp, H., Gastric carcinoma metastasis to a knee with a newly
inserted prosthesis, Acta Scand. Orthop., 61, 369, 1990.
Roe, F.J.C., Introduction, in The Biology of Cancer, Ambrose, E.J. and Roe, F.J.C. (Eds.),
D. Van Nostrand, New York, 1966, 28.
Chemical and Foreign-Body Carcinogenesis 271

Tait, N.P. et al., Malignant fibrous histiocytoma occurring at the site of a previous
total hip replacement, Br. J. Radiol., 61, 73, 1988.
Troop, J.K. et al., Malignant fibrous histiocytoma after total hip arthroplasty, Clin.
Orthop. Rel. Res., 253, 297, 1990.
U.S. Department of Health and Human Services, Proceedings of a workshop/con-
ference on the role of metals in carcinogenesis, Atlanta, GA, March 24–28, 1980,
Environmental Health Perspectives, Vol. 40, 1981.
van der List, J.J.J. et al., Malignant epithelioid hemangioendothelioma at the site of
a hip prosthesis, Acta Orthop. Scand., 59, 328, 1988.
14
Mineral Metabolism

14.1 Introduction
The next chapter will deal with the distribution of metallic ions and some
simple models for their dispersion. This chapter will consider one well
known metal — iron — in detail and a lesser known metal — chromium.
This consideration is important in its own right, as well as being an indicator
of the complexity of the metabolism of metals. As in the case of most other
metals, the details of metabolic pathways and kinetics of these two metals
are not fully known. However, the attempt will be to contrast iron meta-
bolism with chromium metabolism.
As discussed in Section 2.2 and Section 2.4, the human body is primarily
composed of four nonmetallic elements (in declining order of abundance):
oxygen, carbon, hydrogen, and nitrogen, which make up 96.9% of body
tissues by weight. Six additional elements, of which only two are metals
(calcium and sodium), play major physiological roles and contribute a fur-
ther ~3.2% of body weight. All other constituents contribute together no
more than 30 g and are termed trace elements. These include at least 13
metals, of which 10 are used routinely as nontrace constituents in human
implants: iron, copper,* aluminum, vanadium, manganese, nickel, molybde-
num, titanium, chromium, and cobalt.
Most of these trace elements (with the possible exception of titanium) play
vital physiological roles and thus are termed essential trace elements. The
role of each is characterized by three important attributes (Mertz 1981):

• Amplification: all known essential trace elements exert their biolog-

ical actions through a succession of regulatory and/or synthetic
steps that produce a many-fold amplification function and lead to
effects on the whole body.
• Specificity: each essential trace element has a specific role as a moiety
in a molecule or as an enzyme cofactor. This specificity depends on

* Copper is not routinely used in human implants, due to its cytotoxicity, but is a component of
some designs of semipermanent intrauterine contraceptive devices (IUDs).

273
274 Biological Performance of Materials: Fundamentals of Biocompatibility

ionic size and valence. Other ions may interfere with the specific
role of a trace element but may not replace its function.
• Homeostatic regulation: without exception, for each essential trace
element, a panoply of absorption, transport, storage, and excretion
mechanisms regulates concentration at the site of action within an
optimum range.

Our interest in trace elements is related to this third attribute — to the

possibility that the introduction of an endogenous source, i.e., release of
material from an implant, may interfere with homeostatic regulation and
produce adverse effects at the site of normal action or at other sites, directly
or by interference with other trace element-mediated processes.
In discussing the toxic effects of mercury and its organometallic com-
pounds, Schwarz (1977) makes the following point:

Below a certain threshold [of concentration], the organism can maintain

an equilibrium. However, once the threshold level is reached, small in-
creases in doses lead to great increases of toxic effects…. Indeed, this
relationship pertains not only to all metals but anything. It is universal,
with the possible exception of mutagenicity and carcinogenicity, but even
there repair mechanisms are at work which may give a small area of
tolerance (p. 3).

The general relationship between metal concentration level and functional

effect is shown in Figure 14.1. It should be emphasized that this is only
schematic in nature; the details and relative extent of each range depend
upon the nature of the metal involved and, probably, upon individual dif-
ferences between patients. Thus, although calculations may be performed
that predict various levels of metallic ions in blood, tissues, etc. (Section
15.4), the real need is to know the details of the metabolic, storage, and
excretory pathways. The next two major sections outline iron metabolism,
whose details are well known, and chromium metabolism, which is less well
known and understood. Until the comparable systems are as well known
for the other major metallic components of common implant alloys such as
aluminum, chromium, cobalt, nickel, titanium, etc. as they are for iron, great
care must be taken in the interpretation of animal and clinical determinations
of metallic content in vivo.

14.2 Iron Metabolism

14.2.1 Introduction
Iron is a biologically ubiquitous metal that is essential to all higher forms of
life owing to its central role in the heme molecule facilitating oxygen and
Mineral Metabolism 275

Lethal

FUNCTION

Toxic

Normal or
tolerable

Normal Tolerable Toxic

CONCENTRATION

FIGURE 14.1
Relation between metal concentration and its functional consequence.

electron transport. The capacity of porphyrin-Fe–protein complexes to bind

large quantities of oxygen reversibly makes hemoglobin and myoglobin well
suited to the transport and storage of oxygen in higher organisms. Iron-
containing enzymes include the cytochromes, catalase, cytochrome c reduc-
tase, succinic dehydrogenase, and fumaric dehydrogenase. Total body iron
in an average adult ranges from 2 to 6 g, depending on body weight, hemo-
globin concentration, age, sex, and size of the storage compartment. As a
trace element, iron’s presence in the body is exceeded only by magnesium
(see Table 2.3).
On the basis of function, two iron metabolic compartments are recognized:

• An essential compartment containing 70% of the total body iron is

composed of hemoglobin, myoglobin, heme enzymes, cofactor, and
transport iron.
• A nonessential storage compartment accounts for 30% of the total
body iron in a normal individual and consists of iron storage in the
form of ferritin and hemosiderin, primarily in the liver, spleen, and
bone marrow.

The essential compartment can be further subdivided into the following

distribution: 85% in hemoglobin, 5% in myoglobin, 10% in intracellular heme
enzymes and iron cofactors in other enzyme systems, and 0.1% as transport
iron bound to transferrin.
276 Biological Performance of Materials: Fundamentals of Biocompatibility

14.2.2 Absorption
Absorption of iron (see Figure 14.2) represents the single most important
factor maintaining the normal balance of iron in the body. It is influenced
by age, state of health, current body iron status, and conditions within the
gastrointestinal tract, as well as by the amount and chemical form of iron
ingested and by the relative iron-chelating nature of other dietary constitu-
ents (e.g., phosphates, phytates, ascorbic acid, and amino acids). Dietary
intake varies between 10 and 30 mg/day, with a common range of 12 to 5
mg/day. Normally, 0.6 to 1.5 mg/day are absorbed through the gastrointes-
tinal mucosa, representing only 5 to 10% of total dietary iron intake.

Fe2+DUODENUM JEJUNUM ILEUM COLON

ABSORPTION
(0.6 - 1.5 mg/day) Fe3+
12-15 mg
Food Fe
Plasma Iron~3mg
TRANSPORTATION 2+ IMPLANT
(Fe - transferrin)
(30 - 40 mg/day)
All Body Cells

UTILIZATION Bone Marrow

and 20-25 mg/day
CONSERVATION
Hemoglobin Death of Cells

Hemoglobin
Catabolism
STORAGE Fe Storage
RE system, Liver,
(~1000 mg
Spleen, Bone marrow
capacity)

EXCRETION Hemorrhage Urine, Sweat, Desquamation

RBC lost in Bile, Feces of Cells
(0.9 - 2.5 mg/day) trauma, urine (0.5 - 1 mg/day) (0.2 mg/day)
(0.1 - 0.5 mg/day)
Menstruation
(<1.2 mg/day)

FIGURE 14.2
Iron metabolism in adults. (From Fairbanks, V.F. and Beutler, E., in Modern Nutrition in Health
and Disease, 7th ed., Shils, M.E. and Young, V.R. (Eds.), Lea & Febiger, Philadelphia, 1988, 193.)
Mineral Metabolism 277

The actual mechanisms of absorption and transport of iron or iron chelates

at the mucosal cell level remain unclear; however, it appears that iron enters
the mucosal brush border by a passive diffusion process and exits on the
serosal surface to the plasma transferrin by an energy-requiring step. Intra-
cellular absorbed iron in excess of immediate physiological needs is com-
bined with a protein, apoferritin, to form ferritin, a water-soluble iron-
storage complex. As the mucosal cells become laden with ferritin, further
absorption is impeded consistent with diffusion kinetics until ferritin iron is
released to the plasma in response to body needs. Most absorbed iron, in
the form of intracellular ferritin, is lost into the intestinal lumen when the
crypt cells complete their 2- to 3-day maturation and migration to the tips
of the villi and are sloughed. Intraluminal factors that decrease iron absorp-
tion include rapid gastrointestinal transit time; achylia; malabsorption syn-
drome; precipitation by alkalinization by phosphates and phytates; and
ingested alkaline clays and antacid preparations. Despite intensive research,
the systemic factors regulating iron absorption have not been identified. In
general, iron absorption increases whenever erythropoiesis (red blood cell
production) is stimulated, during pregnancy, and in patients with hemochro-
matosis; decreased absorption is associated with depressed erythropoiesis
and iron overload.

14.2.3 Transportation
After an iron atom enters the physiological system, it is virtually trapped,
cycling almost endlessly from plasma to developing erythroblasts. Iron is
then released into the circulation for 100 to 160 days, moved to phagocytic
cells where it is cleaved from hemoglobin, and finally released into the
plasma to repeat the cycle. From the standpoint of distribution of total body
iron, the transport compartment is the smallest (~0.008%). However, kinet-
ically, it is by far the most active, turning over as often as 10 times every 24
hours.
The vehicle of this rapid transport and turnover of iron is transferrin, a
β1-globulin of approximately 8.6 × 104 molecular weight, with a half-life of
8 to 10.5 days. Transferrin is synthesized in the liver, and the total body
transferrin content of 7 to 15 g is nearly equally distributed between the
intra- and extravascular spaces. It functions to accept iron from gut absorp-
tion, storage sites, and phagocytic cells and to deliver iron to erythroid
marrow for hemoglobin synthesis, to cellular reticuloendothelium for stor-
age, to the developing fetus, and to all cells for incorporation into iron
metalloenzymes. Normally, approximately one-third of the total body trans-
ferrin (termed total iron-binding capacity; TIBC = 300 to 360 μg/100 ml) is
saturated with iron. The remaining transferrin represents a latent or unbound
reserve (unbound iron-binding capacity [UIBC]). The degree of saturation
(%) and the TIBC are important parameters in the study of iron metabolism
and related disease syndromes.
278 Biological Performance of Materials: Fundamentals of Biocompatibility

UIBC
TIBC
Serum Iron
500

Iron (mg/100 ml) 400

300

200

100 NORMAL

IRON DEFICIENCY

LATE PREGNANCY

INFECTION
INFLAMMATION
MALIGNANCY
HEMOCHROMATOSIS
HEMOSIDEROSIS

NEPHROSIS

THALASSEMIA MAJOR
LIVER CELL NECROSIS
PROPHYRIA
FIGURE 14.3
Serum iron concentration and the specific size of the transport compartment in a variety of
clinical conditions. (Adapted from Fairbanks, V.F. and Beutler, E., in Modern Nutrition in Health
and Disease, 7th ed., Shils, M.E. and Young, V.R. (Eds.), Lea & Febiger, Philadelphia, 1988, 193.)

For example, increased TIBC is characteristically found in iron deficiency,

in the third trimester of pregnancy, and in response to hypoxic states,
whereas decreased TIBC is evident in infection, protein malnutrition, iron
overload conditions, malignancy, cirrhosis of the liver, nephrosis, and pro-
tein-losing exteropathies. Figure 14.3 illustrates the relationships between
plasma iron and transferrin in a variety of clinical conditions. In general, the
level of plasma iron and saturation of TIBC are determined by the sum of
factors extracting iron from the blood for storage and utilization balanced
against those factors releasing iron into the blood, e.g., absorption, hemolysis,
and storage site release.

14.2.4 Utilization
It can be estimated from adult blood volume (typically ~5 l) and erythrocyte
lifetime (t1/2 ≈ 60 days) that, although more than 2.5 g of iron exists in
hemoglobin within red blood cells, only 20 to 25 mg/day is used in hemo-
globin synthesis in bone marrow. This is supplied through the transferrin
transport pathway from absorption or release from storage. Defects in
absorption, release, and/or transport may produce suppression of hemoglo-
bin synthesis and, over time, lead to an erythrocyte deficiency or anemia.
Mineral Metabolism 279

14.2.5 Storage and Excretion

Iron in excess of metabolic needs is stored intracellularly as ferritin or hemo-
siderin in various tissues of the body. Ferritin is normally found in many
tissues of the body; however, the reticuloendothelial system of the liver and
cells in the intestinal mucosa are the most significant metabolic storage sites.
Hemosiderin, a granular water-soluble compound, is thought to be an
aggregation of ferritin molecules. It can be seen microscopically in unstained
tissue sections of bone marrow as clumps or granules of golden refractile
pigment. Similar material is occasionally found in vivo in association with
iron-bearing implants such as stainless steel device components (Winter
1976) and is thought by some investigators to be locally produced hemosid-
erin.* Although macrophages adjacent to implants can accumulate hemosid-
erin from endogenous sources, the source of this particular material is
thought to be a combination of iron released from the original hematoma
(formed during surgery) and iron oxides and hydroxides formed during
corrosion. Primary storage sites are the hepatic parenchymal cells and retic-
uloendothelial cells of the bone marrow, liver, and spleen.
The relationship governing deposition of iron as ferritin or as hemosiderin
is unclear. However, it is postulated that the relative content of iron in either
storage form is a function of the total storage iron concentration. By chem-
ically binding or shielding iron from the surrounding intracellular environ-
ment, ferritin and hemosiderin serve to reduce its inherent toxicity. Iron in
its ionic form in excess of the TIBC of the blood is extremely and instanta-
neously toxic, as demonstrated by Gitlow and Beyers (1952). Slow intrave-
nous injection of only 10 mg ferric ammonium citrate totally saturated the
blood’s iron-binding capacity of all patients tested, and the excess randomly
diffused into all body tissues. The toxic response was manifested by cough-
ing, sneezing, nausea, and, occasionally, vomiting.
However, the human body tenaciously conserves its content of iron, losing
less than 0.01% of the total amount daily through excretory and other mech-
anisms (see Figure 14.2). Iron loss from the body consists of routine excretion
(in urine, bile, feces, sweat), periodic loss (desquammation of cells and men-
strual flow [in women]), and occasional loss (hemorrhage secondary to
trauma or disease). Iron excretion is essentially passive and increases only
very slowly with increases in body stores of iron.

14.2.6 Iron Overload

Considering the precisely regulated intestinal absorption of iron coupled
with limited physiological excretory capabilities, one can easily envision the
development of an iron overload state if iron, by pathological or iatrogenic
route, gained access to the endogenous system. For example, excessive

* Titanium-containing hemosiderin-like complexes are seen adjacent to titanium alloy implants,

further suggesting that hemosiderin is a nonspecific precipitate.
280 Biological Performance of Materials: Fundamentals of Biocompatibility

absorption of iron occurs in idiopathic hemochromatosis, a situation in which

iron is continually deposited in the parenchymal cells of various organs,
often resulting in arthritis, liver disease, cardiac failure, and diabetes. Exces-
sive intake of iron may lead to hepatic and reticuloendothelial involvement
manifesting as portal cirrhosis. This is demonstrated by the Bantu tribesmen
of Africa, who have traditionally consumed large quantities of food cooked
and stored in iron pots. Additionally, a few reported instances of parenteral
iron administration to treat misdiagnosed anemias have resulted in iatro-
genic iron overload states with consequent toxic signs indistinguishable from
those of hemochromatosis.
Therefore, evidence indicates that excessive amounts of intracellular iron
stemming from an iron overload condition may predispose to a variety of
liver disorders (Bacon 1998), primarily by causing progressive destruction
of parenchymal cells and subsequent fibrotic replacement. The positive rela-
tionship of cardiomyopathies and diabetes mellitus in iron overload to the
deposition of hemosiderin in the myocardium and the pancreas further
substantiates the toxicity of iron even in its bound storage form (Sullivan
2004). In an iron overload state, the serum iron and transferrin saturation
are usually increased and the TIBC is somewhat depressed (Figure 14.3).

14.2.7 Iron and Susceptibility to Infectious Disease

For nearly half a century, it has been recognized that an element of host
response to bacterial invasion is a reduction in the iron content of the blood
serum (reduction in SI; see Figure 14.2) (Weinberg 1974; Ward et al. 1996).
The mechanism of this reduction has been identified as a suppression of
intestinal absorption of iron concurrent with an increase in storage of iron
in the liver and a concomitant reduction in transferrin saturation. The net
effect is that growth-essential iron is made less available to microbial invad-
ers, thus producing a so-called “nutritional immunity” (Kochan 1973) for
the host.
To illustrate the strength of this argument, patients with infection and
inflammation are unable to mobilize iron from reticuloendothelial cell depots
irrespective of a normal total body iron content (Shils et al. 1993) and thus
experience a transient anemia. This suggests that the physiological system
would rather endure a short period of iron deficiency than risk a microbial
invasion. In a survey of pathogen–host metal interrelationships, Weinberg
(1971) concluded that “…in the contest between the establishment of a bac-
terial or mycotic disease and the successful suppression of the disease by
animal hosts, iron is the metal whose concentration in host fluids appears
to be most important.”
To acquire iron, microbial cells must often synthesize siderophores, phe-
nolates, or hydroxamates, whose function is to solubilize ferric iron at neutral
pH and assimilate the metal. In the presence of small iron-containing metallic
particles, such as wear debris from stainless-steel implants up to 10 to 15
Mineral Metabolism 281

μm in major dimension, macrophages may assist this process by attempts

to digest or dissolve iron and iron hydroxide. In addition, many microorgan-
isms have the potential to produce powerful iron-binding ligands that com-
pete with the host compounds, e.g., transferrin, for the available iron.
Organisms that have the ability to solubilize, assimilate, and bind iron inde-
pendently are termed autosequesteric.
Because most bacteria and fungi require only 0.3 to 4.0 μM concentrations
of iron for growth, human blood plasma with a concentration of 10 to 65
μM would appear suitable to support bacterial growth and multiplication,
leading to bacteremia. That bacteremias are the exception rather than the
rule illustrates the profound role plasma transferrin plays in resistance to
disease by production of “nutritional immunity.” Transferrin has an associ-
ation constant for iron of approximately 1030. This indicates that, in the
normal physiological situation in which transferrin is 25% saturated with
iron, the equilibrium free, ionic iron concentration is approximately 6 × 10–9
μM or 108–fold less than that required for microbial growth. Thus, for micro-
bial invaders to have any chance of survival in a foreign host, they must
have evolved the capability to synthesize siderophores with association con-
stants for iron of 1030 or greater in order to compete for the available iron.
Indeed, many bacterial invaders have iron-binding ligands capable of
extracting iron from host transferrin that is saturated 30% or more. Kochan
(1973) has demonstrated the microbiostatic action of various mammalian
serums in culture media with respect to bacterial growth of tubercle bacilli
(Table 14.1). For example, human serum containing 30% saturated transferrin
inhibited bacterial multiplication of the BGG strain of Mycobacterium tuber-
culosis. Addition of 36 μM iron neutralized this bacteriostatic activity.
Similar results were observed for bovine, mouse, and rabbit sera. However,
owing to their normally high transferrin saturation, guinea pigs had sera
equally susceptible to tuberculosis before and after iron addition. Other
microorganisms demonstrating similar behavior include species of Candida,

TABLE 14.1
Correlation of Transferrin Saturation (TR) with Bacterial Growtha in Mammalian
Sera in Vitro
Bacterial Growtha
Source of No. Fe Concentration in Saturation No 36 μm
Sera Samples Serum (μ
μm) TR (%) Added Fe Added Fe
Human 10 17 30.0 0–1 10–14
Cow 4 34 39.0 0–1 10–14
Mouse 10 41 60.2 1–5 9–15
Rabbit 8 36 64.3 1–5 10–15
Guinea pig 20 49 84.2 9–14 9–14
a Bacterial growth expressed as the number of generations of M. tuberculosis in 14 days.
Source: Adapted from Kochan, I., Curr. Top. Microbiol. Immunol., 60, 1, 1973.
282 Biological Performance of Materials: Fundamentals of Biocompatibility

Clostridium, Escherichia, Pasteurella, Shigella, and Staphylococcus. In vivo

studies using strains of Pseudomonas aeruginosa, Staphylococcus typhimurium,
Listeria monocytogenes, and E. coli have corroborated the in vitro results.
Weinberg (1974) summarized the role that iron plays in nutritional immunity:

A very consistent finding is that the intricate checks and balances be-
tween the iron chelators of the microbes and of hosts are readily and
markedly upset by changes in the environmental concentration of iron.
If the metal is added, microbial growth is enhanced; if the metal is
deleted, host defense is strengthened. This situation obtains not only in
experimental systems in vitro and in vivo, but also in clinical disease
situations.

Weinberg (1974) asks, “Might cryptic disturbances in iron metabolism

during the lifetime of individual hosts permit resurgence of latent infections
such as tubercular lesions?” In light of a report concerning tuberculosis in
dialysis patients (Pradhan et al. 1974), perhaps this question has been
unknowingly addressed. Pradhan and colleagues have attributed the 15
times higher incidence of tuberculosis in dialysis patients to an increased
susceptibility stemming from the uremic state and a consequently decreased
immunological responsiveness. These conditions notwithstanding, it may be
possible that sufficient iron (owing to the inherent iron concentration of the
dialysate) enters the blood stream during repeated dialysis and constitutes
a “cryptic disturbance,” thus rendering the patient more susceptible to tuber-
cular infection and/or relapse. Weinberg (1996) and Walter et al. (1997)
provide modern reviews of this topic.

14.2.8 Role of Implants in TIBC Saturation

As mentioned previously, Gitlow and Beyers (1952) have shown that 10 mg
of intravenous ferric-ammonium citrate was sufficient to saturate the TIBC
of the blood and produce immediate signs of toxicity. Weinberg (1974) has
expressed the belief that even small additions of iron may increase the
transferrin saturation of the blood, rendering an individual more susceptible
to infection. The use of stainless steel in joint replacement and fracture
fixation applications, coupled with the evidence of Lux and Zeisler (1974)
demonstrating the nature and relative proportions of corrosion products in
metallotic tissue, would seem at this junction to warrant investigation of the
possibility that the well recognized (and accepted) local corrosion may even-
tually elicit subtle long-term systemic consequences.
A single calculation can illustrate the minute quantities of iron involved
in these considerations and, similarly, the amount of corrosion that these
quantities represent. For a standard 316L stainless steel orthopaedic total hip
replacement prosthesis with a total surface area of approximately 200 cm2,
corrosion equivalent to 10 mg of iron release would constitute a general
surface dissolution to a depth of 625 Å — an amount beyond light
Mineral Metabolism 283

microscopic resolution. In terms of local corrosion, a pit of 1 mm depth

would release 10 mg of iron. Naturally, for bilateral hip implantation or for
devices with a porous sintered stainless-steel surface, the increased surface
area would mean that proportionately less iron would need to corrode per
unit area for equivalent effects. As suggested in Figure 14.2, the most likely
mode of release is into the transferrin transport system.
It is clear, however, that such corrosion would not be equivalent to the
clinical experiment of Gitlow and Beyers (1952). Corrosion would release
iron over a period of time rather than in a single “dose.” Furthermore, other
implant-derived metals such as chromium and aluminum can also bind to
transferrin, further reducing the UIBC and contributing to a higher apparent
transferrin saturation (TR).

14.3 Chromium Metabolism*

14.3.1 Introduction
Chromium differs from iron by only two atomic numbers (24 vs. 26) and by
less than 7% in average atomic weight. Both elements can form divalent and
trivalent ions, although chromium can additionally form a hexavalent ion.
However, there are profound differences in their biological roles. Iron is well
known for its primary role in hemoglobin, the primary oxygen transport
molecule in mammals; chromium plays a no less vital role in regulating
metabolism. These differences in biological roles and metabolic pathways
are an excellent example of the principle of specificity in essential trace
element function. Figure 14.4 compares and contrasts these roles and also
illustrates the principle of amplification of function: extremely small daily
intakes of each element have led to vital physiological roles in the whole
organism.
It would be desirable to be able to depict a complete and systemic overview
of chromium uptake, transportation, utilization and conservation, storage,
and excretion, as shown in Figure 14.2 for iron; however, this is not possible.
Based in part on estimates, Table 14.2 summarizes the overall picture of the
situation.

14.3.2 Absorption
Absorption of Cr+3 takes place in the gastrointestinal tract. Although the diet
may contain Cr+2 and Cr+6 as well as Cr+3, the latter valence is the predom-
inant form in the acidic conditions of the stomach and upper digestive tract
(Mertz 1983). Because Cr+3 is essentially excluded from cellular contents due

* This section draws strongly on the work of Langård and Norseth (1986).
284 Biological Performance of Materials: Fundamentals of Biocompatibility

IRON CHROMIUM 3+,6+

3+
Dietary: Fe Dietary: Cr
Absorbed: Fe 2+ Absorbed: Cr 3+
Turnover: 2 x 10-3g/day Turnover: 10-9g/day

Hemoglobin Glucose
tolerance factor

Turnover: 3 x 10-2g/day Potentiation

of insulin
Oxygen Turnover: 2 x 10-3g/day
transport

Oxidative Regulation of
phosphorylation energy metabolism
(1500 - 4000 kcal/day)

FIGURE 14.4
Comparison of iron and chromium amplification. (Adapted from Mertz, W., Science, 213, 1332,
1981.)

TABLE 14.2
Chromium Metabolism
Daily intake
Dietary supplya 200 μg
Absorbed (@1.5% absorbance)b 3 μg
Transport
Serum content (3.2 L @ 0.16 ppbc) 0.5 mg
Daily utilization Unknown, leads to 2 mg insulin synthesis/dayd
Storage
Daily addition 1 μg
Total body burden (70-kg individual) 7 mg
Daily excretion
Urinary (1.5 l @ 0.2 ppbe) 0.3 μg
Fecal, desquammation, etc. (balance) ~ 1.7 μg
a Source: Table 14.3, maximum recommended.
b Source: Table 14.3, mean value.
c Source: Versieck, J. and Cornelis, R., Anal. Chim. Acta, 116, 217, 1980.
d Source: Mertz, W., Science, 213, 1332, 1981.
e Source: Cornelis, R. and Wallaeys, B., in Trace Element — Analytical Chemistry in Medicine
and Biology, Vol. 3., Walter de Gruyter & Co., Berlin, 1984, 219.

to an inability to cross cellular membranes (Sanderson 1976), it is very poorly

absorbed. However, chromium can be found within cells in nonimplanted
individuals, suggesting that mechanisms exist in vivo to oxidize dietary Cr+3
to Cr+6 (Rogers 1984). Detection of chromium in nonparticulate form, espe-
cially in nonphagocytic cells such as erythrocytes, must be considered as
strong evidence of the presence of Cr+6.
Mineral Metabolism 285

Although daily absorption of chromium is very small compared to that of

iron (3 μg vs. 1 to 2 mg), the serum concentration is even lower (5 μg [0.16
ppb] vs. 3 mg [0.94 ppm]). As in the case of iron, serum chromium is
primarily bound to transferrin (Hertel 1986), although it can also be nonspe-
cifically bound to albumin. The turnover of biosynthetic chromium as an
element in a glucose tolerance factor (GTF) leads to the synthesis of insulin
and to the ability to metabolize carbohydrates (Schroeder 1966).

14.3.3 Storage and Excretion

About one third of the daily absorption (perhaps that portion reduced to
Cr+6) is stored in the reticuloendothelial structures in cells and nonspecifically
in other cells, such as erythrocytes. The balance is excreted, primarily fecally
and through desquammation of cells. In comparing chromium metabolism
to iron metabolism, the overall picture is one of radical differences, despite
the similarity of the elements, even in the absence of implants. This should
highlight the difficulties in understanding release, storage, and excretion of
metals from implants in which less knowledge of the metabolic pathways
is involved.

14.3.4 Biological Consequences of Excess Chromium

As indicated elsewhere in this work, chromium has historically been viewed
benignly, due to its essential role in sugar metabolism and the widely held
belief that the dietary form, Cr+3, was the only or predominant form encoun-
tered in vivo, no matter what the route of release. More recent views suggest
roles for chromium and chromium-bearing molecules in a wide variety of
toxic, carcinogenic, and allergenic effects. The latter two phenomena are
discussed in Chapter 13 and Chapter 12, respectively; see Dayan and Paine
(2001) for a comprehensive historical review of all three effects.

14.4 Human Dietary Metal Intake

The only significant source of essential trace elements, including metals, for
human metabolic processes is through the gastrointestinal tract. Transdermal
absorption and inhalation, although capable of causing local host responses,
rarely can be shown to contribute to internal metal concentrations or remote
storage. Primarily, humans depend upon dietary sources for essential trace
elements. Table 14.3 lists the recommended dietary allowances (RDAs) and
suggested safe and adequate intakes (SAIs) for two physiological and ten
essential trace elements, as determined by the U.S. National Research
286 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 14.3
Daily Recommended and Safe Mineral Intake
Dietary Absorbed
Mineral Content (%) Internalc

Recommended dietary allowances (RDA) a,b

Phosphorus 800 mg 50–70 400–560 mg

Calcium 800 mg 20–40 160–320 mg
Magnesium 315 mg 40–60 125–190 mg
Zinc 14.5 mg 2–38 0.3–5.1 mg
Iron 10 mg Heme: 20d
Nonheme: 6–18 0.5–1.0 mg
Iodine 150 μg 30–50 45–75 μg
Selenium 62.5 μg 80 50 μg

Safe and adequate intake (SAI)a

Fluorine 1.5–4.0 mg 35–100 0.5–4.0 mg

Manganese 2.0–5.0 mg ? <5.0 mg
Copper 1.5–3.0 mg 36 0.5–1.1 mg
Molybdenum 75–250 μg ? <250 μg
Chromium 50–200 μg 0.5–2e 0.25–1.0 μg
a Source: National Research Council, Recommended Dietary Al-
lowances, 11th ed., National Academy Press, Washington,
D.C., 1992.
b Adjusted for average of male (79 kg) and female (63 kg), 25
to 50 years old.
c Calculated; presumably equal to loss by all routes.
d Decreases with increasing nonheme iron in meal.
e Decreases with increasing daily chromium intake.

Council.* Note that, because of its strong action in preventing tooth decay,
fluorine is included, although its natural concentration in the body is very
low and despite the absence of a known normal biological role in mammals.
This table also lists ranges of percent absorption and resulting internal avail-
ability. It is against this latter amount that release by an implant should be
judged.
Three final points deserve to be made to complete this discussion of normal
mineral metabolism. In the first place, even in the absence of the routine
common consumption of the “one-a-day” type of vitamin and mineral sup-
plements (which usually contain 50 to 150% of the RDA or SAI of all essential
trace elements), modern diets in the developed nations generally provide all
essential minerals required for normal physiological functions in healthy
individuals. The widespread “enrichment” of foodstuffs such as milk, bread,
and breakfast cereals; the increasing amounts of fresh foods consumed; and

* The roles of trace metallic elements in human nutrition are coming under much more scrutiny
than in previous years. Perhaps this is a partial explanation for a more recent replacement of
RDAs and SAIs with the concept of daily adequate intake (DAI) (Vincent 2004).
Mineral Metabolism 287

modern food-preservation and -preparation techniques virtually guarantee

this result.* Furthermore, the homeostatic systems that control metal absorp-
tion, transport, storage, and excretion render the occasional practice of con-
suming “megadoses” of essential trace elements meaningless: in the general
case, the excess will not be absorbed and will simply be excreted directly.
For some easily absorbed minerals, excess consumption may lead to transient
toxicity. If defects in regulation exist, excessive intakes may lead to one or
another of various metal storage diseases (Underwood 1977; Mertz 1988).
Fortunately, these are rare and, in many cases, have familial, presumably
genetic, predispositions.
In the second place, it must be emphasized that concern about in vivo metal
released from implants centers primarily on two situations:

• The possibility of very elevated (10- to 100-fold normal) concentra-

tions occurring near implants or in remote storage sites
• The possibility of release of metal in a different valence state than
normally occurs in the body with subsequent formation of biologi-
cally active organometallic species

As will be discussed at further length in Chapter 15, measurements of

serum concentrations and urinary excretion of metals, although useful, do
not provide a true picture of either of these effects. Unfortunately, with the
advent of modern, highly sensitive techniques for determining metal content
of biological materials (atomic absorption spectroscopy, neutron activation
analysis, inductively coupled plasma-mass spectroscopy, etc.), a number of
commercial concerns have become involved in diagnostic studies of trace
element profiles in patients and in normal individuals. Homeostasis permits
a broad range of concentrations about the optimal concentration for any
essential trace element (see Figure 14.1); thus, in the absence of a metallic
implant, which could produce elevated concentrations and/or different
valence states, small changes in serum metal concentrations within a normal
range may only reflect daily variations in intake, utilization, etc.
It is highly unlikely that such studies can lead to primary diagnosis of
actual metal deficiencies or overloads, with a concomitant need for corrective
therapy, that have not already been detected by clinical indications
(Kruse–Jarres 1987). Thus, the validity of such studies in healthy individuals
or in nonimplant patients for other than verification of prior diagnoses must
be viewed with great skepticism. In current practice, the appropriate use of
such studies is for large-scale, prospective epidemiology to determine
whether relationships exist between group mean values and incidence or
severity of metal induction or storage diseases.

* One exception to this assertion may be chromium; increasing dietary use of refined (white)
sugar, which contains less chromium than raw sugars but requires insulin for metabolic conver-
sion (Mertz 1983), may lead to progressive chromium deficiency. See Baran (2004) for a more
complete discussion.
288 Biological Performance of Materials: Fundamentals of Biocompatibility

Finally, because implants can release metal in various forms, including

inorganic ions, organometallic soluble complexes, and fine particles, prima-
rily or secondarily by precipitation (Jacobs et al. 1995), the mere presence of
metal in a biological fluid or tissue cannot be placed in context without an
understanding of the bioavailability of that metal. That is, the discussion of
the role of bacterial access to iron (Section 14.2.8) can and should be gener-
alized to address the question of which cells encounter metals in their peri-
cellular environment and whether the form permits chemical and
biochemical interaction with extra- or intracellular process. Unfortunately,
traditional approaches to the issue of bioavailabilty (see Caussy et al. 2003)
neglect release from implants. Thus, the question raised by MacDonald
(2003) as to whether a safe level of metal ions is released from a particular
class of orthopaedic implants simply cannot be answered at this time.
Notwithstanding these caveats, the next chapter will take up in detail more
general issues of release, distribution, and excretion of ions from implants
for which some data and models exist.

References
Bacon, B.R., Metabolic liver disease. Iron overload states, Clin. Liver Dis., 2(1), 63, 1998.
Baran, E.J., Trace elements supplementation: recent advances and perspectives, Mini
Rev. Med. Chem., 4(1), 1, 2004.
Caussy, D. et al., Lessons from case studies of metals: investigating exposure, bio-
availability, and risk, Ecotox. Environ. Safety, 56, 45, 2003.
Dayan, A.D. and Paine, A.J., Mechanisms of chromium toxicity, carcinogenicity, and
allergenicity: review of the literature from 1985 to 2000, Hum. Exp. Toxic., 20,
439, 2001.
Fairbanks, V.F. and Beutler, E., Iron, in Modern Nutrition in Health and Disease, 7th ed.,
Shils, M.E. and Young, V.R. (Eds.), Lea & Febiger, Philadelphia, 1988, 193.
Gitlow, S.E. and Beyers, M.R., Metabolism of iron, J. Lab. Clin. Med., 39, 337, 1952.
Hertel, R.F., Sources of exposure and biological effects of chromium, in Environmental
Carcinogens: Selected Methods of Analysis, Vol. 8, O’Neill, I.K., Schuller, P. and
Fishbein, L. (Eds.), IARC Scientific Publication 71. International Agency for
Research on Cancer, Lyon, 1986, 63.
Jacobs, J.J. et al., Local and distant products of modularity, Clin. Orthop. Rel. Res., 319,
94, 1995.
Kochan, I., The role of iron in bacterial infections with special consideration of host-
tubercle bacillis interaction, Curr. Top. Microbiol. Immunol., 60, 1, 1973.
Kruse–Jarres, J.D., Clinical indications for trace element analysis, J. Trace Elem. Elec-
trolytes Health Dis., 1, 5, 1987.
Langård, S. and Norseth, T., Chromium, in Handbook on the Toxicology of Metals, 2nd
ed. Vol. II: Specific Metals, Friberg, L., Nordberg, G.F. and Vouk, V.B. (Eds.),
Elsevier, Amsterdam, 1986, 185.
Mineral Metabolism 289

Lux, F. and Zeisler, R., Investigations of the corrosive deposition of components of

metal implants and the behavior of biological trace elements in metallosis tissue
by means of instrumental multielement activation analysis, J. Radioanal. Chem.,
19, 289, 1974.
MacDonald, S.J., Can a safe level for metal ions in patients with metal-on-metal total
hip arthroplasties be determined? J. Arthropol., Suppl. 3, 19(8), 71, 2004.
Mertz, W., The essential trace elements, Science, 213, 1332, 1981.
Mertz, W., Chromium: an ultra-trace element, Chemica Scripta, 21, 145, 1983.
Mertz, W. (Ed.), Trace Elements in Human and Animal Nutrition, 5th ed. Academic Press,
New York, 1988.
National Research Council, Recommended Dietary Allowances, 11th ed., National Acad-
emy Press, Washington, D.C., 1992.
Pradhan, R.P. et al., Tuberculosis in dialyzed patients, JAMA, 229, 798, 1974.
Rogers, G.T., In vivo production of hexavalent chromium, Biomaterials, 5, 244, 1984.
Sanderson, C.J., The uptake and retention of chromium by cells, Transplantation, 21,
526, 1976.
Schroeder, H.A., Chromium deficiency in rats: a syndrome simulating diabetes mel-
litus and retarding growth, J. Nutrition, 88, 439, 1966.
Schwarz, K., Essentiality vs. toxicity of metals, in Clinical Chemistry and Chemical
Toxicology of Metals, Brown, S.S. (Ed.), Elsevier/North–Holland, Amsterdam,
1977, 3.
Shils, M.E., Olson, J.A. and Moshe, S. (Eds.), Modern Nutrition in Health and Disease,
8th ed., Lea & Febiger, Philadelphia, 1993.
Sullivan, J.L., Is stored iron safe? J. Lab. Clin. Med., 144(6), 280, 2004.
Underwood, E.J., Trace Elements in Human and Animal Nutrition, 4th ed., Academic
Press, New York, 1977.
Vincent, J.B., Recent developments in the biochemistry of chromium (III), Biolog. Trace
Elem. Res., 99, 1, 2004.
Walter, T. et al., Iron, anemia, and infection, Nutr. Rev., 55(4), 111, 1997.
Ward, C.G., Bullen, J.J. and Rogers, H.J., Iron and infection: new developments and
their implications, J. Trauma, 41(2), 356, 1996.
Weinberg, E.D., Roles of iron in host–parasite interactions, J. Infect. Dis., 124, 401, 1971.
Weinberg, E.D., Iron and susceptibility to infectious disease, Science, 184, 952, 1974.
Weinberg, E.D. Iron withholding: a defense against viral infections, Biometals, 9(4),
393, 1996.
Winter, G.D., Wear and corrosion products in tissues and the reactions they provoke,
in Biocompatibility of Implant Materials, Williams, D. (Ed.), Sector Publications,
London, 1976, 28.

Bibliography
Brown, S.S. and Savory, J. (Eds.), Clinical Chemistry and Chemical Toxicology of Metals,
Academic Press, Amsterdam, 1984.
Burrows, D., Chromium: Metabolism and Toxicity, CRC Press, Boca Raton, FL, 1983.
Cohen, M.D. et al., Mechanisms of chromium carcinogenicity and toxicity, Crit. Rev.
Toxicol., 23(3), 255, 1993.
290 Biological Performance of Materials: Fundamentals of Biocompatibility

Cornelis, R. and Wallaeys, B., Chromium revisited, in Trace Element — Analytical

Chemistry in Medicine and Biology, Vol. 3., Walter de Gruyter & Co., Berlin, 1984,
219.
Crichton, R.R. and Ward, R.J., Iron homeostasis, Met. Ions Biol. Syst., 35, 633, 1998.
Davies, I.J.T., The Clinical Significance of the Essential Biological Metals, Charles C
Thomas, London, 1972.
Ducros, V., Chromium metabolism. A literature review, Biol. Trace Elem. Res., 32, 65,
1992.
Friberg, L., Nordberg, G.F. and Vouk, V.B. (Eds.), Handbook on the Toxicology of Metals,
2nd ed. Vol. I: General Aspects. Vol. II: Specific Metals. Elsevier, Amsterdam, 1986.
Gibson, R.S., Essential trace elements and their nutritional importance in the 1990s,
J. Can. Diet. Assoc., 51, 292, 1990.
Goldenberg, H.A., Regulation of mammalian iron metabolism: current state and need
for further knowledge, Crit. Rev. Clin. Lab. Sci., 34(6), 529, 1997.
Mertz, W., Chromium in human nutrition: a review, J. Nutr., 123(4), 626, 1993.
Mu, Y. et al., Causes of titanium release from plate and screws implanted in rabbits,
J. Mater. Sci.: Mater. Med., 13, 583, 2002.
Okazaki, Y. et al., Comparison of metal concentrations in rat tibia tissues with various
metallic implants, Biomaterials, 25, 5913, 2004.
Schroeder, H.A., The Trace Elements and Man: Some Positive and Negative Aspects,
Devin–Adair, Old Greenwich, CT, 1973.
da Silva, J.R.R.F. and Williams, R.J.P., The Biological Chemistry of the Elements, 2nd ed.,
Oxford University Press, Oxford, 2001.
Theil, E.C., Iron, ferritin, and nutrition, Annu. Rev. Nutr., 24, 327, 2004.
Versieck, J. and Cornelis, R., Normal levels of trace elements in human blood plasma
or serum, Anal. Chim. Acta, 116, 217, 1980.
Von Schroeder, H.P. et al., Titanemia from total knee arthroplasty, J. Arthropol., 11,
620, 1996.
Wang, Y.T. and Shen, H., Bacterial reduction of hexavalent chromium, J. Ind. Micro-
biol., 14(2), 159, 1995.
Williams, D.R., The Metals of Life: The Solution Chemistry of Metal Ions in Biological
Systems, Van Nostrand Reinhold, London, 1971.
Xiu, Y.M., Trace elements in health and diseases, Biomed. Environ. Sci., 9(2–3), 130,
1996.
Zaffe, D. et al., Accumulation of aluminium in lamellar bone after implantation of
titanium plates, Ti-6Al-4V screws, hydroxyapatite granules, Biomaterials, 25,
3837, 2004.
15
Systemic Distribution and Excretion

15.1 Introduction
The traditional approach to consideration of biological performance has been
to focus on the implant–host interface. Thus, material response studies have
dealt with degradation of implant properties and host response studies have
focused on formation of a capsule and other events within the adjacent tissue.
More modern considerations recognize that a mammal, such as a test animal
or a human patient, is an interconnected structure with various mechanisms
permitting exchange between all of its tissues and organs. The systemic and
remote site results of such exchanges involving implants and implant deg-
radation products will be dealt with in Chapter 16.
Chapter 3 through Chapter 5 and Chapter 7 have considered mechanisms
that can modify native proteins or release materials from implants. This
chapter examines some aspects of the distribution and excretion of these
implant-related products. Their distribution through the various systems of
the body can take place in a number of different ways:

• Movement of solid bodies

• Movement of particulate materials, passively or actively (cell
mediated)
• Movement of dissolution or corrosion products by passive diffusion
or by active circulatory transport
• Movement of modified cells or native proteins

15.2 Movement of Solid Bodies

15.2.1 Large Particles
Large particles or portions of implants can move through soft tissue if they
possess a certain degree of structural asymmetry. A sphere, such as a

291
292 Biological Performance of Materials: Fundamentals of Biocompatibility

shotgun* pellet, will stay in its initial position for an indefinite time. An
asymmetric “needle,” such as a sewing needle or a porcupine quill, will
move point first and may travel for long distances due to the action of muscle
forces on it.
It is also possible for large material particles to become involved in blood
circulation. Wear particles from vascular prostheses will move “down-
stream” until they are trapped in reduced vessel diameters on the arterial
side of capillary beds or in the lungs on the venous side of the circulatory
path. Much larger particles can also be transported. A report of four cases
of shell fragments transported into the cerebral circulation is a dramatic
illustration of this possibility (Kapp et al. 1973).
More common is the finding of extracellular particles too large to be
phagocytosed, such as some wear debris, precipitated corrosion products,
or fibrillar fragments from tendon prostheses, in the lymphatic drainage, in
regional lymph nodes, or in remote medullary locations or organs. Such
observations have been made in animals (Margevicius et al. 1996) as well as
in patients (Case et al. 1994; Jacobs et al. 1995; Urban et al. 2004) with
functioning implants of various types. Note that all “foreign” particles found
in remote sites in patients with implants may not have been released from
implants (Gatti and Rivasi 2002), so care should be taken in their iden-
tification and analysis.
Pins, wires, and other implants used for internal fixation of fractures and
for adjunctive tissue immobilization during placement of permanent
implants can also become dislodged and migrate. Lyons and Rockwood
(1990) reviewed reports of 47 such occurrences after surgery in the vicinity
of the shoulder. Smooth pins and wires were more likely to be reported as
migrating than threaded ones; screws and staples were not reported as
migrating. Eight of the patients died (six of them suddenly) due to damage
to heart and blood vessels near the heart by the migrating implants. In a
majority of the reports (35 of 39) in which a postoperative time course could
be determined, migration apparently occurred within 8 months of implan-
tation, although the mean time to diagnosis was 22 months.
Migrating device fragments are no respecter of organs; they have been
reported to enter the lungs (Aalders et al. 1985) and the heart (Lyons and
Rockwood, 1990) and to move as far as from the shoulder to the spleen
(Potter et al. 1988). Therefore, thoughtful design of materials and the
implants using them should minimize or eliminate the possibility of release
of macroscopic fragments.

* It was once a reasonable assumption that such pellets were composed primarily of lead; how-
ever, since 1981 lead shot has been illegal for use by hunters over wetlands in the U.S. Thus, pel-
lets that have been in situ for less than 25 years may be copper- or nickel-coated lead or made of
copper-coated steel, tungsten, or bismuth. This should be taken into account when host response
is considered.
Systemic Distribution and Excretion 293

15.2.2 Phagocytic Transport

As Section 8.2.3 discussed, particles that are sufficiently small are phagocy-
tosed by a variety of cells. Cellular phagocytosis has four possible results:

• The phagocytic cell (PC)* can successfully digest the particle. Partial
digestion and externalization (exocytosis) of the particle may also
occur, but rarely so.
• The PC attempts to digest the particle but the degradation products
prove to be cytotoxic. Then the PC dies, its phagosomes and cell
membrane lyse, and another PC may attempt to phagocytose the
particle and digest it. If this progression continues through many
repetitions, dead PCs accumulate, resulting in caseation; the result-
ing mass of dead cells resembles cheese.
• The PC may transport the particle by passing into the blood or
lymphatic circulation, but most usually to regional lymph nodes
where particle-loaded cells accumulate and may produce granulo-
mas, such as the “teflonomas” reported by Charnley (1961) after the
use of a poly(tetrafluoro)ethylene as a bearing surface in total hip
replacement. In some cases, this may lead to a secondary histiocytic
response in the vessels or lymph nodes (Albores–Saavedra et al.
1994).
• The PC may be able to transport the particle to the lungs. There it
is possible for the particle to be extruded through the lung wall and
exhaled through the airway (Styles and Wilson 1976).

It should additionally be noted that all of these outcomes may result in

activation of the PC (Schnyder and Baggiolini 1978), with concomitant
release of biologically active agents, which may alter local host response to
the implant (see Section 8.2.3)
In the context of this chapter, it would be very desirable to make some
statements about the rates of transport of particles by phagocytes in the third
and fourth situations listed previously. It is probably not possible to gener-
alize, but some extension of the comments in Section 8.2.3 is desirable.
The transport of particles by phagocytic cells (primarily macrophages
because neutrophils are short lived and FBGCs tend to remain near the
implant site) consists of at least two major steps: uptake and transport. Very
little is known about transport rates in the lymphatic system, primarily
because most uptake studies have been done in vitro or by systemic injection
of a colloid of particles in vivo followed by sequential sampling of the PC
population in the arteriovenous circulatory system.

* The term phagocytic cell (PC) is used here for generality, rather than the more common term
phagocyte, because it now appears that a number of different cell types may display phagocytic
behavior.
294 Biological Performance of Materials: Fundamentals of Biocompatibility

Some details are known about the first step, uptake. Two general
approaches have been taken to describe the uptake process. The first is
to fit the kinetics of phagocytic removal of particles to the Michae-
lis–Menten model used in studies of enzyme activity (Normann 1974). In
this approach, uptake is considered to have two phases: attachment to the
PC and engulfment.
Attachment is modeled as consisting of a reversible attachment step and
an irreversible engulfment step:

k1 k3
P + PC ←⎯⎯⎯
k
→ ⎡⎣ PC − P ⎤⎦ ⎯engulfmen
⎯ ⎯⎯⎯ t
→ PCp (15.1)
2
attachment

where P is an extracellular particle, and p is an intracellular one. Previous

studies have shown that, for a wide variety of animal species, V, the velocity
of clearance of circulating particles (uptake), follows a first-order propor-
tional absorption law dependent upon particle concentration, C (Normann
1974):

dC
V= = − kC (15.2)
dt

The following equation is typical of the results of this approach:

1 1 ⎡ K ⎤ 1
= +⎢ c ⎥ (15.3)
V k3 Eo ⎣ k3 Eo ⎦ ⎡⎣C ⎤⎦

where
V = clearance velocity (uptake rate of particles by PCs)
C = concentration of extracellular particles
Eo = total available attachment sites
Kc = overall kinetic constant = (k2 + k3/k1)

A saturation effect (presence of a fixed maximum rate) is attributed to the

fact that the number of sites on the phagocyte membrane that can initiate
invagination and engulfment (Eo) is limited.
This result faces two major criticisms. The first is that the uptake velocity
is not a first-order process for all particle concentrations, and the second is
that it seems unlikely that specific limited numbers of membrane loci exist.
The second approach (Stiffel et al. 1970) is to describe the observed kinetics
as exponential in the general form given by Equation 15.4:

C = Co10–Kt (15.4)
Systemic Distribution and Excretion 295

where
K = total body phagocytic index (essentially, the particle clearance
velocity)
Co = initial particle concentration
t = time

The major difficulty with this result is that, once again, it is not a good
description of the kinetics of phagocytic behavior. What is actually observed
is a complex uptake velocity behavior with three domains, provided that the
particles are within a defined size range and that the initial concentration is
high enough so that the kinetics are not dictated by flow processes (Ver-
non–Roberts 1972). The initial phase seems to be first order and dictated by
the adsorption of serum opsonins to the particles or the adsorption of the
coated particles to the phagocyte. The second phase is exponential, dictated
by the dose (concentration of particles) as in any other dose–response situ-
ation. The third phase is a slowly disappearing component seen when a
heterogeneous distribution of particles is injected. What this “tail” actually
represents is the removal of smaller particles.
The possible explanation for this strange kinetic pattern is that there are
two opposing effects. One is a saturation effect attributed to a limited number
of binding sites by some and, more appropriately, to a limited concentration
of serum opsonins by others (Jenkin and Rowley 1961). The second, coun-
tervailing effect is the increased efficacy of clearance as the blood (or pre-
sumably tissue) concentration of particles goes down. The limit on the
particle size range mentioned in Chapter 8 restricts these investigations to
particles on the order of the size of leukocytes, approximately 4 to 7 μm.
A somewhat more pragmatic approach (Korn and Weisman 1967; Weisman
and Korn 1967) has led to the conclusion that the kinetics of uptake are
determined by a constant (absorption) vesicle volume. Careful studies with
well controlled particle size ranges led to the conclusion that, in an amoeba
model, although larger particles are taken up singly, small particles are
accumulated external to the cell until a critical volume is reached, whereupon
the “cemented” mass is absorbed simultaneously. Typical results obtained
by earlier investigators in a mammalian model that also lead to this conclu-
sion are given in Table 15.1. In this table, it is also useful to note that oxygen
consumption is required during phagocytosis by PMNs (because the process
requires energy and PMNs are aerobic cells) and that it increases with particle
size.
There are however, additional difficulties. All particles are not equal: com-
position, morphology, and surface charge may play a role in uptake velocity.
As discussed in Section 8.4.1, a variety of dissolved metal ions, such as Ni+2
and Cr+3, suppress phagocytic efficiency (Graham et al. 1975). Therefore, one
might expect a slower uptake of nickel- or chromium-bearing particles, due
to high local metal concentrations, than of polymeric particles of the same
size, tissue concentration, etc. Kawaguchi et al. (1986) have shown addition-
ally that phagocytosis of polystyrene (as measured by cellular oxygen
296 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 15.1
Effect of Particle Size on Phagocytosis by Guinea Pig PMNs
Diameter of Polystyrene Uptake O2 Consumption No.
Particles ( μm) (μ
μg/mg wet wt. PMNs) (μ
μl/mg PMN/min) Particles/PMN
0.088 7.4 0.0144 24,000
0.264 30.1 0.0372 3,600
0.557 28.1 0.0412 360
0.871 36.9 0.0396 102
1.305 34.3 0.0427 34
3.04 35.9 0.0422 3
>7 0 <0.012 0
No polystyrene 0 0.0124 0
Source: Adapted from Roberts, J. and Quastel, J.H., Biochem. J., 89, 150, 1963.

consumption) depends strongly upon surface potential and thus upon fixed
surface charge. Kapur et al. (1996) have also shown that surface charge
heterogeneity can greatly affect phagocytic ability.
Although these models, calculations, and experiments tell something
about the relatively likelihood of uptake (clearance velocity) as a function
of particle size and properties, they shed little light upon the question of net
removal rates from implant sites. This remains an area for further investiga-
tion. Far more is known about active and passive removal of dissolved
species.

15.3 Transport of Dissolved Species

15.3.1 Leaching of Monomers
Leaching or dissolution of polymers into circulatory system fluids results in
rapid dispersion throughout the body. This is a result of an arteriovenous
circulatory rate of approximately 1 min–1. That is, the normal blood volume
(about 6 to 8% of body weight or about 5 l) passes through the lungs once
a minute. An illustration of the rapidity of this circulation is the study of
Homsy et al. (1972) concerning release of monomer from poly(methyl)meth-
acrylate bone cement upon its insertion into the body. In a canine model,
insertion of a dose of freshly mixed cement as a femoral transcortical plug
in a dose of less than 2 g/kg body weight produced monomer levels of up
to 1 mg/100 ml in the inferior vena cava within 2 min of implantation. The
peak concentration was reached in 3 to 4 min, followed by a decline ascribed
to monomer clearance by evaporation through the lungs as well as progres-
sion of polymerization of the cement that reduced the source concentration.
Similar results were obtained in patients receiving PMMA-cemented femoral
endoprostheses: peak monomer concentrations occurred in the vena cava by
Systemic Distribution and Excretion 297

2 minutes postimplantation and 99%+ of (integrated) exhaled monomer was

detected in the airway within 6 min.

15.3.2 Corrosion of Metals

15.3.2.1 Local Effects
The corrosion of metals has been discussed in Chapter 4. It is appropriate
to inquire how metallic corrosion and dissolution products are distributed
in the body.
It is clear that, usually, an accumulation of corrosion products is found
around a metal implant. These products include membrane-bound ions,
particles released by intergranular and fatigue processes, locally precipitated
products resembling hemosiderin, and insoluble reaction products such as
metal hydroxides. The combination of these leads to the familiar tissue
discoloration termed metallosis, particularly in older reports. In addition to
tissue discoloration, consequences of passive diffusion and distribution adja-
cent to an implant may also be seen histologically as a varying degree of cell
reaction. Figure 15.1 shows the variation of effect with material “reactivity”
— that is, corrosion and/or dissolution rate combined with tissue response
— for needles inserted in the cerebral cortex of rabbits for periods of up to
1 to 1.5 years. Typical materials used were aluminum; platinum (nonreac-
tive); molybdenum; tantalum (reactive); and silver, iron, and cobalt (highly
reactive and toxic).
The picture presented by Figure 15.1 reflects corrosion and diffusion of
metal; however, it is complex and difficult to analyze because the breadth
and type of response about each implant depend upon the rate of corrosion,
the valence (speciation) of released metal, its diffusion constant, and its
toxicity. Furthermore, the release may be affected by anatomical location

Normal Central Nervous System Tissue

Necrosis
Macrophages

Leptomeningeal Connective Tissue

Glia Limitans
Implant
Giant
Zone of Astrocytosis Cells

FIGURE 15.1
Histological changes around implants of increasing reactivity (left to right) in the rabbit cerebral
cortex. (Adapted from Stensaas, S.S. and Stensaas, L.J., Acta. Neuropath. (Berl.), 41, 145, 1978.)
298 Biological Performance of Materials: Fundamentals of Biocompatibility

because local physicochemical conditions vary (see Section 2.2), resulting in

different rates of corrosion (Oron and Alter 1984). Implant site infection may
also affect the corrosion rate (Hierholzer et al. 1984) as well as tissue response
to the corrosion products.
A more accurate view of passive diffusion may be obtained by direct
analysis of the tissues in and near the implant site. Energy-dispersive
microanalysis has been widely used to study situations such as metal diffu-
sion into bone from implanted dental devices (Arvidson and Wroblewski
1978). One of the more graphic studies is that of Lux and Zeisler (1974), who
used neutron activation analysis to examine the spatial distribution of cor-
rosion products in soft tissues adjacent to a steel fracture fixation device.
Their results, obtained by analysis of tissues obtained at device retrieval, are
shown in Figure 15.2.

AFFECTED UNAFFECTED
TISSUE TISSUE
Metal concentration ( mg/kg = ppm)

100 Fe

Fe
Fe
minus
background
10

1
Cr

Cr

Mo: < 0.1 ppm Mo not detectable

0.1
1 2 3 4
Distance from implant (cm)

FIGURE 15.2
Tissue metal content surrounding an implant. (Adapted from Lux, F. and Zeisler, R., J. Radioanal.
Chem., 19, 289, 1974.)
Systemic Distribution and Excretion 299

In a study of 38 patients with histologically diagnosed metallosis at

retrieval of fracture fixation hardware, Lux and Zeisler found mean concen-
trations of iron, chromium, and molybdenum near a stainless-steel–tissue
interface to be two orders of magnitude above background values previously
determined in tissues from patients without metallic implants. These con-
centrations declined logarithmically with distance from the implant, reach-
ing background (“normal”) levels at distances of greater than 4 cm from the
implant. The ratio of Fe:Cr:Mo in the alloy (V4A) was 66:17.5:2.3 (≅ 29:7.5:1)
and typical tissue concentration ratios were 56.5:8:1 (at the tissue–implant
interface) and 117:9:1 (1 cm away from the interface). Nickel, which formed
12% of the implant alloy, was detected at the interface but not in the sur-
rounding tissue.
A number of conclusions may be drawn from these results:

• Ion concentrations (and persuadably leachate concentrations from

polymeric materials) may be very high in the vicinity of implants
when compared to systemic and remote site concentrations.
• The local accumulation of released products depends upon the
nature of their reactions with surrounding tissues and their diffusion
concentrations. In this example, nickel diffuses away rapidly and is
seen only at the interface, chromium and molybdenum diffuse less
rapidly, and iron is accumulated locally, presumably due to
the formation of precipitates that contribute to the discoloration of
metallosis.
• Tissue concentrations do not necessarily reflect alloy proportions.
Finding metals near implants in their alloy constituent proportions
is more evidence of wear debris accumulation than of burdening of
tissues with soluble or precipitated corrosion products (Michel
1987).*
• Finally, it should be pointed out that the mere detection of metal in
tissues does not reflect its biological availability. Metal may be
present as free ions (unlikely), bound to specific carrier molecules
(e.g., Fe–transferrin), nonspecifically bound (e.g., to albumin), in the
form of a wear particle, or as a precipitate. Any of these forms may
be extracellular or intracellular.

This pattern observed by Lux and Zeisler (1974) reflects the end stage of
an equilibrium between the implant and the tissue. In a later study utilizing
a rabbit implant model, Lux et al. (1976) reported a nonspecific decrease in

* One of the difficulties in understanding such results arises in distinguishing among corrosion
products in solution, local precipitates, and insoluble particles, whether intra- or extracellular. If
the local composition matches the composition of the implant, then the evidence is fairly reliable
that the material is present in particulate form (as Michel, 1987, suggests), even if not resolvable
by light microscopy. However, the reverse is not true because selective corrosion may change the
apparent composition of small particles.
300 Biological Performance of Materials: Fundamentals of Biocompatibility

transport rate with time for all detectable corrosion products; this was
ascribed to maturation of the fibrous capsule about the implant. Further-
more, they showed an inverse correlation between tissue iron and zinc
content, even though zinc was not released by the implant. These findings
underline the great complexity of consideration of release and distribution
of implant degradation products, even in the near vicinity of the implant.
In a traditional sense, as would be observed in an in vitro corrosion study,
equilibrium would nevertheless be reached when an equilibrium concentra-
tion of ions was reached in the surrounding fluid. However, in vivo, the
bathing medium is dynamic, and a fractional excretion process is always in
competition with corrosion processes as they approach equilibrium.

15.3.2.2 Distribution of Body Water

If maximum implant corrosion rates (as determined by formation of soluble
species) can be measured, then equilibrium times should depend only upon
the volume of the medium and the fractional excretion rates. The medium
under discussion, water, has the relative volumes and distribution in the
human body as shown in Figure 15.3.
If no input/output of fluid occurs, a volume of 42 l would be under con-
sideration. Similarly, if no compartmental exchange took place, smaller vol-
umes of fluid might be considered — as small as 3.2 l in the case of an implant
bathed in blood. However, the situation is dynamic, as shown in Figure 15.4.
In this figure, note that some water entering the gastrointestinal tract never
exchanges directly with the internal water compartments but passes straight
through as a portion of fecal excretion (dashed line to right of plasma pool).
The 1.3 l/day excretion identified as “other” includes this volume as well as
sweat (lost from the fast interstitial pool) and the water content of sloughed
or desquammated cells (lost from the intracellular pool).

BY COMPARTMENT BY TISSUE

NON-WATER
COMPONENTS

EXTRA- MUSCLE 22.1 (52.6%)

CELLULAR 15.7 (37.3%)

PLASMA 3.2 (7.6%)

INTRA- SKIN 9.1 (21.6%)

23.1 (55.1%)
CELLULAR BLOOD 4.7 (11.1%)
OTHER 6.1 (14.7%)
TOTAL BODY WATER: 42 LITERS
60% OF BODY WEIGHT

FIGURE 15.3
Water content of the human body (70 kg).
Systemic Distribution and Excretion 301

2.5 liters/24 hr

SLOW INTERSTITIAL
7.3 liters

Implant

FAST INTERSTITIAL PLASMA

0.3 liters/24 hr 8.4 liters 3.2 liters
(metabolism)

INTRACELLULAR
23.1 liters

2.8 liters/24 hr
(1.5 urine,
1.3 other)

FIGURE 15.4
Daily water input/output/exchange for the human body (70 kg).

One can now see that volumes vary considerably depending upon the
details of consideration. An implant placed in soft or hard tissue is directly
in contact with a fast interstitial pool of 8.4 l, as shown. This contacts slower
exchange pools of 7.3 l (interstitial) and 23.1 l (intracellular). It also contacts
a fast exchange pool of 3.2 l (plasma water). This pool passes through the
kidneys at a rate of 180 l/24 h and experiences a net throughput (intake =
output) of 2.5 to 2.8 l/24 h. Thus, on an annual basis, an implant is in contact
with a pool size exceeding 1000 l. The point of equilibrium of corrosion or
leached products depends on the overall kinetics of release from the implant,
exchange, and excretion. These kinetics are difficult to analyze. The previous
chapter considered the details of one system, the iron system, in some depth.

15.4 Distribution and Excretion of Dissolved Species

15.4.1 Metallic Ion Distribution
However, some of the elements of these distribution systems can be studied
in more detail. One factor to examine is urinary excretion, which is probably
the major route for clearance from the body of metals released from implants,
although biliary excretion may also play a role for some metals (Brauer 1959).
Of the metals of interest as components of implants, only iron is known to
be excreted (incorporated in porphyrins, which are degradation products
of hemoglobin) predominantly through a biliary route (Ishihara and
302 Biological Performance of Materials: Fundamentals of Biocompatibility

Matsushiro 1986). Biliary excretion is mediated by concentration in and/or

excretion by the liver of low molecular weight (<10,000) organometallic
species known collectively as metallothioneins (Cherian and Goyer 1978;
Klaassen 1976). These are released through the bile duct, stored in the gall
bladder, and released as bile at a rate of 0.4 to 1.2 l/day into the jejunum. A
significant proportion of the ionic content of bile, perhaps as much as 95%,
is reabsorbed in the ileum, with only 200 mg/day excreted. Biliary excretion
serves primarily to aid in digestion of fats and metal excretion appears to
be a secondary role (Ganong 1989).
Urine, on the other hand, is formed primarily to regulate plasma compo-
sition and pH. It is formed in the mammalian kidney by a combination of
three processes: glomerular filtration, tubular reabsorption, and tubular
secretion. In the glomeruli, all formed cellular elements of blood and any
molecules with a molecular weight above approximately 5000 are retained
while all small molecules and a considerable amount of water are filtered
into the proximal renal tubules. In a 70-kg person, between 115 and 125 ml/
min of plasma are filtered. Normally nearly all of the water is also filtered
out. If this were excreted, it would result in a 24-h urine volume of as much
as 180 l.
This does not happen because over 99% of the water is reabsorbed in the
proximal renal tubules, resulting in a nominal rate of urine production near
1 ml/min. In addition, other materials such as the physiological metal ions
(Na+, K+, etc.), essential anions (Cl–, OH–, etc.) and organic molecules (glu-
cose, urea, etc.) are reabsorbed. Some reabsorption processes are passive but
others are active (energy requiring). As shown in the left portion of Figure
15.5, each active reabsorption process — for example, for glucose — has an
asymptotic limit (different for each reabsorbed species) termed the tubular
reabsorption limit (Tm). This has the effect of limiting and controlling the
maximum concentration of that particular species in plasma.

REABSORPTION SECRETION

ed
ret
Clearance Rate

Clearance Rate

c
Ex
ed
ed

ter
Fil
ter

Reabsorbed Tm
Fil

Secreted Tm
ted
cre
Ex

Plasma Concentration Plasma Concentration

FIGURE 15.5
Renal reabsorption and secretion. (Adapted from Pitts, R.F., in Physiology of the Kidney and Body
Fluids, An Introductory Text, Year Book Medical Pub. Inc., Chicago, 1963, 69, 116.)
Systemic Distribution and Excretion 303

TABLE 15.2
Urinary Secretion of Implant Alloy Componentsa
Plasma Conc. Urine Conc. Permeability Excretion
Element (μ
μg/l) (μ
μg/l) Ratio (Kx) Ratio (Ex)
Al 2.2 6.4 2.9 0.040
Co 0.05 0.33 6.6 0.092
Cr 0.06 0.13 2.2 0.030
Ni 0.2 1.0 5.0 0.069
Ti 3.3 0.41 0.12 0.006
V 0.16 0.61 3.8 0.053
Reference: creatinine 10 mg/l 1.5 g/l 150 2.08
a These data differ considerably from those in the first edition due to significant
improvements in collection and analysis techniques. However, it is widely believed
that urine collection, especially from female subjects, is subject to contamination.
Therefore, Kx and Ex may be too high; values should be considered in relative rather
than absolute terms.
b Reconstructed, assuming 1.5 g creatinine/liter.
Sources: Al, Ti, V: Jacobs, J.J. et al., J. Bone Joint Surg., 73A, 1475, 1991; Co, Cr, Nib:
Sunderman, F.W., Jr. et al., J. Orthop. Res., 7, 307, 1989; creatinine: Ganong, W.F., in
Review of Medical Physiology, 14th ed., Appleton & Lange, Norwalk, CT, 1989, 593.

Finally, materials may be removed from plasma by secretion through the

walls of the distal renal tubules. This process may be passive or active; in
the latter case (right portion of Figure 15.5), the effect is also to demonstrate
a limit. Metals in plasma are bound specifically to transferrin or other carrier
proteins (nickeloplasm, etc.) or nonspecifically to albumin, all of which have
too high a molecular weight to be filtered in the glomeri. Therefore, urinary
excretion of essential and trace metals must be primarily through a tubular
secretion pathway.* Because this represents an interaction between post-
glomerular plasma and distal tubule urine, it is instructive to look at the
relative concentrations of metal in these fluids (Table 15.2).
The permeability ratio, Kx, is the ratio of concentration in urine to that in
plasma. Values greater than 1 suggest positive secretion, values near 1 indi-
cate simple equilibrium between urine and plasma, and values less than 1
suggest a barrier to secretion. The permeability ratio times the relative vol-
umes of urine and plasma (= 0.78) is the proportion of normal plasma content
secreted in 24 h. The excretion ratio,** Ex, based upon a 24-h urine volume of
2.5 l and a renal filtration of 180 l/24 h, reflects the efficiency of secretion or
the proportion of the renal filtrate secreted in 24 h. The higher the value of
Ex is, the greater is the probability of secretion of an ion in any one pass
through the kidneys. The clearance rate, Cx, of a substance, x, is given by:

* However, some metal–protein dissociation may occur, leading to direct excretion; tubular reab-
sorption is also possible (Araki et al. 1986a).
** The term “excretion ratio” is used irrespective of renal mechanism involved.
304 Biological Performance of Materials: Fundamentals of Biocompatibility

Cx = KxV (ml/min) (15.5)

where V = rate of urinary output.

Creatinine is a degradation product released by cell death and is widely
used as a concentration marker for studying ionic concentration in urine.
Urine may be more or less concentrated, depending on fluid intake, perspi-
ration loss, etc., but the serum concentration and amount of urinary creati-
nine excretion in 24 h remains remarkably constant in normal individuals
(Ganong 1989; Araki et al. 1986b).

15.4.2 Distribution Models: One Compartment

Projections of the equilibrium accumulations of metals in the body have been
made using this type of data (Taylor 1973). Taylor proposed that, for a given
rate of continuous release of corrosion products, R, per day, the increase in
an organ or the total body content of a metal can be given as

R
Qt = Qo +
k
(
1 − e − kt ) (15.6)

where
Qo = the normal metal content
Qt = the content after “t” days
k = the fractional rate of excretion of the metal

That is, in a given day, R metal is released and kQt is excreted. Letting t go
to infinity (when equilibrium is presumably achieved), one then finds that

R
Qe = Qo + (15.7)
k

where Qe is the equilibrium metal content. Note that this is a first-order analysis
that treats the body as a single homogeneous compartment. For a hypothetical
cobalt–chromium implant (60% Co, 30% Cr, 8% Mo, 1% Ni, 1% Fe) with a
surface area of 200 cm2 and a corrosion rate of 30 mg/cm2/day, Taylor obtains
the results given in Table 15.3. Taylor concluded that modest elevations of
cobalt and nickel and large elevations of chromium content should occur in
this case. Similar calculations for implantation of stainless steel predict a mod-
est elevation of nickel and a large increase in chromium content.
Taylor’s calculations are probably in error on two fundamental grounds.
Although admittedly using high corrosion rates, he does not point out that
these rates are high by at least an order of magnitude. Second, the fractional
excretion rates used are based upon urine/plasma concentration ratios and
do not take into account exchange with slower compartments, particularly
situations in which precipitated storage is possible, as in the liver. Reducing
Systemic Distribution and Excretion 305

TABLE 15.3
Secretion and Accumulation Rates of Alloy
Components of a Cobalt–Chromium Alloy
Qo k R Qe
Element (mg) (day–1) (mg/day) (mg) Qe/Qo
Co 3 0.07 3.6 54 18
Cr 6 0.0011 1.8 1636 273
Mo 5 0.139 0.48 8 1.7
Fe 4000 0.0010 0.06 4060 1.02
Ni 10 0.0010 0.06 70 7

the corrosion rate by an order of magnitude and, continuing to use his

assumption, calculating fractional excretion rates based upon whole-body
content (Qo) and urine concentrations (Sunderman et al. 1989) for Co, Cr,
and Ni yields the results given in Table 15.4. The very large Qe/Qo ratios for
Co and Cr suggest that concentrations of these elements would never come
to equilibrium but should be observed to increase steadily with time postim-
plantation. Such effects have been reported in animals (Woodman et al. 1983)
and humans (Michel et al. 1991) for long periods of time.
The fundamental error in all of these considerations is the idea that the
metals distribute freely and do not concentrate or bind preferentially in any
site. In perhaps the first attempt to study systemic distribution of metals
released from implants in animals, Ferguson and his coworkers (1962a, b)
observed the following patterns:

• A large regional variation of metal ion concentration can be found

in normal rabbit and human tissue. Furthermore, different metals
have different patterns. Thus, the nickel concentration is higher in
the liver than in other tissues, and the molybdenum concentration
is higher in liver and kidney than in lung or spleen.
• After implantation of metals, organ concentrations of ions rise. The
spleen has a broad ability to retain metals; nickel and cobalt are
preferentially retained by the kidney.

TABLE 15.4
Recalculation of Secretion and Accumulation Rates of Alloy
Components of Cobalt–Chromium Alloy of Table 15.3
Qo Secretion R Qe
Element (mg) (24 h, μg) k (day–1) (mg/day) (mg) Qe/Qo
Co 3 0.82 0.00027 0.36 1336 445
Cr 6 0.32 0.00005 0.18 3606 601
Ni 10 2.5 0.00025 0.006 34 3.4
306 Biological Performance of Materials: Fundamentals of Biocompatibility

Continuing studies of accumulation and distribution of corrosion products

released by implants up to and including autopsy studies on patients with
long-term implants (Michel et al. 1991) support these observations.

15.4.3 Distribution Models: Multicompartment

For technical and ethical reasons, it is extremely difficult to perform whole-
body studies of metal metabolism in humans. However, a methodology has
been developed that involves study of the distribution and excretion of a single
dose of radioactive metal ions administered intravenously in animals. It is
possible that this method may be applied selectively to humans in the future.
Although the technique was originally developed by Sunderman at the
University of Connecticut, Greene et al. (1975) provide the best early report.
An experimental animal, such as the rat or rabbit, is used as a model. A
single intravenous injection of nickel as 63NiCl2 is given at a dose of 0.24
mg/Ni/kg body weight. The animals are housed in metabolic cages so that
urine and feces can be collected, and serum specimens are obtained period-
ically over a period of time.
For nickel in the rabbit, an expression for the concentration of nickel in
serum takes the form of:

( )
S μg / liter = A1e − a1t + A2 e − a2t (15.8)

The first term is large, corresponding to an early rapid disappearance rate,

and the second term is small, corresponding to a reduced disappearance rate
from 3 to 7 days after injection. This can be interpreted in terms of a large,
fast exchange compartment (the intercellular space) exchanging with a
smaller, slower compartment of undetermined identity. The four constants
assume different values for rats and rabbits.
Greene et al. (1975) made the following extrapolations:

From what is known about nickel corrosion, one can estimate that in
humans who have implants made of a nickel-containing alloy, the rate
of nickel release from the device can range between 5 and 500 mg/year
per individual. This corresponds to a range of 0.81 to 0.0081 μg/h per
kg body weight on the basis of 70 kg for humans. The following table
gives the estimated steady state values of nickel concentration in plasma
resulting from three different input rates within that range.

Steady State Ni Concentration

Infusion Rate (μ
μg/l)
(μ
μg/h per kg of Body Weight) Rabbit Rat
0.81 45 19.9
0.081 4.5 1.99
0.0081 0.45 0.20
Systemic Distribution and Excretion 307

These figures have to be compared with the normal range of nickel

concentration in human plasma…2.6* ± 0.08 μg/l.

Taylor’s calculations (Taylor 1973) for an alloy that released 22 mg per year
predict a 7× increase (3.4× corrected calculation based upon his secretion
data) in total body burden of nickel. Because Greene made no assumption
on the partitioning of metallic ions between various compartments, the
elevation that he predicts for plasma must be taken as the total body eleva-
tion. For a 2 mg per year release rate, Greene would predict a 1.76× increase
(by rabbit data) or a 1.34× increase (by rat data). These are somewhat smaller
than Taylor’s original calculations (but near to the corrected calculations of
Table 15.3); however, they lend credence to the idea that net concentrations
of metal ions will rise in the presence of an endogenous source of ions, such
as a corroding implant.
Sunderman’s approach has a number of difficulties, including the partition
question and the inability to identify internal compartments except by infer-
ence from calculations. This work has been extended and is reported by
Onkelinx (1977). He has developed a more general multicompartment
model, as shown in Figure 15.6. Here, V1 represents the intracellular com-
partment, and V2 and V3 represent other compartments with net (reversible)
interchange flow rates f2 and f3, respectively. Again, the assumption of a
partition coefficient of 1 is made so that ion fluxes can be represented as
fluid flow rates at constant concentration. Excretion is represented by a flow

INJECTION

f3 f2
V3 V1 V2

Ft
S
fs

fd fu

FIGURE 15.6
General multicompartment model for metal metabolism. (Adapted from Onkelinx, C., in Clinical
Chemistry and Chemical Toxicology of Metals, Brown, S.S. (Ed.), Elsevier North–Holland, Amster-
dam, 1977, 37.)

* Note that this is significantly higher than typical modern values; see, for instance, Table 15.2.
The probably lower true value makes the predicted increases even more striking.
308 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 15.5
Comparison among Compartment Volumes,
Interchange, and Excretion for Ni, Co, and Cr in
Rats
63 Ni2+ 57 Co2+ Cr3+
51

Age (days) 85 60 60
Compartment volumes
(ml/100 g body wt.)
V1 36.1 46.4 30.8
V2 4.0 78.2 16.7
V3 — 65.6 7.7
Apparent excretory flow rates
(ml/h)
Ft 3.91 7.29 1.42
fu 3.07 5.97 0.91
fd 0.62 0.72 0.07
fs 0.22 0.60 0.44
Ratios of excretory flow rates
fu/Ft 0.79 0.82 0.64
fd/Ft 0.16 0.10 0.05
fs/Ft 0.06 0.08 0.31
Compartment net interchange
flow rates (ml/h)
f2 0.06 13.11 2.47
f3 — 0.56 0.13
Source: Adapted from Onkelinx, C., in Clinical Chemistry
and Chemical Toxicology of Metals, Brown, S.S. (Ed.), Elsevier
North–Holland, Amsterdam, 1977, 37.

rate, Ft, composed of collection in irreversible sinks (fs), urinary excretion

(fu), and fecal excretion (fd). Loss of tissue, desquammation, etc. are lumped
with fs.
Table 15.5 reports some results obtained by injection of nickel, cobalt, and
chromium into rats between 2 and 3 months old. Onkelinx (1977) reported
results for other ages, suggesting age dependence, especially for chromium.
Examination of this table graphically shows the differences in metal meta-
bolism for these ions. Although it is difficult to apply physical identities to
compartment volumes, the differences are obvious. Nickel does not penetrate
V3, as previously pointed out, and the relative penetration of each ion is
quite different. Of particular interest is the ratio fs/Ft. Here, modest amounts
of nickel and cobalt are trapped in a tissue sink, but over 30% of chromium
is retained. Again, with reference to Taylor (1973), if this material is bound
irreversibly, the effect would be to lower transient tissue concentrations and
lengthen the time until steady state is reached, but raise the final body burden
above that originally predicted.
It is inviting to equate secondary compartments (in this case, V1, V2, and
S), as derived from Onkelinx’ analysis with specific tissue types or organs,
on the basis of apparent volumes. This is simplistic; metabolic pools and
Systemic Distribution and Excretion 309

storage depots can only be identified by following the course of metal, in

the form of a suitable radioactive isotope, from the implant site to its eventual
destination. Unfortunately, such studies may not ethically be performed in
humans, so compartment identities may remain unclear for a long time
to come.

15.4.4 Equilibrium Models

Another experimental approach to this problem (Smith and Black 1977;
Smith 1982) is to implant metallic devices with varying surface areas and
track the plasma levels as a function of time. In these reports, an experiment
was performed in a short-term rabbit model to investigate whether implan-
tation of 316L stainless steel is accompanied by elevated plasma levels of
iron and chromium.
For a given alloy system with a uniform processing history, the parameter
that would appear to govern the rate of corrosion product delivery to the
body in any particular implant site is the ratio of implant surface area to
body weight (SA/BW). For a 70-kg patient receiving a typical total hip joint
replacement (surface area = 200 cm2), this ratio is approximately 2.86 cm2/
kg. This value is termed “1×” and multiples of it reflect higher relative
exposures. Note that the use of such a ratio probably underestimates the
exposure in small-animal models. Renal clearance is roughly proportional
to basal metabolic rate (per kilogram of body weight). Because individual
metabolic rate in adult mammals, including humans, is proportional to
(BW)0.734 (Brody 1945), a 1.5-kg rabbit has a basal metabolic rate/kg 2.78
times that of a 70-kg patient. Thus, 100× for the patient is approximately 36×
for the rabbit, if adjusted to reflect basal metabolic rates, or, conversely, 100×
for the rabbit is 278× for the patient. The concept of SA/BW ratio can be
extended to in vitro studies by recognizing that a 70-kg patient has a water
content of 42 l (Figure 15.4), forming the ratio of surface area to body fluid
volume (SA/BFV) and assigning a value of 4.76 cm2/l.
In Smith and Black’s (1977) studies, New Zealand white rabbits received
implants of passivated surgical grade 316L stainless steel (ASTM F 55, type
B) in two forms (Steinmann pin segments and 40-μm spherical powders) at
two anatomic sites (paraspinal musculature and femoral medullary canal)
as shown in Table 15.6. Blood specimens were obtained at intervals of up to
7 months, and tissue specimens were obtained at sacrifice. The serum was
analyzed for chromium, free (nonheme) iron (PI), total iron-binding capacity
(TIBC), and percent of TIBC saturation (% sat.). The tissues were analyzed
for iron and chromium.
No significant differences (experimental vs. control) were detected in PI,
TIBC, or % sat. until 20 weeks postoperative. Group V plasma iron concen-
tration was 14% elevated (p < 0.05) by 20 weeks postoperative. None of the
groups exhibited significant changes in TIBC or % sat.; however, the latter
showed a “tendency” toward elevation, particularly in Group V. The kidney
310 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 15.6
Animal Groups for Evaluation of SA/BW Effects
SA/BW
Group N= Implant Site (Unadjusted)
I 10 Pin Muscle 1
II 10 Pin Bone 1
III 15 Microspheres Bone 1
IV 10 Microspheres Bone 10
V 15 Microspheres Bone 100
VI 14 None — 0

showed an ability to accumulate iron; however, Group V liver iron concen-

trations were elevated 30% on dry weight and total protein basis.
The results for chromium are shown in Table 15.7. Plasma iron elevations
in Group V at 20 and 28 weeks indicated that iron release (corrosion into
the blood circulation) exceeded the inherent iron turnover rate of the
Fe–transferring-binding system. The consequent trend toward elevation in
% sat., if observed in humans, might occasion a reduction in a patient’s
disease resistance (see Section 14.2.8). Plasma iron and, especially, plasma
chromium concentrations appear to reflect the duration of implantation and
the SA/BW ratio. Sporadic elevations in plasma chromium in Groups I
through IV toward the last 2 months of the study may suggest periods of

TABLE 15.7
Serum and Tissue Chromium Content after Stainless Steel Implantation as a
Function of Implantation Time and SA/BW Ratio in Rabbit
Control
Sample (Group VI) % Change from Control (Significance)a
(ng/ml) Group I Group II Group III Group IV Group V

Serum

Pre-op 13.1 ± 0.9 –11.4 (p < 0.05)

4 wk PO 11.4 ± 0.4 +16.7 (p < 0.001)
8 wk PO 12.9 ± 0.1 Lost group +12.8 (p < 0.08)
20 wk PO 11.2 ± 0.4 +6.2 (p < 0.05) +21.4 (p < 0.01)
24 wk PO 10.3 ± 0.7 +15.5 (p < 0.01) +17.5 (p < 0.01) +24.3 (p < 0.01) +23.3 (p < 0.01)
28 wk PO 15.3 ± 0.7 +17 (p < 0.01) +10.4 (p < 0.01)

Tissue (ng/mg dry wt.)

Kidney 1.97 ± 0.5

Liver 1.07 ± 0.7 +64.2 (p < 0.02) +85 (p < 0.01)
a Only significant changes are shown.
Source: Adapted from Smith, G.K., Ph.D. thesis, University of Pennsylvania, Philadelphia, 1982.
Systemic Distribution and Excretion 311

accelerated corrosion or release into the circulation. Because plasma chro-

mium concentrations are probably not in equilibrium with storage compart-
ments, these chromium elevations are not likely to be an indicator of body
stores but rather a dynamic measure of serum transport at the moment of
sampling. The kidney demonstrated no capacity to accumulate iron
or chromium.
On the other hand, the liver exhibited elevated iron and chromium accu-
mulations, which were apparently a function of SA/BW ratio, in contrast to
Ferguson’s study (Ferguson et al. 1962a, b), which reported no elevations in
iron or chromium in the liver or kidney. From persistently elevated plasma
iron and chromium concentrations, it can be expected that liver accumula-
tions would have continued beyond the experimental period and that, with
time, groups I through III may have exhibited elevations.
One of the weaknesses of this approach is that it is impossible to distin-
guish the metal released from the implant from that available from dietary
sources or storage depots. The presence of the implant may alter the binding
and storage mechanisms that normally handle metal (Woodman et al. 1983).
An alternate approach would be to use radioactive isotopes as injected
materials (Bergman et al. 1980) or, preferably, as components of implant
alloys, followed by periodic sacrifice and autoradiography. However, the
rarity of appropriate isotopes, difficulties in fabricating suitable implants,
and safety concerns about handling the animals and their waste products
severely handicaps this approach.
It is clear from these equilibrium studies that trace metal metabolism is
highly complex and that detailed studies of each metal of implant impor-
tance should be carried out in the future.

15.5 Final Comment

Systemic distribution, storage, and excretion of corrosion and degradation
products from implants has not attracted great attention as a subfield of study
within biomaterials science and engineering. Initially, this was due to an inabil-
ity to detect normal or elevated levels of such materials, especially metal-
bearing ions, and a parallel failure to recognize their biological importance.
More recently, studies of host response have continued largely to focus on
circumimplant effects and have continued to equate low corrosion/elution
rates and small plasma concentration elevations with absence of accumulation
of degradation products and, thus, absence of biological effects. The weakness
of this assumption may be illustrated by the following argument.
A proposal may be to establish how many cars are in a particular turnpike
rest area by only examining the flow of traffic on the roadway. First consid-
erations suggest a possible positive correlation between the number of cars
passing a marker nearby on the main roadway and the number to be found
312 Biological Performance of Materials: Fundamentals of Biocompatibility

in the rest area. However, the rest area might be closed for repairs (large
number on roadway, none in rest area) or the driving conditions may have
recently become treacherous, perhaps due to icing, causing drivers to decide
to delay their further travel (small number on roadway, large number in rest
area). Thus, on second consideration, the conclusion is that the only way
reliably to determine the number of cars in a turnpike rest area at any one
time is to count them. The situation for detection of debris and corrosion
products in remote tissues and organs is parallel by analogy.
What is required is the recognition that any implant, whether designed
specifically for that purpose or not, acts as a slow-release system in vivo.
Thus, the appropriate approach to understanding the total host response to
implants must parallel that of biological fate studies in pharmacology and
environmental fate studies in ecology.
From this viewpoint, three important questions should be answered when
a new implantable biomaterial is evaluated:

• What is the nature of the degradation products released from the

implant in vivo?
• Where do they go within the body?
• What is the host response to release, distribution, remote concentra-
tion, and excretion of these degradation products?

These are extremely difficult questions to answer. Although there are the-
oretical bases for the answers to the first and second questions, the third
remains conjectural until actual human clinical data can be obtained. This
can only be done within the context of carefully controlled prospective
studies in well identified cohorts of patients with implants and suitable
controls. Jacobs et al. (1991) offer an example of this approach in a study
that has now extended some 15 years and is gradually being correlated with
studies of retrieved tissues and fluids as patients reach the end of life.
Previously, some aspects of local host response have been considered. The
next chapter will take up the more global issue of systemic and remote site
host response produced by the transport phenomena discussed in this chapter.

References
Aalders, G.J. et al., An exceptional case of pneumothorax — “a new adventure of the
K wire,” Injury, 16, 564, 1985.
Albores–Saavedra, J. et al., Sinus histiocytosis of pelvic lymph nodes after hip re-
placement. A histiocytic proliferation induced by cobalt–chromium and titani-
um, Am. J. Surg. Pathol., 18, 83, 1994.
Araki, S. et al., Filterable plasma concentration, glomerular filtration, tubular balance,
and renal clearance of heavy metals and organic substances in metal workers,
Arch. Environ. Health, 41, 216, 1986a.
Systemic Distribution and Excretion 313

Araki, S. et al., Comparison of the effects of urinary flow on adjusted and nonadjusted
excretion of heavy metals and organic substances in “healthy” men, J. Appl.
Toxicol., 6, 245, 1986b.
Arvidson, K. and Wróblewski, R., Migration of metallic ions from screwposts into
dentin and surrounding tissues, Scand. J. Dent. Res., 83, 200, 1978.
Bergman, B. et al., The distribution of nickel in mice, an autoradiographic study, J.
Oral Rehabil., 7, 319, 1980.
Brauer, R.W., Mechanisms of bile secretion, JAMA, 169, 1462, 1959.
Brody, S., Bioenergetics and Growth, with Special Reference to the Efficiency Complex in
Domestic Animals, Reinhold, New York, 1945, 352.
Case, C.P. et al., Widespread dissemination of metal debris from implants, J. Bone
Joint Surg., 76B, 701, 1994.
Charnley, J., Arthroplasty of the hip. A new operation, Lancet, 1, 1129, 1961.
Cherian, M.G. and Goyer, R.A., Minireview, metallothioneins and their role in the
metabolism and toxicity of metals, Life Sci., 23, 1, 1978.
Ferguson, A.B. et al., Trace metal ion concentration in the liver, kidney, spleen, and
lung of normal rabbits, J. Bone Joint Surg., 44A, 317, 1962a.
Ferguson, A.B. et al., Characteristics of trace ions released from embedded metal
implants in the rabbit, J. Bone Joint Surg., 44A, 323, 1962b.
Ganong, W.F., Renal function and micturation, in Review of Medical Physiology, 14th
ed., Appleton & Lange, Norwalk, CT, 1989, 593.
Gatti, A.M. and Rivasi, F., Biocompatibility of micro- and nanoparticles. Part I: in
liver and kidney, Biomaterials 23(11), 2381, 2002.
Graham, J.A. et al., Effect of trace metals on phagocytosis by alveolar macrophages,
Infect. Immunol., 11, 1278, 1975.
Greene, N.D. et al., Engineering and biological studies of metallic implant materials,
in Biomaterials, Horowitz, E. and Torgesen, J.L. (Eds.), NBS Special Publication
415, U.S. Government Printing Office, Washington, D.C., 1975, 45.
Hierholzer, S. et al., Increased corrosion of stainless steel implants in infected plated
fractures, Arch. Orthop. Trauma Surg., 102, 198, 1984.
Homsy, C.A. et al., Some physiological aspects of prosthesis stabilization with acrylic
polymer, Clin. Orthop. Rel. Res., 83, 317, 1972.
Ishihara, N. and Matsushiro, T., Biliary and urinary excretion of metals in humans,
Arch. Environ. Health, 41, 324, 1986.
Jacobs, J.J. et al., Local and distant products from modularity, Clin. Orthop. Rel. Res.,
319, 94, 1995.
Jacobs, J.J. et al., Release and excretion of metal in patients with titanium-base alloy
total hip replacement components, J. Bone Joint Surg., 73A, 1475, 1991.
Jenkin, C.R. and Rowley, D., The role of opsonins in the clearance of living and inert
particles by cells of the reticuloendothelial system, J. Exp. Med., 114, 363, 1961.
Kapp, J.P. et al., Metallic fragment embolization to the cerebral circulation, J. Trauma,
13, 256, 1973.
Kapur R. et al., Human monocyte morphology is affected by local substrate charge
heterogeneity, J. Biomed. Mater. Res., 32, 133, 1996.
Kawaguchi, H. et al., Phagocytosis of latex particles by leukocytes, I, Dependence of
phagocytosis on the size and surface potential of particles, Biomaertials, 7, 61,
1986.
Klaassen, C.D., Biliary excretion of metals, Drug Metab. Rev., 5, 165, 1976.
Korn, E.D. and Weisman, R.A., Phagocytosis of latex beads by acanthamoeba. II.
Electron microscopic study of the initial events, J. Cell. Biol., 34, 219, 1967.
314 Biological Performance of Materials: Fundamentals of Biocompatibility

Lux, F. and Zeisler, R., Investigations of the corrosive deposition of components of

metal implants and of the behavior of biological trace elements in metallosis
tissue by means of instrumental multi-element activation analysis, J. Radioanal.
Chem., 19, 289, 1974.
Lux, F. et al., A mechanistic model for the metabolism of corrosion products and of
biological trace elements in metallosis tissue based on results obtained by
activation analysis, J. Radioanal. Chem., 32, 229, 1976.
Lyons, F.A. and Rockwood, C.A., Current concepts review. Migration of pins used
in operations on the shoulder, J. Bone Joint Surg., 72A, 1262, 1990.
Margevicius, K.J. et al., Identification and distribution of synthetic ligament wear
particles in sheep, J. Biomed. Mater. Res., 31, 319, 1996.
Michel, R., Trace metal analysis in biocompatibility testing, CRC Crit. Rev. Biocompat.,
3, 235, 1987.
Michel, R. et al., Systemic effects of implanted prostheses made of cobalt–chromium
alloys, Arch. Orthop. Trauma Surg., 110, 61, 1991.
Normann, S.J., Kinetics of phagocytosis. II. Analysis of in vivo clearance with dem-
onstration of competitive inhibition between similar and dissimilar foreign
particles, Lab. Invest., 31, 161, 1974.
Onkelinx, C., Whole-body kinetics of metal salts in rats, in Clinical Chemistry and
Chemical Toxicology of Metals, Brown, S.S. (Ed.), Elsevier North–Holland, Am-
sterdam, 1977, 37.
Oron, U. and Alter, A., Corrosion in metal implants embedded in various locations
of the body in rats, Clin. Orthop. Rel. Res., 185, 295, 1984.
Pitts, R.F., Tubular reabsorption; tubular secretion, in Physiology of the Kidney and Body
Fluids, An Introductory Text, Year Book Medical Pub. Inc., Chicago, 1963, 69, 116.
Potter, F.A. et al., The migration of a Kirschner wire from shoulder to spleen: a brief
report, J. Bone Joint Surg., 70B, 326, 1988.
Schnyder, J. and Baggiolini, M., Role of phagocytosis in the aviation of macrophages,
J. Exp. Med., 148(6), 1449, 1978.
Smith, G.K., Systemic transport and distribution of iron and chromium from 316l
stainless steel implants, Ph.D. thesis, University of Pennsylvania, Philadelphia,
1982.
Smith, G.K. and Black, J., Elevation of Fe and Cr concentrations in blood plasma after
stainless steel implantation, Trans. ORS, 2, 281, 1977.
Stensaas, S.S. and Stensaas, L.J., Histopathological evaluation of materials implanted
in the cerebral cortex, Acta. Neuropath. (Berl.), 41, 145, 1978.
Stiffel, C. et al., Kinetics of the phagocytic function of reticuloendothelial macroph-
ages in vivo, in Mononuclear Macrophages, van Furth, R. (Ed.), Academic Press,
New York, 1970, 335.
Styles, J.A. and Wilson, J., Comparison between in vitro toxicity of two novel fibrous
mineral dusts and their tissue reactions in vivo, Ann. Occup. Hyg., 19, 63, 1976.
Sunderman, F.W., Jr. et al., Cobalt, chromium, and nickel concentrations in body
fluids of patients with porous-coated knee or hip prostheses, J. Orthop. Res., 7,
307, 1989.
Taylor, D.M., Trace metal patterns and disease, J. Bone Joint Surg., 55B, 422, 1973.
Urban, R.M. et al., Accumulation in liver and spleen of metal particles generated at
nonbearing surfaces in hip arthroplasty, J. Arthroplasty, 19 (8 Suppl. 3), 94, 2004.
Vernon–Roberts, B., The kinetics of phagocytosis stimulation and depression of the
phagocytic activity of macrophages, in Vernon–Roberts, B., The Macrophage,
Cambridge University Press, Cambridge, 1972, 92.
Systemic Distribution and Excretion 315

Weisman, R.A. and Korn, E.D., Phagocytosis of latex beads by acanthamoeba I. Bio-
chemical properties, Biochemistry, 6, 485, 1967.
Woodman, J.L. et al., Release of cobalt and nickel from a new total finger joint
prosthesis made of vitallium, J. Biomed. Mater. Res., 17, 655, 1983.

Bibliography
Anderson, J.M. and Miller, K.M., Biomaterial biocompatibility and the macrophage,
Biomaterials, 5, 5, 1984.
Friedman, M.H., Principles and Models of Biological Transport, Springer–Verlag, Berlin,
1986.
Guyton, A.C., The kidneys and body fluids, in Textbook of Medical Physiology, 8th ed.,
Guyton, A.C. (Ed.), W.B. Saunders, Philadelphia, 1991, 273.
Harrison, P.M. and Treffry, A., Storage and transport of transition-metal ions, in
Inorganic Biochemistry, A Review of the Recent Literature Published up to Late 1977,
Vol. 1, H.A.O. Hill (Ed.), The Chemical Society, London, 1979, 120.
Ingham, E. and Fisher, J., The role of macrophages in osteolysis of total joint replace-
ment, Biomaterials, 26, 1271, 2005.
Iyengar, G.V., Review: reference values for the concentrations of As, Cd, Co, Cr, Cu,
Fe, I, Hg, Mn, MO, Ni, Pb, Se, and Zn on selected human tissues and body
fluids, Biol. Trace Elem. Res., 12, 263, 1978.
Koushanpour, E. and Kriz, W., Renal Physiology, Principles, Structure, and Functions,
2nd ed., Springer–Verlag, New York, 1986.
Langkamer, V.G. et al., Systemic distribution of wear debris after hip replacement.
A cause for concern? J. Bone Joint Surg., 74B, 831, 1992.
Lote, C.J., Principles of Renal Physiology, 4th ed., Chapman & Hall, London, 2000.
Pitts, R.F., Physiology of the Kidney and Body Fluids, An Introductory Text, 3rd ed., Year
Book Medical Pub. Inc., Chicago, 1974.
Roberts, J. and Quastel, J.H., Particle uptake by polymorphonuclear leucocytes and
Erlich ascites-carcinoma cells, Biochem. J., 89, 150, 1963.
da Silva, J.J.R.F. and Williams, R.J.P., The Biological Chemistry of the Elements, 2nd ed.,
Clarendon Press, Oxford, 2001.
Solomon, A.K., Compartmental methods of kinetic analysis, in Mineral Metabolism,
Vol. 1A, Comar, C.L. and Bronner, F. (Eds.), Academic Press, New York, 1960,
119.
Valtin, H. and Schafer, J.A., Renal Function, 3rd ed., Little, Brown and Co., Boston,
1994.
16
Effects of Degradation Products on Remote
Organ Function

16.1 Introduction
This chapter will complete the discussion of the effects of materials on
biological systems (host response). With the exception of a number of rec-
ognized systemic effects, this chapter will simply recapitulate and emphasize
areas already discussed in Chapter 8 through Chapter 15.
Systemic effects of foreign materials, such as implants, are now well rec-
ognized. It is more correct to distinguish between remote effects (involving
actions, perhaps secondary to deposition and concentration of degradation
products) on a target tissue or organ and systemic effects (those affecting
large-scale systems such as the cardiovascular or neurological systems) How-
ever, for the sake of brevity, they will be grouped together here under the
common title of systemic effects.
The effects of drugs, collectively pharmacological effects, are by and large
necessarily systemic or remote effects. Drugs are given or injected in one
portion of the body and directly or indirectly affect cellular and systemic
physiology in other areas, even if a purely local or topical effect is intended.
Discussion of such effects is beyond the scope of this book; however, many
of the effects that have been discussed are essentially pharmacological effects
secondary to the primary or intended effect. Nonetheless, drug release mate-
rials provide a hypothetical example that illustrates the complexity of pos-
sible host response to an implant:

• The drug and/or carrier material is designed to evoke a local

(implant site) response, such as alleviation of pain or reduction of
fibrosis.
• The drug and/or degradation products from the carrier may evoke
desirable or undesirable systemic effects, such as alteration of arterial
pressure.
• The drug may have a specific remote organ target, such as the heart.

317
318 Biological Performance of Materials: Fundamentals of Biocompatibility

• The degradation products of the carrier might produce adverse

effects in other remote organs, such as the kidney.

16.2 Examples of Systemic Effects

16.2.1 Polymers
An example of these types of effects is that associated with release of methyl
methacrylate monomer during the polymerization of PMMA-type cements
in vivo. Homsy et al. (1972) (see Section 15.3.1) were attracted to this problem
not by the observation of the rapid transport of the monomer to the lungs,
but by the observed systemic effects. Early in the use of these cements, it
was observed that systemic hypotension developed immediately after the
insertion of the cement into the prepared bone cavity. Arterial pressure drops
of more than 15 mmHg were observed associated with transient cyanosis,
and a number of cardiac arrests drew medical attention to the effect (Keret
and Reis 1980). Further studies in dogs showed that this problem is accen-
tuated by marked fluid and/or blood loss (McMaster et al. 1974).
The corrective therapy now used to prevent the development of centrally
mediated hypotension after PMMA insertion is maintenance of the patient
in a state of positive hydration. In addition, it has been shown that venting
the bony (medullary) cavity with a drain line during device insertion reduces
the driving pressure differential that aids in monomer take-up by blood and
may reduce the embolization of fat to the lungs as encountered in intramed-
ullary fixation of fractures (Giannoudis et al. 2002). Polymers released from
implants have few if any known systemic effects beyond this hypotensive
effect. The rates of release and the normal routes of molecular catabolism
apparently combine to keep concentrations of products from common poly-
meric implants below levels required for direct pharmacological activity.
This is not to suggest that elevated blood levels of polymer degradation
products do not exist or that they may be harmless in the long term. In the
short term, host responses may be indistinguishable from disturbances of
homeostasis associated with surgery (Dahl 1997). However, in cases of high
surface area and/or repetitive exposure, as in hemodialysis (as a treatment
for renal insufficiency or failure), significant deviations from normal condi-
tions may be encountered. Lewis et al. (1977) examined the blood of indi-
viduals undergoing chronic hemodialysis, looking explicitly for breakdown
products or plasticizers that might be released from the polyvinyl chloride
(PVC) polymer used as a blood conduit material. Significant levels, up to a
mean level of 751 ng/ml serum of bis (2-ethylhexyl) phthalate, a common
PVC plasticizer, were found in these patients after dialysis. Although the
catabolism of this material was rapid and thus led to its being undetectable
in blood 5 to 6 hours after completion of dialysis, the yearly dose for the
Effects of Degradation Products on Remote Organ Function 319

chronic dialysis patient was estimated to be 150 to 250 mg. The long-term
effects of this are unknown, as is the case for the majority of long-term, low-
level exposures to foreign materials and their degradation products.
A final interesting example is provided by the early (now abandoned)
practice of using injected fluid silicone materials for cosmetic tissue augmen-
tation. In addition to producing local fibrosis, these fluids can migrate
through tissue and produce a variety of remote effects, including tissue mass
formation, adenopathy, and, possibly, pulmonary failure (Kossovsky and
Heggers 1987). However, broader claims of connective tissue and immune
disorders associated with silicone gel-filled breast augmentation devices
appear to have no firm basis (Gabriel 1998; Noone 1997).

16.2.2 Metals
The situation with metals is somewhat different. In addition to the “physi-
ological” metals (Ca, Na, K, and Fe) and despite very low normal plasma
and tissue concentrations, a large number of metals, including Co, Cr, Mg,
Zn, and Cu, have normal roles in metabolism and are thus classified as
essential trace elements. Therefore, it should come as no surprise that natu-
rally occurring diseases of inherited as well as acquired etiology involve
imbalances in the metabolism of these metals.
In addition to anemia (iron deficiency), iron overload diseases also exist.
One such disease, hemochromatosis (Elinder 1986), results in the accumula-
tion and deposit of iron in the form of the compound hemosiderin in tissues
with rough endoplasmic reticulae. This accumulation has a number of phys-
iological effects. One example is the development of diffuse arthritis in
widely separated joints. This also occurs secondarily to the internal bleeding
associated with hemophilia. Hemochromatosis involves skin pigmentation,
liver failure, and diabetes. Elinder (1986) points out that although iron is a
physiological element, it is potentially toxic in all doses and forms and is
capable of producing a variety of local and systemic toxic effects in animals
and humans.
A less well-known disease is Wilson’s disease (Aaseth and Norseth 1986),
or the so-called copper man syndrome. This is an accumulation disease,
primarily hereditary, in which copper accumulates in a variety of tissues,
including liver, cornea, and skin, instead of being maintained in balance. A
green skin color develops and a high incidence of mortality due to intravas-
cular hemolysis and liver failure results from the cytotoxicity of copper and
its compounds.
Metals that are normally foreign to the body, such as Pb, Be, and As, can
combine competitively with enzymes that normally use other trace metals
as cofactors. Even normally present metals, such as Al and Cr, may do so,
if present in sufficiently high concentrations. The abnormal cofactor–enzyme
combinations may have higher stability than the normal cofactor bonds.
Thus, the effect is to inactivate a portion of the enzyme pool without
320 Biological Performance of Materials: Fundamentals of Biocompatibility

stimulating additional enzyme production. In low concentrations, the net

effect will be to inhibit enzyme activity. This can be seen in a reduction of
the efficiency or effectiveness of an enzyme process. Examples of this are the
peripheral neuropathy observed in the case of long-term, low-level ingestion
of lead and the suppression of hemoglobin synthesis by chromium. At higher
concentration levels, metals can be highly toxic poisons through enzyme
inactivation as in the familiar case of arsenic.
Beyond these specific mechanisms, a wide variety of systemic medical
problems has been suggested as associated with imbalances in trace metal
levels. As noted in Section 14.4, care must be taken in dealing with this
literature because a less than totally scientific nutritional school of thought
ascribes virtually all unexplained physical and mental disabilities to such
effects. The mere detection of a foreign metal with adverse biological effects,
such as mercury, is insufficient to reach conclusions concerning possible
clinical problems; for example, recent studies associate autoimmunity in
patients with dental amalgam more with the silver content of the restoration
than with mercury or its compounds (Enestrom et al. 1995). There are, how-
ever, legitimately recognized associations, such as the increased incidence
of cardiomyopathy associated with elevated cobalt intake (Alexander 1972)
and the recognized association of arteriosclerosis with hardness (primarily
Ca content) of ground water (Perry 1973).*
Consideration of the effects of metals must address all aspects of physiol-
ogy. Perhaps one should be more observant of systemically detectable devi-
ations from mean values of clinical parameters, such as liver transaminase
serum concentrations in patients with chronic implants (Chopra 1988), even
when they remain within “normal” limits. Concentrations of metals well
below those needed to produce externally measurable changes in such phys-
iological variables can produce profound behavioral abnormalities and men-
tal disorders (Weiss 1978).

16.2.3 Ceramics
As noted in Section 4.11, ceramics used in implants may be insoluble or
soluble. Insoluble ceramic biomaterials, such as alumina, titania, zirconia,
etc., pose no systemic challenge, at least in nonparticulate form. To avoid
unwanted local site responses, resorbable ceramics, such as tricalcium phos-
phate, etc., are chosen so that their elemental cations and anions (including
Cl–, SO4–2, CO3–2, and PO4–3) lie primarily within the range of physiological
compositions. However, the possibility of soluble mineral components such
as Sr** in the latter generally argues for the use of fully defined (synthetic)

* As well as, presumably, the general levels of trace element intake.

** 90Sr, a radioactive element with a half-life of 29 years, was present as a natural impurity in zir-
conium and zirconia in their early use as biomaterials (Burger et al. 1997). Although they are now
fully removed from such materials, its presence presented an early example of an unwanted
source of systemic effects.
Effects of Degradation Products on Remote Organ Function 321

materials rather than those obtained from natural sources to avoid possible
systemic effects caused by their dissolution products.

16.3 A Review of Systemic Aspects of Host Response

Host responses will now be reviewed briefly in terms of systemic or remote
site effects.

16.3.1 Interaction of Molecules with Surfaces

Proteins and enzymes released from surfaces in an irreversibly denatured
state will possibly elicit remote effects directly or indirectly through the
action of the immune system. Depletion of pools of unactivated coagulation
or complement factors, as frequently occurs during hemodialysis or blood
oxygenation, may suppress coagulation effects at remote sites of injury.
Conversely, the increased circulating concentration of surface-activated fac-
tors may also have systemic or remote sequelae. It is also possible that
denatured molecules bound to wear debris subject to passive or active trans-
port can elicit systemic or remote effects.

16.3.2 Inflammation
Inflammation would be expected to be a local effect restricted to the
vicinity of the implant. It is possible that an implant can release pyrogenic
agents directly or produce them indirectly through denaturation pro-
cesses. Some evidence indicates that denatured molecules or released
products of unknown identity may produce long-term systemic hallmarks
of inflammation, as in the chronic erythrocyte sedimentation rate elevation
observed by Shih et al. (1987) in patients after PMMA-cemented total hip
replacement. In addition, friction, wear, and some dissolution processes
that release particulate material may result in an inflammatory response
at a site of remote accumulation. An example of this is the abdominal
“teflonoma” frequently seen after use of poly(tetrafluoro) ethylene as the
material for fabrication of acetabular cups in Charnley’s early efforts at
hip replacement (Charnley 1979). More subtle may be the effects associ-
ated with particulate transport and accumulation through venous or lym-
phatic return pathways (Langkamer et al. 1992). Precipitation/
redissolution of corrosion products in remote sites may also be expected
to produce inflammation due to response to the resulting particulate
material or to increased local concentrations of ions.
322 Biological Performance of Materials: Fundamentals of Biocompatibility

16.3.3 Coagulation and Hemolysis

Similarly, one can expect that primary coagulation problems would be local-
ized to the vicinity of a cardiovascular system implant or external blood
conduit or treatment device. The ability of implants to surface activate factors
in the coagulation cascade (as noted earlier), as well as to shed thrombi,
renders this a systemic effect with remote site manifestations. Similarly,
whether through blood–surface interactions, turbulent shear, or by direct
mechanical damage in valves, pumps, etc., hemolysis is a systemic problem
due to the reduction in viable erythrocytes and the rapid dispersion of
hemoglobin and cell fragments. It is clear that the current factors limiting
successful long-term left ventricular assist and total (artificial) heart replace-
ment are remote effects (primarily cerebral and pulmonary infarcts) second-
ary to systemic distribution of shed thrombi.

16.3.4 Adaptation
The discussion of adaptation emphasized the effects at the biomaterial–tis-
sue interface. This is certainly the most important adaptive remodeling
site, and it is expected that systemic or remote effects would be secondary
to this and thus not directly related (if they occur at all). However, one
should not rule out, a priori, the possibility of cytokines, growth factors,
etc. released from sites of adaptive remodeling having effects on remote
tissues or organs.

16.3.5 Chemical Carcinogenesis

Of necessity, chemical carcinogenesis must be considered a systemic problem
because of the variable sensitivity of cells to chemically induced neoplastic
transformation and the possibility of metastases. If chemically mediated
carcinogenesis occurs in humans due to the use of implant materials, it can
thus be expected to have a significant systemic manifestation. It is worth
pointing out again that possibilities of systemic and remote-site tumorgen-
esis associated with implants tend to be overlooked in patient populations
for two reasons:

• The tumor types expected are no different from those that would
already exist in a comparable patient population without implants.
• The specialization of medicine makes the connection of a tumor in
a remote tissue or organ system to the presence of an implant in
another tissue or organ system unlikely (Black 1984).
Effects of Degradation Products on Remote Organ Function 323

16.3.6 Foreign-Body Carcinogenesis

One would expect foreign-body carcinogenesis that occurs in patients to be
a local problem. However, the ability of particulate materials to move in the
body and the inherent ability of many neoplasms to metastasize render it a
potential systemic problem. It must be emphasized that this is a putative
consideration because the presence of primary foreign body neoplasias (at
the implant site) has not been reliably detected in humans (see Section 13.3).

16.3.7 Infection
Acceptance of Weinberg’s (1974) arguments concerning nutritional immu-
nity (see Chapter 14) means that it is necessary to recognize the possibility
of problems associated with elevated iron concentrations in the vicinity of
implants and with elevated concentrations in remote storage sites. Further-
more, suppression of the immune system may also predispose to infection
at distant sites as well as at the implant–tissue interface.
The possibility of the inverse — that is, of hematogenous “seeding” of an
implant site infection from a distant site such as a dental abscess or urinary
tract infection — must not be discounted, although clinical data remain
equivocal at this time (Thyne and Ferguson 1991). In most surgical special-
ties, pre- and perioperative precautions are now employed for implantation
procedures when there is foreknowledge of infection present at a remote
site, particularly in the oral cavity (Carmona et al. 2002).

16.3.8 Allergic Foreign-Body Response

The discussion of this subject (Chapter 12) emphasized the systemic nature
of the response. Therefore, in addition to possible problems in the vicinity
of the implant such as pain, loosening, etc., a wide variety of allergic
responses at remote sites can potentially be associated with the presence of
metallic, and possibly polymeric, implants as sensitizing or challenge agents.
Sensitization is a matter of particular concern because it may evoke a later
local or systemic response apparently spatially unrelated to the original site
of implantation.

16.4 A Final Comment

It is important to re-emphasize that studies of host response have focused
primarily on the implant site and adjacent tissues. In the future, a much
broader view must be taken and the response of the entire host, whether
experimental animal or human patient, must be studied in detail.
324 Biological Performance of Materials: Fundamentals of Biocompatibility

However, a major pitfall in such considerations must be avoided. Much is

made in the lay press about individuals who develop an apparent “environ-
mental allergy”: an elevated sensitivity, with immune response symptoms,
to a very broad variety of agents that have in common only that they are
man-made. An analogous situation exists in the field of clinical application
of biomaterials with lay concern over the relationship between release of
mercury from mercury-based dental amalgams and a broad variety of patient
symptoms. Without passing judgment on either of these situations,* I would
like to suggest, somewhat in the spirit of Furst’s requirements for accepting
the carcinogenicity of a metal in animal models (see Section 13.2.4), that the
following criteria should be met before the existence of a remote or systemic
effect in humans is taken as proven:

• The basic mechanism of the biological response must be demon-

strated in at least one in vitro situation or in an animal biological
model.
• After the causative implant-related species has been identified, its
release by a functional implant and systemic distribution in an ani-
mal model or in patients (preferably both) must be shown.
• The putative biological response must be identified in an animal
model or in patients (preferably both) with functional implants.
• If it is demonstrated in patients, the biological response must be
recognized on the basis of a statistically sound epidemiological
study, with suitable nonexposed controls and, unless the response
is of a threshold type, must demonstrate a dose–response or expo-
sure–incidence relationship.

In my mind, these are the necessary and sufficient conditions to conclude

that an implant-related systemic (and/or remote site) effect exists in patients.
Their presence does not settle the issue of the clinical importance of the effect;
this depends upon other considerations such as treatment alternatives
(including no treatment) and the benefit of the use of the device in which
the biomaterial is incorporated.
However, I would suggest that if the first criterion is met — that is, a
biological mechanism leading to a putative systemic or remote site effect is
identified — then an index of suspicion should be attached to the biomaterial
in question. Similarly, isolated case reports of local or distant adverse
implant-related responses should also arouse a degree of suspicion; what
they lack in numbers they make up for to a degree in specificity. A lack of
sound knowledge in any of the latter three areas should not lead to an
inference of safety; this would be morally equivalent to the statement that
“What I don’t know can’t hurt me.”

* See Section 12.4 for a more complete discussion of immune responses to implants and Section
14.4 for a critique of environmental allergy and related issues.
Effects of Degradation Products on Remote Organ Function 325

For this reason, I have suggested that a biomaterial can never be considered
safe or unsafe, but merely biocompatible (or not) in a specific application
(Black 1995). Rather, satisfaction of the first criterion should lead to changes
in behavior in consideration of that biomaterial for specific current device
designs (Black 1988) and in planning future basic and applied studies of the
biomaterial’s safety and efficacy in present as well as proposed applications.

References
Aaseth, J. and Norseth, T., Copper, in Handbook on the Toxicology of Metals, Vol. II.,
Friberg, L., Nordberg, G.F. and Vouk, V.B. (Eds.), Elsevier, Amsterdam, 1986,
233.
Alexander, C.S., Cobalt-beer cardiomyopathy, Am. J. Med., 53, 395, 1972.
Black, J., Systemic effects of biomaterials, Biomaterials, 5, 11, 1984.
Black, J., Does corrosion matter? J. Bone Joint. Surg., 70B, 517, 1988.
Black J., “Safe” biomaterials, J. Biomed. Mater. Res., 29, 791, 1995.
Burger, W. et al., New Y-TZP powders for medical grade zirconia, J. Mater. Sci. Mater.
Med., 8, 113, 1997.
Carmona, I.T. et al., An update on the controversies in bacterial endocarditis of oral
origin, Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod., 93, 660, 2002.
Charnley, J., Low Friction Arthroplasty of the Hip, Springer–Verlag, Berlin, 1979, 6.
Chopra, S., Disorders of the Liver, Lea & Febiger, Philadelphia, 1988.
Dahl, O.E., Cardiorespiratory and vascular dysfunction related to major reconstruc-
tive orthopedic surgery, Acta Orthop. Scand., 68, 607, 1997.
Elinder, C.-G., Iron, in Handbook on the Toxicology of Metals, Vol. II., Friberg, L., Nor-
dberg, G.F. and Vouk, V.B. (Eds.), Elsevier, Amsterdam, 1986, 276.
Enestrom, S. and Hultman, P., Does amalgam affect the immune system? A contro-
versial issue, Int. Arch. Allergy Immunol., 106(3), 180, 1995.
Gabriel, S.E., Soft tissue response to silicones, in Handbook of Biomaterial Properties,
Black, J. and Hastings, G. (Eds.), Chapman & Hall, London, 1998, 556.
Giannoudis, P.V. et al., Review: systemic effects of femoral nailing: from Kuntscher
to the immune reactivity era, Clin. Orthop. Rel. Res., 404, 378, 2002.
Homsy, C.A. et al., Some physiological aspects of prosthesis stabilization with acrylic
polymer, Clin. Orthop. Rel. Res., 83, 317, 1972.
Keret, D. and Reis, D.R., Intraoperative cardiac arrest and mortality in hip surgery.
Possible relationship to acrylic bone cement, Orthop. Rev., IX(7), 51, 1980.
Kossovsky, N. and Heggers, J.P., The bioreactivity of silicone, CRC Crit. Rev. Biocom-
pat., 3, 53, 1987.
Langkamer, V.G. et al., Changes in the proportions of peripheral blood lymphocytes
in patients with worn implants, J. Bone Joint Surg., 74B, 831, 1992.
Lewis, L.M. et al., Determination of plasticizer levels in serum of hemodialysis
patients, Trans. Am. Soc. Artif. Intern. Organs, XXIII, 566, 1977.
McMaster, W.C. et al., Blood pressure lowering effect of methylmethacrylate mono-
mer, Clin. Orthop. Rel. Res., 98, 254, 1974.
Noone, R.B., A review of the possible health implications of silicone breast implants,
Cancer, 79(9), 47, 1997.
Perry, H.M., Jr., Minerals in cardiovascular disease, J. Am. Diet. Assoc., 62, 631, 1973.
326 Biological Performance of Materials: Fundamentals of Biocompatibility

Shih, L.-Y. et al., Erythrocyte sedimentation rate and C-reactive protein values in
patients with total hip arthroplasty, Clin. Orthop. Rel. Res., 225, 238, 1987.
Thyne, G.M. and Ferguson, J.W., Antibiotic prophylaxis during dental treatment in
patients with prosthetic joints, J. Bone Joint Surg., 73B, 191, 1991.
Weinberg, E.D., Iron and susceptibility to infectious disease, Science, 184, 952, 1974.
Weiss, B., The behavioral toxicology of metals, Fed. Proc., 37(1), 22, 1978.

Bibliography
Davies, I.J.T., The Clinical Significance of the Essential Biological Metals, Charles C Tho-
mas, Springfield, IL, 1972.
Debelian, G.J. et al., Systemic diseases caused by oral microorganisms, Endod. Dent.
Traumatol., 10, 57, 1994.
DiCarlo, E.F. and Bullough, P.G., The biological responses to orthopedic implants
and their wear debris, Clin. Mater., 9(3–4), 235, 1992.
Friberg, L., Nordberg, G.F. and Vouk, V.B. (Eds.), Handbook on the Toxicology of Metals,
Vols. I and II. Elsevier, Amsterdam, 1986.
Ling, R.S.M., Systemic and miscellaneous complications, in Complications of Total Hip
Replacement, Ling, R.S.M. (Ed.), Churchill–Livingstone, London, 1984, 201.
McCall, J.T. et al., Implications of trace metals in human diseases, Fed. Proc., 30(3),
1011, 1971.
Pier, S.M., The role of heavy metals in human disease, Tex. Rep. Biol. Med., 33(1), 85,
1975.
Rothman, R.H. and Hozack, W.J., Complications of Total Hip Arthroplasty, W.B. Saun-
ders, Philadelphia, 1988.
Schroeder, H.A., Trace metals and chronic diseases, Adv. Intern. Med., 8, 259, 1956.
Schierholz, J.M. and Beuth, J., Implant infections: a haven for opportunistic bacteria,
J. Hosp. Infect., 49(2), 87, 2001.
Webb, M., Metabolic targets of metal toxicity, in Clinical Chemistry and Chemical
Toxicology of Metals, Brown, S.S. (Ed.), Elsevier North–Holland, Amsterdam,
1977, 51.
Williams, D.F. (Ed.), Systemic Aspects of Biocompatibility, Vols. I and II, CRC Press, Boca
Raton, FL, 1981.
Interpart 2
Implant Materials: Clinical Performance*

I2.1 Introduction
Elaine Duncan (1990) once posed the question of whether biomaterials are
at risk of becoming endangered species. Her query was motivated, in part,
by a letter distributed by Dow Corning, Inc. (Midland, MI) warning of the
company’s intention of withdrawing an old standby polyurethane biomate-
rial, Pellethane™, from the market — at least for applications intended to
last longer than 30 days in vivo. Citing published reports of cracking of the
material after longer times in vivo (Stokes and Chem 1988), a company
representative, J.R. Stoppert (1989), asserted that no data support long-term
use of the material.
My immediate reaction to this statement was incredulity. The use of mate-
rials similar to Pellethane had been reported by Boretos and Pierce (1968)
more than 20 years earlier. Discussing such materials, Boretos (1973) said
that “[they] possess a combination of properties not available in other mate-
rials, outstanding of which [is]...excellent stability over long implant period.”
So, how can there be no data to support long-term human implantation of
Pellethane?
I believe that what Stoppert (1989) meant is that there were, literally, no
data. That is, there was evidence neither to support nor to contradict bio-
medical device designers’ decisions to use Pellethane™ in long-term appli-
cations. Boretos’ (1973) comments are not data and neither are the majority
of papers published about this material or, for that matter, about virtually
any other biomaterial in use in long-term clinical applications.** Most of
these papers, especially those dealing with clinical observations, are not
studies in the strict scientific sense, and the failure to conduct studies results
in the absence of data, whether positive or negative.

* An earlier version of this interpart was published in Black (1990). Table I2.1 is adapted from
Table 14.2 in Black (1988).
** The first competent study that I am aware of documenting in vivo stress-related degradation
of such materials, albeit in an animal model, did not appear until 1990 (Zhao et al. 1990).

327
328 Biological Performance of Materials: Fundamentals of Biocompatibility

What has happened is that workers in the field of biomaterials have been
blinded by success. The techniques used in the 1960s to qualify materials
(limited in vitro studies, 12- to 104-week animal studies, 2-year human clin-
ical studies; see Chapter 18 and Chapter 19) are still current practice in the
early 2000s. This is despite the widespread use of biomaterials in long-term
clinical applications that, in at least one device type (total hip replacement;
Malchau et al. 2002), exceed 25 years in individual and group patient expe-
rience. It appears that two critical aspects of the study of biomaterials have
been neglected: the epidemiology and human physiology of biomaterials.

I2.1.1 Epidemiology of Biomaterials

Continuing to study the success and failure of implanted materials in exper-
iments with animals numbering in the tens and twenties, biomaterials
researchers have largely overlooked the vast clinical “experiments” under
way in which thousands, in many cases tens or hundreds of thousands, of
human patients receive virtually identical biomaterials as chronic implants.
A Center for Disease Control survey performed in 1988 (Moss et al. 1991)*
suggests that as many as 14.5 million people, or nearly 1 in 20 in the U.S.,
had permanent implants. With the exception of occasional reports of clinical
failures and studies of the materials aspects of retrieved devices, almost
nothing is known about biomaterials' performance in these implants in the
human clinical environment. Dependable incidence and prevalence data on
the devices are hard to obtain; such data on the materials from which they
are made, including their exact (not merely specified) composition and pro-
cessing, are still essentially nonexistent.
Today, making the same mistake as certain penologists who, wishing to
know about crime, study only failed criminals (that very small nonrandom
proportion of the criminal population actually apprehended, convicted, and
incarcerated), one persists in studying only random device failures. Even
these limited studies, based upon clinical or postmortem retrievals, are fre-
quently incomplete, focusing on the clinical features of the failure or upon
the physical attributes of the failed device, depending upon the background
and interests of the principal investigator, but rarely dealing with both
aspects in a balanced way.
Recent increased interest in studying outcomes of surgical procedures
(change in patient lifestyle, satisfaction level, relative cost, etc.) rather than
merely the success of the procedure (rated as excellent, good, etc. on largely
subjective bases) may improve matters, but only if accurate information on
the implant and its materials of construction is made part of the permanent

* As this is written in 2005, it is odd, bordering on the bizarre, that Moss et al. 1991, which is
based upon data that is now 17 years old (!), remains the most reliable source of such data for the
U.S. experience with permanent implants. This stands in stark contrast to many other countries
in which device registries and resultant up-to-date statistics are now available for periods
exceeding two decades of clinical use.
Implant Materials: Clinical Performance 329

clinical record. Suggestions have been repeatedly made concerning the need
for registration systems for implants in the U.S. so that this information may
follow patients as they move from place to place. Countries with national
health services, such as the U.K., have had some success in developing such
national registries. National systems exist for registration of certain classes
of implants such as hip replacements — for example, in Sweden (from 1979)
and Norway (from 1987).
However, countries with predominantly private health care systems, such
as the U.S., have encountered difficulties in establishing such systems.
Numerous proposals have been made for a national system for all permanent
implants in the U.S. (Black 1996; see also Chapter 22). An important contem-
porary effort is one sponsored by the American Academy of Orthopedic
Surgeons (Maloney 2001), but it has been very slow to get under way, even
on a pilot basis, due to concerns about protection of patient confidentiality
(Maloney 2004). Some medical device manufacturers maintain registries of
their products, but access to these is severely restricted. Efforts by profit and
nonprofit private concerns have floundered after a few years due to their
voluntary and incomplete nature.

I2.1.2 Human Physiology of Biomaterials

One would be quick to discount the knowledge of a nephrologist who based
his or her entire understanding of human renal function on the study of
healthy rats, rabbits, dogs, and an occasional human autopsy specimen.
Fortunately, professional nephrologists have a vast armamentarium of in vivo
tests that permit them to study the physiology of the functioning human
kidney in health and in disease. The biomaterials scientist, when addressing
functioning human implants, lacks all but the most rudimentary of these
capabilities: clinical imagining, using primarily single-plane x-rays. The pH,
pO2, interfacial stresses, etc. in the vicinity of a functioning implant and the
ranges of values that assure long-term success or predict imminent failure
cannot be stated with any certainty.
However, it is possible to obtain such data, at least in animal models.
Baranowski and Black (1987) reported repetitive in vivo measurement of pH
and pO2, associated with successful and unsuccessful stimulation of bone
growth, near active and inactive implanted stainless steel electrodes in the
tibial medullary canal of rabbits. With care and well-designed protocols, such
techniques could be extended to studies in human subjects, especially with
the increasing development of microcatheters and arthroscopes. Already, as
discussed in Chapter 15, studies are being performed to detect and quantify
metal-bearing species associated with implants in human patients. Early
studies even suggest a correlation between elevations in serum concentra-
tions and loosening of implants (Jacobs et al. 1991), although the relation-
ships between cause and effect remain unclear.
330 Biological Performance of Materials: Fundamentals of Biocompatibility

I2.2 An Example: Total Hip Replacement

Even in the absence of clinical epidemiology and the development of clinical
tests of biomaterials’ performance, it is still possible to gain some knowledge
from current clinical experience. A clinical internship and, if possible, con-
tinuing contact with a clinical population can be extremely valuable for a
biomaterials scientist or engineer, especially if he or she is prepared to be
observant and analytical in approach. It is probably possible in any clinical
implant application to produce a list of symptoms (radiographic, clinical, or
histological findings) and associated putative mechanisms leading to impli-
cations or conclusions concerning the clinical performance of the implanted
materials. Table I2.1 presents such a triple listing for a frequent orthopaedic
procedure: total replacement of the hip. Note that other device-related clin-
ical findings are possible; those listed are believed to be directly referable to
the biomaterials used in or in conjunction with the device rather than to
device design, surgical technique, or patient use factors.

TABLE I2.1
Materials-Associated Findings in Total Hip Replacement
Finding Mechanism Implication

Radiographic

PMMA fragments (early) Operative debris Third-body wear; single-cycle

fracture
PMMA fragments (late) Fatigue Third-body wear; cup loosening
PMMA mantle fracture Inadequate bony support; Stem subsidence; loosening
(early) single-cycle fracture
PMMA mantle fracture Inadequate bony support; Stem subsidence; loosening
(late) fatigue
Broken cerclage wire Fatigue (early) Trochanteric
dislodgement, nonunion, wire
migration; (late) wire migration
Stem deformation Plastic deformation Change in bony support;
inadequate stem size or yield
point; impending failure
Stem, cup, or cup screw Fatigue Manufacturing defect; chronic
fracture mechanical overload; (early)
inadequate bony support; (late)
change in bony support
Eccentric cup-head Plastic deformation; wear UHMWPE creep; uniform wear;
centers spalling?
Loose metallic debris Wear Third-body wear; fretting (loose
component)
Fatigue ± corrosion Inadequate processing of porous
coating

(continued)
Implant Materials: Clinical Performance 331

TABLE I2.1 (CONTINUED)

Materials-Associated Findings in Total Hip Replacement
Finding Mechanism Implication
Loose ceramic debris Cracking Component impingement,
subluxation, recurrent
dislocation
Focal lytic lesiona Particle phagocytosis Excessive wear debris; endotoxin?
Immune response? Metal sensitivity? (both)
Progressive failure?
Progressive dissecting Osteoclasis Excessive wear debris; metal
lesiona sensitivity? Neoplasm?

Clinical

Intraoperative Central control Methyl methacrylate (monomer)

hypotension sensitivity? Fat embolism?
Hip paina Immune response Metal sensitivity?
Venous blockade Excessive wear
Bursa formation Local inflammation Metal sensitivity?
Ectopic calcification Wear debris nucleation? Excessive wear
Dermatitis Delayed hypersensitivity? Metal sensitivity?
Eczema Delayed hypersensitivity? Metal sensitivity?
Bronchospasm Delayed hypersensitivity? Metal sensitivity?

Histologic

Fibrous capsule Local host response Normal response

Histiocytosis with Chronic inflammation Manufacturing defect;
multinuclear cells inappropriate material; excessive
wear
Lymphocytic infiltration Delayed hypersensitivity? Metal sensitivity? Polymer
with plasma cells sensitivity?
Fibrosarcoma Neoplastic Chemical neoplasia?
transformation
Lymphoma Neoplastic Chemical neoplasia?
transformation
Rhabdomyosarcoma Neoplastic Chemical neoplasia?
transformation
Malignant fibrous Neoplastic Chemical neoplasia?
histiocytoma transformation
Osteosarcoma Neoplastic Chemical neoplasia?
transformation
a In absence of infection.
Note: ? = possible mechanism or implication.
Source: Adapted from Black, J., Orthopedic Biomaterials in Research and Practice, Churchill–
Livingstone, New York, 1988, 319.
332 Biological Performance of Materials: Fundamentals of Biocompatibility

I2.3 A Final Word

Many of the basic sciences underlying the field of biomaterials science and
engineering have been neglected. When issues of long-term survival of bio-
materials in vivo are considered, the lack of attention to epidemiology and
physiology still remains a key issue. Unless researchers become more careful
and observant of clinical performance, biomaterials may, indeed, as Elaine
Duncan suggested, become endangered species and one will hear more
often, “‘And there are no data,’ he said.”

References
Baranowski, T.J., Jr. and Black, J., The mechanism of faradic stimulation of osteogen-
esis, in Mechanistic Approaches to Interactions of Electric and Electromagnetic Fields
with Living Systems, Blank, M. and Findl, E. (Eds.), Plenum Press, New York,
1987, 399.
Black, J., Orthopedic Biomaterials in Research and Practice, Churchill–Livingstone, New
York, 1988, 319.
Black, J., “And there are no data, he said,” Biomater. Forum, 12(4), 9, 1990.
Black, J., Overview of PMS in an international perspective: global developments and
global cooperation, Int. J. Risk Safety Med., 8, 3, 1996.
Boretos, J.W., Concise Guide to Biomedical Polymers: Their Design, Fabrication, and Mold-
ing, Charles C Thomas, Springfield, IL, 10, 1973.
Boretos, J.W. and Pierce, W.S., Segmented polyurethane: a polyether polymer, J.
Biomed. Mater. Res., 2, 121, 1968.
Duncan, E., Editorial: endangered species? Biomater. Forum, 12(3), 4, 1990.
Jacobs, J.J. et al., Release and excretion of metal in patients who have a total hip-
replacement component made of titanium-base alloy, J. Bone Joint Surg., 73A,
1475, 1991.
Malchau, H. et al., The Swedish Total Hip Replacement Register, J. Bone Joint Surg.,
84A Suppl 2, 2, 2002 (Erratum: J. Bone Joint Surg., 86A, 363, 2004).
Maloney, W.J., National Joint Replacement Registries: has the time come? J. Bone Joint
Surg., 83A, 1582, 2001.
Maloney, W.J., Personal communication, 2004.
Moss, A.J. et al., Advance Data No. 191, (PHS) 91-1250. U.S. Government Printing
Office, Washington, D.C., 1991.
Stokes, K.B. and Chem, B., Polyether polyurethanes, biostable or not? J. Biomater.
Appl., 3, 228, 1988.
Stoppert, J.R. (1989), Letter cited in Duncan, E., Biomater. Forum, 12(3), 4, 1990.
Zhao, Q. et al., Cellular interactions with biomaterials: in vivo cracking of prestressed
Pellethane 2363-80A, J. Biomed. Mater. Res., 24, 621, 1990.
Implant Materials: Clinical Performance 333

Bibliography
Finerman, G.A.M. et al. (Eds.), Total Hip Arthroplasty Outcomes, Churchill–Living-
stone, New York, 1998.
Fraker, A.C. and Griffin, C.D. (Eds.), Corrosion and Degradation of Implant Materials:
Second Symposium, ASTM Special Technical Publication 859, American Society
for Testing and Materials, Philadelphia, 1985.
Hench, L.L. and Wilson, J. (Eds.), Clinical Success of Skeletal Prostheses, Chapman &
Hall, London, 1995.
Improving medical implant performance through retrieval information, Current Bib-
liographies in Medicine, 99-5, National Library of Medicine, Washington, D.C.,
2000.
Improving medical implant performance through retrieval information: challenges
and opportunities, National Institutes of Health Technology Assessment Con-
ference Summary, Technology Assessment Statement 19, 2000.
Leir, H.K., Casebook: Alien Implants, Dell, New York, 2000.
Syrett, B.C. and Acharya, A. (Eds.), Corrosion and Degradation of Implant Materials,
ASTM Special Technical Publication 684 American Society for Testing and
Materials, Philadelphia, 1979.
Weinstein, A., Horowitz, E. and Ruff, A.W. (Eds.), Retrieval and Analysis of Orthopedic
Implants, NBS Special Publication 472, U.S. Government Printing Office, Wash-
ington, D.C., 1977.
Weinstein, A. et al. (Eds.), Implant Retrieval: Material and Biological Analysis, NBS
Special Publication 601, U.S. Government Printing Office, Washington, D.C.,
1981.
 Part IV

Methods of Testing for

Biological Performance
17
In Vitro Test Methods

17.1 Test Strategies

Biological performance, as defined in Chapter 1, has two aspects: material
response and host response. Because the traditional approach has been to
define biological performance in terms of biocompatibility (host response),
and then to observe evidence of material degradation that arises during in
vitro and in vivo testing, the development of tests has concentrated upon
those that measure host response. In practice this has not been a bad course
of events. Despite the uncertainties in both areas, it is still easier to predict
material response than to predict host response. Thus, it remains a good
strategy to proceed through the following five steps in early biomaterial
qualification studies:

• Select candidate materials based upon engineering properties and

previous material and host response information.
• Experimentally determine whether the host response is satisfactory
(acceptable) for the intended application.
• Carefully watch for evidence of unacceptable material response dur-
ing the conduct of host response studies.
• Verify satisfactory material response (for the material selected) dur-
ing long-term in vivo implant studies if it is relevant to the applica-
tion.
• Be responsive to clinical reports that may reflect changes in material
response in actual service.

That is, given an adequate a priori level of expected material response, it is

still more appropriate to expend a larger effort on determining host rather
than material response. The selection of in vitro host response tests for mate-
rials qualification will be dealt with in Chapter 19.

337
338 Biological Performance of Materials: Fundamentals of Biocompatibility

17.2 In Vitro Test Types

A broad range of materials is available and only a very small percentage has
been used in biological environments; thus, there has been a continual need
for quick screening methods that can be used in vitro. These can be divided
into two general classes:

• Tissue culture methods

• Blood contact methods (for blood contact applications)

Within each of these classes are a wide variety of tests and variations of test
methods. Some variations are traditional for particular applications and
others are the practice of a particular laboratory. Thus, these types of methods
can only be considered in a general way.

17.3 Tissue Culture Tests

17.3.1 Generic Methods
Tissue culture refers collectively to the practice of maintaining portions of
living tissue in a viable state in vitro. There are three generic methods:

• Cell culture: the growth of initially matrix-free, disassociated cells.

This method closely parallels the methods used to grow bacteria in
vitro. Cells may be grown in solution or on agar or other media
substrates. Exposure to biomaterials may be through direct contact
with bulk materials, diffusional contact through an agar (or other
gel) intermediate layer, or by inclusion of particles or extracts or
elutants from materials in the culture media.
• Tissue culture: the growth of portions of intact tissue without prior
cellular dissociation. This method usually utilizes a substrate rather
than a suspended technique. Exposure to biomaterials is similar to
that for true cell culture.
• Organ culture: the growth of intact organs in vitro. This may vary from
the use of fetal bone explants that can survive without external support
systems to the use of whole, adult, perfused organs such as the kidney
or heart. Again, a variety of exposures to biomaterials is possible.

As Rae (1980) has pointed out, each of these generic methods involves
compromises and deviations from in vivo conditions. These problems will
be discussed further in the next section.
In Vitro Test Methods 339

17.3.2 Problems Inherent in Tissue Culture

Several key points concern the use of tissue culture techniques. The principal
difference between in vivo and in vitro conditions, after noting the obvious
humoral one (lack of access to pathways to and from remote locations in the
intact animal) is the failure of local blood circulation and lymphatic drainage.
Tissue in culture media can interchange materials only by diffusion. Diffu-
sion distances must be quite short or otherwise toxic effects will be seen;
this is obvious because cells in vivo are never more than a few hundred
microns distant from a capillary. Even quite modest volumes of tissue will
show central or focal necrosis due to a diffusional limit on nutrients or waste
products. For this reason, tissue culture as a field is often referred to as a
study of cell death. The use of cell culture (rather than tissue or organ culture)
minimizes these effects until cell densities rise to levels associated with
frequent intercellular contact.
Associated with diffusional problems and lack of humoral components is
the general observation that, perhaps as a consequence, metabolic rates in
vitro are always lower than those in vivo. Even in the absence of necrosis,
cell, tissue, and organ culture conditions are inhibitory, with an initial max-
imum effect and a significant residual effect. Some of this inhibition may be
an accidental result of growth in 21% oxygen because many cells in vivo are
acclimated to much lower oxygen tensions. Although less active in vitro than
in vivo, all cells display cell type specific growth rates. Unfortunately, fibro-
blasts retain quite high replication rates in vitro, so even a small contamina-
tion of specimen with a slower replication rate, such as a neural cell culture,
may result in a “weedy” overgrowth of these unwanted cells.
Most tissue culture models attempt to maximize cell metabolism by grow-
ing cells in monolayers (to provide easy exchange with large fluid volumes)
and by adding various stimulatory agents (fetal or neonatal serum, vitamin
C, etc.) and preservation agents (α-tocopherol [vitamin E]), antibiotics, fun-
gicides, etc.). Fortunately, most cells are contact inhibited and thus form
monolayers easily in culture; this condition is called confluence and is char-
acterized by a relatively constant cell number. Cell monolayers are not, in
general, subject to diffusion limits. Although these agents improve the via-
bility and longevity of tissue culture models, their formation produces still
greater differences between in vitro and in vivo conditions.
The morphology and the function of cells in vivo depend to a degree on
the chemistry of the pericellular environment and the nature of the substrate
(matrix) supporting and surrounding the cells (see Section 11.3). Lack of
these environmental factors may cause cells to become inactive or to revert,
at least functionally, to more primitive (less committed) types. An example
is the chondrocyte; if it is grown in an agar gel, it retains its appearance and
biosynthetic function, but if it is cultured on a typical matrix-free surface, it
rapidly takes on fibroblastic shape and synthetic function. Related to this
problem is the previously noted recognition that cells frequently exist in vivo
in regions of mixed interdependent cell types, such as in bone marrow or
340 Biological Performance of Materials: Fundamentals of Biocompatibility

the cerebral cortex. In such situations, cell shape and function may depend
in part on interaction with other cell types; this may be reproduced in cell
culture by coculture: growing more than one cell type in a common container
— in some cases, with permeable barriers to isolate individual cell types
geographically but not diffusionally.* Such coculture may be necessary to
demonstrate in vitro complex processes such as bone matrix mineralization.
Another, less well appreciated difference from in vivo conditions is the
absence of external mechanical strain on cells. Particularly in the musculo-
skeletal, respiratory, and cardiovascular systems, in vivo cells are subjected
to a broad spectrum of mechanical forces and deformations, which may be
transduced directly or perhaps through electromechanical intermediate sig-
nals. It is possible to reproduce the effects of these mechanical stimuli in a
limited way by culturing cells on membranes subject to various cyclic defor-
mations. This is now a common enough practice that commercial devices
have been developed.
Finally (and this cannot be overemphasized), cell culture results cannot be
useful or relevant, even in preliminary screening for local host response,
unless care is taken in selection of the biomaterial samples being evaluated.
Earlier (Section 2.6), I pointed out the need to replicate processing, cleaning,
and sterilization conditions expected in the clinical application. However, in
designing tissue culture studies, one must include other important consid-
erations. These include:

• Selection of the appropriate cell type (especially for implants

intended to be placed internal to specific organs such as heart, liver,
etc.)
• Consideration of the dosage aspect of exposure because cells may
tolerate materials at low exposures but be adversely affected at high
dosages (Shanbhag et al. 1994)
• Usage of appropriate forms of the biomaterial and its expected deg-
radation products in the proposed application

17.3.3 Cell Types and Observations

Cell types may vary in activity and properties between species — within
species, between individuals, and within a specific cell line — between
generations. In the scientific literature, one finds references to cell types with
designations such as “P1534 leukemic cells.” These citations refer to cells
from established cell lines available from the American Type Culture Col-
lection (ATCC)** (Hay 1983) or similar sources. These organizations isolate,
characterize, and preserve (by ultrafreezing) uniform cell populations. One
may order specimens from particular generations (passages) of these cells
* Although such arrangements vary widely between experimenters, they are referred to generi-
cally as Boyden chambers, after their original developer (Boyden 1962).
** http://www.atcc.org/.
In Vitro Test Methods 341

and grow them in the laboratory for use in testing. In general, the lower the
generation number is (often referred to as the passage number), the more
the cells will resemble those in vivo in appearance and function. With care,
the use of established cell line techniques screens out a large component of
biological variability.
An alternate approach is to use freshly derived cells or so-called primary
cultures. Such cultures are more vigorous than maintained cell lines; how-
ever, they are difficult to prepare in a repeatable manner, are usually poorly
characterized, and (as previously noted) may suffer from overgrowth of
faster growing or “weed” cells, such as fibroblasts.
Tissue culture techniques, as a group, may be used to study the following
aspects of host responses:

• Cell survival: toxicity; organelle and membrane integrity

• Cell reproduction: growth inhibition
• Metabolic activity: energetics, synthesis, and catabolism
• Affective activity: inhibition of locomotion, chemotaxis, and phago-
cytosis; alteration of cellular size, shape, and appearance
• Cell damage: chromosomal aberration, mutagenicity, and carcinoge-
nicity

Tissue culture techniques for screening materials may use one or more
normal mammalian cell lines such as murine macrophages, abnormal cells
such as HeLa or lymphoma cells, or bacterial cell lines such as Staphylococcus
aureus or Escherricia coli. Each test that has been developed uses selected cells
suitable for the particular questions posed.

17.3.4 Examples of the Use of Tissue Culture in Materials Evaluation

A classic example of such studies (Galin et al. 1975) illustrates the general
approach. In this case, cultures of rabbit kidney cells were used to investigate
the toxicity of several types of ocular implants. The implants were exposed
to the cells directly by placement on a gel-supported cell monolayer and
indirectly by inclusion in media in which the cells were cultured. Toxicity
was assayed by observations of progressive cell death in the monolayer and
by cellular changes in culture.
Galin’s study examined the host response to a fabricated implant. Similar
tests may be used to study biomaterials in unfabricated forms or to observe
host response to degradation products such as wear debris. An early example
of the latter is the work of Cohen and coworkers, who examined tissue
culture response to finely divided particles of implant alloys as models for
wear debris and corrosion product release. A number of cell lines were used
and the effects of the metallic debris and debris supernatant on cell counts
in the replication phase (Pappas and Cohen 1968) and in the lag phase (Mital
342 Biological Performance of Materials: Fundamentals of Biocompatibility

and Cohen 1968) were obtained. The authors concluded that using lag phase
cell counts to detect direct toxicity was more sensitive than lumping toxicity
and growth and replication inhibition together.
It has been suggested that the primary effects seen in the exposure of
polymeric biomaterials to cells in tissue culture are due to low molecular
weight species that diffuse out of the biomaterial. Thus, many techniques
use the material as well as extracts of the material prepared with various
hydrophilic and hydrophobic solvents. A study by Homsy et al. (1970) exam-
ined this issue by autoclaving a number of specimens in a pseudoextracel-
lular fluid (PECF) for 62 hours at 115° and then using the PECF supernatant
as a challenge for cultures of cells derived from newborn mouse hearts. In
addition, specimens of the supernatant were analyzed quantitatively to
determine the total concentration of –CH3, –CH2, and –CH radicals present.
A total of 22 polymers were studied (Table 17.1). A comparison of cellular
toxicity of the supernatant with the concentration of organic radicals eluted
from the polymer showed a strong positive correlation that apparently over-
whelms the details in differences between the chemistries of individual
polymers. Thus, Homsy and colleagues suggested that this tissue culture
method and/or an analysis of a PECF supernatant may be used as rapid,
reliable screening techniques for new polymers.

17.3.5 Use of Tissue Culture for Host Response Screening

As previously noted, Rae (1980) discussed the pros and cons of various
approaches to cell, tissue, and organ culture methods in the general deter-
mination of biocompatibility of implant materials. The present consensus is
that these techniques are best used to measure early, acute response to
materials.
However, the real utility of in vitro culture techniques depends to a great
degree on their correlation with in vivo local host response. Many attempts
have been made to correlate in vitro cell culture studies with results of animal
exposure studies. For example, Wilsnack (1976) used an ATCC human fibro-
blast cell line, WI-38, to examine the in vitro cytotoxicity of a large number
of polymeric materials and compared the results with responses in rabbits
to systemic and local injections of soluble material eluted with saline solution
and with cottonseed oil, according to USP methods (see Section 20.2.1). He
concluded that the in vitro test was more sensitive than the in vivo one
selected (yielding more examples of cytoxicity); however, correlation
between in vitro and in vivo results was poor when both were positive. In
this case, one can point to reasons other than inherent inadequacies in cell
culture to explain disparities between in vitro and in vivo results: different
species, single cell line vs. multiple cell types, solid materials vs. eluted
materials, etc. This example illustrates the difficulties in extrapolation from
in vitro biological tests. However, Johnson and Northup (1983) attempted a
similar correlation, but with fewer differences between techniques: they
In Vitro Test Methods 343

TABLE 17.1
Correlation between Tissue Culture Response to PECF and PECF Chemical
Analysisa
Total -CH3 , -CH2 , and
Tissue Culture -CH in PECF by IR Analysis
Polymerb Responsec (mpmd equivalent n-hexanol)
Silicone (Silastic™ 372) +1 5
Polyethylene (U. of Texas) +1 17
Fluorinated ethylene propylene (type 1) +1/+2 nde
Polyphenylene oxide (type 1) +1/+2 27
Polyethylene (type 1) +2 nde
Acrylic molding powder +2 nde
Polyphenylene oxide (type 2) +2 17
Polyethylene (type 2) +2 17
Fluorinated ethylene propylene (type 1) +2 23
Ionomer (type 1) +2 142
Polypropylene (food grade) +2 198
Vinylidene fluoride +2/+3 3
Nylon™(grade 101) +2/+3 14
Ionomer (type 2) +2/+3 30
Cellulose proprionate +3 81.7
Polystyrene +3 168
Nylon™ (grade 38) +4 12
Poly(vinyl)chloride +3/+4 277
Polyurethane (type 1) +4 89
Polyurethane (type 2) +4 328
Poly(vinyl)chloride (U. of Texas) +4 514
Acrylonitrile, butadiene, styrene (ABS) +4 516
a Polymer exposed to PECF for 62 h @ 115°C, 30 psia.
b See Homsy, C.A. et al., J. Macromol. Sci.-Chem., A4(3), 615, 1970, for a more complete
description of polymers.
c Scale: +1 = some vacuolization, morphological changes, and growth inhibition; +2 =
moderate vacuolization, morphological changes; +3 = severe growth inhibition and vac-
uolization; +4 = total growth inhibition.
d Moles per million.
e Not detected.

conducted a “battery” of four in vitro tests using an ATCC murine fibroblast

cell line, L-929, and compared the results with the local response to 7-day
rabbit intramuscular implantation. Each test outcome was ranked 1+ to 4+
and pass/no pass (fail) criteria were established. The results of several of
their test series are summarized in Table 17.2.
Wieslander et al. (1990) extended these studies and examined the effects
of elution conditions. They suggested that testing using a combination of
water extracts obtained at elevated temperatures and physical contact
between cells and material produced the most reliable results.
In vitro tests have come to be seen as screening tests serving as precursors
for more involved, costly, and time-consuming animal implantation. A sec-
ondary use, resulting from extensive correlative investigations such as those
344 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 17.2
Correlation between In Vitro Screening
Tests and Short-Term Implantation
Implantation
Pass Fail
In vitro: Pass 94.7% —
Fail — 5.3%
N = 687.
Source: Adapted from Johnson, H.J. and North-
up, S.J., in Cell-Culture Test Methods, ASTM STP
810, Brown, S.A. (Ed.), American Society for Test-
ing and Materials, Philadelphia, 1983, 25.

of Johnson and Northup (1983), is for batch-to-batch quality control by

materials or device manufacturers.
These three studies and similar ones make use of a single in vitro cell line.
In fact, the ATCC L-929 cell line has come to be a de facto standard for in
vitro testing of biomaterials and is used in several ASTM in vitro host
response test methods, such as F-813 and F-895 (see Chapter 20). However,
a more favored approach is that of generating a cell culture spectrum (CCS).
In this technique, a number of different cell lines are used and are challenged
by the biomaterial (or extracts of the biomaterial) being studied in one or
more standard exposure techniques. The cell types are selected to represent
a variety of origins and sensitivities. The response of each of perhaps six cell
types is then quantified. An index based upon these quantitative findings is
thus a CCS reflection of relative acute host response. A consensus CCS would
be more sensitive in revealing differences between acute host response of
different materials. It can be hoped that, in the future, this method can be
fully developed, tested, and adopted by the ASTM and CCS indices devel-
oped for all current biomaterials. Then new materials can be compared,
insofar as acute response is concerned, by application of this test protocol
and reference to a table of standard results.
Although not reflecting consensus, a number of screening protocols in use
today utilize cell culture methods, with biomaterials or eluted materials as
the challenge, in evaluation of a quantitative host response to implant mate-
rials. For instance, primary testing (level 1) as utilized by Autian (1977) uses
five tissue culture tests as a portion of a 14-test initial battery that results in
the derivation of a cumulative toxicity index (CTI). The culture tests involve
one direct exposure model and four tests using extracts. Dillingham (1983)
has shown that these five tests produce significant degrees of correlation
with the in vivo tests incorporated in the CTI.
Northup (1986) has provided a useful summary of early efforts to use cell
culture techniques for the study of the generalized features of local host
response.
In Vitro Test Methods 345

17.3.6 Delayed Hypersensitivity Tests

Immune system (specific) response to biomaterials is extremely difficult to
evaluate in vitro. Numerous claims have been made about the ability to
screen individuals for hypersensitivity to foreign materials by challenging
cell-free serum with foreign materials and detecting the resulting anti-
gen–antibody complexes. These claims are probably groundless in a biom-
aterials host response context for two reasons: first, freely circulating
antigens are associated with humoral rather than the expected T-cell-medi-
ated response to foreign nonproteinaceous materials (see Section 12.3); sec-
ond, any response that occurs in vivo is to a complexed or opsinized material
rather than to the material itself.
In vitro testing for T-cell-mediated type 4 delayed hypersensitivity is pos-
sible, again depending upon use of an appropriate challenge agent. The most
reliable technique is the leukocyte migration inhibition factor (MIF) test
(Merritt and Brown 1980) (Section 12.4.2). Like the supposed humoral
response tests, however, this test is more a test of the prior sensitization of
a test animal or individual subject than of the ability of a specific material
to induce immune sensitization. Hallab et al. (2000) have summarized the
use of in vitro tests based upon leukocyte migration and concluded that those
based upon the transmembrane cell migration, as in the Boyden (1962)
chamber approach, appear best suited for studies in which multiple com-
parisons need to be made among subjects and/or materials and involving
sequential repetitions of measurements.
Proper conduct of this type of in vitro test requires considerable care and
the availability of known nonsensitive, nonimplanted animals or individuals
as longitudinal control donors. Care must be taken with all of these tests to
avoid spurious results due to possible inhibitory and/or cytotoxic properties
of the challenge agents used (see also Chapter 12). Other in vitro tests have
been used to examine immune sensitivity to biomaterials with varying
degrees of success (Merritt 1986).

17.3.7 Mutagenicity and Carcinogenicity Tests

The use of tissue culture in evaluation of biomaterials has centered on studies
of acute toxicity and of inhibition of growth or activity. Considerations of
carcinogenic potential of materials discussed in Chapter 13 suggest the need
for a rapid screening test for neoplastic transformation. Such tests exist, but
are quite controversial.
It is now generally accepted that chemical carcinogenesis proceeds by a
mutagenic route, with the neoplastic attributes heritable. Thus, any con-
firmed carcinogens are expected to be mutagens as well. However, because
the converse is not yet proven — that is, that all mutagens have the potential
of being carcinogens — the controversy over tests for carcinogenicity tends
to focus on interpretation of the results.
346 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 17.3
Results of Validation of Ames Test
Number Number Mutagenic
Status of Agent Tested (%)
Reported Animal carcinogens 176 158 (90)
Reported Human carcinogens 18 16 (89)
Reported Animal noncarcinogens 108 13 (12)
Source: Adapted from Ames, B.N., Science, 204, 587, 1979.

The most widely used test method depends upon evidence of mutagenesis
as an index of carcinogenic potential. This test, the Ames test (Ames et al.
1975), is conducted in the following way. A culture is prepared with a mutant
bacterial cell line (usually Salmonella typhimurium) that requires histidine for
growth. It is grown in histidine-free culture with the material to be tested
and a fresh preparation of rat liver cell homogenate. This provides enzyme
systems that will detect potential mutagenic materials that require metabolic
conversion to mutagens. Only cells that then mutate back to the more normal
histidine-independent state can multiply.
This test has been used widely and its results have been shown to have
very strong correlations with carcinogenic potential. Ames (1979) reported
on a study of 302 chemicals; the results are summarized in Table 17.3. The
finding of 12% “false” positives was discussed by Ames, who suggested that
a number of these may be weak carcinogens that were not previously
detected by (less sensitive) animal tests. This high degree of correlation
between mutagenicity and carcinogenicity (McCann et al. 1975) has sug-
gested to many workers that the Ames test has a useful role in early material
screening.
The Ames test has a number of attractive attributes. It is inexpensive, quick,
and relatively safe. By exposing approximately 106 organisms to a mutagen,
it can detect transformation rates far below those possible in any reasonable
animal test. Furthermore, it has been shown to be able to detect any of the
recognized mutagenic mechanisms. However, it has been strongly criticized.
Ashby and Styles (1978) had difficulty in reproducing Ames’ results, espe-
cially at low mutagen concentrations. They suggest that the sensitivity of
the test is lower than Ames asserted and also feel that correlations between
mutagenic potency and carcinogenic potency should be viewed with extreme
caution.
More recent general criticism of this and other tests of potential neoplastic
transforming agents has focused on the practice of the use of high doses of
challenge materials, which may also be directly cytotoxic. High doses are
frequently used to improve statistical response and reduce cost, followed by
extrapolation of effects to low doses. Many workers have suggested that
cytotoxicity encourages unnaturally elevated rates of cell replication, thus
artificially enhancing the rate of production of transformed cells and leading
to a too high risk after extrapolation (Epstein and Swartz 1988).
In Vitro Test Methods 347

It is clear that the Ames test, in common with other tissue culture tests,
must be done with great care. Precautions (DeSerres and Shelby 1979) should
include the use of multiple bacterial strains, multiple suspension systems, a
wide range of doses of material under evaluation, and use of positive and
negative controls in each study. The need for careful control of the form and
chemical composition of the material being studied was demonstrated by
Abbracchio et al. (1982), who showed that crystallinity plays a key role in
the biological response to particulate metal sulfides.
The need for great care in conducting the test in order to duplicate Ames’
results, as well as more fundamental criticisms concerning the relationship
between mutagenicity and carcinogenicity, led Purchase et al. (1978) to com-
pare six in vitro and in vivo tests for carcinogenic potential of organic com-
pounds. They concluded that the Ames type of test is highly accurate (93%
accurate in predicting animal carcinogenesis in their study vs. 90% for Ames,
1979) (Table 17.3); however, they also concluded that

The inclusion of the [other] four tests in a screening battery predictably

resulted in a great increase in overall inaccuracy and loss of discrimina-
tion, even though the detection of carcinogens is increased. All tests were
shown to generate both false positive and false negative results, a situ-
ation which may be controlled by the use, where possible, of chemical-
class controls, to identify the test which is optimal for the class of chem-
ical under test.

That is, they suggest that the use of compounds of similar structure as
positive and negative controls permits validation and selection of specific
tests from a battery of tests.
Forster (1986) has reviewed in vitro mutagenicity testing with special ref-
erence to evaluation of biomaterials. Despite the inherent shortcomings of
the test methods and the need to adapt them to the material in question and
the potential application, he suggests that in vitro mutagenicity tests should
play a role in any screening study of biomaterials. The strategy, then, should
be much the same as that for the use of tissue culture tests of toxicity.

17.4 Blood Contact Tests

17.4.1 General Comments
Materials problems in cardiovascular devices are primarily those of initial
or short-term inadequate biological performance due to the acute nature of
the host response to foreign materials in the cardiovascular system. Chapter
9 previously considered the principal problems of coagulation of blood and
lysis of blood-borne cells. These problems not only are intrinsic to materials
but also are extrinsically related to design features such as the presence of
348 Biological Performance of Materials: Fundamentals of Biocompatibility

interfaces, inappropriate flow rates, and localized turbulence conditions, as

well as overall device function.
Thus, it is hard to divorce the problems of determining host response to
materials for blood contact applications from the details of device design
and qualification. Materials testing, insofar as host response is concerned,
can be separated into three types of tests: in vitro static, ex vivo dynamic, and
in vivo dynamic. These are usually applied sequentially to a new material
or to a material being considered for a new application. It is important to
note with regard to this last comment that considerable regional variations
exist in environment in the cardiovascular system. It is well recognized that
a material that works well in a high-flow arterial application often fails in a
static venous site.
Initial testing for host response to blood contact materials is usually done
in vitro. Despite the wide recognition of the inadequacy of these tests, they
continue to be used on a screening basis for two reasons:

• They are relatively inexpensive compared with in vivo tests.

• They are not known to yield false negatives; that is, a material that
does poorly in these screening tests will not be a satisfactory implant
material in cardiovascular applications.

These tests are generally of a comparative type and examine coagulation

times or hemolysis rates in static or dynamic systems during or after contact
with foreign materials.

17.4.2 Static Tests

Of the static tests, the Lee–White and the Lindholm tests are best known
(Lindholm et al. 1973; Mason 1973; Mason et al. 1974). The reference material
is usually silicone-coated soft glass. Although this is not a satisfactory
material for implant use due to its mechanical properties, it is perhaps the
best inexpensive test material with a low and reproducible host response.
Few materials in use as cardiovascular implant materials exceed the biolog-
ical performance of this material in vitro. Repeated efforts have been made
to develop and widely distribute standard reference materials, such as low-
density polyethylene (LDPE) and polymethylsiloxane (PDMS) (Bélanger et
al. 2001), but they have generally been unsuccessful. Thus, the tests described
in this and the next section are more useful for comparing new materials
with older materials within a single study, using selected positive and neg-
ative “reference” materials, rather than providing intrinsic measurements of
host response.
The thrombotic potential of a surface is usually evaluated by timing the
development of clot in a static situation and comparing this time for clot
development, to the same point of maturity, for blood drawn simultaneously
from the same source (donor) and exposed to one or more reference surfaces.
In Vitro Test Methods 349

Adequate testing requires considerable replication in specimens in a single

test and in repetition of tests on different days, with different donors, etc. A
canine donor is usually used, although some investigators have used human
volunteer donors.
Hemolysis can be determined simultaneously in such a static test by cen-
trifugation of the clotted blood followed by optical spectrophotometric mea-
surement of the hemoglobin content of the supernatant serum. Again, it is
usual to compare the results obtained with a test material and reference
materials using blood drawn at one time (ASTM International [2004]: F756-
00). If the hemolysis rate is deemed acceptable for the intended application
on an in vitro screening basis, then more detailed static testing may be
performed, such as determination of complement activation (ASTM Inter-
national [2004]: F1984-99 ; F2065-00) and alteration of leukocyte morphology
(ASTM International [2004]: F2151-01) by biomaterial contact. Unfortunately,
like all blood response studies, these are effectively restricted to evaluation
of materials in solid (nonparticulate, nonsoluble) form.

17.4.3 Dynamic Tests

It is extremely difficult to evaluate coagulation effects of biomaterials on
blood ex vivo because the growth of the thrombus interferes with flow
dynamics in the system external to the donor and may have potential lethal
effects on the donor. A short-term solution to this problem that is only
applicable to relatively thromboresistant materials is to determine the den-
sity of adhered platelets after a brief exposure under controlled flow condi-
tions (Skarja et al. 1997). Arguing from knowledge of the role of platelets in
the intrinsic and common pathways (Figure 9.1), one can then predict relative
coagulation rates (Grabowski et al. 1977).
A further sophistication can be afforded by the wide variety of standard
in vitro quantitative assays available for the vast majority of factors involved
in the coagulation cascade. Such a study is performed by obtaining a blood
specimen, testing it for the presence of a clotting factor, and retesting after
brief, timed exposure to the test material (and a reference material), in much
the same manner as testing for complement activation (see previous section).
Although these tests are very precise, their specificity is unclear because it
is difficult to predict coagulation outcomes from the change in concentration
of any one factor.
If the material to be evaluated does not promote significant coagulation,
hemolysis can be quantified by exposing the surface to blood briefly, in a
transient mode, or over a longer period, in a stable ex vivo system. Erythro-
cyte damage is determined by direct analysis of released hemoglobin or by
counting the percentage of erythrocyte “ghosts” in a blood smear. In animal
ex vivo models, this test can be improved by prelabeling erythrocytes with
51Cr+6. This method takes advantage of the easy penetration of Cr+6 through

cell membranes and its intracellular conversion to Cr+3 that is then only
350 Biological Performance of Materials: Fundamentals of Biocompatibility

released if the cell wall is disrupted. Then, the 51Cr activity in cell-free serum
samples is a direct measure of hemolysis. Again, adequate testing requires
comparison with a standard surface and repetition of the test with a number
of donors. This approach is based on historic use of 51Cr+6 in clinical deter-
mination of blood volumes; however, modern concerns about the biological
activity of Cr[VI] now limit it to animal subjects.
In vitro and ex vivo tests share a number of problems. The biggest of these
is the nonstandard nature of blood. Blood is an extremely variable material.
Its coagulation and lytic response depend upon species, health of the indi-
vidual donor, diet, medication, and age, among other factors. Even the act
of drawing blood may change the host response capabilities of the donor’s
blood, thus rendering the use of a standard donor extremely misleading.
Therefore, all in vitro tests for blood–biomaterial interactions must lean
heavily on isochronal comparisons of the test material with a widely used
negative reference material such as siliconized glass and repeated tests using
different donors.
Another problem is the inability to reproduce implant site blood dynamics
ex vivo. Even the use of reasonable vessel sizes and flow rates based on study
of the planned implant site may prove misleading. The details of the prop-
erties of the vascular intima near the implant site have a profound influence
on the actual flow conditions, as do the features of the final implant/host
configuration. Thus, although some success is encountered in static and
single, or one-pass, in vitro tests, those based upon continued flow and
observations over a period of hours show poor correlation with in vivo
performance.
Nonetheless, these one-pass and continued-flow tests, in which a direct
connection between an animal’s or patient’s circulatory system is created,
are collectively termed ex vivo tests and, as the second type of test, are
becoming increasingly popular. In vitro or ex vivo tests for screening purposes
are no more than screening tests. They must be followed up by well-designed
in vivo testing, which is the subject of the next chapter.
A major effort to evaluate and compare in vitro (as well as ex vivo and in
vivo) tests was made by a working group originally assembled by the
National Heart Lung Blood Institute in 1985 (Harker et al. 1993).

17.5 Final Comments

The tone of this chapter may seem overly pessimistic, especially when read
against the background of the previous interpart on clinical study of biom-
aterials. Certainly, all biomaterials workers wish for the ease and precision
of the medical diagnostic tools seen on StarTrek but then return to the
laboratory to struggle with the inadequacies of established in vitro test meth-
ods. It is probably the case that development and improvement of test
In Vitro Test Methods 351

methods for in vitro or ex vivo prediction of host response has continued to

be neglected in recent years because this topic is not especially novel or
attractive to public or private funding agencies. Courtney et al. (1994),
Charissoux et al. (1996), and Kirkpatrick et al. (1998) have discussed these
general issues in more detail; Hanks et al. (1996) and Schmalz (1997) have
addressed issues specific to applications in the oral cavity.
However, even the crude methods presently available have much to com-
mend them due to low cost, relative speed and ease of conduct, and low
risk to participants. Therefore, I suggest — especially given an increased
societal sensitivity to the use of animals in medical research — that renewed
interest and effort in this area might be richly rewarded.

References
Abbracchio, M.P., Heck, J.D. and Costa, M., The phagocytosis and transforming
activity of crystalline metal sulfide particles are related to their negative surface
charge, Carcinogenesis, 3, 175, 1982.
ASTM International, Standard practice for assessment of hemolytic properties of
materials, F756-00, in 2004 Annual Book of ASTM Standards, Vol. 13.01: Medical
Devices; Emergency Medical Services, ASTM International, West Conshohocken,
PA, 2004.
ASTM International, Standard practice for testing for whole complement activation
in serum by solid materials, F1984-99, in 2004 Annual Book of ASTM Standards,
Vol. 13.01: Medical Devices; Emergency Medical Services, ASTM International,
West Conshohocken, PA, 2004.
ASTM International, Standard practice for testing for alternate pathway complement
activation in serum by solid materials, F2065-00, in 2004 Annual Book of ASTM
Standards, Vol. 13.01: Medical Devices; Emergency Medical Services, ASTM Inter-
national, West Conshohocken, PA 2004.
ASTM International, Standard practice for assessment of white blood cell morphol-
ogy after contact with materials, F2151-01, in 2004 Annual Book of ASTM Stan-
dards, Vol. 13.01: Medical Devices; Emergency Medical Services, ASTM
International, West Conshohocken, PA, 2004.
Ames, B.N., Identifying environmental chemicals causing mutations and cancer,
Science, 204, 587, 1979.
Ames, B.N., McCann, J. and Yamasaki, E., Methods for detecting carcinogens and
mutagens with the salmonella/mammalian-microsome mutagenicity test, Mu-
tat. Res., 31, 347, 1975.
Ashby, J. and Styles, J.A., Does carcinogenicity potency correlate with mutagenic
potency in the Ames assay? Nature, 271, 452, 1978.
Autian, J., Toxicological evaluation of biomaterials: primary acute toxicity screening
program, Artif. Organs, 1(1), 53, 1977.
Bélanger, M.-C. et al., Hemocompatibility, biocompatibility, inflammatory, and in vivo
studies of primary reference materials low-density polyethylene and polydim-
ethylsiloxane: a review, J. Biomed. Mater. Res. (Appl. Biomater.), 58, 467, 2001.
Boyden, S., The chemotactic effect of mixtures of antibody and antigen on polymor-
phonuclear leukocytes, J. Exp. Med., 115, 453, 1962.
352 Biological Performance of Materials: Fundamentals of Biocompatibility

Charissoux, J.L. et al., Development of in vitro biocompatibility assays for surgical

material, Clin. Orthop. Rel. Res., 326, 259, 1996.
Courtney, J.M. et al., Biomaterials for blood-contacting applications, Biomaterials, 15,
737, 1994.
DeSerres, F.J. and Shelby, M.D., The Salmonela mutagenicity assay: recommenda-
tions, Science, 203, 563, 1979.
Dillingham, E.O., Primary acute toxicity screen for biomaterials, in Cell-Culture Test
Methods, ASTM STP 810, Brown, S.A. (Ed.), American Society for Testing and
Materials, Philadelphia, 1983, 51.
Epstein, S.S. and Swartz, J.B., Carcinogenic risk estimation, Science, 240, 1043, 1988.
Forster, R., Mutagenicity testing and biomaterials, in Techniques of Biocompatibility
Testing, Vol. II, Williams, D.F. (Ed.), CRC Press, Boca Raton, FL, 1986, 137.
Galin, M.A., Chowchuvech, E. and Galin, A., Tissue culture methods for testing the
toxicity of ocular plastic materials, Am. J. Opthalmol., 79, 665, 1975.
Grabowski, E.F. et al., Platelet adhesion to foreign surfaces under controlled condi-
tions of whole blood flow, Trans. Am. Soc. Artif. Intern. Organs XXIII, 141, 1977.
Hallab, N. et al., Hypersensitivity to metallic biomaterials: a review of leukocyte
migration inhibition assays, Biomaterials, 21, 1301, 2000.
Hanks, C.T. et al., In vitro models for biocompatibility, Dent. Mater., 12(3), 186, 1996.
Harker, L.A., Ratner, B.D. and Didisheim, P. (Eds.), Cardiovascular Materials and Bio-
compatibility, Cardiovas. Pathol. 2(3) (Suppl.), 1993.
Hay, R.J., Availability and standardization of cell lines at the American Type Culture
Collection, in Cell-Culture Test Methods, ASTM STP 810, Brown, S.A. (Ed.),
American Society for Testing and Materials, Philadelphia, 1983,114.
Homsy, C.A. et al., Rapid in vitro screening of polymers for biocompatibility, J.
Macromol. Sci.-Chem., A4(3), 615, 1970.
Johnson, H.J. and Northup, S.J., Tissue-culture biocompatibility testing program, in
Cell-Culture Test Methods, ASTM STP 810, Brown, S.A. (Ed.), American Society
for Testing and Materials, Philadelphia, 1983, 25.
Kirkpatrick, C.J. et al., Current trends in biocompatibility testing, Proc. Inst. Mech.
Eng., 212(H), 75, 1998.
Lindholm, D.D., Klein, E. and Smith, J.K., Relative thrombogenicity of blood interface
materials: methods and results, Proc. Dialysis Transplant Forum, 3, 39, 1973.
Mason, R.G., Blood compatibility of biomaterials: evaluation of a simple screening
test, Biomater. Med. Dev. Artif. Organs, 1(1), 131, 1973.
Mason, R.G. et al., Blood compatibility of biomaterials: further evaluation of the
Lindholm test, Biomater. Med. Dev. Artif. Organs, 2(1), 21, 1974.
McCann, J. et al., Detection of carcinogens as mutagens in the Salmonella/microsome
test: assay of 300 chemicals, Proc. Nat. Acad. Sci., 72(12), 5135, 1975.
Merritt, K., Immunological testing of biomaterials, in Techniques of Biocompatibility
Testing, Vol. II, Williams, D.F. (Ed.), CRC Press, Boca Raton, FL, 1986, 123.
Merritt, K. and Brown, S.A., Tissue reaction and metal sensitivity, Acta Orthop. Scand.,
51, 403, 1980.
Mital, M. and Cohen, J., Toxicity of metal particles in tissue culture. Part II: a new
assay method using cell counts in the lag phase, J. Bone Joint Surg., 50A, 547,
1968.
Northup, S.J., Mammalian cell culture methods, in Handbook of Biomaterials Evalua-
tion., von Recum, A.F. (Ed.), Macmillan, New York, 1986, 209.
In Vitro Test Methods 353

Pappas, A.M. and Cohen, J., Toxicity of metal particles in tissue culture. Part I: a new
assay method using cell counts in the phase of replication, J. Bone Joint Surg.,
50A, 535, 1968.
Purchase, I.F.H. et al., An evaluation of six short-term tests for detecting organic
chemical carcinogens, Br. J. Cancer, 37, 873, 1978.
Rae, T., A review of tissue culture techniques suitable for testing biocompatibility of
implant materials, in Advances in Biomaterials, Vol. 1: Evaluation of Biomaterials,
Winter, G.D., Leray, J.L. and de Groot, K. (Eds.), John Wiley & Sons, Chichester,
U.K., 1980, 289.
Schmalz, G., Concepts in biocompatibility testing of dental restorative materials, Clin.
Oral Invest., 1(4), 154, 1997.
Shanbhag, A.H. et al., Macrophage/particle interactions: effect of size, composition,
and surface area, J. Biomed. Mater. Res., 28, 81, 1994.
Skarja, G.A. et al., A cone-and-plate device for the investigation of platelet biomaterial
interactions, J. Biomed. Mater. Res., 34, 427, 1997.
Wieslander, A., Magnusson, Å. and Kjellstrand, P., Use of cell culture to predict
toxicity of solid materials in blood contact, Biomater. Artif. Cells Artif. Org., 18(3),
367, 1990.
Wilsnack, R.E., Quantitative cell culture biocompatibility testing of medical devices
and correlation to animal tests, Biomater. Med. Dev. Artif. Org., 4(3–4), 235, 1976.

Bibliography
Anderson, D., An appraisal of the current state of mutagenicity testing, J. Soc. Cosmet.
Chem., 29, 207, 1978.
Autian, J., The new field of plastics toxicology — methods and results, CRC Crit. Rev.
Toxicol., 2, 1, 1973.
Brown, S.A. (Ed.), Cell-Culture Test Methods, ASTM STP 810, American Society for
Testing and Materials, Philadelphia, 1983.
Douglas, J.F., Carcinogenesis and Mutagenesis Testing, Humana Press, Totowa, NJ, 1984.
Grandjean–Laquerriere, A. et al., Importance of surface area ratio on cytokines pro-
duction by human monocytes in vitro induced by various hydroxyapatite par-
ticles, Biomaterials, 26, 2361, 2005.
Hanson, S.R., Blood–material interactions, in Handbook of Biomaterial Properties, Black,
J. and Hastings, G. (Eds.), Chapman & Hall, London, 1998, 545.
Helgason, C.D. and Miller, C.L., Basic Cell Culture Protocols, 3rd ed., Humana Press,
Totowa, NJ, 2004.
Kitchin, K.T., Carcinogenicity: Testing, Predicting, and Interpreting Chemical Effects, Mar-
cel Dekker, New York, 1999.
Kleinschmidt, J.C. and Hollinger, J.O., in Bone Grafts & Bone Substitutes, Habal, M.B.
and Reddi, A.H. (Eds.), W.B. Saunders, Philadelphia, 1992, 133.
Moroff, G., Methods for evaluating alterations in platelet and red cell properties, in
von Recum, A.F. (Ed.), Handbook of Biomaterials Evaluation, 1st ed., Macmillan,
New York, 1986, 233.
Rice, R.M. et al., Biocompatibility testing of polymers: in vitro studies with in vivo
correlations, J. Biomed. Mater. Res., 12, 43, 1978.
354 Biological Performance of Materials: Fundamentals of Biocompatibility

Venitt, S., Crofton–Sleigh, C. and Forster, R., Bacterial mutation assays using reverse
mutation, in Venitt, S. and Perry, J.M. (Eds.), Mutagenicity Testing: A Practical
Approach, IRL Press, Oxford, 1984, 45.
Venitt, S. and Perry, J.M. (Eds.), Mutagenicity Testing: A Practical Approach, IRL Press,
Oxford, 1984.
von Recum, A.F. (Ed.), Handbook of Biomaterials Evaluation, 1st ed., Macmillan, New
York, 1986; 2nd ed., Taylor & Francis, Boca Raton, FL, 2004.
Waters, R., DNA repair tests in cultured mammalian cells, in Venitt, S. and Perry,
J.M. (Eds.), Mutagenicity Testing: A Practical Approach, IRL Press, Oxford, 1984,
99.
Wennberg, A. et al., A method for toxicity screening of biomaterials using cells
cultured on millipore filters, J. Biomed. Mater. Res., 13, 109, 1979.
Wilson, R.S. et al., Blood–material interactions: assessment of in vitro and in vivo test
methods, in Techniques of Biocompatibility Testing, Vol. II, Williams, D.F. (Ed.),
CRC Press, Boca Raton, FL, 1986, 151.
Yuspa, S.H., Tumor promotion in epidermal cells in culture, in Mechanisms of Tumor
Promotion, Vol. III: Tumor Promotion and Carcinogenesis in Vitro, Slaga, T.J. (Ed.),
CRC Press, Boca Raton, FL, 1984, 1.
18
In Vivo Implant Models

18.1 Introduction
18.1.1 Approaches to In Vivo Tests
After acute screening by physical or biological in vitro techniques, it is the
practice to test new implant materials, or old materials in significantly dif-
ferent applications, in extended-time, whole-animal tests. Although the use
of nonhuman species involves many limitations and compromises, it is the
common judgment that such tests involving the exposure of materials to
systemic physiological processes are a necessary practical and ethical prece-
dent to human clinical testing.
With the exception of materials for application in the cardiovascular sys-
tem, the site chosen for initial nonfunctional (see Section 18.2.1) testing is
usually in soft tissue. This decision is based on the assumption that cytotoxic
effects have a generality of action and because soft tissue sites can be
approached in animals with relatively minor surgery. For these reasons, the
peritoneal cavity (Wortman et al. 1983) and subcutaneous “air pouch” (Wil-
loughby et al. 1986) have been widely used as sites for acute in vivo screening
studies. For joint replacement or fracture fixation applications, initial implan-
tation is in cortical and, occasionally, cortico-cancellous bone. Specialized
sites such as the cornea and the cerebral cortex are used for materials
intended for specific, limited applications.
Chick embryo and fetal rodent models have been used; however, most
testing now uses mature, higher animals. Although still used in multispecies
tests, rodents are rarely used alone since the experience of the Oppenheimers
(see Chapter 13). A variety of sites have been used for chronic testing; the
most popular ones are:

• Subcutaneous
• Intramuscular (e.g., supraspinatus)
• Intraperitoneal

355
356 Biological Performance of Materials: Fundamentals of Biocompatibility

• Transcortical (e.g., femur, tibia, and cranium)

• Intramedullary (e.g., femur and tibia)

Functional testing (see Section 18.2.2) requires a wider range of species

due to the specific requirements of the material or device configuration
selected. No single species presents an ideal general model for the human
species, and some human structural functions, such as patellar motion, are
not attainable in animal models. Anderson and Hughes (1986) provide a
still relevant overview to the problems of selection of appropriate animal
models.

18.1.2 Animal Welfare

In recent years, the use of animals in medical research has come under strong
criticism from a number of organizations, particularly People for the Ethical
Use of Animals (PETA).* This criticism, strident at times, tends to ignore two
key features of animal use:

• Investigators are generally sensitive individuals concerned about the

well-being of their test subjects.
• Sound research applicable to problems of human disability and dis-
ease requires that the test subjects, of whatever species, generally be
healthy and not subject to avoidable stress (unrelated to the exper-
imental design).

Nevertheless, there is now a well-developed system for monitoring and

controlling animal research in the U.S. The enabling legislation is the Ani-
mal Welfare Act (7 USC 2131, December 23, 1985) and a final regulation
issued by the U.S. Department of Agriculture regarding inspection and
compliance (Dept. of Agriculture 1991). The requirements of the Animal
Welfare Act are summarized in an NRC publication (National Research
Council, 1996) that should be on the desk** of any investigator involved
in research using animals. The central points of this system of regulation
and inspection are:

• Research involving animals may be performed only under a pro-

spective protocol that describes the purpose and significance of the
experiment; the species, source, and number of animals to be stud-
ied; the procedures to be performed; and the general care and hous-
ing conditions of the experimental animals and identifies the
personnel involved in animal experimentation and animal care and
the qualifications of these personnel.

* http://www.peta.org.
** Also available online: http://oacu.od.nih.gov/regs/guide/guidex.htm.
In Vivo Implant Models 357

• The protocol must be reviewed and approved before initiation by

an institutional animal care committee, although this review applies
only to the animal welfare aspects of the protocol and not the sci-
entific content or objectives of the study.
• The facilities in which surgical and experimental procedures are
performed and animals are housed must meet specific minimum
requirements.
• The personnel, including students, who conduct the procedures,
must be adequately trained and prepared.
• The animal housing facilities and all procedures involving animals
must be under the supervision of a veterinarian.

Additionally, standards are set forth concerning specific housing require-

ments (cage type, size, etc.) for each species and for the provision of veter-
inary care to minimize pain and treat conditions unrelated to the experiment.
The vast majority of U.S. fund-granting agencies and organizations now
require that research proposals be reviewed before submission to determine
that they meet the requirements of the Animal Welfare Act. This is a worth-
while and humane system that may at times seem burdensome to the inves-
tigator, particularly because some institutional review committees have
experienced confusion when trying (in many cases mistakenly) to balance
concerns about animal welfare against the scientific and technological require-
ments of the proposed studies. This occasionally produces the unusual result
of withholding permission to perform procedures on animals that are routine
in human clinical practice, such as multiple procedures on the same individual.
However, such reviews are necessary because of societal requirements and
ethical considerations, and their conduct emphasizes the investigator’s pro-
fessional and ethical responsibilities for experimental animals.
If the in vivo testing is intended to support future claims of safety in the
clinical use of the material used in fabricating the animal implants, additional
requirements are imposed by the Good Laboratory Practices (GLPs) regula-
tion of the U.S. Food and Drug Administration.* Although these practices
generally describe good procedures for conducting research, three sections
explicitly relate to animal care facilities and the provisions made for care
within them. Although these generally duplicate the provisions of the Ani-
mal Welfare Act, the overall requirements imposed by the GLPs make safety
studies difficult to perform in university or other educational settings. Such
studies are often more easily carried out by one of the commercial enterprises
that have sprung up in recent years.
Three additional ethical responsibilities must be emphasized at this point:

• The general obligation is to avoid unnecessary use of animals as

experimental subjects. Part of the decision to use an animal model
* http://www.access.gpo.gov/nara/cfr/waisidx_01/21cfr58_01.html. 2001 CFR Title 21, Vol. 1,
Part 58.
358 Biological Performance of Materials: Fundamentals of Biocompatibility

should involve careful a priori investigation (and rejection) of alter-

native approaches (cell culture; mechanical, electronic, or computa-
tional simulations; etc.).
• A determination should be made that the experiment has not been
done before or that reliable data are not available from other sources.
• Once a determination has been made to use animals and a surgical
model has been selected, the experiment should be designed to use
the minimum but sufficient number to provide a reasonable assur-
ance of a statistically valid result. Failure to provide adequate pro-
spective statistical design in an animal experiment can lead to the
need to replicate the experiment and, in some cases, to use a signif-
icantly larger total number of animals than would have been
required by a better thought out approach. In this regard, sequential
designs that use a few animals to work out experimental design and
procedure uncertainties before larger statistically valid main exper-
iments are done are strongly recommended.

The National Institutes of Health (U.S.) now provide an extramural com-

petitive grant program for development of alternatives to the use of animals
in drug and biomaterials evaluation. In some areas, this has resulted in
significant progress, as in the development of a number of cell and tissue
culture screening techniques as candidates to replace the Draize rabbit ocular
irritation test (Curren and Harbell 1998). However, more complex clinical
applications, especially those involving blood contact and/or chronic
implantation, will probably always require the use of animal testing before
there is sufficient confidence in safety to conduct initial human studies.

18.2 Test Types

18.2.1 Nonfunctional Tests
Tests divide globally into two types: functional and nonfunctional. In the
nonfunctional type, the implant is of an arbitrary shape, perhaps in a form
required for later mechanical tests of material response, and “floats” pas-
sively in the tissue site. Ferguson et al. (1960) popularized the prototype for
such tests. This is a supraspinatus implantation in the rabbit in a cavity
produced by blunt dissection, followed by a 16-week observation period.
This test series formed the basis for a peer-reviewed, national-consensus
protocol now in use, the F 981 method of the American Society for Testing
and Materials (ASTM, 2004b) (see Chapter 18, Appendix 1).
Nonfunctional tests focus on the direct interactions between the substance
of the material and the chemical and biological species of the implant envi-
ronment. The absence of mechanical loads and other elements of function
In Vivo Implant Models 359

limits the usefulness of such tests. Thus, they tend to be of short to interme-
diate duration, usually a few weeks to 24 months in length. For these results
and those of later functional tests to be useful predictors of biological per-
formance in patients, the implants must be in a condition as close as possible
to the physical and chemical condition proposed to be used in the final
clinical application. Care must be taken, especially in cleaning and steriliza-
tion, because surface contaminants may affect the host response (Greenfield
et al. 2005).
A typical short-term study of this type is described in the British Standards
Institute Method BS 5736, part 2 (1981). This test protocol utilizes 1- × 10-
mm strips of material, with positive and negative controls, implanted by
blunt stylet in the paravertebral (supraspinatus) muscle of the rabbit and
recovered by sacrifice after 7 days. Thus, although it is an in vivo method, it
evaluates only acute soft-tissue response. Turner et al. (1973) have shown
that such a 7-day model, although capable of producing false negatives,
produces no false positives (when compared with local host response to 12
weeks), thus justifying the use of this model for acute in vivo screening. A
similar updated procedure for 7-, 30-, and 90-day studies is F 763 (ASTM,
2004a). To examine more fully the effects of systemic physiology on material
and host response, a longer term study of materials that pass such a short-
term in vivo screen is needed.
F 981 (Chapter 18, Appendix 1) illustrates the typical course of such a
longer term study. Standard-sized implants fabricated from the material in
question, as well as one or more well-characterized comparative control
materials, are used. They are selected so as to come close to SA/BW ratios
for major implantation in humans (see Section 15.4.4), at least for soft-tissue
sites. The method uses muscle sites in the rat, rabbits, and larger animals,
as well as femoral transcortical sites, if indicated by the proposed application.
Rats are maintained and sacrificed at 12, 26, and 52 weeks postimplantation.*
Evaluation consists primarily of inspection of the test and control implants
and the sites of implantation.
At present, no control materials for the F 981 protocol are universally
available. The method specifies the use of a number of metals and one
ceramic and one polymer that have a long history of experimental evaluation
and human clinical use as reference or control materials. However, individ-
ual variations in composition, surface properties, etc. in these materials make
intertest comparisons difficult. There is considerable continuing interest in
developing positive and negative metallic and polymeric controls that might
be made available through a national program such as the Standard Refer-
ence Material program administered by the National Institute of Standards
and Testing (NIST). No reference standards for studies of porous implants
are currently recognized.
* Previous practice had been to use dogs when larger species were needed and to utilize a 2-year
implantation period in addition. Improved understanding of host response has permitted the
routine use of alternative large animal species such as sheep and goats and eliminated the 104-
week sacrifice period as redundant in most cases. See SectionX1.9, Appendix 1 (Chapter 18).
360 Biological Performance of Materials: Fundamentals of Biocompatibility

Major defects in the design and practice of F 981 have been recognized.
Notwithstanding its strong position as a consensus test method providing
standardized results that can be compared with others (Escalas et al. 1975)*
even when variants from the method are used, these shortcomings must be
recognized. In its concentration on study of the implant site, F 981 overlooks
systemic or remote site effects unless they are extreme enough to produce
visible morbidity or mortality. Thus, remote site neoplastic transformation
will not be seen in this test even if adequate implantation time has elapsed.
Additionally, implantation times are probably too short in the rabbit or dog,
or even the rat, to overcome latency effects (see Section 13.2.5). Furthermore,
for practical reasons, the numbers of test animals used are so small that even
observation of a tumor at the implant site or at a remote site cannot be
evaluated statistically with any acceptable level of certainty. Thus, the
method is unsuited for yielding information beyond the magnitude of the
acute and chronic inflammatory processes and subsequent fibrotic responses.
A question can be raised about the quantitative significance of the fibrotic
response. Measurement of capsule thickness in muscle-site implant studies
tends to show large variances. In most studies, differences must exceed 25
μm to be statistically significant. This result is usually ascribed to “biological”
variation. A study (Kupp et al. 1982) suggests a different explanation. In this
study, sites were identified in the rat hind quarter that produced considerable
motion between implant and muscle (intermuscular) when the leg was
moved passively from full extension to full flexion and that produced neg-
ligible relative motion (intramuscular). The relative degrees of motion in the
two sites were verified radiologically. Spherical capped cylindrical implants
of a polyester (PE) and a polyacetal (PA) were placed in each of these two
sites, and capsule mean thicknesses were studied at 21 days. The results are
given in Table 18.1.
A one-way analysis of variance (ANOVA) was performed, using the
method of contrasts. A near significant effect of motion may be seen in the
response to PA (25.1 vs. 14.1 μm) (H2). However, the greater intrinsic reac-
tivity of PE masks this effect (H1). This greater reactivity may be seen by
comparison of the minimal motion sites, a more usual place for implant
evaluation (24.3 vs. 14.1 μm) (H3). An apparent additive effect of motion in
the maximal motion implant site masks this difference (H4). Thus, capsule
thickness seems to reflect the additive response to two factors: an intrinsic
(chemical) activity and an extrinsic (mechanical) activity.
It is clear from this study that relative tissue-implant motion may have a
large influence on the observed capsule thickness. It may be that specimen
geometries should be changed in this type of test or the specimen secured
to the surrounding tissue to minimize such motion variables. Experiments
of this sort are called two-factor experiments because they examine effects
related to chemical and mechanical factors. For more complete evaluations,

* Escalas et al. (1975) used a precursor method, ASTM F 361. See Rationale (X1), Appendix 1
(Chapter 18) for a more complete discussion.
In Vivo Implant Models 361

TABLE 18.1
Interaction of Intrinsic Reactivity and Motion in Implant Capsule
Thickness
Motion
Minimala Maximalb
Material Thickness of Capsule (mean ± S.E.M.)
PE 24.3 ± 2.9 μm 31.4 ± 2.7 μm
PA 14.1 ± 2.8 μm 25.1 ± 3.5 μm

ANOVA Contrasts
Hypothesis F Significance
H1: Tpe,min = Tpe,max 2.05 p > 0.2
H2: Tpa,min = Tpa, max 3.07 p = 0.125
H3: Tpe,min = Tpa, min 3.01 p = 0.125
H4: Tpe,max = Tpa,max 0.60 p > 0.5
a 0.00 ± 0.07 cm.
b 0.18 ± 0.07 cm.
Source: Kupp, T. et al., in Advances in Biomaterials, Vol. 3: Biomaterials 1980,
Winter, G.D., Gibbons, D.F. and Plenk, H., Jr. (Eds.), John Wiley & Sons,
Chichester, 1982, 787.

one must move to three-factor experiments that include (or control for)
electrical effects. Thus, in this study, it may have been the case that differing
electrical charge densities at the polymer–tissue interfaces could account in
part for the observed effects.
The findings of Kupp and colleagues (1982) are particularly important
when one is considering host response to the implant’s degradation prod-
ucts, such as wear debris or precipitated corrosion products, rather than to
the implant. Several models utilizing deliberate imposed motion at the
implant–tissue interface demonstrate effects on materials response (Over-
gaard et al. 1996) and host response (Howie et al. 1988).
In any case, implant shape must be carefully selected and standardized
for each test protocol to avoid effects on capsule thickness associated with
surface features of the biomaterial. It has long been recognized that sharp
edges on an implant produce a locally increased capsule thickness. Thus,
when a flat-ended cylinder is used (Wood et al. 1970), the familiar dog-bone
shaped capsule results from a local thickening over the circular edges of the
rod ends. This effect has been more generally studied by Matlaga et al. (1976),
and it has been shown that tissue response is inversely related to the included
angle at the edge of the implant. Finally, all soft-tissue sites, in any given
species, are probably not equivalent in terms of the fibrous response evoked
by an implant (Bakker et al. 1988).
One might further point out that observation of capsule thickness is an
indirect measure of cellular response. Dillingham (1983) has shown a good
correlation between capsular response and in vitro cytotoxicity; however, a
useful addition to this type of test might be to evaluate the concentration
362 Biological Performance of Materials: Fundamentals of Biocompatibility

and spatial distribution of cellular enzymes directly (Salthouse and Willigan

1972; Salthouse and Matlaga 1975).
Even if a refined histochemical evaluation system is not used, it might be
a good idea to evaluate tissue around animal implants in a more detailed
way than simply by measuring capsule thickness. Salthouse (1980) has sug-
gested using a scheme that incorporates a number of observations into an
overall index, analogous to the CTI approach of Autian (1977) in his level 1
test scheme. The scheme suggested by Salthouse is a modification of one
proposed by Gourlay et al. (1978), which is based upon the earlier work
of Sewell et al. (1955). An outline of the method is given in Chapter 18,
Appendix 2. The advent of modern image analysis systems now permits the
use of quantitative measures of tissue response (Hunt and Williams 1995);
however, earlier semiquantitative grading scales still remain in use.
In the long run, the most severe criticism of F 981 is its nonfunctional
aspect. Even when this drawback is offset by selecting tissue sites more
typical of the application — for instance, in the rabbit cornea for intraocular
applications — the results remain inferior in providing confidence in pre-
diction of human clinical experiences to those obtained from functional tests.
An alternate approach is to isolate the material from the tissue so as to
remove the mechanical features and examine the chemical (and presumably
electrical) effects on cellular aspects of local host response directly. Perhaps
the best known example of this approach is the cage implant system devel-
oped by James Anderson and his students (Marchant 1989; Kao and Ander-
son 1998). In this case, the implanted material is placed within a wire or
mesh enclosure and, after appropriate sterilization, implanted. This “cage”
will rapidly fill with a mixture of extracellular exudate and inflammatory
cells. The fluid may be sampled from time to time by transdermal puncture
and analyzed for enzymatic and cellular content, and a chronological picture
can be built of the local response to the implant. Although elegant in con-
ception and widely applied by its originators, the major criticism of this
approach is that the local host response to the implanted material is that of
cells already exposed to the cage material, which is usually different in
composition and surface condition. Although it is possible to overcome this
problem by fabricating the enclosure from the same material as that of the
test specimen; in the general case this is difficult and costly.

18.2.2 Functional Tests

18.2.2.1 General Requirements and Problems
Functional tests require that, in addition to being implanted, the material,
at least in some degree, be placed in the functional mode that it would
experience in human implant service. This is required to study, for instance,
tissue ingrowth into porous materials for fixation purposes, formation of
neointima in vascular processes, and production of wear particles in load-
bearing devices (and possible clinically relevant tissue response to them).
In Vivo Implant Models 363

Functional tests are obviously of much greater complexity and cost than
nonfunctional ones. Simple devices have been designed to exert dynamic
loads typical of the desired application on the material. An example of this
approach is the dynamic aortic patch (von Recum et al. 1978) for evaluating
materials for cardiac assist devices in the canine aorta.
Tests of materials for partial or total joint replacements present different
problems. In some cases, small human implants may be used directly, as in
the case of insertion of a phalangeal interphalangeal (PIP) joint in the fore
leg of a cat (Woodman et al. 1983). On the other hand, the design of a
completely functional animal version of the proposed device is frequently
required. Design, fabrication, mechanical testing, and implantation of these
devices may be more difficult than the final production of the device for
human use. In both cases, problems arise from the small size of economically
priced animals, differences in animal anatomy, and the inability of the animal
“patients” to cooperate actively with the experiment.
Despite these difficulties, total hip joint replacement designs, for example,
have been made and tested in rats, cats, dogs, sheep, and goats. Figure 18.1
illustrates the evaluation of a fibrous material for fixation by bony ingrowth.
Initial tests utilized nonfunctional, cylindrical transcortical plugs. This was
followed by implantation of a proximal medullary device that permitted
mechanical “pull-out” testing to evaluate the shear strength of the interface,
albeit under essentially unloaded conditions. Finally, the material was incor-
porated into a custom designed tibial component for a fully functional canine
total knee replacement (Berzins et al. 1994; Rivero et al. 1988).
In some cases, veterinary models of implants exist that can be used as
material test devices. Figure 18.2 is a radiograph of a total hip implanted in
a cat. The femoral component is a modified small size of a canine femoral
(Gorman) prosthesis; the UHMWPE cup was custom designed and is held
in a metallic retainer attached to the iliac crest with bone screws.
In other cases, a human clinical design can simply be scaled down to a suitable
size for an animal. Figure 18.3 illustrates a metallic femoral component designed
for cemented total hip replacement in 450-g rats, maintained at constant skeletal
dimensions by dietary control (Powers et al. 1995). The head diameter is 2 mm;
advantage was taken of rodentine anatomy to implant the devices functionally
(head placed medially) and nonfunctionally (head placed laterally). One of the
advantages of such small implants and test animals is that full histological
sections may be made with the components in place.
Fracture fixation designs have been tested in all of the species mentioned
earlier, as well as in larger ones such as cows and horses. For such relatively
small devices in larger species, it is often possible to use actual human
dimension prototype models rather than having to produce modified
designs.
Particular problems arise in each of these test methods. Some difficulties,
however, are common to all of them. Considerable interspecies variation is
found in the histological appearance of tissue. Additionally, local tissue
conditions that are abnormal in some species may be chronic in others.
364 Biological Performance of Materials: Fundamentals of Biocompatibility

FIGURE 18.1
Specimens for canine evaluation of fiber metal (titanium) as an ingrowth material. Clockwise
from bottom: transcortical plug, medullary pullout specimen (Ti6Al4V substrate), tibial com-
ponent for total knee replacement (Ti6Al4V substrate and pins, ultra high molecular weight
polyethylene articular surface). (From Berzins, A. et al., J. Appl. Biomater., 5(4), 349, 1994; Rivero,
D.P. et al., J. Biomed. Mater. Res., 22, 191, 1988; previously unpublished, used by permission.)

Misinterpretations of test results, insofar as local tissue response and remote

organ condition are concerned, have arisen from the involvement of clinical
(rather than veterinary) pathologists in evaluation of histology from implant
studies. Satisfactory and reliable evaluation of tissue from multispecies tests
can only be achieved by an individual specifically trained in comparative
histology and animal pathology.
The inability of animals to cooperate with treatment has already been
mentioned. Specifically, it is not possible to return an animal to full activity
In Vivo Implant Models 365

FIGURE 18.2
Radiography of feline THR. (From author’s research in collaboration with D. Nunamaker.
Implants provided by Richards Manufacturing, now Smith+Nephew Richards, Memphis, TN.)

on a planned schedule. The animal will move as it sees fit and set its own
schedule. Similarly, an animal cannot indicate or describe internal problems
of discomfort or pain. Experienced animal handlers may be able to detect
early signs of pain accompanying infection or tissue reaction. More com-
monly, these problems are not detected until the animal is systemically ill
or loses function in a limb, or, incidentally, at autopsy. The presence of
culturable infection of any origin at an implant site invalidates any obser-
vations on that animal (unless an infectious agent was deliberately intro-
duced as part of the experimental plan). Systemic infection imposes a
366 Biological Performance of Materials: Fundamentals of Biocompatibility

FIGURE 18.3
Section of femoral component of a rodentine total hip replacement (~25× magnification). Ma-
terial: stem/head F-75 type cast CoCr alloy; cement: F 451 acrylic bone cement. (Component
fabricated by DePuy, Inc., Warsaw, IN; section by L. Smith, Clemson University.)

significant stress on experimental animals, may cause weight loss, and casts
doubts on the validity of observations.
Test animals such as cats, dogs, rats, etc. have shorter life spans and higher
metabolic rates than humans (Brody 1945). Over and above particular vari-
ations in physiology, these factors introduce other problems of unknown
magnitude. For instance, what is the appropriate factor by which to scale
down an implant for an animal to experience the same apparent body load
of foreign material as man does (see Section 15.4.4)? Because lifetime as well
as neoplastic transformation induction times are shorter in these animals
In Vivo Implant Models 367

than in humans, how can these be scaled up to expected human experience?

The importance of this latter point is underlined by occasionally finding
implant site sarcomas in dogs in conjunction with high corrosion rate stain-
less steel implants in clinical veterinary practice (see Section 13.5). In dogs,
these have occurred after an average implantation time of 5.8 years. What
is the comparable period in man after which one should expect to see this
type of tumor, if the assumption is that the transformation mechanisms are
the same in both species?
Questions of this type are unresolved and are the subject of continuing
research. A final question related to these interspecies differences is how
long to test before true chronic conditions are realized. Although current
practice is to limit chronic implant models to 1-year duration, this question
remains unresolved on a universal basis. The initial expense of animals and
an annual holding cost for dogs frequently exceeding $2000 per animal make
this last question one of vital importance.

18.2.3 Cardiovascular Functional Tests

18.2.3.1 Material Tests
The previous comments have applied primarily to implants in locations
other than in the cardiovascular system. The nature of such implants is
largely functional and, as in the case of in vitro testing, must be discussed
separately.
In vivo testing in the cardiovascular system is an extension of the ex vivo
or dynamic type of testing previously discussed in Section 17.4.3. The simpler
form of such testing is the use of an implant of an idealized geometry rather
than a full-scale working device. Instead of reproducing the exact functional
design of the implant, a standard design of simple geometry is used to
introduce the material into a vascular process. Patches or daggers may be
attached to the vascular wall in various locations — in some cases with
provision for mechanical loading (von Recum et al. 1978). Sections of blood
vessels may be replaced, often with devices with deliberate flow-disturbing
defects in them such as in the Gott and Kusserow tests.
The Gott or canine vena cava test (Gott and Furuse 1971) is the better
known of these two methods. In this technique, rings 9 mm long by 8 mm
(OD), with a 7-mm-diameter lumen, are surgically inserted to replace a
portion of the inferior vena cava in the dog. The rings may have a small
internal constriction or web to increase blood turbulence. Groups of five
animals are used with implantation periods of 2 hours and, if the initial
group remains patent, 2 weeks. Evaluation is by examining patency of the
rings and degree of coverage by thrombosis on removal.
The Kusserow or renal embolus test (Kusserow et al. 1970) involves replac-
ing a portion of the suprarenal aorta with a similar ring. An infrarenal
constriction is produced by partial ligation to force a large portion of the
aortic circulation (an estimated 90%) through the renal arteries into the
368 Biological Performance of Materials: Fundamentals of Biocompatibility

kidneys. Groups of five to eight animals are used, with implantation times
of 3 days to 2 weeks. Evaluation is by examination of the rings and histo-
logical quantization of kidney infarcts secondary to emboli “shed” by the
implants.
The Gott test appears more severe with respect to adherent thrombi due
to lower flow rates in the canine vena cava than in the aorta, even after
partial ligation. However, the Kusserow test provides better overall evalua-
tion because it permits examination of adherent thrombus and remote emboli
(in the kidney primarily), thus more closely modeling human clinical expo-
sure.

18.2.3.2 Transitional Tests

A transitional form of this sort of testing may involve the use of a portion
of a clinical design implanted in a different location in an animal. Sawyer et
al. (1976) described an early example. In this study, sections of cardiac cath-
eters intended for human clinical use were implanted as segmental replace-
ments in jugular and femoral veins of dogs as well as being placed in the
more usual location in the right atrium through right jugular insertion.
This study illustrates a feature common to most cardiovascular tests of the
idealized type; that is, the response at 2 hours after implantation mirrors
and closely predicts that seen at 2 weeks. It is this acute response of the
cardiovascular system to foreign materials that makes functional testing so
difficult. The early events are complicated by the establishment of equilib-
rium between material and host and by the events of trauma associated with
implantation, and they are difficult to study due to the requirements to
support the test animal clinically. However, from the point of view of the
test and of the eventual clinical response, the acute response may be the
most important. Thus, it is often said of materials tested in vivo for cardio-
vascular applications: “If they will last 2 hours, they will last 2 weeks; if they
will last 2 weeks, they will last 2 years.”

18.2.3.3 Device Tests

The more complex form of functional testing for cardiovascular application
is the evaluation of a material fabricated into a final device design. The costs
and difficulties associated with such tests can be easily appreciated. A major
problem is the same as that which faces evaluation of any human clinical
device: the animal implant site is not the same as the human site, so problems
of comparison and scaling arise. These scaling problems are particularly
acute in the case of blood contact materials because of the relative interspe-
cies constancy of the viscosity of mammalian blood. Calves are often used
for such studies; however, their continuing growth usually limits such tests
to 9 months’ duration. In the final analysis, materials for cardiovascular
applications can only achieve qualification by tests in actual clinical appli-
cations. On the face of it, this may seem to be an insupportable position;
In Vivo Implant Models 369

however, it must be recognized that such testing is always based upon two
prerequisites:

• The material must first perform sufficiently well in vitro and in

animal tests in which likelihood of success is good and chance of
failure is small.
• There must be a real potential benefit to the specific patients involved
in this study. (Chapter 19 deals with this point at greater length.)

Functional device tests using typical clinical geometries and methods of

construction and implantation are essential to examine the effects of the
details of construction on response to materials. For instance, for a new
design, it is difficult to predict the rate of hemolysis that will result from
relative movements of parts in a heart valve and the portion of this that can
be ascribable to materials selection. In this respect, full clinical testing is
highly desirable despite its risks and inherent ethical problems.

18.2.4 Human Tests

In the final analysis, clinical testing is the only technique by which the true
biological performance of implantable biomaterials can be determined. In
short, the only completely valid subject for study is the human being. When
necessary preconditions are met (Chapter 19), human implantation will
begin; the challenge is to obtain data from this experience.
The second consideration cited in the previous section — that is, that any
human clinical experiment must provide a potential benefit to the patients
involved — essentially prevents the use of humans as test subjects for bio-
materials per se. There are very rare exceptions to this rule, as in the study
of Hofmann and colleagues (1990) involving patients who were to receive
staged bilateral total knee replacement arthroplasties (TKAs). During the
first surgery, a TKA was performed on one side and plugs of porous implant
materials were inserted in the medial femoral condyle of the opposite knee.
At the second surgery 9 weeks later, the plugs were retrieved and the second
TKA performed without deviation from the technique that would have been
employed in the absence of the experimental implants. This study, which
was approved by a local review committee (see Section 19.2.1) and for which
each patient gave informed consent, probably represents the extreme limit
to which the “potential benefit” principle can be stretched. The need for such
tests is underlined by the observation of the investigators that their results
(ingrowth of bone into porous cobalt- and titanium-base alloy plugs) did not
replicate the responses seen earlier in a canine model (Hofmann 1993).
Although there is a marked lack of study of device function (and biological
performance of the materials involved) in patients (as noted in Interpart 2),
opportunities exist for examination of these issues during device retrieval
subsequent to clinical failure or at autopsy. References to a number of such
370 Biological Performance of Materials: Fundamentals of Biocompatibility

studies are provided in Interpart 2; Chapter 22 discusses them in further

detail. Although engineering tests on the device to ascertain, among other
data, the material response are routine, examination of the local host
response is somewhat more difficult. Efforts have been made to develop
comparative overall scores for local host response in parallel to those used
in animal studies. The “Mirra scale” (Mirra et al. 1976) (Chapter 18, Appendix
3) or modifications of it are still in general use to describe the largely mate-
rials-mediated response of patient tissues to partial and total hip and knee
replacement devices. Although it is highly desirable that this scale continue
to be used (so that historical comparisons can be made), more sophisticated
ones are needed, especially as materials and material configurations used in
implanted clinical devices continue to change.

18.3 A Final Comment

As consensus emerges concerning qualification of new materials, better judg-
ments about the role that each of the generic test methods will play can be
made. However, animal tests of the functional and nonfunctional type play
a vital part in determining material and host response in the application of
biomaterials. Despite their expense, complexity, and difficulty of interpreta-
tion, they will continue to be used for the foreseeable future.
Nevertheless, it is necessary to recognize that the limitations of animal
testing are becoming an increasing problem as clinical experience with estab-
lished biomaterials extends. It may be that their use in development and
selection of materials for novel devices and applications should be empha-
sized, rather than depending upon them for preclinical qualification for older
materials in established or new devices. For the latter situations, an increased
dependence on studying the actual clinical experience seems preferable.

References
American Society for Testing and Materials, Standard practice for short-term screen-
ing of implant materials, F 763-04, in, 2004 Annual Book of ASTM Standards, Vol.
13.01: Medical Devices; Emergency Medical Services, ASTM International, West
Conshohocken, 2004a.
American Society for Testing and Materials, Standard practice for assessment of
compatibility of biomaterials for surgical implants with respect to effect of
materials on muscle and bone, F 981-04, in, 2004 Annual Book of ASTM Standards,
Vol. 13.01: Medical Devices; Emergency Medical Services, ASTM International, West
Conshohocken, pp. 290ff, 2004b.
In Vivo Implant Models 371

Anderson, L.C. and Hughes, H.C., Experimental animal selection, in, Handbook of
Biomaterials Evaluation, 1st ed., von Recum, A.F. (Ed.), Macmillan, New York,
1986, 255.
Autian, J., Toxicological evaluation of biomaterials: primary acute toxicity screening
program, Artif. Organs, 1(1), 53, 1977.
Bakker, D. et al., Effect of implantation site on phagocyte/polymer interaction and
fibrous capsule formation, Biomaterials, 9, 14, 1988.
Berzins, A. et al., Effects of fixation technique on displacement incompatibilities at
the bone-implant interface in cementless total knee replacement in a canine
model, J. Appl. Biomater., 5(4), 349, 1994.
British Standards Institute, Evaluation of medical devices for biological hazards, BS
5736, Part 2: method of testing for tissue implantation, BSI, London, 1981.
Brody, S., Basal energy and protein metabolism in relation to body weight in mature
animals of different species, in Bioenergetics and Growth, with Special Reference
to the Efficiency Complex in Domestic Animals, Reinhold, New York, 1945, 352.
Curren, R.D. and Harbell, J.W., In vitro alternatives for ocular irritation, Environ.
Health Perspect., 106 (Suppl 2), 485, 1988.
Department of Agriculture, Animal welfare; standards; final rule, 9 CFR Part 3, Fed.
Reg. 56(Feb. 15), 6426, 1991.
Department of Health and Human Services, Guide for the Care and Use of Laboratory
Animals, NIH Publication 85-23, U.S. Government Printing Office, Washington,
D.C., 1985.
Dillingham, E.O., Primary acute toxicity screen for biomaterials: rationale, in vitro/
in vivo relationship and interlaboratory performance, in Cell-Culture Test Meth-
ods, ASTM STP 810, Brown, S.A. (Ed.), American Society for Testing and Ma-
terials, Philadelphia, 1983, 51.
Escalas, F. et al., MP35N: a corrosion resistant, high strength alloy for orthopedic
surgical implants: bioassay results, J. Biomed. Mater. Res., 9, 303, 1975.
Ferguson, A.B., Jr. et al., The ionization of metal implants in living tissues, J. Bone
Joint Surg., 42A, 77, 1960.
Gott, V.L. and Furuse, A., Antithrombogenic surfaces, classification, and in vivo eval-
uation, Fed. Proc., 30(5), 1679, 1971.
Gourlay, S.J. et al., Biocompatibility testing of polymers: in vivo implantation studies,
J. Biomed. Mater. Res., 12, 219, 1978.
Greenfield, E.M. et al., Does endotoxin contribute to aseptic loosening of orthopedic
implants? J. Biomed. Mater. Res., Appl. Biomater., 72B, 179, 2005.
Hofmann, A.A, Bachus, K.N. and Bloebaum, R.D., In vivo implantation of identically
structured and sized titanium and cobalt alloy porous coated cylinders into
human cancellous bone, Trans. Soc. Biomater., 13, 80, 1990.
Hoffman, A.A., Response of human cancellous bone to identically structured com-
mercially pure titanium and cobalt chromium alloy porous-coated cylinders,
Clin. Mater., 14, 101, 1993.
Howie, D.W. et al., A rat model of resorption of bone at the cement-bone interface
in the presence of polyethylene wear particles, J. Bone Joint Surg., 70A, 257, 1988.
Hunt, J.A. and Williams, D.F., Quantifying the soft tissue response to implanted
materials, Biomaterials, 16, 167, 1995.
Kao, W.J. and Anderson, J.M., The cage implant system, in Handbook of Biomaterials
Evaluation, 2nd ed, von Recm, A.F. (Ed.), Taylor & Francis, Philadelphia, 1998,
659.
372 Biological Performance of Materials: Fundamentals of Biocompatibility

Kupp, T. et al., Effect of motion on polymer implant capsule formation in muscle, in

Advances in Biomaterials, Vol. 3: Biomaterials 1980, Winter, G.D. Gibbons, D.F.
and Plenk, H., Jr. (Eds.), John Wiley & Sons, Chichester, U.K.,1982, 787.
Kusserow, B. et al., Observations concerning prosthesis-induced thromboembolic
phenomena made with an in vivo embolus shunt test system, Trans. Am. Soc.
Artif. Intern. Organs, XVI, 58, 1970.
Marchant, R.E., The cage implant system for determining in vivo biocompatibility of
medical device materials, Fund. Appl. Toxicol., 13, 218, 1989.
Matlaga, B.F. et al., Tissue response to implanted polymers: the significance of sample
shape, J. Biomed. Mater. Res., 10, 391, 1976.
Mirra, J.M. et al., The pathology of the joint tissues and its clinical relevance in
prosthesis failure, Clin. Orthop. Rel. Res., 117, 221, 1976.
National Research Council, Guide for the Care and Use of Laboratory Animals, National
Academy Press, Washington, D.C. 1996.
Nunamaker, D.M. and Black, J., Tissue response associated with ingrowth into porous
stainless steel, Trans. Orthop. Res. Soc., 3, 160, 1978.
Overgaard, S. et al., Role of different loading conditions on resorption of hydroxya-
patite coating evaluated by histomorphometric and stereological methods, J.
Orthop. Res., 14, 888, 1996.
Powers, D.L. et al., The rat as an animal model for total hip replacement arthroplasty,
J. Invest. Surg., 8, 249, 1995.
Rivero, D.P. et al., Calcium phosphate-coated porous titanium implants for enhanced
skeletal fixation, J. Biomed. Mater. Res., 22, 191, 1988.
Salthouse, T.N., Personal communication, 1980.
Salthouse, T.N. and Matlaga, B.F., An approach to the numerical quantitation of acute
tissue response to biomaterials, Biomater. Med. Dev. Artif. Org., 3(1), 47, 1975.
Salthouse, T.N. and Willigan, D.A., An enzyme histochemical approach to the eval-
uation of polymers for tissue compatibility, J. Biomed. Mater. Res., 6, 105, 1972.
Sawyer, P.N. et al., Experimental and clinical evaluation of a new catheter material,
Trans. Am. Soc. Artif. Intern. Organs, XXII, 527, 1976.
Sewell, W.R., Wiland, J. and Craver, B.N., A new method of comparing sutures of
ovine catgut with sutures of bovine catgut in three species, Surg. Gynecol.
Obstet., 100, 483, 1955.
Turner, J.E., Lawrence, W.H. and Autian, J., Subacute toxicity testing of biomaterials
using histopathologic evauation of rabbit muscle tissue, J. Biomed. Mater. Res.,
7, 39, 1973.
von Recum, A.F. et al., Biocompatibility tests of components of an implantable cardiac
assist device, J. Biomed. Mater. Res., 12, 743, 1978.
Willoughby, D.A. et al., The use of the air pouch to study experimental synovitis and
cartilage breakdown, Biomed. Pharmacother., 40(2), 45, 1986.
Wood, N.K., Kaminski, E.J. and Oglesby, R.J., The significance of implant shape in
experimental testing of biological materials: disc vs. rod, J. Biomed. Mater. Res.,
4, 1, 1970.
Woodman, J.L., Black, J. and Nunamaker, D.N., Release of cobalt and nickel from a
new total finger joint prosthesis made of vitallium, J. Biomed. Mater. Res., 17,
655, 1983.
Wortman, R.S., Merritt, K. and Brown, S.A., The use of the mouse peritoneal cavity
for screening for biocompatibility of polymers, Biomater. Med. Dev. Artif. Org.,
11(1), 103, 1983.
In Vivo Implant Models 373

Bibliography
Anderson, J.M., Biological response to materials, Annu. Rev. Mater. Sci., 31, 81, 2001.
Berry, C.L. (Ed.), The Pathology of Devices, Springer–Verlag, Berlin, 1994.
Department of Health and Human Services, Guidelines for Blood–Material Interactions,
NIH Pub. 85-2185, U.S. Government Printing Office, Washington, D.C., 1985.
Gad, S.C., Safety Evaluation of Medical Devices, 2nd ed., Marcel Dekker, New York, 2001.
Greco, R.S. (Ed.), Implantation Biology: The Host Response and Biomedical Devices, CRC
Press, Boca Raton, FL, 1994.
Homsy, C.A. et al., Surgical suture–canine tissue interaction for six common suture
types, J. Biomed. Mater. Res., 2, 215, 1968.
Horowitz, E. and Torgesen, J.L. (Eds.), Biomaterials, NBS Special Pub. 415, Washington,
D.C., 1975.
Leininger, R.L., Polymers as surgical implants, CRC Crit. Rev. Bioeng., 1, 333, 1972.
Loomis, T.A. and Hayes, A.W., Loomis’s Essentials of Toxicology, 4th ed., Academic
Press, New York, 1996.
von Recum, A.F. (Ed.), Handbook of Biomaterials Evaluation, 2nd ed., Macmillan, New
York, 1998.
Revell, P.A., Pathology of Bone, Springer–Verlag, Berlin, 1986, 215.
Schmidt–Nielsen, K., Scaling: Why Is Animal Size so Important? Cambridge University
Press, Cambridge, U.K., 1985.
Skurla, C.P. and James, S.P., Assessing the dog as a model for human total hip
replacement: analysis of 38 postmortem-retrieved canine cemented acetabular
components, J. Biomed. Mater. Res., Part B. Appl. Biomater., 73B, 260, 2005.
Williams, D. (Ed.), Biocompatibility of Implant Materials, Sector Publishing, London,
1976.
Williams, D.F. (Ed.), Techniques of Biocompatibility Testing, Vols. I and II, CRC Press,
Boca Raton, FL, 1986.
374 Biological Performance of Materials: Fundamentals of Biocompatibility
In Vivo Implant Models 375
376 Biological Performance of Materials: Fundamentals of Biocompatibility
In Vivo Implant Models 377
378 Biological Performance of Materials: Fundamentals of Biocompatibility
In Vivo Implant Models 379
380 Biological Performance of Materials: Fundamentals of Biocompatibility
In Vivo Implant Models 381
19
Clinical Testing of Implant Materials

19.1 Goal of Clinical Trials

After material selection, device design, in vitro tests, and implantation in
animals, a material must eventually be tested in humans. Such tests are
necessary because the goal of implant materials development, selection, and
testing is the alleviation of human disability and disease and because knowl-
edge about biological performance of materials is insufficient to predict
clinical success with confidence on the basis of only laboratory and animal
testing. In this chapter, some aspects of the design and conduct of clinical
trials will be briefly considered.
Before considering a clinical trial, it is well to consider the goal of such
evaluation. Unless an implant is designed for acute use, trials cannot extend
beyond a short fraction of intended device life. Furthermore, testing a new
or novel biomaterial in a particular device cannot result in qualification of
the material. Therefore, clinical trials must be regarded as serving primarily
as detectors of bad news; in much the same way that canaries in coal mines
warned of impending life-threatening gas concentrations, a clinical trial rep-
resents a limited, controlled, well-observed introduction of a new material
and/or design. The use of the term introduction is deliberate because, unlike
in the case of animal trials, the experimental subjects will continue to be
exposed to the device and its material components even after the period of
observation. The longer the time elapsed is and the greater the number of
patients studied during such limited introduction is, without detection of
adverse results, one is entitled to have greater confidence in acceptable
biological performance following general introduction.

383
384 Biological Performance of Materials: Fundamentals of Biocompatibility

19.2 Design of Clinical Trials

19.2.1 General Requirements
Burdette and Gehan (1970) identify four types or sequential phases of clinical
trials:

• Phase I — early trial: selecting a new treatment from among several

options for further study
• Phase IIA — preliminary trial: if the new treatment is not effective
in the early trial, this phase examines whether further studies should
be performed or the treatment abandoned
• Phase IIB — follow-up trial: estimating the effectiveness of a new
treatment that appears promising based upon phase I or phase IIA
trials
• Phase III: comparison of the effectiveness of the new treatment with
a standard method of management or some other treatment

In the human testing of materials within devices, phases I, IIA, and IIB
are rarely planned in a formal sense. Their function is usually fulfilled by
the use of individual custom devices for selected patients under the direct
care or supervision of the surgeon member of the research group. Only when
the new material/device (usually in comparison to other material/device
arrangements) is perceived to have relative benefit does formal clinical test-
ing begin with a phase III trial. This phase in implant development may be
further subdivided into two subphases (Burdette and Gehan 1970):

• Phase IIIA: examination of clinical outcome of a defined new treat-

ment for a group of patients with defined indications
• Phase IIIB: following success in a subphase IIIA trial, examination
of the clinical outcome for a defined (refined) new treatment for a
group of patients with defined (refined) indications, usually involv-
ing multiple investigators and institutions*

It is standard practice in new drug trials to employ the double-blind

method; that is, a drug and a harmless inactive material (placebo) are used
in a treatment plan for a defined group of patients with a common set of
symptoms. Which patients receive the drug and which the placebo is pre-
determined at random. Neither the patients (single blind) nor the treating
physician (double blind) knows whether they are receiving the active drug.
When the experimental trial is complete, an identifying code assigned to the
drug and placebo doses is deciphered and an analysis of effectiveness of the
* A very useful source for up-to-date information on clinical trials in the U.S. can be found at
http://www.clinicaltrials.gov.
Clinical Testing of Implant Materials 385

treatment is made. A further sophistication is employed in some designs:

the treatments for the two groups are “crossed over” or interchanged half-
way through the trial. Thus, each subject has a near equal chance of benefit
from the drug and of adverse effects from the drug or the placebo (because
administering the placebo prevents the use of other [previous] drugs known
to have beneficial effects on the patient’s condition).
When an implant is surgically inserted, whatever the phase of the trial, it
is not possible to pair the implanted patient with a placebo-treated patient.
The case of no insertion is clearly not blind to patient or doctor, and it would
not be ethically or practically acceptable in longer term trials. A study com-
paring identical devices made of different materials is at best blind to the
patient. Differences in appearance, weight, shape, etc. between devices made
of different materials usually render them easily distinguishable to the phy-
sician and the patient. Furthermore, as is noted in Chapter 21, the interrela-
tionship between materials selection and device design is such that it would
be unlikely that the two devices would differ only in respect to materials of
construction. It is more usual that several factors, including surgical tech-
nique, are changed. Although this still permits comparison between the old
and new treatments, the effect of materials change is statistically confounded
and thus cannot be observed with any certainty.
Therefore, clinical trials of implant materials must be based upon different
experimental designs. Comparisons may be made as follows:

• Between the condition of the patient before and after implant sur-
gery. This is useful to detect acute changes that may take place in
an individual with underlying disease (for which the implant is
indicated) that may be associated with response to the implant
material.
• Among patients with similar implants made of different materials.
When possible, this is useful to investigate acute and chronic differ-
ences in material and host responses.
• Between patients with implants and nondiseased (control) individ-
uals of the same age and sex, and with a similar home/workplace
environment. This may be useful for detection of subtle chronic
effects of materials (host response). Spousal or partner controls are
used in many such studies.

A number of efforts have been made to set standards for selection and
treatment of patients in clinical trials. Some of the general rules that have
emerged are:

• Medical care must be under the direction of a medical professional.

• The patients must give informed consent to any experimental pro-
cedure. This consent can only be obtained after a full explanation of
386 Biological Performance of Materials: Fundamentals of Biocompatibility

possible benefits and risks of the proposed procedure in comparison

to alternatives, including no treatment.
• The identity of the patients must be protected and the confidentiality
of their medical records must be preserved.
• Perhaps the most important point is that, for the trial to be justified,
whatever the phase, there must be a reasonable possibility of specific
benefit to the patients in the trial combined with reasonable assur-
ance of the absence of unusual risk.

The governing ethical considerations of which these are a part are the 12
basic principles of the Declaration of Helsinki II, revised and extended by
the 29th World Medical Assembly (Silverman 1985).
It is clear that clinical trial protocols for drugs or devices are difficult to
design and implement. As an aid in such efforts, virtually all medical
research and treatment facilities involved with patients maintain Human
Subjects Committees (also known as Internal Review Boards [IRBs]). These
committees are available to help in preparing protocols and generally must
review the procedures and safeguards in any experimental program involv-
ing human subjects before the clinical trial is started. Federal agencies now
make such reviews by IRBs a prerequisite to public funding of clinical
research. In addition, many of the Device Classification Panels of the Center
for Medical Devices and Radiological Health (CDRH) of the Food and Drug
Administration (FDA) have developed guidelines for design of clinical trials
and for statistical treatment and format for reporting results. If relevant
guidelines exist in the area under consideration, they should be consulted
at an early point of protocol development.
Although these detailed guidelines are of use, they have now generally
been supplanted by an FDA control document. This arises from the need to
obtain an exemption from certain provisions of the Medical Device Amend-
ments (1976) (see Section 20.2) in order to manufacture, ship interstate, and
implant the quantities of implants required for phase III clinical trials.* The
necessary authorization is obtained through a successful (approved) appli-
cation for an investigational device exemption (IDE).** The referenced por-
tions of the Code of Federal Regulations describe the procedure for
application for an IDE and the responsibilities of the sponsor, investigators,
and Human Subjects Committees (here called IRBs), as well as set standards
for informed consent, protection of patient confidentiality, and reporting of
study results.
It should be further noted that an IDE would not be granted unless the
supporting tests of the type to be discussed later in this chapter, as well as
others, meet the requirements of the regulations on good laboratory practice

* Devices required for earlier phases are manufactured individually at the surgeon’s or physi-
cian’s prescription and are permitted to be used under the custom devices provisions of the
Medical Device Amendments (1976).
** Federal Register 45(13):3732, 1980. See also Dobelle et al. (1980).
Clinical Testing of Implant Materials 387

(GLP).* These regulations set forth standards concerning design and docu-
mentation of preclinical trials, qualification of personnel, and preservation
and presentation of experimental results. Although the GLP regulations
apply only to aspects of testing to support claims of safety at this time, it is
not improbable that they will eventually be extended to apply to all aspects
of preclinical testing, as well as to a wide variety of biomedical research
efforts not directly associated with direct material and device development.
Before clinical trials of a new material (in a device configuration) can be
countenanced, at least two preliminary types of nonclinical tests of the
material seem imperative. These are in addition to tests that may be required
to demonstrate the safety and effectiveness of the material and the device
design before phase III clinical trials may begin. However, if GLPs are
observed, these preliminary test results may be used as part of a later IDE
application submission.

19.2.2 Preclinical Tests

The first of these two preliminary tests is acute screening based upon in vitro
and tissue culture techniques as outlined in Chapter 17. This area has no
general standards and many different protocols are in use. Completion of
these acute screening trials should lead to a second type of preliminary test,
the chronic animal demonstration test. The ASTM F 981 protocol (Chapter
18) is such a test. Chronic animal tests should essentially meet or exceed the
requirements of F 981. Equivalency to F 981 involves, as a minimum, the use
of:

• Multiple species
• The same control (reference) materials used in acute testing
• Group sizes and sacrifice schedules substantially equal to or greater
than those required by F 981
• Operative sites similar to those of the intended human application

If a material demonstrates that it is equal or superior to materials in present

use in both of these types of tests and no extraordinary hazards specifically
associated with it arise, the planning and execution of phases I and II clinical
trials seems warranted. The requirements for such clinical trials are probably
more stringent than those needed to demonstrate performance of new device
designs. This is the case due to the subtlety of many materials’ problems
and the general “endorsement” that a new material may achieve inferentially
after successfully completing its first clinical trial series.
The question of what tests are necessary and sufficient before phase III
clinical trials of new materials are warranted is extremely controversial in
all aspects. It would be very inviting to develop a consensus viewpoint or

* Federal Register 43(247):59986, 1978.

388 Biological Performance of Materials: Fundamentals of Biocompatibility

matrix into which all new materials, material combinations, and material
applications could be classified, thus settling the generic problem once and
for all. A number of groups, including working groups within various U.S.
and foreign national governmental agencies and national standards-making
organizations, are examining this approach to the question. Progress has
been slow, and the end product is clearly a long way away.
The first step was the development of ASTM F 748: Practice for Selecting
Generic Biological Test Methods for Materials and Devices. The standard
contains a recommended test matrix that distinguishes among external
devices, external communicating devices, and implants, and between tissue
types and contact periods (Figure 19.1). When originally written, generic
host response tests, including an F 981 type chronic implantation test, were
recommended for each exposure class. Although the tests were defined
generically, the ASTM F-4 committee has gone on to recommend specific test
procedures in each area that are now incorporated by reference (Table 19.1).
The success of this voluntary practice, originally adopted in 1982, led to
the so-called Tripartite Biocompatibility Guidance (TBG) (Kammula 1991),
which was ratified on April 24, 1987. This document was developed by a
joint U.S., Canada, and U.K. working group and was intended to assist
manufacturers and government health agencies in the three countries in
anticipating the information necessary for preclinical evaluation of new
materials. The TBG retained the matrix approach of F 748, deleted the dis-
tinction between contact periods (intraoperative, short term, or chronic) and
added several additional possible tests, including determination of the phar-
macokinetics of released material (“biological fate”) and of reproductive and
developmental toxicity. Unlike F 748, the TBG does not refer to specific
recommended test methods but simply puts forward recommended aspects
of such tests.
The formation of the European Economic Union (EU), effective at the end
of 1992, led to interest within the International Standards Organization (ISO)
in producing a systematic approach to selection of tests for biological eval-
uation of materials. This standard (ISO 10993-1, 1991) was drafted by Tech-
nical Committee 194 and draws very strongly on the TBG, which it has
effectively superseded. In fact, it uses a similar matrix and recommended
tests but restores the exposure classes of F 748 by distinguishing among three
conditions:

• Limited exposure (≤24 hours; includes intraoperative)

• Prolonged or repeated exposure (>24 hours but <30 days) (= short
term)
• Permanent contact (>30 days) (= chronic)

In addition, the ISO guidance document is somewhat subtler in its

approach in that it distinguishes among nine initial and four supplementary
evaluation tests. This approach recognizes the criticality of nine initial tests,
Clinical Testing of Implant Materials 389

Recommended Applicable Tests

Sys. Toxicity — Acute/Subchronic

Skin Irritation/Intracutaneous

Implanatation — Short-Term
Mucus Membrane Irritation

Implantation — Long-Term
Cell Culture Cytoxicity

Blood Compatibility

Immunogenicity

Carcinogenicity
Sensitization

Genotoxicity
Pyrogenicity
Hemolysis
Classification of Material or
Device and Application
External devices
Intact surfaces (all time periods) X X X
Breached surfaces
Intraoperative X X X
Short term X X X X
Chronic X X X X X
External devices communicating with
Intact natural channels
Intraoperative X X X X
Short term X X X X X X
Chronic X X X X X X X X X
Body tissues and fluids
Intraoperative X X X X A
Short term X X X X A X X
Chronic X X X X A X X X X
Blood path, indirect
Intraoperative X X X X X X X
Short term X X X X X X X
Chronic X X X X X X X X X
Blood path, direct
Intraoperative X X X X X X X
Short term X X X X X X X X X
Chronic X X X X X X X X X X X
Implanted devices principally contacting
Bone/tissue/tissue fluid
Intraoperative X X X X
Short term X X X X X X
Chronic X X X X X X X X X X
Blood
Intraoperative X X X X X X X
Short term X X X X X X X X X X
Chronic X X X X X X X X X X X X
Notes: X: recommended; A: may be considered (especially for central nervous system); intraop-
erative: < 24 hours; short term: up to and including 30 days; chronic: >30-days. Consult
standard for definitions of tissue types.
FIGURE 19.1
Selection of preclinical tests (ASTM F 748 Recommended Practice) (Adapted from ASTM F
748-04, ASTM Annual Book of Standards, Vol. 13.01, ASTM International, West Conshohocken,
PA, 2004 (see Table 20.3).
390 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 19.1
Test Methods Referred to in F 748-04
Test Type (per F 748-04) Recommended Test Protocol
Cell culture cytoxicity F 813, F 895, F 1027, F1903
Sensitization F 720 F 2147, F 2148
Skin irritation or intracutaneous F 719
Mucus membrane irritation F 749a
Systemic toxicity, acute or subchronic USP, F 750
Blood compatibility F 2151b
Hemolysis F 756
Pyrogenicity USP, LAL
Short-term implantation F 736, F 1408, F 1904
Long-term implantation F 981, F 1983
Immunogenicity F 1905, F 1906
Genotoxicity c

Carcinogenicity F 1439
Notes: For USP, see Section 20.2.1, for F XXX, see Table 20.3. LAL: lim-
ulus amebocyte lysate test.
a In suitable animal/tssue.
b See also tests for complement activation (F 1984, F 2065).
c No single test agreed.
Sources: Ross, V.C. and Twohy, C.W., Prog. Clin. Biol. Res., 189, 267, 1985;
Munson, T.E., Prog. Clin. Biol. Res., 231, 143, 1987.

excluding chronic toxicity and carcinogenicity, for the majority of short- and
intermediate-term applications; it makes no recommendations concerning
testing for reproductive and developmental toxicity and the biological fate
of degradation products, except for an application-by-application consider-
ation.
Comparing ISO 10993-1 and ASTM F 748, one observes that the former
emphasizes cytotoxicity, sensitization, and intracutaneous irritation more
than the latter. In addition, ISO 10993-1 is less tailored to differences in
exposure class. This probably reflects the overall EU regulatory outlook that,
in comparison to the U.S. approach, depends less on preintroduction testing
and more on clinical observation of outcomes of materials (and device) use.
Finally, ISO is following the example of F 748 and developing specific test
methods for many of the 13 generic test categories used in its selection matrix
(published as additional parts of standard ISO 10993).*
Although the matrix approaches taken to date in ISO 10993 and ASTM F
748 are extremely beneficial, the ideal generic test selection matrix should
incorporate the following criteria:

1. Separation of test requirements by the following technical aspects:

a. Type of tissue that the implant will contact (muscle, blood, etc.)

* Subsequent revisions over the years suggest that eventually no useful distinctions will be pos-
sible between F 748 and IS 10993-1.
Clinical Testing of Implant Materials 391

b. Duration of implant, by classes of time intervals (short-term,

intermediate-term, etc.)
c. Relative exposure of materials to the patient’s body (SA/BW
ratio, etc.)
2. Selection of tests by generic description (with minimum require-
ments) rather than by detailed specification of procedures
3. Specification of levels of certainty (“confidence levels”) rather than
setting specific sample or group sizes in individual tests

The ISO standard addresses some but not all of these concerns. In particular,
it fails to deal with 1c and 3. Great care must be taken in defining specific
test methods for host response in this context. Although the intent is clear
to set minimum requirements, the high cost of testing often converts these
minima into maxima. In the case of new classes of materials, this may permit
subtle but deleterious aspects of host response to be overlooked.
In practice, it appears that F 748, the TBG, and ISO 10993 have not been
strictly adhered to; that is, they appear to serve a useful role as benchmarks
by defining the consensus minimum preclinical testing required. Industrial
sponsors tend to develop their own test matrices, involving additional test-
ing, based upon these views (Stark 1991).

19.2.3 Clinical Trials

Finally, the time will arrive when confidence in a new material has risen
sufficiently that clinical trials can be begun with caution. Clinical trials must
be performed under the control of a defined (written) prospective protocol
that includes the following provisions:

• Description of the implant device (note that the implant site must
be that of the proposed application)
• Outline of indications for the surgical procedure
• Outline of the uniform surgical procedure used
• Outline of postoperative treatment
• Outline of follow-up schedule and postoperative evaluation tech-
niques

The following information is needed for adequate consideration and eval-

uation of clinical testing results (with respect to biological performance of
materials):

• Protocol as listed previously

392 Biological Performance of Materials: Fundamentals of Biocompatibility

• Identification of 200 patients, by code number, who constitute con-

secutive individuals seen by the treating physician and meet item 2
of the protocol
• Results of follow-up of these patients for a minimum of 5 years and
average of 7 to 10 years for those not lost to follow-up at an earlier
date (minimum of 100)*
• Summary of all adverse results (note that a statistical summary is
desirable; individual results with code should not be reported)

The last point should be dwelt upon. Presumably, at the time at which the
trial protocol was being developed, consideration was given to each of the
questions to be asked and the statistical measures to be used in answering
them. At the end of the trial, it is thus appropriate to suggest that statistical
measures be employed. Therefore, reports of clinical trials should take care
to:

• Discuss accuracy and precision of all measurements, where possible.

• Define a minimum confidence level for all statistical measures of the
data, usually p <0.05.
• Report confidence intervals or other measures of significance asso-
ciated with all derived parameters.
• Indicate the significance of any conclusion arrived at by analysis of
the trial.

19.3 Conclusions from Clinical Trials

19.3.1 Introduction
As pointed out in Interpart 2, no clinical trial can approach the numbers and
period of exposure that will be experienced when a material enters into
general use. Thus, in a sense, it is imperative that clinical testing never end.
The treating physician and his consultants should distinguish between new
and old materials and should continue to be sensitive to possible biological
performance problems associated with the use of either kind.
A well-designed and conducted clinical trial does not serve to qualify a
material. In addition to design studies, physical measurements, and acute
and animal tests, it provides data that are required for a decision to release
a product for general use. The decisions made along the path to release

* In practice, it is difficult to distinguish material trials from device trials. Unfortunately the de
facto standard for device trials is only 2 years’ minimum follow-up, and no additional follow-up
is generally recommended in the case of new materials. This is grossly inadequate for the eval-
uation of new materials.
Clinical Testing of Implant Materials 393

should all be based on appropriate statistical tests at a minimum confidence

level of 95%. Such release, when it occurs with the approval of the appro-
priate regulating agencies, is necessarily a risk. That is, the benefits attendant
to the use of the device incorporating the material in question are felt at the
moment of decision to outweigh the risks involved.
Throughout this book, aspects of biological performance, including
material and host response, have been considered. The latter sections of this
work have begun to examine the details of test methods for examining
biological performance leading to clinical use. Two factors are common to
all of these methods:

• Large investments of time and money are required to produce results

with reasonable levels of reliability and statistical significance due
to the variability of the biological systems involved.
• Large quantities of inductive reasoning must be used to apply the
results of in vitro and animal testing to the prediction of clinical
performance.

These two general conclusions overshadow all considerations of clinical

testing and introduction into routine clinical use of new materials or old
materials in new designs or applications.

19.3.2 Complication Incidence Rates

In discussions of clinical evidence for materials incompatibility (failure to
exhibit adequate levels of biological performance), time and again, incidence
rates of complications are small. For instance, the “variant poppet” problem
in early heart valve designs probably did not affect more than 3% of patients
receiving the prosthesis. The more adverse experience with an early design
of an anterior cruciate ligament replacement prosthesis (Chen and Black
1980) resulted in failure in a larger percentage of cases, but still probably
less than 20%. In many high-use applications, such as total hip and knee
joint replacement, device-related failures for all causes, including failure of
biological performance, are now appreciably less than 1% per year after
surgery.
Why then is the concern expressed here for examining biological perfor-
mance? A manufacturer could fairly argue that modern implant materials
have proven beneficial for 95 to 99%+ of patients receiving them. Similarly,
a study by the Carnegie–Mellon Institute of costs and benefits associated
with research to improve then current orthopaedic prosthetic devices con-
cluded that failure/complication rates, even in the 1970s, were acceptably
small and that the investment required to reduce these rates significantly
would be costly out of proportion to the resulting benefits (Piehler 1978).
This remains clearly the case today, with an additional problem: in some
cases, such as total hip replacement, present technology is so successful that,
394 Biological Performance of Materials: Fundamentals of Biocompatibility

even if they are actually superior to current choices, newer developments,

such as changes in the materials of the articulating wear pair, cannot rea-
sonably be expected to demonstrate statistical superior clinical outcomes due
to practical limitations on size and duration of clinical trials (Black 1996)
Thus, one is faced with the problem that one cannot define a failure or
complication rate during clinical trials and that such a rate, a priori, could
not be judged to be acceptable except in extreme cases. How then is one to
decide when a material has proven itself sufficiently, with respect to biolog-
ical performance, to move into general use?

19.4 Aspects of the Decision for General Clinical Use

19.4.1 Current Concerns
I believe that it is necessary to address this question in two frames of refer-
ence. The first is the current medical/legal environment. The general argu-
ments cited earlier concerning cost of additional testing and the “acceptable”
level of performance of current devices were raised time and again during
the hearings on the Kennedy (U.S. Senate) and Rogers (House of Represen-
tatives) bills that resulted in the Medical Device Amendments of 1976 (see
Section 20.2). However, they were more than offset by the force of individual
testimony concerning the human and financial costs to individuals as a result
of device malfunction and failure. Thus, although statistical failure rates are
low, the failure rate in a given individual with a defective device is perceived
as 100%. Despite the implementation of the 1976 legislation, continuing
concerns about the impact of individual device malfunction and failure led
to two subsequent safe medical devices acts (1990, 1992) (see Section 20.2).
This humane view of individuals rather than average statistics has resulted
in a dramatic rise in medical malpractice and damage suits associated with
device failure. A typical early report* summarizes five suits concerning
“defective” heart valves. Only one of the five alleged defects was connected
with the death of a patient; the damage claimed in the other four cases was
disability and the need for additional surgery. Total damages sought in the
five cases were $55 million. Although the final outcome of such cases will
not be known for some time, malpractice/device failure awards have already
exceeded $1 million in individual nonfatal cases and, by 1990, new cases
were being filed at a rate of 900/day, with average final awards of $300,000
(Kiplinger 1991).
As reflected in congressional testimony and in the results of malpractice
suits, public opinion contributed in no small part to the formula adopted in
the Medical Device Amendments (1976). The test of adequate performance
laid out in this law is that the device be “safe and effective” and expose the
* The New York Times, Dec. 4, 1977.
Clinical Testing of Implant Materials 395

patient to no “unreasonable” risk or hazard. Thus, decisions on what will

be acceptable failure rates remain subjective and depend on the continued
interaction of public opinion, expert advice, and administrative action.
Individuals are torn between two views. The heart says that no failure is
acceptable, especially if it were to happen to oneself or to one dear to one.
The head says that such a goal may be unattainable and approachable only
at prohibitive cost.

19.4.2 Response to Current Concerns

The search for a defect-free materials technology for medical and surgical
implants and devices has a precedent. Early in the development of intercon-
tinental ballistic missiles, it was recognized that, in the normal course of
events, the complexity of the electronic control systems required would lead
to nonfunctioning systems. That is, the level of reliability of individual elec-
tronic components, then exceeding 99.9%, was insufficient to permit systems
with 106 to 108 components to have any real level of satisfactory performance.
Two approaches were taken to deal with this problem. The first was the
idea of redundancy. Each section of a control system was to have one or
more “backup” sections that operated in support or in parallel and could
take over if the primary system failed. This led to the practice in the present
space shuttle program of having three (and in some cases four) parallel
systems serving each critical function.
The second approach was to adopt the position that no absolute level of
performance for an individual component was definitively acceptable. This
latter view led to a highly successful initiative, first instituted by the Mar-
tin–Marietta Company (Denver) and later by the U.S. Air Force and NASA,
called the Zero Defects Program. The basic concept is to test devices con-
tinually, even after they pass into active service, and feed the results back
into product improvement with a view to eventual “100%” satisfactory
performance through evolutionary change. The combination of these two
approaches contributed to the success of the Apollo Lunar Program and
the continued high level of performance of civilian and military aerospace
hardware.
I suggest that both of these concepts and an additional idea of a fail-safe
product can be applied to considerations of biological performance in the
current environment. The idea of redundancy is hard to apply directly
toward the design of devices for a number of reasons. It is applicable to
active devices, such as heart pacers, but far less so to joint replacements,
sutures, etc. However, redundancy can be incorporated into materials and
device testing. The F 981 protocol already does this to a degree in its use of
multiple animal species. Despite efforts to the contrary, multiple, parallel,
and overlapping tests should be continued, driven in part by financial con-
siderations and in part by greater, perhaps misplaced, confidence in short-
term test results.
396 Biological Performance of Materials: Fundamentals of Biocompatibility

The concept of zero defects can be incorporated into clinical evaluation

and experience by refusing to accept any given level of performance as
permanently satisfactory. Thus, one can insist that performance equal to or
better than that of materials in use today be demonstrated before new mate-
rials can go into clinical trials and routine use. However, this should not
blind one to the need to examine the performance of current (old) and future
(new) materials constantly with a view toward continual evolutionary
improvements where and when possible — thus the emphasis in Interpart
2 on increased attention to the human epidemiology of biomaterials.
The third idea of a fail-safe product can also be adopted. This is the concept
that the ill effects of the failure of a device to function in its intended mode
may be minimized by design* provisions. An example of fail-safe design in
everyday life is the Westinghouse type AB air brake used on railroad cars.
If the train separates at a coupling or the brake control system fails, the
system is designed to apply the brakes automatically in the individual cars
and to maintain braking until each system is manually released and reset.
In fracture fixation applications, such a concept might lead to the choice of
a material with lower strength and higher ductility, such as a stainless steel,
over alloys with higher strength but limited ductility. Here the fail-safe
feature is that the observation of permanent deformation of an internal
fixation device can lead to a change in external support prescribed by the
physician. Even though it may result in an angulation in the healed fracture,
the “failed” situation for a device fabricated from a ductile material (that is,
angulation) is far more acceptable to patient and physician than the “failed”
situation for the less deformable device — acute device fracture — that may
lead to additional disability, surgery, and, possibly, legal action.

19.4.3 Future Concerns

The second frame of reference that should be briefly examined with respect
to cost/risk/benefit aspects of material introduction is that of the future.
How can one judge when a new material or device is ready to enter use and
perhaps supplant present, apparently less effective products?
I think the answer is rather simple. If the ideas of safe failure modes in
design, redundancy in testing, and, especially, continual evolutionary per-
formance improvement are adopted, a real distinction between old and new
materials will no longer be present. No one would knowingly substitute an
inferior new product for a current product unless driven by inhumane
motives. With this in mind, I hope that progressive attitudes on the parts of
researchers, manufacturers, surgeons, and regulatory authorities will lead to
a continual upgrading of the performance of current materials and the grad-
ual introduction of new materials when subjective levels of safety and effi-
cacy, most probably defined by then-current experience, are reached.

* See Chapter 21 for a discussion of the materials’ design process.

Clinical Testing of Implant Materials 397

Critics would suggest the need for some absolute (minimum) level of
safety that must be obtained before introduction of a new material. With
respect to devices, it has been proposed that the following definition be used:
“A device is safe enough to use when it is no worse than others in use and
presents no greater hazard than the condition it is to be used to treat.” This
appears clear enough in instances in which large improvements in devices
(and materials) can be demonstrated and the conditions treated are life
threatening. In situations in which a new material represents an evolutionary
change in composition and/or processing and in subsequent behavior and
the aim is to improve the quality of life of the patient by alleviating a
condition of low mortality and morbidity, such a statement is a poor guide.
Hazards may be of a new and noncomparable type. Meaningful comparison
of hazards, in any case, is possible only when potential outcomes differ
greatly. Thus, I suggest that decisions on device and materials introductions
must be made on the individual merits and demerits of each situation and
not shackled by a set of rigid rules.
Except at an early point in this discussion, I have said nothing about the
costs associated with this approach to materials application in the medical
and surgical field. This was deliberate. I suggest that the analyses of the type
made by Piehler (1978) and others fail in the face of the human and emotional
aspects of this field. As long as the financial costs of devices remain a
relatively small component of the true cost of disease and disability, includ-
ing the cost of health care (as they currently are in the U.S.), money should
not be an important factor in these considerations. In individual cases, it is
clear that the increased cost of research, development, and testing of mate-
rials and devices that is the legacy of the Medical Device Amendments and
the Safe Medical Devices Acts, the increased number and size of malpractice
and product liability suits, and increased public attention will act to stifle
innovation. A situation parallel to that in the drug field has developed: an
improvement in the nature of products newly introduced and a tendency to
move research and development activities “offshore.” It is hard to pass
judgment on this continuing development. Whether it is good or bad, it is
coming about in response to a clear public demand for safe and effective
materials for medical and surgical implants and devices.
One can pass judgment, however, on the increasing trend towards market-
driven rather than technology-driven introduction of new devices and mate-
rials. Clinical experience with many existing materials now exceeds three
decades; thus, it is very difficult to argue that short-term (2- to 5-year) testing
of new materials is capable of revealing subtle or long-term defects in them
or, more directly, of providing the information needed to determine whether
the new material is equal to or exceeds the performance of the older material
that it may replace simply on a novelty basis. A Gresham’s law appears to
be operating in the development of medical and surgical devices and their
materials through which novelty has a market value. The drive to use new
materials is depriving patients of the proven performance of older ones. In
a free market system that provides many benefits and maximizes individual
398 Biological Performance of Materials: Fundamentals of Biocompatibility

freedom, it is hard to see how such a situation can be corrected. Physician

and patient education, more sophisticated regulatory approaches, and eco-
nomic restrictions imposed by widespread recognition of the need to curtail
the growth of medical expenses may help. However, one can expect to avoid
future problems associated with inadequate biological performance of mate-
rials in patients only through the bioengineer’s endorsement of the Hippo-
cratic injunction to, in the first place, do no harm.

19.5 Final Comments

One of the chronic problems that the materials scientist or engineer encoun-
ters in the medical device clinical literature is an inability to determine from
what materials devices were made and, even if that is possible, the details
of composition, processing, etc. selected within the generic material. For this
reason, I suggest that a radical change is needed in how the clinical use of
materials is viewed. The clinical introduction of a material should be viewed
more as the beginning of the qualification process rather than the end of it.
Studies of retrieved devices have helped to provide information concerning
biological performance in actual clinical settings. In the last two decades,
such studies have slowly been extended from examination of clinically
retrieved failed devices to study of successful devices obtained at autopsy
or, in some cases, still in use. However, lacking reliable incidence and prev-
alence data, such studies tend to be isolated examples of numerators for
which reliable denominators are, as yet, unavailable. Therefore, it seems
advisable to be at least as serious about durable medical devices, especially
chronic (>30-day) implants as one is about motor vehicles and devise some
form of internationally acceptable registration and tracking system. Chapter
22 discusses early efforts towards this goal in the U.S.
To further illuminate this need, consider two situations: the first use of an
adapted or new material as a biomaterial and the subsequent use of that
material in a second design or application. In the first case, the device
designer, surgical developer, IRB, and regulatory agency are all tempted to
ask, “Is the material biocompatible?” The main thrust of this work is to
suggest that the more appropriate question is, “What is the predicted bio-
logical performance of the material in the intended application?” The answer
to this question is then approached through theoretical and practical con-
siderations, laboratory testing, and a progression of in vitro and in vivo
biological testing before clinical evaluation, as described in this chapter,
begins. However, the second case is different. The material has been in
clinical use now for some years and the questions that need to be asked and
answered are different. Now actual experience can be drawn upon — unfor-
tunately not with the material but with devices fabricated in part or in whole
from it.
Clinical Testing of Implant Materials 399

Each component of an implant has three elements: a functional element,

a connectional element, and a structural element. That is, each component
has a desired function in relation to its site of implantation, a method of
connecting it to the surrounding tissues so that it remains in the intended
site and orientation, and a structural aspect to preserve the spatial relation-
ship between the sites of connection and of function. In some cases, compo-
nents may consist of a single material that can constitute all three elements,
such as a monofilament surgical suture. In other cases, different materials
may be joined in a single component, with each contributing one or more
of these elements. For instance, a femoral component of a hip replacement
may have an articulating ceramic head and a metal intramedullary stem
with a porous ingrowth fixation coating to attach it to bone.
Now suppose that a material has been in use as an articulating (functional)
element in the hip, but using it as an articulating element in the shoulder is
desired. The temptation is to revert to the initial question of biological per-
formance. However, this is not logical because there is (or should be) infor-
mation on the actual, rather than conjectural, performance of the material in
the biological environment of a human joint — in this case, one fairly similar
to that of the new application (hip vs. shoulder). Then the appropriate initial
question becomes, “What data and clinical experience support the assertion
that the material in question will prove safe and effective in the proposed
(new) clinical application?”
A study leading to answering this question now should have the following
charge:

Considering the material in question (a ceramic used for femoral heads

in the hip), in the proposed design in the proposed application (prosthetic
replacement of the humeral head), find and analyze the laboratory and
clinical predicates that support the assertion that its use will meet ap-
propriate regulatory standards for safety and effectiveness.*

The analysis then proceeds through the following secondary questions:

• For each component of the proposed design fabricated in whole or

in part from the material in question, the question is “What are the
laboratory and clinical predicates for friction and wear, structural
integrity, and fixation (including possible adverse local and systemic
reactions to wear debris and other degradation products) that would

* Safe and effective are foundational descriptors in medical device regulation. However, as is the
case for biocompatibility, they cannot be defined or determined on absolute bases. Therefore, sat-
isfaction of locally prevailing regulatory definitions and standards provides the usual test, rather
than any intrinsic de novo considerations. Note, however, that in U.S. experience, no legal con-
nection exists between a regulatory decision that a device is safe and effective for a given set of
indications and the actual observed clinical performance. This issue has been extensively liti-
gated, pro and con. However, this complex topic is beyond the scope of this work. For general
regulatory considerations, see Chapter 20.
400 Biological Performance of Materials: Fundamentals of Biocompatibility

lead one to conclude that this design in the proposed clinical appli-
cation will be safe and effective?”*
• For interfaces between such components and other components of
the proposed design, the question is “What are the predicates to
support the assertion that the use of the material in question will
not produce clinical outcomes inferior to those experienced with
materials now in general clinical use in the proposed market?”

It should be clear that the ability to answer this progression of questions

depends acutely on the quality and quantity of data that exist concerning
the actual material in question, its material and host responses in patients,
and the overall clinical outcomes as a function of time. In this situation, it
should be remembered and taken to heart that data are not the plural of
anecdote.

References
F 748-04 Standard Practice for Selecting Generic Biological Test Methods for Materials
and Devices, ASTM Annual Book of Standards, Vol. 13.01, ASTM International,
West Conshohocken, PA, 2004.
F 981-04 Standard practice for assessment of compatibility of biomaterials for surgical
implants with respect to effect of materials on muscle and bone, ASTM Annual
Book of Standards, Vol. 13.01, ASTM International, West Conshohocken, PA, 2004.
Black, J., Metal on metal bearings: a practical alternative to metal on polyethylene
bearings? Clin. Orthop. Rel. Res., 329S, S244, 1996.
Burdette, W.J. and Gehan, E.A., Planning and Analysis of Clinical Studies, Charles C
Thomas, Springfield, IL, 1970.
Chen, E.H. and Black, J., Materials design analysis of the prosthetic anterior cruciate
ligament, J. Biomed. Mater. Res., 14, 567, 1980.
Dobelle, W.H. et al., How to comply with the Food and Drug Administration’s new
“investigational device exemption (IDE)” regulations, including an application
form, Artif. Organs, 4(4), 1, 1980.
International Standards Organization, Biological testing of medical and dental ma-
terials and devices, Part 1: guidance on selection of tests. ISO/DIS 10993-1:1994.
ISO, Switzerland.
Kammula, R.G., Tripartite biocompatibility guidance for medical devices, in Biocom-
patibility Workshop Notebook, Duncan, P.E. and Wallin, R.F. (Eds.), Society for
Biomaterials, San Antonio, TX, 1991.
Kiplinger, A., The Kiplinger Washington Letter, 68(20), 4, 1991.
Munson, T.E., FDA LAL guideline — update, Prog. Clin. Biol. Res., 231, 143, 1987.

* Note that the question has been specialized for the application: friction and wear reflect the
functional element of this application; structural integrity the structural element, and fixation
the connectional element; the listing of possible adverse observations reflects a general under-
standing of clinical performance of joint replacements.
Clinical Testing of Implant Materials 401

Piehler, H.R., Regulating orthopedic surgical implants, Orthopaedic Rev., 7(1), 75, 1978;
effect of FDA Medical Device Amendments on the benefit and cost of implants,
(2), 65; better data acquisition and analysis are needed to pinpoint device failure
sources, (3), 97; orthopedic implant retrieval studies document a part of failure
story, (4), 79; orthopedic surgeon and patient play important roles in success
of implant, (5), 99; FDA Medical Device Amendments regulation of orthopedic
implants is misdirected, (7),103, 1978.
Ross, V.C. and Twohy, C.W., Endotoxins and medical devices, Prog. Clin. Biol. Res.,
189, 267, 1985.
Silverman, W.A., Human Experimentation: A Guided Step into the Unknown, Oxford
University Press, Oxford, 1985.
Stark, N.J., How to organize a biocompatibility testing program: a case study, Med.
Dev. Diag. Ind., 13(6), 68, 1991.

Bibliography
Fleiss, J.L., The Design and Analysis of Clinical Experiments, John Wiley & Sons, New
York, 1986.
Friedman, L.M. et al., Fundamentals of Clinical Trials, 3rd ed., Springer, New York, 1999.
Plantadosi, S., Clinical Trials: A Methodological Approach, Wiley-InterScience, New
York, 1997.
Peto, R. et al., Design and analysis of randomized clinical trials requiring prolonged
observation of each patient. I. Design, Brit. J. Cancer, 34, 585, 1976; Part II.
Analysis and examples, Brit. J. Cancer, 35, 1, 1976.
Rozovsky, F.A. and Adams, R.K., Clinical Trials and Human Research: A Practical Guide
to Regulatory Compliance, Jossey–Bass (John Wiley), New York, 2003.
Whitehead, J., The Design and Analysis of Sequential Clinical Trials, John Wiley & Sons,
New York, 1997.
20
Standardization and Regulation of
Implant Materials

20.1 Historical Perspective

The manufacture and sale of drugs in the U.S. has been under gradually
increasing federal regulation since passage of the Wiley Act in 1897 and the
first Pure Food and Drug Act of 1906. These acts, as well as subsequent ones,
were adopted against a background of the sale of patent medicines with
exaggerated claims and the production of food with extensive and deliberate
contamination. There are many horror stories about the effects of patent
medicines from the pre-Wiley Act era and the later period of weak legislation
up to the 1930s (see Lamb 1936; Mintz 1965). Perhaps the strongest single
factor in the initiation of federal regulation of food additives and purity was
the publication of The Jungle by Upton Sinclair (1906). This novel describes
the conditions in the processed meat industry in Chicago at the time in
horrifying detail and caused widespread revulsion to and rejection of pro-
cessed meats such as sausage, ham paste, etc.
Various legislative acts directed towards regulation of content and safety
of food, drugs, and cosmetics brought the U.S. Food and Drug Administra-
tion (FDA) into being. Although legislative authority probably existed from
1923 to regulate implants, practical regulation did not begin until the 1970s.
A series of amendments to the Food, Drug, and Cosmetic Act (1976), collec-
tively termed The Medical Device Amendments, was adopted and signed
into law on May 28, 1976. These amendments gave the then recently orga-
nized Bureau of Medical Devices and Diagnostic Aids* of the FDA broad
powers to regulate implants, surgical instruments, and medical devices as
articles of commerce. These powers generally parallel the powers afforded
in the regulation of drugs; differences in the law and the regulatory arrange-
ments reflect some of the differences between devices and drugs. These
amendments have been supplemented and modified to minor degrees by

* Now called the Center for Devices and Radiological Health.

403
404 Biological Performance of Materials: Fundamentals of Biocompatibility

legislative action; however, they underwent significant recent extension

beginning in 1990 by adoption of the Safe Medical Devices Act.
Numerous efforts at standardization and thus control of medical and sur-
gical devices and materials predate these legislative efforts and continue in
a supplementary and parallel fashion today.

20.2 Drug Standardization Activities

20.2.1 The U.S. Pharmacopeia
The idea of standardization of drugs and, more recently, of medical and
surgical devices and materials is quite an old one. The motives involved are
usually the related desires to assure reproducible effect (efficacy) while pro-
tecting the patient against hazards associated with adulteration, mislabeling,
misuse, etc. (safety).
The first concrete effort in this area in the U.S. was the proposal by Dr.
Lyman Spalding in January 1817 to establish a national pharmacopeia. The
pharmacopeia was seen as a widely agreed upon and accepted document
that would set out the composition, identity, properties, and, to some extent,
clinical behavior of drugs and other medical substances shown to be useful
— that is, beneficial in action. In response to Dr. Spalding’s proposal, the
First United States Pharmacopeial Convention (USPC) assembled in Wash-
ington, D.C., on January 1, 1820. The First U.S. Pharmacopeia was published
on December 15, 1820, in Latin and English. Its 272 pages listed some 217
drugs considered worthy of recognition. At that time, provisions were made
to hold subsequent meetings of the convention and to issue a revised phar-
macopeia every 10 years.
The first USPC and the First Revision Committee were composed exclu-
sively of physicians. By 1830, pharmacists had been invited to join the con-
vention and numbers of them have continued to join over the years. The
present bylaws of the United States Pharmacopeia require, however, that at
least one-third of the members of the Board of Trustees and the Committee
of Revision continue to represent the medical profession.
The initial policy of the USPC was to select the most fully established and
best understood substances from among those that possess medicinal power.
Over the years and through its various revisions, this principle had been
adhered to. The last independent version, The Pharmacopeia of the United States
of America (USP XIX 1975), was the 19th revision, published subsequent to
the USPC of April, 1970. It contains 1284 articles describing a somewhat
lower number of drugs and other medical agents. Implants and other med-
ical devices are not discussed in USP XIX, with two exceptions. The first
and more important of these exceptions is that provision is made for the
Standardization and Regulation of Implant Materials 405

definition and testing of glass and plastic containers for drugs. The methods
of test for containers outlined in USP XIX are:*

• Light transmission
• Chemical resistance (glass containers)
• Biological tests (plastic containers): injection of extracts and exami-
nation of 72-hour implants in rabbits and mice
• Physiochemical tests (plastic containers): extraction, residue iden-
tification, residue ignition, heavy metal content, and buffering capacity

The other medical device described is the absorbable surgical suture. This
is the so-called “catgut” suture, although the basic material is now derived
from other sources. USP XIX sets out methods of test and standards for
length, diameter, tensile strength, content of soluble chromium compounds,
and color of extracts, as well as describing methods of needle attachment
for these sutures.
Although neither of these device areas is directly applicable to implant
materials, many of the methods, particularly those used to qualify container
materials, have been used extensively by biomaterials investigators. Of inter-
est is the provision for the use of a standard implant reference material for
evaluation of the 72-hour animal tests. The material is a low-molecular-
weight polyethylene fiber that can be inserted through a hypodermic needle.
It is stocked in a supply maintained by the USPC.
In 1974, the USP and the National Formulary (NF) (see next section) were
combined. The current edition (USP 28 2005) continues the USP series as the
28th revision and includes the 23rd revision of the NF. Although they are
now published together, an internal distinction is maintained, with the USP
articles addressing drug composition and dosage (as well as general issues
of testing and packaging) and NF articles dealing with pharmaceutical ingre-
dients other than drugs. The combined USP/NF has been published every
5 years since 1975 and is enlarged by annual supplements and by a bimonthly
magazine, Pharmacopeial Forum; with the 2002 edition, it will now be pub-
lished annually. The rate of growth can be appreciated by noting the addition
of 112 chapters and monographs as well as 637 revisions of previous ones
in the 28th edition.
The current revision, USP 28 (2005), in addition to continuing the nonbi-
ological tests of previous revisions, now lists a total of six host response tests
(Table 20.1). It is of great interest that the in vitro test methods now cite ASTM
standards as references.

* USP XIX, p. 642.

406 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 20.1
Host Response Test Methods in USP 28
General article <82>: biological reactivity tests, in vitro
Agar diffusion
Direct contact
Elution
General article <83>: biological reactivity tests, in vivo
Systemic injection
Intracutaneous injection
Implantation
Source: USP 28, The Pharmacopeia of the United States of
America, 28th revision, incorporating The National Formu-
lary, 23rd revision, The United States Pharmacopeial Con-
vention, Inc., Washington, D.C., 2005.

20.2.2 The National Formulary

As mentioned earlier, another compilation of drugs and their properties is
the National Formulary, which first appeared in 1888. This is prepared and
published by the American Pharmaceutical Association, which was orga-
nized in 1852. The stated goals of the NF are similar to those of the USP,
with the exception that not only the drugs of the greatest therapeutic merit
are to be included, but also drugs of any demonstrated merit. This factor
and the domination of the NF by the manufacturers rather than the users of
drugs, leads to a somewhat different format and emphasis. The NF was
originally published at 10-year intervals, more recently in 5-year intervals
and, since 2002, annually. The last independent edition (see previous section)
was the 14th edition (NF 14 1975) and includes 1009 articles defining and
describing a somewhat greater number of drugs and medical materials.
NF 14 describes materials, in the nondrug sense, in only two areas. It makes
provisions for examination and qualification of glass containers for drug
packaging that are similar to and depend upon USP 19 provisions. In addi-
tion, special provisions are made for qualification of containers for oph-
thalmic preparations. These provisions include a previously mentioned eye-
irritation test using saline and cottonseed oil extracts in the eye of the albino
rabbit.

20.3 Biomaterials Standardization Activities

20.3.1 The American Dental Association
A number of efforts have been made in the standardization of biomaterials
(in the sense of materials without primary pharmacological effects). Begin-
ning in 1926, the American Dental Association (ADA) has sponsored and
Standardization and Regulation of Implant Materials 407

conducted a program to define the physical and chemical properties of

materials used in restorative dentistry. Over the years a large number of
specifications have been developed for various metal alloys, cements,
impression materials, casting and investment waxes, plaster, resin, and elas-
tomeric products, as well as cutting instruments and equipment for radiation
diagnosis and therapy.
In addition to these specifications, the ADA maintains a program to certify
specific dental material products and manufacturers of these certified prod-
ucts. The results of this program, carried out through a cooperative effort
with the National Bureau of Standards (now the National Institute for Stan-
dards and Technology [NIST]), was a periodic publication entitled Guide to
Dental Materials and Devices most recently published in a seventh edition
(ADA 1974).* This contains a great deal of technical information as well as
some 25 materials specifications. Today there are about 67 standards; how-
ever, they may be obtained only individually from the ADA or from the
American National Standards Institute, Washington, D.C. Table 20.2 lists
dental materials and materials test standards currently in use.

20.3.2 The American Society for Testing and Materials**

An effort of greater generality is that on the part of the American Society for
Testing and Materials (ASTM). This organization was founded in 1898 and
is the principal scientific and technical organization for the voluntary devel-
opment of standards on characteristics and performance of materials, prod-
ucts, systems, and services in the U.S. It performs its work through 130 main
technical committees with more than 20,000 active members.*** These com-
mittees function in prescribed fields under regulations that ensure balanced
representation by producers, users, and general interest participants.
In 1962, the Committee F4 on Medical Devices was organized (Brown and
Cook 1982). More recently, this committee was reorganized and renamed the
Committee F4 on Medical and Surgical Materials and Devices. It includes
within its organization a resources subcommittee with individual sections
devoted to specific materials classes such as polymeric materials, metallur-
gical materials, etc., as well as biocompatibility, and a series of subcommit-
tees in various surgical specialties such as orthopaedics, cardiovascular
surgery, neurosurgery, etc. The division of areas addressed by these latter
medical subcommittees approximately parallels that of the Device Classifi-
cation Panels established by the Food and Drug Administration subsequent
to the passage of the Medical Device Amendments (1976).

* The reason for discontinuation of this publication is unclear; however, the subsequent approval
of these standards by the American National Standards Institute (ANSI) and their joint publica-
tion renders the decision moot.
** Since 2004, called ASTM International; however, I have preserved the older name here because
it is more familiar to readers.
*** http://www.astm.org.
408 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 20.2
ANSI/ADA Dental Biomaterials Standards and Test Methods

Biomaterials

1 Alloy for dental amalgam

5 Dental casting alloys
6 Dental mercury
11 Agar impression materials
12 Denture base polymers
13 Denture cold-curing repair resins
14 Dental base metal casting alloys
15 Synthetic polymer teeth
16 Dental impression paste — zinc oxide-eugenol type
17 Denture base temporary relining resins
18 Alginate impression materials
19 Dental elastomeric impression material
20 Dental duplicating material
22 Intraoral dental radiographic film
24 Dental baseplate wax
27 Resin-based filling materials
30 Dental zinc oxide–eugenol and zinc oxide–noneugenol cements
32 Orthodontic wires
37 Dental abrasive powders
38 Metal-ceramic dental restorative systems
39 Pit and fissure sealants
42 Polymer-based crowns and bridges
57 Endodontic sealing material
69 Dental ceramic
75 Resilient lining materials for removable dentures — part 1: short-term materials
82 Dental reversible/irreversible hydrocolloid impression material systems
87 Dental impression trays
88 Dental impression alloys
96 Dental water-based cements

Test Methods

41 Biological evaluation of dental materials

82 Dental materials — determination of color stability
97 Corrosion test methods
Source: American Dental Association (ADA) http://www.ada.org.

The scope of this committee is the development of definitions of terms

and nomenclature, methods of test, specifications, and performance require-
ments for medical and surgical materials and devices. By 2004, the committee
had adopted and approved through society vote more than 75 biomaterials
specifications and 50 methods of test for biological response (Table 20.3), as
well as device standards and other methods of test. The biomaterials spec-
ifications are consensus documents that describe the results of present prac-
tice in the fabrication of these materials. In that they are adhered to and the
incorporated standardized materials are used as reference materials, the
methods of test can be considered as standard tests, within the meaning of
Standardization and Regulation of Implant Materials 409

TABLE 20.3
ASTM Biomaterials Standards and Methods of Testing for Host and Material
Responsea

Biomaterials

F0067-00 Specification for Unalloyed Titanium, for Surgical Implant Applications (UNS
R50250, UNS R50400, UNSR50550, UNS R50700)
F0075-01 Specification for Cobalt-28 Chromium-6 Molybdenum Alloy Castings and Casting
Alloy for Surgical Implants (UNS R30075)
F0086-04 Practice for Surface Preparation and Marking of Metallic Surgical Implants
F0090-01 Specification for Wrought Cobalt-20Chromium-15Tungsten-10Nickel Alloy for
Surgical Implant Applications (UNS R30605)
F0136-02A Specification for Wrought Titanium-6Aluminum-4Vanadium ELI (Extra Low
Interstitial) Alloy for Surgical Implant Applications (UNS R56401)
F0138-03 Specification for Wrought 18Chromium-14Nickel-2.5Molybdenum Stainless Steel
Bar and Wire for Surgical Implants (UNS S31673)
F0139-03 Specification for Wrought 18Chromium-14Nickel-2.5Molybdenum Stainless Steel
Sheet and Strip for Surgical Implants (UNS S31673)
F0451-99AE01 Specification for Acrylic Bone Cement
F0560-05 Specification for Unalloyed Tantalum for Surgical Implant Applications (UNS
R05200, UNS R05400)
F0562-02 Specification for Wrought 35Cobalt-35Nickel-20Chromium-10Molybdenum Alloy
for Surgical Implant Applications (UNS R30035)
F0563-00 Specification for Wrought Cobalt-20Nickel-20Chromium-3.5Molybdenum-
3.5Tungsten-5Iron Alloy for Surgical Implant Applications (UNS R30563)
F0602-98AR03 Criteria for Implantable Thermoset Epoxy Plastics
F0603-00 Specification for High-Purity Dense Aluminum Oxide for Medical Application
F0604 Specification for Silicone Elastomers Used in Medical Applications
F0620-00 Specification for Alpha plus Beta Titanium Alloy Forgings for Surgical Implants
F0621-02 Specification for Stainless Steel Forgings for Surgical Implants
F0639-98AR03 Specification for Polyethylene Plastics for Medical Applications
F0641-04 Specification for Implantable Epoxy Electronic Encapsulants
F0648-00E01 Specification for Ultra-High-Molecular-Weight Polyethylene Powder and
Fabricated Form for Surgical Implants
F0665-98R03 Classification for Vinyl Chloride Plastics Used in Biomedical Application
F0688-05 Specification for Wrought Cobalt-35 Nickel-20 Chromium-10 Molybdenum Alloy
Plate, Sheet, and Foil for Surgical Implants (UNS R30035)
F0702-98AR03 Specification for Polysulfone Resin for Medical Applications
F0745-00 Specification for 18Chromium-12.5Nickel-2.5Molybdenum Stainless Steel for Cast
and Solution-Annealed Surgical Implant Applications
F0754-00 Specification for Implantable Polytetrafluoroethylene (PTFE) Polymer Fabricated
in Sheet, Tube, and Rod Shapes
F0755-99R05 Specification for Selection of Porous Polyethylene for Use in Surgical Implants
F0799-02 Specification for Cobalt-28Chromium-6Molybdenum Alloy Forgings for Surgical
Implants (UNS R31537, R31538, R31539)
F0899-02 Specification for Stainless Steels for Surgical Instruments
F0961-03 Specification for 35Cobalt-35Nickel-20Chromium-10Molybdenum Alloy Forgings
for Surgical Implants (UNS R30035)
F0983-86R05 Practice for Permanent Marking of Orthopedic Implant Components
F0997-98AR03 Specification for Polycarbonate Resin for Medical Applications

(continued)
410 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 20.3 (CONTINUED)

ASTM Biomaterials Standards and Methods of Testing for Host and Material
Responsea

Biomaterials (continued)

F1058-02 Specification for Wrought 40Cobalt-20Chromium-16Iron-15Nickel-7Molybdenum

Alloy Wire and Strip for Surgical Implant Applications (UNS R30003 and UNS R30008)
F1088-04A Specification for Beta-Tricalcium Phosphate for Surgical Implantation
F1091-02 Specification for Wrought Cobalt-20Chromium-15Tungsten-10Nickel Alloy
Surgical Fixation Wire [UNS R30605]
F1108-04 Specification for Titanium-6Aluminum-4Vanadium Alloy Castings for Surgical
Implants (UNS R56406)
F1185-03 Specification for Composition of Hydroxylapatite for Surgical Implants
F1251-89R03 Terminology Relating to Polymeric Biomaterials in Medical and Surgical
Devices
F1295-05 Specification for Wrought Titanium-6 Aluminum-7 Niobium Alloy for Surgical
Implant Applications (UNS R567000)
F1314-01 Specification for Wrought Nitrogen Strengthened 22 Chromium/N 13 Nickel/N 5
Manganese/N 2.5 Molybdenum Stainless Steel Alloy Bar and Wire for Surgical Implants
(UNS S20910)
F1341-99 Specification for Unalloyed Titanium Wire UNS R50250, UNS R50400, UNS R50550,
UNS R50700, for Surgical Implant Applications
F1350-02 Specification for Wrought 18Chromium-14Nickel-2.5Molybdenum Stainless Steel
Surgical Fixation Wire (UNS S31673)
F1377-04 Specification for Cobalt-28 Chromium-6 Molybdenum Powder for Coating of
Orthopedic Implants (UNS-R30075)
F1472-02A Specification for Wrought Titanium-6Aluminum-4Vanadium Alloy for Surgical
Implant Applications (UNS R56400)
F1537-00 Specification for Wrought Cobalt-28 Chromium-6 Molybdenum Alloy for Surgical
Implants
F1538-03 Specification for Glass and Glass Ceramic Biomaterials for Implantation
F1579-02E01 Specification for Polyaryletherketone (PAEK) Polymers for Surgical Implant
Applications
F1580-01 Specification for Titanium and Titanium-6 Aluminum-4/tVanadium Alloy Powders
for Coatings of Surgical Implants
F1581-99 Specification for Composition of Anorganic Bone for Surgical Implants
F1586-02 Specification for Wrought Nitrogen Strengthened 21 Chromium/M10 Nickel/M3
Manganese/M2.5 Molybdenum Stainless Steel Alloy Bar for Surgical Implants (UNS
S31675)
F1609-03 Specification for Calcium Phosphate Coatings for Implantable Materials
F1713-03 Specification for Wrought Titanium-13Niobium-13Zirconium Alloy for Surgical
Implant Applications (UNS R58130)
F1813-01 Specification for Wrought Titanium/N 12 Molybdenum /N 6 Zirconium/N 2 Iron
Alloy for Surgical Implant (UNS R58120)
F1839-01 Specification for Rigid Polyurethane Foam for Use as a Standard Material for
Testing Orthopedic Devices and Instruments
F1855-00R05 Specification for Polyoxymethylene (Acetal) for Medical Applications
F1873-98 Specification for High-Purity Dense Yttria Tetragonal Zirconium Oxide Polycrystal
(Y-TZP) for Surgical Implant Applications
F1876-98R03E01 Specification for Polyetherketoneetherketoneketone (PEKEKK) Resins for
Surgical Implant Applications

(continued)
Standardization and Regulation of Implant Materials 411

TABLE 20.3 (CONTINUED)

ASTM Biomaterials Standards and Methods of Testing for Host and Material
Responsea

Biomaterials (continued)

F1925-99R05 Specification for Virgin Poly(L-Lactic Acid) Resin for Surgical Implants
F2005-00 Terminology for Nickel-Titanium Shape Memory Alloys
F2026-02 Specification for Polyetheretherketone (PEEK) Polymers for Surgical Implant
Applications
F2038-00R05 Guide for Silicone Elastomers, Gels and Foams Used in Medical Applications
Part I/M Formulations and Uncured Materials
F2042-00R05 Guide for Silicone Elastomers, Gels, and Foams Used in Medical Applications
Part II/M Crosslinking and Fabrication
F2063-00 Specification for Wrought Nickel-Titanium Shape Memory Alloys for Medical
Devices and Surgical Implants
F2066-01 Specification for Wrought Titanium-15 Molybdenum Alloy for Surgical Implant
Applications (UNS R58150)
F2146-01 Specification for Wrought Titanium-3Aluminum-2.5Vanadium Alloy Seamless
Tubing for Surgical Implant Applications (UNS R56320)
F2210-02 Guide for Processing Cells, Tissues, and Organs for Use in Tissue Engineered
Medical Products
F2211-02 Classification for Tissue-Engineered Medical Products (TEMPs)
F2224-03 Specification for High-Purity Calcium Sulfate Hemihydrate or Dihydrate for
Surgical Implants
F2229-02 Specification for Wrought, Nitrogen-Strengthened 23Manganese-21Chromium-
1Molybdenum Low-Nickel Stainless Steel Alloy Bar and Wire for Surgical Implants (UNS
S29108)
F2257-03 Specification for Wrought Seamless or Welded and Drawn 18 Chromium-14Nickel-
2.5Molybdenum Stainless Steel Small Diameter Tubing for Surgical Implants (UNS S31673)
F2311-03 Guide for Classification of Therapeutic Skin Substitutes
F2312-04 Terminology Relating to Tissue-Engineered Medical Products
F2313-03 Specification for Virgin Poly(glycolide) and Poly(glycolide-co-lactide) Resins for
Surgical Implants with Mole Fractions Greater than or Equal to 70% Glycolide
F2315-03 Guide for Immobilization or Encapsulation of Living Cells or Tissue in Alginate
Gels
F2386-04 Guide for Preservation of Tissue-Engineered Medical Products (TEMPs)
F2393-04 Specification for High-Purity Dense Magnesia Partially Stabilized Zirconia (Mg-
PSZ) for Surgical Implant Applications

Methods of Test for Host and Material Response

F0561-05 Practice for Retrieval and Analysis of Implanted Medical Devices and Associated
Tissues
F0619-03 Practice for Extraction of Medical Plastics
F0624-98AR03 Guide for Evaluation of Thermoplastic Polyurethane Solids and Solutions for
Biomedical Applications
F0719-81R02E01 Practice for Testing Biomaterials in Rabbits for Primary Skin Irritation
F0720-81R02E01 Practice for Testing Guinea Pigs for Contact Allergens: Guinea Pig
Maximization Test
F0732-00 Test Method for Wear Testing of Polymeric Materials Used in Total Joint Prostheses

(continued)
412 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 20.3 (CONTINUED)

ASTM Biomaterials Standards and Methods of Testing for Host and Material
Responsea

Methods of Test for Host and Material Response (continued)

F0746-04 Test Method for Pitting or Crevice Corrosion of Metallic Surgical Implant Materials
F0748-04 Practice for Selecting Generic Biological Test Methods for Materials and Devices
F0749-98R02E02 Practice for Evaluating Material Extracts by Intracutaneous Injection in the
Rabbit
F0750-87R02E01 Practice for Evaluating Material Extracts by Systemic Injection in the Mouse
F0756-00 Practice for Assessment of Hemolytic Properties of Materials
F0763-04 Practice for Short-Term Screening of Implant Materials
F0813-01 Practice for Direct Contact Cell Culture Evaluation of Materials for Medical Devices
F0895-84R01E01 Test Method for Agar Diffusion Cell Culture Screening for Cytotoxicity
F0897-02 Test Method for Measuring Fretting Corrosion of Osteosynthesis Plates and Screws
F0981-04 Practice for Assessment of Compatibility of Biomaterials for Surgical Implants with
Respect to Effect of Materials on Muscle and Bone*
F1027-86R02 Practice for Assessment of Tissue and Cell Compatibility of Orofacial Prosthetic
Materials and Devices
F1408-97R02E01 Practice for Subcutaneous Screening Test for Implant Materials
F1439-03 Guide for Performance of Lifetime Bioassay for the Tumorigenic Potential of
Implant Materials
F1635-04 Test Method for in Vitro Degradation Testing of Hydrolytically Degradable Polymer
Resins and Fabricated Forms for Surgical Implants
F1801-97 Practice for Corrosion Fatigue Testing of Metallic Implant Materials
F1830-97 Practice for Selection of Blood for in Vitro Evaluation of Blood Pumps
F1841-97 Practice for Assessment of Hemolysis in Continuous Flow Blood Pumps
F1877-98R03E01 Practice for Characterization of Particles
F1903-98R03 Practice for Testing for Biological Responses to Particles in vitro
F1904-98R03 Practice for Testing the Biological Responses to Particles in vivo
F1905-98R03 Practice for Selecting Tests for Determining the Propensity of Materials to Cause
Immunotoxicity
F1906-98R03 Practice for Evaluation of Immune Responses in Biocompatibility Testing Using
ELISA Tests, Lymphocyte Proliferation, and Cell Migration
F1926-03 Test Method for Evaluation of the Environmental Stability of Calcium Phosphate
Coatings
F1983-99R03 Practice for Assessment of Compatibility of Absorbable/Resorbable
Biomaterials for Implant Applications
F1984-99R03 Practice for Testing for Whole Complement Activation in Serum by Solid
Materials
F2003-02 Practice for Accelerated Aging of Ultra-High Molecular Weight Polyethylene after
Gamma Irradiation in Air
F2025-00 Practice for Gravimetric Measurement of Polymeric Components for Wear
Assessment
F2027-00E01 Guide for Characterization and Testing of Substrate Materials for Tissue-
Engineered Medical Products
F2064-00 Guide for Characterization and Testing of Alginates as Starting Materials Intended
for Use in Biomedical and Tissue/Engineered Medical Products Application
F2065-00E01 Practice for Testing for Alternative Pathway Complement Activation in Serum
by Solid Materials

* See Appendix 1 in Chapter 18.

(continued)
Standardization and Regulation of Implant Materials 413

TABLE 20.3 (CONTINUED)

ASTM Biomaterials Standards and Methods of Testing for Host and Material
Responsea

Methods of Test for Host and Material Response (continued)

F2102-01E01 Guide for Evaluating the Extent of Oxidation in Ultra-High-Molecular-Weight

Polyethylene Fabricated Forms Intended for Surgical Implants
F2103-01 Guide for Characterization and Testing of Chitosan Salts as Starting Materials
Intended for Use in Biomedical and Tissue-Engineered Medical Product Applications
F2129-04 Test Method for Conducting Cyclic Potentiodynamic Polarization Measurements
to Determine the Corrosion Susceptibility of Small Implant Devices
F2131-02 Test Method for in Vitro Biological Activity of Recombinant Human Bone
Morphogenetic Protein-2 (rhBMP-2) Using the W-20 Mouse Stromal Cell Line
F2147-01 Practice for Guinea Pig: Split Adjuvant and Closed Patch Testing for Contact
Allergens
F2148-01 Practice for Evaluation of Delayed Contact Hypersensitivity Using the Murine
Local Lymph Node Assay (LLNA)
F2149-01 Test Method for Automated Analyses of Cells/The Electrical Sensing Zone Method
of Enumerating and Sizing Single Cell Suspensions
F2150-02E01 Guide for Characterization and Testing of Biomaterial Scaffolds Used in Tissue-
Engineered Medical Products
F2151-01 Practice for Assessment of White Blood Cell Morphology after Contact with
Materials
F2183-02 Test Method for Small Punch Testing of Ultra-High Molecular Weight Polyethylene
Used in Surgical Implants
F2212-02 Guide for Characterization of Type I Collagen as Starting Material for Surgical
Implants and Substrates for Tissue Engineered Medical Products (TEMPs)
F2255-05 Test Method for Strength Properties of Tissue Adhesives in Lap-Shear by Tension
Loading
F2256-05 Test Method for Strength Properties of Tissue Adhesives in T-Peel by Tension
Loading
F2258-05 Test Method for Strength Properties of Tissue Adhesives in Tension
F2259-03 Test Method for Determining the Chemical Composition and Sequence in Alginate
by Proton Nuclear Magnetic Resonance (1H NMR) Spectroscopy
F2260-03 Test Method for Determining Degree of Deacetylation in Chitosan Salts by Proton
Nuclear Magnetic Resonance (1H NMR) Spectroscopy
F2347-03 Guide for Characterization and Testing of Hyaluronan as Starting Materials
Intended for Use in Biomedical and Tissue-Engineered Medical Product Applications
F2382-04E01 Test Method for Assessment of Intravascular Medical Device Materials on
Partial Thromboplastin Time (PTT)
F2392-04 Test Method for Burst Strength of Surgical Sealants
Note: F-999-02R03E02 means standard F999, adopted 2002, revised 2003, two editorial
changes (to 2003 version).
a ASTM standards are routinely revised on a 5-year cycle. A number of those listed here
have been reprinted more recently with minor or editorial changes only. Please refer to
most recent editions of ASTM annual standards books for current texts and explanatory
notes.
Source: 2005 Annual Book of ASTM Standards, Vol. 13.01. ASTM International, West Consho-
hocken, PA, 2004, xi–xv.
414 Biological Performance of Materials: Fundamentals of Biocompatibility

Section 1.2. Individual standards are available from ASTM, as well as annual
collections. Most technical libraries maintain recent full sets of ASTM stan-
dards in their reference collections; biomaterials standards are in annual
volume 13.01.
It is worth noting that, although ASTM F-4 standards for test methods are
consensus documents and have wide support in government, academia, and
industry, they are rarely used. That is, most investigators derive variations
of these procedures; however, when care is taken to meet the requirements
of the parent procedure, the revised and extended procedure is properly said
to adhere to the ASTM standard or recommended method of test.

20.3.3 Other Efforts

A number of other organizations have entered into the specification and
standardization of medical materials and devices. However, none are as
advanced in their efforts as the ADA and the ASTM. Perhaps the best known
of the remainder of these organizations is the Association for Advancement
of Medical Instrumentation (AAMI). This organization has been involved
for a number of years in developing specifications for active medical devices
such as heart pacers and neurostimulators. A number of these specifications
are coming into general use.
Outside the U.S., other countries have made progress in this field. Some,
like Canada, have decided to follow the progress of American groups such
as the ASTM and ADA. As standards have been adopted as American
National Standards (by the American National Standards Institute [designa-
tion: ANSI]), they are also being adopted after review as Canadian standards.
For instance, all of the ADA standards listed in Table 20.2 are now ANSI
standards, as are many ASTM standards. Some countries, such as Germany,
France, and England, have developed, relatively independently, their own
national standards for implant materials and devices. An international stan-
dards-making group, the International Standards Organization (ISO),* has
organized two committees with broad international representation:

• ISO TC 150 is evaluating national device and material standards and

attempting to adopt common international versions. This effort is
under way in the fields of orthopaedic, cardiovascular, and neuro-
surgery and will eventually spread to encompass all medical disci-
plines.
• ISO TC 194 is conducting similar activities in the area of measure-
ment of biological response.

Subject to approval by the European Commission, ISO standards began

to supplant or override national standards with the formation of the Euro-

* http//www.iso.org.
Standardization and Regulation of Implant Materials 415

pean Common Market in 1992 and, as a result, are becoming de facto stan-
dards for firms involved in international trade in medical and surgical
materials and devices.* Table 20.4 lists the ISO standards currently in force
for biomaterials and methods of test for host response. Individual AAMI
and ISO standards are available from the Association for Advancement of
Medical Instrumentation.**
Trade associations such as the Orthopedic Surgical Manufacturers Associ-
ation (OSMA) and the Health Industry Manufacturers Association (HIMA)
have taken active roles in developing standards on their own or through
activity of their representatives in standards-writing organizations such as
ASTM, ISO, etc. Traditional professional organizations in the health and
engineering professions also have standards committees that act as focal
points for technical input into standards preparation by standards-making
organizations.

20.4 U.S. Federal Regulation of Medical Devices and

Biomaterials
20.4.1 Medical Device Amendments (1976)
Although the various standardization efforts described in previous sections
continue through today, they were not generally perceived to be sufficient
to provide safe and effective medical and surgical materials and devices for
the public and to control unsafe and ineffective materials and devices. The
result in the U.S. was a legislative mandate for the executive branch of
government to control and regulate medical device manufacturing in the
public interest. The initial chosen legislative tool was the Medical Device
Amendments (1976) whose overall goal is to assure the safety and efficacy
of devices. This legislation provides for classification of devices, which will
be considered later in this chapter. They also lay out a scheme of general
controls including provisions for dealing with adulterated and misbranded
devices, for registering device types and device manufacturers, for premar-
ket notification of the introduction of new devices, and for dealing with
banned devices. Details of manufacturers’ obligations to repair, replace, or
refund in the case of defective devices are also included. The amendments
also permit the establishment of regulations to define “good manufacturing
practices” and of performance standards as well as premarket product devel-
opment protocols for new materials and devices.

* For a detailed idea of how these various standards and specifications interact generically with
the process of federal regulation of medical materials and devices, the reader is referred to Every-
thing You Always Wanted to Know about the Medical Device Amendments…and Weren’t Afraid to Ask,
HHS, FDA, Rockville, MD, FDA 92-4173.
** http://www.aami.org.
416 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 20.4
ISO Biomaterials Standards and Methods of Testing for Host Responsea

General Biomaterials

ISO 5832-1:1997 Implants for surgery — metallic materials — part 1: wrought stainless steel
ISO 5832-2:1999 Implants for surgery — metallic materials — part 2: unalloyed titanium
ISO 5832-3:1996 Implants for surgery — metallic materials — part 3: wrought titanium
6–aluminum 4–vanadium alloy
ISO 5832-4:1996 Implants for surgery — metallic materials — part 4:
cobalt–chromium–molybdenum casting alloy
ISO 5832-5:1993 Implants for surgery — metallic materials — part 5: wrought
cobalt–chromium–tungsten–nickel alloy
ISO 5832-6:1997 Implants for surgery — metallic materials — part 6: wrought
cobalt–nickel–chromium–molybdenum alloy
ISO 5832-7:1994 Implants for surgery — metallic materials — part 7: forgeable and cold-
formed cobalt–chromium–nickel–molybdenum–iron alloy
ISO 5832-8:1997 Implants for surgery — metallic materials — part 8: wrought
cobalt–nickel–chromium–molybdenum–tungsten–iron alloy
ISO 5832-9:1992 Implants for surgery — metallic materials — part 9: wrought high-nitrogen
stainless steel
ISO 5832-10:1996 Implants for surgery — metallic materials — part 10: wrought titanium
5–aluminum 2,5–iron alloy
ISO 5832-11:1994 Implants for surgery — metallic materials — part 11: wrought titanium
6–aluminum 7–niobium alloy
ISO 5832-12:1996 Implants for surgery — metallic materials — part 12: wrought
cobalt–chromium–molybdenum alloy
ISO 5833:2002 Implants for surgery — acrylic resin cements
ISO 5834-1:1998 Implants for surgery — Ultrahigh molecular weight polyethylene — part 1:
powder form
ISO 5834-2:1998 Implants for surgery — ultrahigh molecular weight polyethylene — part 2:
molded forms
ISO 10334:1994 Implants for surgery — malleable wires for use as sutures and other surgical
applications
ISO 13356:1997 Implants for surgery — ceramic materials based on yttria-stabilized tetragonal
zirconia (Y-TZP)
ISO 13779-1:2000 Implants for surgery – part 1 — ceramic hydroxyapatite
ISO 13781:1997 Poly (L-lactide) resins and fabricated forms for surgical implants — in vitro
degradation testing
ISO 13782:1996 Implants for surgery — metallic materials — unalloyed tantalum for surgical
implant applications

Dental Materials

ISO 1561:1995 Dental casting wax

ISO 1563:1990 Dental alginate impression material
ISO 1564:1995 Dental aqueous impression materials based on agar
ISO 1567:1988 Dentistry — denture base polymers
ISO 6871-1:1994 Dental base metal casting alloys — part 1: cobalt-based alloys
ISO 6871-2:1994 Dental base metal casting alloys — part 2: nickel-based alloys
ISO 6872:1995 Dental ceramic
ISO 6874:1988 Dental resin-based pit and fissure sealants
ISO 6876:2001 Dental root canal sealing materials
(continued)
Standardization and Regulation of Implant Materials 417

TABLE 20.4 (CONTINUED)

ISO Biomaterials Standards and Methods of Testing for Host Responsea

Dental Materials (continued)

ISO 6877:1995 Dental root-canal obturating points

ISO 7491:2000 Dental materials — determination of color stability of dental polymeric
materials
ISO 8891:1998 Dental casting alloys with noble metal content of 25% up to but not including 75%
ISO 9333:1990 Dental brazing materials
ISO 9693:1997 Metal-ceramic dental restoration systems
ISO 9694:1996 Dental phosphate-bonded casting investments
ISO 9917:1991 Dental water-based cements
ISO 9917-2:1998 Dental water-based cements — part 2: light-activated cements
ISO 10139-1:1991 Dentistry — resilient lining materials for removable dentures — part 1:
short-term materials
ISO 10139-2 Dentistry — resilient lining materials for removable dentures — part 2: long-
term materials
ISO 11244:1998 Dental brazing investments
ISO 11245: 1999 Dental restorations — phosphate-bonded refractory die material
ISO 11246:1996 Dental ethyl silicate-bonded casting investments
ISO 12163:1999 Dental baseplate modeling waxes
ISO 16744:2003 Dentistry — base metal materials for dental restorations
ISO 24234:2004 Dentistry — mercury and alloys for dental amalgam

Methods of Test for Host Response

ISO 7405:1997 Dentistry — preclinical evaluation of biocompatibility of medical devices used

in dentistry — test methods for dental materials
ISO 10993–1:2003 Biological evaluation of medical devices — part 1: evaluation and testing
ISO 10993-2:1992 Biological evaluation of medical devices — part 2: animal welfare
requirements
ISO 10993-3:2003 Biological evaluation of medical devices — part 3: tests for genotoxicity,
carcinogenicity and reproductive toxicity
ISO 10993-4:2002 Biological evaluation of medical devices — part 4: selection of tests for
interactions with blood
ISO 10993-5:1999 Biological evaluation of medical devices — part 5: tests for in vitro
cytotoxicity
ISO 10993-6:1994 Biological evaluation of medical devices — part 6: tests for local effects after
implantation
ISO 10993-10:2002 Biological evaluation of medical devices — part 10: tests for irritation and
delayed-type hypersensitivity
ISO 10993-11:1993 Biological evaluation of medical devices — part 11: tests for systemic toxicity
ISO 10993-12:2002 Biological evaluation of medical devices — part 12: sample preparation
and reference materials
ISO 10993-16:1997 Biological evaluation of medical devices — part 16: toxicokinetic study
design for degradation products and leachables
ISO 10993-17:2002 Biological evaluation of medical devices — part 17: establishment of
allowable limits for leachable substances
ISO 11979-5:1999 Opthalmic Implants — intraocular lenses — part 5: biocompatibility
ISO 12891-1-4:1998, 2000 Retrieval and analysis of surgical implants
a In some cases, ISO standards have multiple designations, reflecting texts identical with
national standards (such as ANSI); however, only ISO designations are listed here.
Source: International Standards Organization (ISO) http://www.iso.org.
418 Biological Performance of Materials: Fundamentals of Biocompatibility

If the Medical Device Amendments can be said to have a central theme,

it is one that closely parallels the ideas of the safety and effectiveness of
drugs, cosmetics, and food additives. That is, the legislation foresaw a pat-
tern in which materials and devices would be developed, tested, and dem-
onstrated to be safe and efficacious before being offered for sale. Once this
point was reached, their future safety and effectiveness would then be con-
trolled by the institution of general standards and controls or by the provi-
sions of specific performance standards.

20.4.2 Safe Medical Devices Act (1990)

The Safe Medical Devices Act (1990) was the first major revision of the
Medical Device Amendments and occurred largely as a result of public
dissatisfaction with the implementation (rather than the content) of medical
and surgical device regulations envisioned at the time of passage of the 1976
legislation. The many provisions of the act (Kahan et al. 1991) are intended
largely to strengthen, streamline, better define, and speed up regulatory
activities. Thus, the provisions primarily enlarge on and modify rather than
replace those of the earlier Medical Device Amendments. However, several
new provisions are introduced, including lifetime tracking of permanently
implanted life-supporting or life-sustaining devices, more and improved
reports of life-threatening device malfunction, regulatory authority to order
mandatory recalls and allow seizure of defective devices, rules for postmar-
ket introduction surveillance of experience with permanent implants, and
creation of a humanitarian device exemption, similar to an “orphan” drug
provision within drug regulation, to simplify and reduce the cost of devel-
opment of devices for diseases or conditions affecting fewer than 4000 indi-
viduals in the U.S.

20.4.3 FDA Modernization Act (1997)

At the time of the adoption of the Medical Device Amendments (1976), no
one could predict the vast increase in number and complexity of medical
and surgical devices that would occur in the 1980s and 1990s. The result was
a gradual slowdown of FDA review and approval activities and a growing
frustration on the part of manufacturers and the public. Various small
changes were made to FDA procedures, including a second Safe Medical
Devices (1992) act; however, the major change came about in 1997 with the
adoption of a widely heralded and long awaited “modernization” act. This
act (Kahan et al. 1998a, b) is complex and its effects are still just beginning
to be felt. Perhaps the most important aspects of its intentions are:

• To reduce the arbitrary and unpredictable nature of FDA device regu-

latory activities by clarifying previsions of the 1976 amendments and
spelling out procedures previously left to administrative definition
Standardization and Regulation of Implant Materials 419

• To make the FDA a partner with industry and physicians during the
device design, development, and qualification process, rather than
merely a judge of the end product performance, thus improving
chances of a new device’s release for use after an initial application;
this is to be brought about by development and approval of a prod-
uct design protocol early in the design process
• To require a higher level of professional achievement within the FDA
regulatory staff by providing for better training, liaison, and reporting

Review times, which had in some cases stretched out to years, are appar-
ently shortening significantly and the FDA’s role in medical device devel-
opment seems less threatening to manufacturers as a consequence of this
act. More recently, subsequent to yet another effort to provide legislative
relief,* various efforts have been made to speed up the review process fur-
ther, in some cases by raising fees or imposing special charges on the man-
ufacturers. However, these have come under significant criticism from
industry because they appear to embody the old political principle of “pay
to play.” The longer term consequences of these many changes and the
continuing influence of the political leadership of the FDA remains to be
seen.

20.5 Regulation of Materials for Implants

20.5.1 Requirements of the Medical Device Amendments
The need for regulatory standards arises from the requirements of the Med-
ical Device Amendments (1976) and the Safe Medical Devices Act (1990).
These are embodied in a classification system that attempts to distinguish
among various generic types of devices on the basis of risk to the patient.
The general pattern of device classification is as follows. Devices are classi-
fied into three categories by one of 14 specialty-oriented Device Classification
Panels:

• Class I, general controls: a device for which controls other than

standards and premarket approval are sufficient to assure safety and
effectiveness
• Class II, performance standards: a device for which general controls
are insufficient to assure safety and effectiveness but for which infor-
mation is sufficient for the establishment of a performance standard
to provide such assurance

* Medical Device User Fee and Modernization Act, PL 107-250.

420 Biological Performance of Materials: Fundamentals of Biocompatibility

• Class III, premarket approval: a device for which insufficient infor-

mation exists to assure that general controls and performance stan-
dards would provide reasonable assurance of safety and
effectiveness and that is represented to be life sustaining, life sup-
porting, or implanted in the body or that presents a potential unrea-
sonable risk of illness or injury

Devices classified as class III and any new device that comes on the market
after May 28, 1976 must pass through some form of scientific premarket
review before market introduction. At the point at which these products are
judged to be reasonably safe, effective, and controllable by a performance
standard, they may be reclassified into class II. Thus, the existence of stan-
dards can be seen to be critical to the introduction of new materials and
devices into general use. It is hoped that many of the voluntary standards
developed by the various organizations mentioned here, as well as others,
can be adapted to be regulatory standards.
To date, very few regulatory standards have been approved for devices
and none for materials. However, many permanent implants have been
reclassified from class III to class II on the basis of long pre- and postenact-
ment experience. The need for regulatory standards, particularly for mate-
rials of construction, will become more acute as more devices achieve such
reclassification.
As a consequence and in line with similar changes throughout the U.S.
government, the effort to substitute voluntary standards (see Section 20.5.2)
for regulatory standards is continuing. This effort is driven by a directive of
the Office of Management and Budget (OMB-119*) (Kono 1998). The basic
provision of OMB 119 is permission (and encouragement) to substitute a
consensus standard, such as those developed by ASTM or ISO, for a specially
drawn regulatory standard if the provisions of the voluntary standard are
appropriate to meet the needs of the regulatory agency, in this case the FDA.
This directive was recognized in a provision of the FDA Modernization Act
(1997) (section 204) to make conformance with an appropriate (preaccepted)
standard the basis for approval for the sale and use of certain devices.

20.5.2 Voluntary vs. Regulatory Standards

It should be clear to the reader that a voluntary standard and a regulatory
standard are not necessarily the same thing. Voluntary standards, whether
consensus derived or otherwise, are designed to describe the content, design,
construction, and performance of existing devices, as well as to set forth
methods of verifying compliance with these aspects of the standard. By their
nature, regulatory standards are designed to regulate — that is, to assure
specific attributes of products. In the case of the (regulatory) performance

* OMB 119: Federal Participation in Development and Use of Voluntary Consensus Standards
and in Conformity Assessment Activities. Fed. Reg. 61:8548, 1998.
Standardization and Regulation of Implant Materials 421

standards required by the Medical Device Amendments (1976), the specific

attributes to be regulated are safety and efficacy. It is not clear at this time
what combination of standards on content, design, and construction, as well
as simulation tests in vitro and in vivo, are necessary and sufficient to meet
such a general requirement for (in vivo) patient performance. Thus, although
it is possible to prepare a voluntary standard for an existing device or for
its materials of construction, the relationship of this standard to a future
generic regulatory standard is tenuous, at best.
It seems safe to presume that as genuinely new materials come under
consideration as candidate biomaterials, they will be qualified in a similar
way to that envisaged specifically for devices. In the course of selection and
development, they pass through series of tests along the lines of those dis-
cussed in Section 19.2. As their behavior becomes better understood and
devices incorporating them pass into clinical trials, the process of voluntary
standards preparation begins. The results of these tests and the proposed
form of the standard then constitute the body of material, with supporting
clinical reports, that can be submitted for initial review by a regulating
agency such as the FDA.
In that case, the use of the words “performance standard” in the 1976
enabling legislation harmonizes well with the ideas of biological perfor-
mance laid out here in earlier chapters. Thus, a performance standard for
an implant material is one that describes the chemical, physical, and pro-
cessing requirements for a material that are needed to assure a reproducible
level of biological performance, as well as to meet the engineering require-
ments, of a proposed application. As soon as it is possible to prepare such
a document, the material becomes in an important sense a known material.
Its biological performance can be examined objectively and its suitability
(and admissibility) for specific applications can be determined. At this point,
as embodied in devices, it should pass easily into class II and be a natural
competitor for use in future specific medical and surgical device applications.
Perhaps the recent revision of OMB 119 (Kono 1998) will encourage this
evolution and lead to materials “generally recognized as safe” for specific
groups of applications.
As an initial step in responding to OMB-119, the FDA maintains an online
registry of standards that it recognizes in regulatory submissions.* In large
part, these standards remain consensus standards that reflect current indus-
trial practice and carry with them no assurance of safety and efficacy of the
devices designed and constructed utilizing them.**

* http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfStandards/search.cfm.
** For completeness, it is worth noting also that product liability litigation has failed to establish
that FDA approval for sale and use implies a guarantee of safety and efficacy; such regulatory
action only establishes that the then current FDA requirements for such approval have been met.
The converse is also true: lack of FDA approval for a specific indication does not de facto render
a medical device and its materials of construction unsafe and/or ineffective.
422 Biological Performance of Materials: Fundamentals of Biocompatibility

Safety will continue to be a relative rather than an absolute attribute

because it will always be related, as I have pointed out, to a balance between
risk and benefit (Black 1995).

20.6 The Biomaterials Supply “Crisis”

Largely as a consequence of litigation associated with the use of silicone gel
in breast augmentation implants and polytetrafluoroethylene in tempero-
mandibular joint (TMJ) replacements, concern about future availability of
many biomaterials rose during the 1990s (Hallab et al. 1997). The problem
was apparently that materials suppliers, whose products contribute only a
few cents or dollars to the cost of medical devices selling for thousands of
dollars, have repeatedly been named by plaintiffs in actions alleging injury
through malperformance of the device or merely maloutcome of the overall
procedure. For some materials producers, the result has been legal defense
costs — even in the absence of adverse judgments against them — that far
exceed any reasonable profit. As a consequence, some manufacturers have
ceased to provide materials to medical device manufacturers. The end effect
is unclear; in the case of medical-grade silicones, the withdrawal of Dow
Corning from the marketplace led to the entry of several new small compa-
nies that now provide a wider range of well characterized materials, albeit
at significantly higher prices, than were previously available (Hallab et al.
1997).
This situation, still regarded by some as an impending “crisis,” seems
unfortunate on two grounds. In the first place, because the device manufac-
turer makes the selection of the material and then tests it as a material and
indirectly during preclinical and clinical device evaluation, the so-called
“learned intermediary” principle seems sufficient to protect the material
manufacturer as long as the materials supplied are made to identified stan-
dards (Harper 1996). That is, the responsibility of the material manufacturer
is to assure that the material is what it is represented to be and that of the
device manufacturer (the “learned intermediary” between the material sup-
plier and the patient) is to make the judgment that the material selected has
suitable biological performance in the intended application.
Although this concept is clear in U.S. litigation experience, it does not
overcome the problem of paying for legal defense before and until the
material manufacturer is discharged from the plaintiff’s action. Numerous
attempts have been made, led by then Rep. Lieberman (D-Conn), to embody
this principle in a federal statute. Despite widespread resistance to product
liability reform, the Biomaterials Access Assurance Act (1998)* was finally

* PL 105-230. See http://www.advamed.org/publicdocs/legal021599.htm for an extensive legal

analysis.
Standardization and Regulation of Implant Materials 423

passed and signed into law. The provisions of this legislation parallel the
arguments of Harper (1996) and provide for summary (immediate) discharge
from medical product liability litigation for any materials supplier that meets
appropriate standards and conditions of general salability for its medical
grades of materials.
In the second place, the idea that withdrawal of medical grades of materials
will make those materials unavailable is, to a degree, naive. Although some
biomaterials are specifically manufactured for medial and surgical applica-
tions, many, such as ultrahigh molecular weight polyethylene and many
titanium- and cobalt-base alloys, are simply selected batches or modest vari-
ants of very large quantity production commercial materials. Thus, because
the FDA does not explicitly regulate biomaterials manufacture or manufac-
turers, it is quite possible that intermediaries or the device manufacturers
could continue to procure suitable materials from commercial sources and
qualify them (i.e., determine their conformance with relevant standards) for
use in medical devices and implants.
An unforeseen negative result of the 1998 act has begun to emerge.
Recently, biomaterials and process suppliers named with device manufac-
tures as plaintiffs in cases in which failure modes appear to center on mate-
rials’ properties rather than device design have begun to assert the defense
that the device manufacturer is, de facto, a learned intermediary and, as a
result, they share no possible liability for clinical maloutcome as long as their
product meets the representations (specifications, etc.) for which they make
it. This defense theory is novel enough that it has not yet been definitively
tested in U.S. courts.
At this time, the situation remains very unclear. However, more than a
decade after the specter of future biomaterials unavailability was seriously
raised, the U.S does not appear to have any important shortages. Suppliers
have changed and materials substitutions have been made, albeit at
increased cost, but device availability to patients appears not to have been
adversely affected so far.

References
American Dental Association, Guide to Dental Materials and Devices, 7th ed. ADA,
Chicago, 1974.
American Society for Testing and Materials, 2005 Annual Book of ASTM Standards,
Vol. 13.01: Medical Devices and Services, ASTM International, West Conshohock-
en, PA, 2005.
Black, J., “Safe” biomaterials (editorial), J. Biomed. Mater. Res., 29, 791, 1995.
Brown, P. and Cook, A.G., The background, formation, and maturation of committee
F-4, ASTM Standardization News, October, 10, 1982.
Department of Health and Human Services, Everything You Always Wanted to Know
about the Medical Device Amendments…and Weren't Afraid to Ask, HHS, FDA,
Rockville, MD, FDA 92-4173, 1992.
424 Biological Performance of Materials: Fundamentals of Biocompatibility

Hallab, N.J. et al., Biomaterials crisis looms, AAOS Bull., 45(1), 13, 1997.
Harper, G.L., An analysis of the potential liabilities and defenses of bulk suppliers
of titanium biomaterials, Gonzaga Law Rev., 32(1), 195, 1996.
Kahan, J.S., The Safe Medical Devices Act of 1990, Med. Dev. Diag. Ind., 13(1), 66, 1991.
Kahan, J.S. and Holstein, H.M., The FDA Modernization Act of 1997: part 1, Med.
Dev. Diag. Ind., 20(3), 105, 1998a.
Kahan, J.S. and Holstein, H.M., The FDA Modernization Act of 1997: part 2, Med.
Dev. Diag. Ind., 20(4), 77, 1998b.
Kahan, J.S. et al., The implications of the Safe Medical Devices Act of 1990, Med. Dev.
Diag. Ind., 13(2), 44, 1991.
Kono, K., OMB A-119 revised, ASTM Stand. News, June, 1998, 19.
Lamb, R. DeF., American Chamber of Horrors: The Truth about Food and Drugs, Farrar
& Rinehart, New York, 1936.
Mintz, M., The Therapeutic Nightmare, Houghton–Mifflin Co., Boston, 1965.
NF XIV, The National Formulary, 14th ed., American Pharmaceutical Association,
Washington, D.C., 1975.
Sinclair, U., The Jungle, New American Library, New York, 1906.
USP XIX, The Pharmacopeia of the United States of America, 19th revision, The United
States Pharmacopeial Convention, Inc., Washington, D.C., 1975.
USP 28, The Pharmacopeia of the United States of America, 28th revision, incorporating
The National Formulary, 23rd revision, The United States Pharmacopeial Conven-
tion, Inc., Washington, D.C., 2005.
U.S. Congress, Medical Device Amendments, PL 94–-295, 1976.
U.S. Congress, Safe Medical Devices, PL 101–629, 1990.
U.S. Congress, Safe Medical Devices, PL 102–300, 1992.
U.S. Congress, FDA Modernization Act, PL 105–15, 1997.

Bibliography
Black, J. and Hastings, G., Handbook of Biomaterial Properties, Chapman & Hall,
London, 1998.
Cangelosi, R.J., Device standards: the view from the FDA, Clin Eng., Jan–Mar, 5(1),
9, 1980.
Department of Health, Education, and Welfare, Federal Food, Drug, and Cosmetic Act,
as Amended, Food and Drug Administration. U.S. Government Printing Office,
Washington, D.C., 1972.
Food and Drug Administration, Medical Devices Standardization Activities Report.
CDRH, FDA, HHS, Washington, D.C. FDA 94-4219. 1994.
Food and Drug Administration, Standards Survey, National Edition, Bureau of Medical
Devices, Washington, D.C., 1979.
Health Industry Manufacturers Association, Guidelines for the Development of Voluntary
Device Law Standards, Report No. 79-6, Health Industry Manufacturers Associ-
ation, Washington, D.C., 1979.
Health Industry Manufacturers Association, Guideline for Evaluating the Safety of Ma-
terials Used in Medical Devices, Report No. 78-7. Health Industry Manufacturers
Association, Washington, D.C., 1978.
Standardization and Regulation of Implant Materials 425

Morton, W.A. and Veale, J.R., Regulatory Issues in Artificial Organs: A Primer, J.B.
Lippincott, Philadelphia, 1987.
Ratner, B.D. et al., Biomaterials Science: An Introduction to Materials in Medicine, 1st ed.,
Academic Press, San Diego, 1996, 457.
21
Design and Selection of Implant Materials*

21.1 Introduction
21.1.1 What Is Design?
Design is what engineers do: they apply scientific knowledge and principles
to the solution of practical problems. The object of their design may be a
process, a new material, or a novel device. The process of design is artistic
and creative, drawing from the same well at which the painter, sculptor, or
writer does. What distinguishes the objects of engineering design from those
of other artistic activities is the extent to which technological factors come
into play in their realization (Asimow 1962).
As Cross (2000) points out, the separation between design and fabrication
of man-made artifacts is a relatively recent event. When hand artisanship
was the rule, design and fabrication were not separated: the maker designed
as the final form of the artifact emerged. For the artist, there is still no
separation in function: the design is the object. For the surgeon, the separa-
tion is incomplete: although surgical procedures are planned prospectively,
detailed and complex decisions are made during the performance of the
operation. However, for the engineer, the separation has become nearly total:
today those who design rarely make and vice versa. This separation has led
to vocational self-selection that produces significant problems for engineers
involved in design. Engineering has become a linear analytical process,
seeking the shortest distance to a solution. Engineers thus often have great
difficulty in dealing with the creative, synthetic aspects of design that require
attempts to devise as many alternative solutions as possible.
Even more than the creative aspects of design, the concept of a design
process must be emphasized. Solutions to engineering design problems
rarely, if ever, spring full blown from the mind of their creator. On the
contrary, what is required is a systematic, dogged, iterative process stretching
from exploration of initial requirements to evaluation of the preferred solu-
tion. In a sense, the design process and its necessary iterative design cycle
* Portions of this chapter appeared in an earlier form as Chapter 13 in Black (1988) and are repro-
duced by permission (Churchill–Livingstone Inc.).

427
428 Biological Performance of Materials: Fundamentals of Biocompatibility

represent attempts of engineering designers to deal with synthetic problems

in an analytic fashion, rather in the manner of using digital computers to
create analog models of systems. The use of a design process also has ethical
implications: the iterative nature of design conducted in this fashion requires
a continuous and repeated testing of goals and assumptions, thus introduc-
ing a system of checks and balances reflected as an inherent safety factor in
the final product.

21.1.2 Introduction to the Orange

Engineers unfamiliar with design often have the same problem as a begin-
ning art student: faced with a blank sheet of paper and an overall conception,
they have no idea where to start. A useful exercise is a consideration of an
orange in a process often referred to as reverse engineering. That is, given
an object or finished artifact, one attempts to understand its rationale and
determine details of the materials and processes of its construction retro-
spectively rather than design it prospectively. In this case, the use of an actual
orange is a useful aid. The exercise proceeds in three steps:

1. From observation and physical examination, make a list of all the things
that it is possible to know about an orange. In doing so, one usually
begins with simple attributes, such as color, weight, size, etc., and
moves to more complex ones, such as shape, number of seeds,
amount of sugar contained, etc.
2. For as many as possible of the quantitative attributes listed in step 1,
estimate the value. The benefit of this step is particularly seen when
a number of individuals do the exercise separately or in groups and
then compare their answers.
3. For as many as possible of the attributes listed in step 1, propose as many
methods as possible for finding their true or actual value. This step pro-
vides clues to the later stages of design by requiring problem solving
based upon estimates and other incomplete information.

The three steps of this exercise help to prime the creative pump. In form,
they replicate steps 1, 2, and 4 of the design cycle, respectively (see Section
21.2.2). The first step teaches observation, the second estimation, and the
third creation of alternatives.
This exercise also can be used to illustrate another point about design: it
is better played as a team sport than as solitaire. This point may easily be
demonstrated by setting the “orange exercise” for an individual and for a
group to perform; the members of the group, no matter what its makeup,
will always be more productive on average, let alone collectively, than the
individual. Design can and often is performed by a single individual. How-
ever, it is far more productive if it is a group project; the resulting synergy
increases in proportion to the variety of people involved.
Design and Selection of Implant Materials 429

21.2 The Design Process

21.2.1 The Phases of Design
Asimow (1962) defines design as a seven-phase process arising from a prim-
itive need (Table 21.1). Materials design and selection, in the sense in which
they are discussed in this chapter, fall within Asimow’s phases I to III,
depending upon the depth and detail required of the design process. In each
instance, a structured design process or cycle (see the next section) is desir-
able. Device design is more likely to have to deal with all of Asimow’s seven
phases. In either case, a single phase may require a number of design cycles
within its process. Design may be required for the development of manu-
facturing processes and design of devices, of surgical procedures and even
of experiments. With suitable modifications, the same structured process
may be utilized.

21.2.2 The Design Cycle

A structured design process consists of the consecutive execution of a repet-
itive design cycle. The design of a simple device, such as a tongue depressor
and its dispensing container, may be achieved in as little as three or four
such cycles; however, a complex design, such as a powered wheel chair, may
require hundreds of such cycles, some in parallel and others in series.
The design of materials is a simpler problem, in general, than the design
of devices. Materials design is rarely addressed directly because device
designers tend to view themselves as expert in materials and to assume that
materials design is merely a matter of selection from among the options
available. This approach is illustrated by Lewis’ (1990) otherwise excellent

TABLE 21.1
Seven Phases of Design

Primitive Need → Preliminary Phases

I: Feasibility design
II: Preliminary design
III: Detailed design

Phases Related to Production/Consumption Cycle

IV: Planning for production

V: Planning for distribution
VI: Planning for consumption
VII: Planning for retirement
Source: Asimow, M., Introduction to Design, Prentice Hall,
Englewood Cliffs, NJ, 1962, 1.
430 Biological Performance of Materials: Fundamentals of Biocompatibility

discussion of the design of a femoral medullary stem for a total hip replace-
ment prosthesis.
It is true that the selection of materials, with and without modification, has
dominated biomaterials design until recently. It is possible to use a formal
design process for selection and/or modification of materials, but this may
seem clumsy and unwarranted except for teaching purposes. However, today
the advancing popularity of composite materials or, more properly, engineered
materials makes necessary the use of a design process for the prospective
selection of biomaterials properties for medical and surgical devices. The evo-
lution of biomaterials as a field into the prospective design of interactive
materials, such as resorbable ceramics and polymeric matrices subject to
postimplantation cellular remodeling, further emphasizes this point.
The design of materials may require several cycles in series: first, selection
of materials properties; then selection of processing methods and param-
eters, followed by consideration of the interaction of various biomaterials
selected. These cycles cannot take place in isolation from the considerations
involved in the design of the device (for which the biomaterials are
designed/selected) because device requirements impose materials require-
ments and materials selections affect design choices.
A number of models may be utilized to develop a design cycle. The
approach in this chapter is derived from that of Love (1986) and is shown
in schematic form in Figure 21.1. The next section is devoted to a step-by-
step discussion of this cycle.

21.2.3 Steps in Design

21.2.3.1 Beginning the Design
Design within a given cycle arises from a primitive need. This need may be
an external statement (if the cycle is the first in the process) or may be the
output of a previous cycle. The example in this chapter will take as the
primitive need this possible statement by a product salesman for an ortho-
paedic implant company: “My customers are interested in a better total hip
replacement (THR) system for younger patients.” The engineering design
group takes up the challenge and develops a concept for a novel femoral
component. However, the group reports that none of the biomaterials in
their handbooks and reference sources provide the appropriate combination
of stiffness, strength, and fatigue life required to realize the preferred design
approach. In Asimow’s terminology, this finding, a result of a preliminary
phase, becomes the input that begins the material design cycle to be exam-
ined here.

21.2.3.2 Step 1: Analyzing Needs

The first formal step in the design cycle is to examine the input statements,
often in collaboration with those who made them, and to develop an
Design and Selection of Implant Materials 431

From
last cycle

Description

Step 1 Analyze Needs

Step 2 Define Goals

Step 3 Develop Target Specifications

Step 4 Create Alternatives

Step 5 Screen for Feasibility

Re-evaluation

Step 6 Complete the Design

<S
Step 7 Select the Solution

ⱖS

To
next cycle

FIGURE 21.1
The design cycle. (Adapted from Love, S.F., Planning and Creating Successful Engineering Designs:
Managing the Design Process, Los Angeles, Advanced Professional Development, Inc., 1986.)

objective summary statement that expresses the needs in an analytic way

and represents a rational objective. For example, development of copper with
a higher melting point is an irrational objective and seeking a higher
strength-to-modulus ratio in the copper–silver binary alloy system is a
rational one. In this case, after careful consultation and deliberation, the
design objective is stated as: “The objective is the design of a new material
suitable for use in fabrication of THR components that combines optimum
432 Biological Performance of Materials: Fundamentals of Biocompatibility

stiffness (modulus) with greater strength and a higher endurance limit than
that for presently available materials.”
Note that the act of stating the objective limits the inquiry: a new material
will be designed rather than a present one modified. It also completes the
translation of the primitive need to a defined materials need, with three
attributes: optimum (to be defined) modulus, increased strength, and higher
fatigue endurance. In the same way that an experimental question (and its
hypotheses) can be tested, this statement can be tested at the end of the cycle
to see whether the objective has been realized.*

21.2.3.3 Step 2: Defining Goals

Design is not a simple process that leads to a unique output. Thus, objectives
must be refined to limit the number of choices at each step and to guide the
design cycle. This is achieved by selecting goals whose attainment (1) is
necessary to reach the desired objective; or (2) represents generally “good”
attributes of engineering design or reflects the desires of the designers. The
first type are called specific goals (demands; Cross 2000) and the second type
are called general goals (wishes; Cross 2000). An initial list of specific and
general goals that might follow from the previously stated summary objec-
tive statement is given under the heading “initial” in Table 21.2.
The initial list of goals arises from a discussion with the customer — in
this case, the device design group — and represents the definitive starting
point of the material design cycle. It embodies the customer’s concepts of
what is desired as an end product of the design process. The material design
team must now put its talents to work to understand these desires and to
satisfy them. Some of the initial goals may come from other sources: the
engineering manager is always worried about manufacturing costs; the color
was suggested by the marketing manager because yellow is widely used in
the company’s packaging and has come to be identified positively with its
products in the mind of the retail customer, the surgeon.
The material design team must actually go through two substeps to pro-
duce the refined set of goals shown in the lower part of Table 21.2. The first
of these, the production of an initial set of goals, is the first creative act in
the design cycle. The question posed at this point is, “What should a new
material for a THR component look like?” As previously noted, the creation
of ideas is not an easy process for engineers. The tendency is to “freeze”: to
be unable to produce ideas or, more commonly, to have an initial thought
and then to proceed to develop it without consideration of further alterna-
tives. The general solution for the individual designer is to produce a situ-
ation that is stimulatory and nonself-critical.**

* Discussion of the design and conduct of experiments is outside the scope of this work. How-
ever, the reader should note that this phase is identical to the statement of an experimental ques-
tion.
** In this section, I refer to a single designer. In Section 21.2.3.5, the situation of creative effort by
a group will be considered.
Design and Selection of Implant Materials 433

TABLE 21.2
Design Goals: New THR Material

Initial

Specific
Modulus < 0.5 × Ti6Al4V
Strength as high as possible
Endurance limit as high as possible
Corrosion/release rate “low”
No wear against UHMWPEa
Color: yellow
General
Minimum cost
No limit on source of supply
Simplicity of fabrication

Refined

Specific
Modulus < 0.5 × Ti6Al4V (H)
Strength as high as possible (M)
Endurance limit as high as possible (H)
Corrosion/release rate as low as possible (H)
Wear rate (against UHMWPE) as low as possible (M)
Color: yellow (L)
Formability in the operating room (M)
Release of wear particles > 25 μm in size only (H)
General
Minimum cost per kilogram (L)
No limit on source of supply (M)
Simplicity of fabrication (M)
a Ultrahigh molecular weight polyethylene

In this case, the designer may decide that “I’m going to set the problem
aside, go for a 5-km run, and when I come in, write down the first ten things
that come into my head.” Such a procedure, with variants, has been adopted
frequently by many if not most creative persons and is sometimes referred
to as creative avoidance of the problem: undertaking other activities to
distract the conscious mind (probably the analytic left brain function) and
using the products of subconscious deliberation (probably the synthetic right
brain function), without self-criticism or censoring.
The initial list is then reviewed for reasonableness and duplication and
perhaps the process is repeated or extended until the sense is that all of the
immediately possible options — in this case, design goals — have been
acquired. Often the review triggers new ideas not previously considered.
The designer in this example has added two specific goals and no general
goal to previously cited desires. Note that this is an abbreviated example;
step 2 of an actual design cycle might produce dozens of specific and general
goals.
434 Biological Performance of Materials: Fundamentals of Biocompatibility

The second substep is the assignment of a priority to each of these initial

goals to produce a set of refined goals. This is necessary because, in an actual
design case, the number of goals very rapidly grows to a point at which it
is obvious, a priori, that all cannot be met simultaneously. Thus, a ranking
of relative importance is necessary. In this case, the designer employed a
common practice and selected three levels of priority:

• High (H): must be met for successful design

• Medium (M): would like to meet during design cycle
• Low (L): desirable to meet but may be sacrificed

Therefore, the material’s modulus is identified as a much more important

attribute than its color, although the desire to satisfy the marketing manager
is still considered as part of the later steps in the cycle. In a more subtle
distinction, it is recognized that in the intended application, the endurance
limit is a more important material attribute than the tensile strength,
although both are important. During this substep, the sets of goals are also
screened to eliminate absolute statements; statements of goals should not
include the terms “never,” “always,” “none,” etc.

21.2.3.4 Step 3: Developing Target Specifications

Setting specific and general goals and then refining the list and assigning
priorities produce considerable clarification of the problem in hand but do
not provide details necessary for later steps in the design cycle. To achieve
this, it is necessary to translate the refined goals of step 2 into measurable
quantities.
These measurable quantities are called specifications. They must be nec-
essary, thus not setting limits unrelated to performance. They must also be
sufficient: taken as a group, their satisfaction must be sufficient to produce
a successful design and to assure that the formal requirements of the Medical
Device Amendments (U.S. Congress 1976) of safety and efficacy are met.*
Finally, they must be conservative: setting too high values or too stringent
criteria will elevate cost unacceptably or possibly make the design unreal-
izable.
Consider the refined goal (Table 21.2):

• “Corrosion/release rate as low as possible (H)”

This might be translated into these specifications:
• Corrosion rate in vivo shall not exceed 0.1 mg/cm2/year.
• Release rate in vivo shall not produce a concentration of products
≥ 5 ppb at a distance of 1 cm from the implant–tissue interface.
* Note that materials used in medical devices are not explicitly regulated but are subject to this
requirement indirectly because they perform in a device. See Section 20.5 for a further discussion
of this point.
Design and Selection of Implant Materials 435

The attainment of minimum corrosion/release was judged to be of high

importance; this is reflected in the use of design margins, multipliers of
minimum values. In practice, these may vary between 10 (for extremely
critical attributes) and 1.1 (for low importance or optional attributes). In this
case, the actual allowable values might have been 1 mg/cm2/year and 50
ppb, respectively, but were reduced by application of a 10× design margin.
There are no objective criteria for deriving design margins: they reflect cur-
rent practice in similar designs.
This is a good time to conduct a design conference to review the project
because completion of phase 3 marks the boundary between defining the
problem and, in a strict sense, solving it. It is worthwhile determining
whether the team is on the right track before the really hard, time-consuming
and costly parts of the cycle are undertaken.

21.2.3.5 Step 4: Creating Alternatives

Development of design concepts or alternative possible solutions is the heart
of the design process and the point at which most fatal mistakes are made.
It requires, again, a suspension of self-criticism and a source of external
stimulation. For most people, concepts and alternatives evolve more readily
in a group situation in which one person’s ideas trigger another’s imagina-
tion. This is the time in the design process when the prior formation of a
multidisciplinary design team really pays dividends. The goal is the same
as in step 2: to develop as many independent approaches to realization of
the design goals as possible.
As an introduction to this step, the material’s designer begins to gather
supporting information on past and present materials used in THR prosthe-
ses components as well as on current progress in materials design and
processing. Information acquired at this time serves the subconscious as a
source of ideas and, if a design team is in existence, all members should
have access to this resource information. This information will also be needed
for the next step.
The primary tool in creating alternatives is the brainstorming or “blue sky”
meeting. Cross (2000) lists the following essential rules for such a session:

• Offer no criticism during the session.

• A large number of ideas is wanted.
• Seemingly crazy ideas are welcome.
• Keep all ideas short and snappy.
• Try to combine and improve on the ideas of others.

Citing William J. Osborne, Love (1986) provides many practical suggestions

on how to organize and run a successful session to create alternatives.
Table 21.3 presents a list of ideas that might arise from such a step 4
exercise. The initial list was developed in two creative sessions: a break was
436 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 21.3
Design Alternatives: New THR Material

Initial List

Animal tusk
Modified wood
Cloned tree with new properties
Petrified wood
Coral
Metal impregnated coral
Woven ceramic fiber/resin impregnated
Carbon fiber/graphite
Carbon/silicon carbide powder composite
Carbon/polyethylene powder composite
Hydroxyapatite/polyethylene powder composite

Break Taken at this Point

Metal-fiber-reinforced silicon nitride
Alumina/polymer composite
Woven sapphire fiber/metal impregnated
Sapphire beads with spring connectors
Whisker-reinforced polymer
New titanium alloy
Titanium/polymer powder composite

Final List

Modified natural material

Fiber-reinforced composite
Powder composite
New titanium alloy

taken between the sessions, and the first part of the list was reviewed by the
group to initiate the second session. The final list was developed some days
later by review and analysis of the initial list.
In this hypothetical case, the creative sessions produced an initial list of
18 ideas that was then reduced to four concepts. The last of these was
eliminated by reference to the objective summary statement (it was not
judged to be able to lead to a new material) and the other three, which focus
primarily on processing leading to new materials, could each be continued
in parallel through later stages of the process. Trouble arises at this point or
at a later point in the process when alternatives are not fully constrained
and/or decisions are made that circumscribe the later steps too narrowly.
Reduction in scope can occur later; what is needed at this point is to have
created a maximum range of possibilities.

21.2.3.6 Step 5: Screening for Feasibility

It is now necessary to examine the alternatives created in the preceding step
(Table 21.3) and select those with which to continue the process. In the
Design and Selection of Implant Materials 437

example given, after analysis of the possibilities proposed, only three alter-
native approaches emerged; it might be reasonable to continue with all three.
However, each would need to be screened for feasibility. If more than three
approaches had resulted from the previous step, feasibility screening could
be used to select the two or three most likely to lead to success.
Feasibility is the process of applying rational criticism, which was sus-
pended in the previous step, to enable estimates to be made of the relative
chance for success of each proposed approach. The primary aspects of each
idea to be examined are: technical, economic, supply, and parsimony. These
are justified as follows:

• Technical: the designer must avoid attempting to violate laws of

nature.
• Economic: cost and resulting price are powerful considerations, even
in materials design. This may refer to the final cost of the material
and of its design and qualification.
• Supply: there should be no reasonable intrinsic or imposed barrier
(through protection of intellectual property, etc.) to provision of
sufficient material for the intended application.
• Parsimony: the simple is preferred over the complex.

Secondary attributes, including the designer’s intuition, and political,

legal, and/or perceptual issues (which may be unrelated to technical func-
tion) may also come into play. The specifications may be used to drive the
screening process by attempting to estimate values for each of the material’s
physical attributes (identified in step 4) and using the apparent ease of
achieving the specified values as an index of feasibility. The screening process
may be qualitative — each alternative is ranked with respect to each aspect
— or it may be made quantitative, with values assigned to rank and an
overall score derived for each approach.*

21.2.3.7 Step 6: Completing the Design

Completing the design of the alternatives created in step 4 that survive
feasibility screening in step 5 is the final pure design step of the design cycle.
Not much needs to be said in that it involves traditional engineering pro-
cesses of analysis, calculation, and simulation and may even require some
pilot experiments to verify design and manufacturing concepts. Parametric
studies, in which the effects of varying controllable independent variables
are tested, are of great value in later considerations. The specifications must
explicitly drive design choices because they will be the basis for testing the
final design in the next step.
When alternative approaches were selected (step 5), design completion
usually results in a definitive ranking in order of preference or in the elim-
* Ranking the three approaches selected in the last step is left as an exercise for the reader.
438 Biological Performance of Materials: Fundamentals of Biocompatibility

ination of one or more owing to an inability to realize a complete design.

However, cost and time considerations may result in a decision to complete
the design for only the most promising (most feasible) alternative.

21.2.3.8 Step 7: Selecting the Solution

Design selection (or evaluation) is a simple process of comparing the
attributes of the final design to the specified values developed in step 3 and
determining how well they have been met. If no design conferences have
been held (with the “customers”) since the one at the end of step 3, now is
the ideal time to do so. The customers (and outside reviewers, if possible)
may serve as a board of review to complete this step. Ideally, all high- and
medium-priority goals should be met, through satisfaction of their depen-
dent specifications ( ≥S, Figure 21.1), for the design to be said to be acceptable
and for it to advance to the next cycle of the overall (device, etc.) design
process.
If, in the opinion of the reviewers, the design is unacceptable (fails to meet
one or more key specifications, <S, Figure 21.1), several options are open.
These include reviewing design concepts to see whether additional ones can
be developed, reviewing the specifications to see whether they (or their
design margins) can be relaxed, and reviewing the goals to see whether they
are all necessary and have appropriate priorities. If changes can be made
retrospectively at any of these three steps, the cycle can be resumed at that
point to determine whether a satisfactory design results.

21.3 The Value of Prospective Design

21.3.1 Why Have a Design Process Anyway?
The design process or even a single design cycle will not always yield a
satisfactory result. Objectives may be unrealistic or even forbidden by basic
physical principles or the goals selected may not be technologically achiev-
able or financially feasible at the time. However, it is clear that, in the majority
of cases, a structured design process does produce satisfactory results with
well-articulated foundations and justification. In most cases, within the
boundaries of assumptions and choices made at various steps, the resulting
designs will represent optimum solutions. Thus, prospective design is to be
preferred to inspired guesses in designing biomaterials, as in other areas of
engineering.
Design and Selection of Implant Materials 439

21.3.2 Design in the Real World

The process elaborated here, based upon ideas put forward by Asimow
(1962), Love (1986), and Cross (2000) as well as my experience in academic
and industrial settings, reflects the ideal. The names of the seven steps
encapsulate the central ideas: analysis of need, proposals for solution, elab-
oration of proposals, and testing of results against the original need. This
cyclic, iterative approach is critical, whether it exists within a fully articulated
process as described here or merely guides more informal considerations.
In the biomedical context, useful new materials and devices can result
neither from pure analytical considerations of engineers and developers nor
from pure synthetic suggestions of physicians and surgeons. What actually
happens is that there is a continuing, interactive collaboration, made difficult
at times by the necessary conflict between analysis and synthesis. Groups
and companies that have been successful in bringing novel materials and
devices incorporating them into clinical use have managed to preserve bal-
anced collaboration. Robertson and Hyatt (1998) illustrate this interaction in
the idealized development of spinal instrumentation hardware.
However, developments dominated by either party have been seen, in
particular cases, to lead to unfortunate consequences. Engineers often
express frustration in working with medical personnel who generally have
difficulty in providing quantitative measures of their clinical observations.
Clinicians tend to be overly impressed by engineering rigor, as exemplified
by finite element analysis, and then disappointed when performance, which
may have been based on inadequate inputs, fails to meet expectations.
As bioengineering matures professionally and clinical experience with
existing materials now extends to periods of decades in some applications,
additional barriers to successful design of materials have emerged. Although
extensive testing protocols have been devised in the laboratory and in animal
models (Chapter 17 and Chapter 18), it is still extremely difficult to predict
biological performance of materials in the long term. Furthermore, the exist-
ence of an increasing range of materials with known host and materials
responses in highly successful devices raises real ethical and practical issues
concerning substitution of novel materials in such applications.
Success in design of materials or of devices requires an understanding of
history as well as cultural and professional differences and all parties’ will-
ingness to be flexible and remain focused on the real goal: a better outcome
for present and future patients.

References
Asimow, M., Introduction to Design, Prentice Hall, Englewood Cliffs, NJ, 1962, 1.
Black, J., Orthopaedic Biomaterials in Research and Practice, Churchill Livingstone, New
York, 1988, 303.
440 Biological Performance of Materials: Fundamentals of Biocompatibility

Cross, N., Engineering Design Methods: Strategies for Product Design, 3rd ed., John Wiley
& Sons, Chichester, U.K., 2000.
Lewis, G., Selection of Engineering Materials, Prentice Hall, Englewood Cliffs, NJ, 1990,
179.
Love, S.F., Planning and Creating Successful Engineering Designs: Managing the Design
Process, Los Angeles, Advanced Professional Development, Inc., 1986.
Robertson, J.T. and Hyatt, D., Concepts and issues of spine device development and
regulation, in Capen, D.A. and Haye, W. (Eds.), Comprehensive Management of
Spinal Trauma, St. Louis, C.V. Mosby, 1988, 414.
U.S. Congress, Medical Device Amendments. PL 94-295, 21 USC 301, 1976.

Bibliography
Ashby, M.F., Materials Selection in Mechanical Design, 2nd ed., New York, Butterworth
Heinemann (Elsevier), 1999.
Bronikowski, R.J., Managing the Engineering Design Function, Van Nostrand Reinhold,
New York, 1986.
Collins, J.A., Failure of Materials in Mechanical Design: Analysis, Prediction, and Preven-
tion, 2nd. ed., John Wiley & Sons, New York, 1993.
Norman, D.A., The Design of Everyday Things, Basic, New York, 2002.
Petroski, H., The Evolution of Useful Things, Vintage, New York, 1994.
Shackelford, J.F., Alexander, W. and Park, J., CRC Practical Handbook of Materials
Selection, CRC Press, Boca Raton, FL, 1995.
22
Clinical Performance of Biomaterials*

22.1 Historical Aspects

Implants used for the alleviation and treatment of human disability and
disease are derived from natural (biological donor) sources or are manufac-
tured from organic and/or inorganic materials. The failure and success of
live cell, tissue, and organ transplantation have been studied extensively.
However, the study of the consequences of the use of manufactured biom-
aterials in implants, in their actual service setting, has been spotty at best
and based primarily upon complications seen in reported clinical series or
on local, intermittent study of devices retrieved at surgical revision (replace-
ment) or at autopsy.
The use of manufactured implants in medicine has its roots in antiquity;
however, the practice has only become prevalent in the last century and has
gained widespread success only since World War II. Early efforts were dis-
tinguished by high rates of complications and failures and thus the use of
nonbiological implants was long regarded as experimental. Historically,
improved biomaterials and devices emerged from short-term animal studies
and clinical observations that eliminate undesirable material and/or design-
related performance on a case-by-case or small-group-study basis.
The result is that now a small group of biomaterials is, by and large, highly
successful when used in a broad variety of designs; its use for many indi-
cations has become routine (see Interpart 1 for typical examples).** Heart
valve replacements, heart pacemakers, and total hip and knee replacements
are examples of devices incorporating such materials for which clinical expe-
rience of more than 10 years allows prediction of a high likelihood of success
(>90 to 95%) for individual patients who meet appropriate indications. It is
hard to estimate how many chronic (intended to remain in situ for more than
30 days) implants are in use today in the U.S., but national data (Moss 1991)
* Many of the ideas in this chapter were developed and elaborated during contractual studies
for the USFDA, CDRH, whose support is gratefully acknowledged.
** However, there is no list of “generally recognized as safe” biomaterials, due to the modern,
and correct, emphasis on biocompatibility being related to specific application requirements and
to the resulting risk/benefit ratio.

441
442 Biological Performance of Materials: Fundamentals of Biocompatibility

suggest that the number was at least 11 million by 1988 with annual increases
since then most probably of about 10%. Non-U.S. experience probably equals
or slightly exceeds these figures; the 2005 worldwide total of chronic
implants probably now exceeds 75 million.
However, success has produced a new set of problems, which can be
summarized as follows:

• Large-scale, routine clinical use of implants, even with low failure

rates, produces significant numbers of patients whose procedures
fail to meet expectations.
• Clinical confidence in implants results in pragmatic extension of the
indications for their use, especially to more difficult medical prob-
lems and to earlier intervention in disease processes.
• Routine use and earlier intervention produce an increasing mis-
match between typically short development and evaluation cycles
and longer intended (and actual) service periods.

Together, these factors have produced needs for certain types of data con-
cerning the biomaterials from which implants are manufactured:

• Long-term effects of in vivo environments on biomaterials properties

• Chronic (including systemic and remote site) effects of manufactured
biomaterials on human physiological processes
• Comparative service experience of different biomaterials in similar
or different device designs used for the same clinical application/
indication

The response to these needs has been an effort, led primarily by bioengi-
neers, to study implants and explants (retrieved devices) in a field that has
come to be termed device retrieval and analysis (DRA). Early DRA efforts
tended to focus on the disease state and view the device generically (medical
or clinical pathology model) or to study the device closely, with little atten-
tion given to the generic disease or to individual patient conditions and use
(engineering failure analysis model).
Since 1976, at least six major U.S. technical conferences* on DRA have had
a primary, if unstated, goal to bring these two models together and thus
produce a unified approach (using a single analytical model) to the study
of biological performance (host and implant response) of devices (and the
biomaterials from which they are fabricated) in human clinical use. Numer-
ous professional societies, commercial concerns, and U.S. government

* Retrieval and Analysis of Orthopedic Implants, Bethesda, MD, 3/5/76; Corrosion and Degra-
dation of Implant Materials, Kansas City, MO, 5/22-23/78; Implant Retrieval and Biological
Analysis, Bethesda, MD, 5/1-3/80; Corrosion and Degradation of Implant Materials: Second
Symposium, Louisville, KY, 5/9-10/83; Symposium on Implant Retrieval, Snowbird, UT, 8/12-
14/88 and Implant Retrieval Symposium, St. Charles, IL, 9/17-20/92.
Clinical Performance of Biomaterials 443

agencies have been involved in the sponsorship of these conferences and in

support of numerous smaller symposia and workshops as portions of larger
engineering and medical professional meetings.
Of outstanding note have been the efforts of the ASTM F-4 Committee on
Medical and Surgical Materials and Devices, which, since 1972, has been
codifying procedures and practices for the physical retrieval and analysis of
individual implants. Although not believed to be widely used exactly as
written, these procedures, such as F-561-05,* provide important and useful
guidance to DRA activities. The American National Standards Institute
(ANSI) and the International Standards Organization (ISO) have more
recently begun to develop standard procedures for DRA. Of note is the work
of Working Group 5 of ISO Technical Committee 150 that is developing
standards for retrieval and analysis of implantable devices.**
However, despite early and continuing perceptions that the data and the
knowledge that can be gained from study of retrieved devices are of vital
importance, most of these efforts continue to focus on the implant and on
the implications of its physical condition (engineering failure analysis
model). The seven major parties to DRA (patient, physician, manufacturer,
treating institution, insurer, regulatory agency, and society at large) recognize
a shared common interest in DRA and its outcomes; however, individual
benefit/risk calculations by each concerned party are in constant conflict
(Black and Fielder 1992) and have stood in the way of emergence of a unified
analytical model or of comprehensive and/or multi-institutional DRA pro-
grams. Table 22.1 briefly highlights the benefits and risks perceived by each
party involved in DRA.
Notwithstanding this continued conflict, it is important for the advance-
ment of the field of biomaterials and the practice of medicine that DRA
efforts continue to advance and mature. The following sections outline some
applicable methodology.

22.2 Procedures for Device Retrieval and Analysis

DRA today is of necessity a “team sport” brought about by the complexities
of medical practice and of the social and legal setting of the early 21st century.
Before one undertakes such activities, it is necessary to establish a supporting
organization and for the parties involved to agree on a number of assump-
tions. The issues to be dealt with include goals, responsibility, methodology,
and reporting.

* F-561-05 Practice for Retrieval and Analysis of Implanted Medical Devices, in 2005 Annual Book
of ASTM Standards, Vol. 13.01: Medical Devices; Emergency Medical Services, ASTM Interna-
tional, West Conshohocken, PA, 2005.
** ISO/DIS 12891-1: Retrieval and analysis of surgical implants — part 1: retrieval and handling.
444 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 22.1
Benefits and Risks in Device Retrieval and Analysis
Party Benefit Risk
Patient Improved medical care Increased cost
Increased concern related to
device performance
Physician Improved service to patient Possible malpractice liability
Manufacturer Increased knowledge of device Increased administrative burden
performance Possible increased cost
Possible tort liability
Treating institution Improved service to patient Increased cost
Possible liability
Insurer Increased knowledge of device Possible increased cost
performance
Reduced cost through use of
“better” devices
Regulatory agency Increased knowledge of device Increased administrative burden
performance
Society at large Increased knowledge of device Possible alarm about device
performance malfunction
Reduced cost through use of
“better” devices
Improved health care

22.2.1 Goals
DRA is a general term that can cover a wide variety of activities, each of
which has specific goals, such as:

• Premarket approval clinical evaluation: goals may include determin-

ing normal and abnormal device and/or material performance,
meeting regulatory requirements for reporting adverse outcomes,
and providing feedback for device and/or surgical technique mod-
ification.
• Routine clinical use of a device: goals may include monitoring of
appropriateness of device/patient matching, determining specific
device-related “failure” rates as part of survivorship calculations,
establishing mechanisms underlying survivorship estimates, and
planning future treatment for individual patients.
• Autopsy retrieval: goals may include investigation of local and sys-
temic host response, long-term material property changes, and dis-
tribution and storage of implant degradation products.

In any of these or other activities, there should be formal written statement

of and agreement to goals because such goals strongly affect the design and
conduct of DRA studies.
Clinical Performance of Biomaterials 445

22.2.2 Responsibility
Any DRA study should be conducted under the direction of a group of
interested, appropriately trained, and committed individuals. The issues
raised by DRA are such that careless or unplanned activities can produce
extremely adverse outcomes for many of the parties involved. There should
be a written, agreed upon set of procedures and methods. One individual,
preferably a Ph.D.-trained person with experience in DRA, should have final
responsibility for the program and should be designated as the custodian
for the devices between their surgical recovery and their discharge from
DRA study.

22.2.3 Methodology
In its most general form, DRA is an example of discovery science. Thus, it
is inappropriate, except in very small, tightly focused studies, for all recov-
ered devices to undergo a fixed set of procedures. Such an approach, espe-
cially in the usual setting of routine clinical practice, would produce
insupportable costs without returning commensurably valuable informa-
tion. Thus, it is good practice to classify devices prospectively before study.
Three generic classes can be easily recognized. These are briefly described
next, with examples of each type of device and of possible response within
a DRA study:

• Class 3: no frank evidence of physical damage to the device pre- and

postexplantation and no implication of involvement of device mal-
function in clinical outcome
• Example: routine (nonsymptomatic) removal of fracture fixation
hardware
• Response: positive identification of device, cataloging, discharge
from recovery system*
• Class 2: frank evidence of physical damage to the device post- and
possibly pre-explantation but no implication of involvement of
device malfunction in clinical outcome
• Example: component of total hip replacement, showing surface
defects, recovered from site of early postoperative infection

* It is assumed in this chapter that any institution in which a DRA study is planned already has
a working device recovery system in place. Most simply stated, a device recovery system is a set
of procedures, parallel to those used for clinical pathology specimens, that dictate how a device
is collected from the surgical field or clinic, handled postrecovery, examined for routine iden-
tification and evaluation purposes and then “discharged” (given to patient, retained for
research, discarded, etc.) (see Section 22.3.4). It is difficult and ill advised to conduct DRA studies
in the absence of such a recovery system; one of the first steps in designing and implementing a
DRA program may have to be working with the host institution to install or improve a device
recovery system.
446 Biological Performance of Materials: Fundamentals of Biocompatibility

TABLE 22.2
Conduct of DRA Studies
Steps Removal Class
Recovery Procedure
Retrieve
Package
Identify
Sterilizea
Classify
Class 1 Class 2 Class 3
Evaluate
Photographs X X X
Culture reports X X
Histology X X
Metallography X
Mechanical analysis X
Chemical analysis X
Special tests X X
Mechanical testing (device portions) X
Hardness X
Specific histologic stains X
Metal analysis (AAS) (fluids, tissues) X
Case disposition X X X
Prepare report X X X
Store/dispose of device X X X
a Specific studies may require devices to be studied in unsterilized con-
ditions; this may require deviations from routine recovery practice.

• Response: as for class 3 with additional procedures to examine

and document noted device defects (see Table 22.2)
• Class 1: frank evidence of physical damage to the device pre- and
postexplantation and implication of involvement of device malfunc-
tion in clinical outcome
• Example: cardiac pacer lead with broken conductor
• Response: as for class 2 with additional procedures to determine
origin and/or cause of noted effects* (see Table 22.2)

22.2.4 Reporting
Timely reporting of results is the key to sustained and useful DRA studies.
Reporting should take the following forms:

• Rapid reports should be made to the treating physicians, on a

time schedule parallel to that in the treating institution for clinical
* Note: Clinical institutions frequently elect to deal with class 1 cases in a different manner than
class 2 and 3 cases because the former involve devices that may become physical evidence in
malpractice and/or product liability proceedings.
Clinical Performance of Biomaterials 447

pathology studies. In addition to a personal report to the primary

physician, a note should be placed in the hospital chart, over the
signature of the DRA supervising professional. Although it is
unusual for such individual studies to have an impact on the further
treatment of the patient in question, such reporting is simply good
manners and helps to maintain the professional relationships needed
for DRA studies.
• In the case of class 1 and 2 devices, reports should be made directly
to the manufacturer. This is especially necessary if more than one
set of similar findings occurs during a study. Sensitivity should be
shown to studies of class 1 devices: because of the possibility of
litigation in such cases, reports should focus on physical findings
and descriptions but omit theories of causation, unless they are
supportable by peer-reviewed, published studies.
• Depending upon the nature of the findings and, in the case of a
class 1 device, the degree of patient involvement, timely reports
may need to be filed with a regulatory agency, such as the FDA in
the U.S. Because this is a legal requirement, the treating institution
should have already established an approved procedure for sub-
mitting such reports as a part of its device recovery system. DRA
personnel should be aware of such a system and provide rapid
access to their findings for the individuals responsible for making
the report.
• Finally, there is a responsibility to make results of studies of groups
of retrieved devices available to the scientific and engineering com-
munity. Such reports might include publication, presentation at pro-
fessional meetings, and participation in workshops and instructional
courses.

It cannot be too strongly emphasized that, except in reports to physicians

within the treating institution, great care must be taken to preserve the
privacy and identity of the patients involved. This is an ethical and legal
requirement (Black and Fiedler 1992).

22.3 Common Concerns about Device Retrieval and Analysis

22.3.1 Cost
Costs for study and analysis of implants are extremely difficult to identify
due to the varied nature of individual implants and the present indications
for analysis. Survey results and anecdotal experience suggest ranges of direct
costs as follows:
448 Biological Performance of Materials: Fundamentals of Biocompatibility

• Class 3: $5 to 25
• Class 2: $25 to $250
• Class 1: $250 to $10,000

These costs certainly are subject to economy-of-scale effects and can be

expected to be nearer the lower ends of each range in large-scale and/or
routine clinical retrieval and analysis studies.

22.3.2 Device Ownership

Central to any functioning DRA program is the concept of rapid availability
of enrolled (recovered) devices, when selected, for nondestructive and
destructive studies. It is good practice that all custodial and other procedures
for handling implants and other medical devices satisfy minimum legal
“chain of evidence” standards; however, the assumption* that patients may
“own” devices removed from their bodies continues. It is suggested that
concern for property rights may be satisfied in one of three ways:

• Institutions that conduct DRA studies beyond routine logging and

discharge of recovered devices should probably alter their standard
permission for treatment forms to include permission to study
devices, including destructive testing. Possible language, as previ-
ously proposed,* is:

I understand that part of my treatment may require removal of an arti-

ficial implant. If an implant is removed, I give permission for any studies
of it related to my treatment. I understand that I and my doctor will be
informed of the results of these studies in a timely fashion and that I will
be consulted before the implant is discarded or if the implant is desired
for other studies.

• In the case of individual (authorized by IRB) prospective studies, an

additional approved permission form may be developed that
explains the study’s goals and conduct and asks for the patient’s
permission for any necessary destructive testing.
• Alternatively, because devices may already be in institutional cus-
tody at the time that they are selected for a study, treating institu-
tions may elect to obtain permission for study retrospectively, using
a form such as that described in the previous option (second bullet).
This option is less desirable than the first (routine permission)
because it may raise the patient’s concern over possible device
malfunction.

* Black 1996.
Clinical Performance of Biomaterials 449

22.3.3 Patient Confidentiality

There is a clear need to retain a defined linkage between DRA data, which
are devoid of patient identification, and medical records of each patient. This
linkage is required because studies of device function and biomaterials prop-
erties may require correlation with disease state, mechanical environment,
etc. encountered by the device during service. A possible linkage, as pro-
posed in Section 22.4, is a coded identification number generated at the time
of recovery of the device. The manual or electronic ways of generating such
unique codes will not be discussed here. However, the key to the coding
system will be deemed to be confidential and related to the patient, thus not
accessible under Freedom of Information Act inquiry. The keys would be
disclosed only to DRA study supervisors and then only after suitable assur-
ances concerning maintenance of patient confidentiality were obtained and
the “need-to-know” established on a patient-by-patient basis.

22.3.4 Recovery System

Repeated recommendations from many sources make it clear that all clinical
facilities in which devices are explanted should possess and maintain recov-
ery systems to deal with removed devices. The need for recovery systems
has been frequently noted in individual hospital accreditation committee
reports and exists independently of DRA studies or of the proposed imple-
mentation of a NIDRA structure (see Section 22.4).
A recovery system should, at a minimum, provide for the following docu-
mented steps:

1. Collection of the device from the surgical field

2. Entry into the patient’s clinical record of a minimum data set
3. Handling and transport of the device to (1) minimize postexplanta-
tion damage to the device; and (2) minimize health risk* to support
personnel
4. Examination and identification of the retrieved device
5. Discharge of the device from institutional care (adequate decontam-
ination, disposal, or discharge to physician, patient, or third party)

* It is routine practice in DRA studies to assume that all explanted devices are contaminated with
human pathogens unless positively sterilized. Because postexplantation sterilization procedures
may affect materials’ properties, routine sterilization is often not performed, during recovery or
retrieval, and devices may need to be handled through some steps of the recovery process in an
unsterilized state.
450 Biological Performance of Materials: Fundamentals of Biocompatibility

22.4 Proposed National Implant Data Retrieval and Analysis

Program (NIDRA)
A 1981 conference* attempted to focus on data and knowledge resulting
from DRA rather than on material and design aspects of implants. However,
it primarily dealt with codification and standardization processes without
addressing the need for an overall knowledge structure with defined internal
data flows.
Notwithstanding this pioneering effort, little progress has been made in
generalizing and unifying the intellectual product of DRA efforts and in
making it accessible in real time. Contemporary DRA programs are still
scarce and tend to be based in hospitals and academic research groups.
Strongly influenced by liability considerations, the medical device industry,
by and large, has elected to react to individual cases of apparent device
malfunction rather than to study the general successful or unsuccessful
performance of its products. Data interchange reflecting case or selected
group studies performed largely without controls continues to be primarily
by podium presentation and paper publication. Two further defects of many
of these studies are that:

• They are performed in large centers that provide secondary or ter-

tiary care and thus have nonrepresentative patient populations.
• They frequently involve or are directed by surgeons and engineers
involved in development of the devices or device classes under
study, without adequate third-party supervision, oversight, or qual-
ity control.

It was apparent by 1992 that a new effort needed to be made to gain

necessary data from the vast human experience of routine clinical use of
implants. The U.S. Food and Drug Administration commissioned me to
develop a design specification for a national data management system. A
questionnaire was developed, widely distributed and the results codified.
The results of that survey clearly supported the need for development of a
national knowledge structure and contain a number of recommendations
that were drawn upon in preparation of this specification. Five guiding
principles were proposed and, validated by the survey, can now be stated
as system requirements for a national effort to study clinical performance of
biomaterials:

• The emphasis should be on understanding biomaterial performance

in vivo as it relates to biomaterials’ composition and processing,
individual patient variables, and device design classes.
* Medical devices: measurements, quality assurance, and standards, Gaithersburg, MD, 9/24-
25/81.
Clinical Performance of Biomaterials 451

• The program should be structured to provide early and continuing

publicly utilizable data on device survival and biomaterial perfor-
mance.
• The system design should embody low-level uniform data retrieval
(for statistical results) combined with focused case studies related
to perceived or possible clinical problems (for biomaterials’ proper-
ties).
• The resulting database should be prospectively linked, in an inter-
active way, with existing and planned engineering, biological, and
clinical databases.
• The system should be decentralized and extramural (nongovern-
mental) insofar as possible, with central intramural activities involv-
ing only planning, direction, data analysis, and audit functions.

The final design specification described the database and its associated
device- and data-acquisition and management systems (termed collectively
the “NIDRA structure”) needed for low-cost, reliable production of bioengi-
neering data from current clinical implant experience to meet the needs of
expanding use of implants.

22.5 Elements of a NIDRA System

The aim of the NIDRA proposal is to describe a system capable of generating,
in a timely fashion, a number of minimum data sets. The driving idea of this
approach is to make the implantation and removal of chronic (>30 days)
implants statistical events much as births and deaths or, for that matter,
purchase and scrapping of automobile tires are. A secondary goal is to make
recovered implants available for larger scale DRA studies than are now
possible within single institutions.

22.5.1 Data Sets

Three data sets are proposed:

• A minimum explantation data (EDATA) set containing no more than

six items:
• Date of removal
• Identification number (ID number) of medical facility where ex-
plantation occurred (per HSS FDA 92-4247*)

* HSS FDA 92-4247: Medical device reporting for user facilities, December 1991.
452 Biological Performance of Materials: Fundamentals of Biocompatibility

• Coded ID number creating link to patient hospital records (may

include previous two items)
• Device identity (per HSS FDA 91-4246*)
• Retrieval grade: recommended codes (referring to device rather
than to clinical outcome)**:
• RG1: no pre- (diagnostic imaging, functional test result, etc.)
or postremoval (naked eye, functional test result, etc.) defects
• RG2: preremoval defects; no postremoval defects
• RG3: pre- and postremoval defects
• RG4: no preremoval defects; postremoval defects
• Device serial/model number (when available) (optional)
• An expanded explantation data set (expanded EDATA): this data set
would include the EDATA set as well as information on implantation
(e.g., original anatomical location), service (e.g., implantation dura-
tion), and analysis (e.g., lipid content of silicone rubber) as well as
the minimum IDATA set (when available). Although it is clear that
the use of standardized analytical procedures would simplify com-
parison of expanded EDATA sets, the careful definition of test meth-
ods and the use of control (reference) materials would permit full
data merger in the envisioned data base.
• A minimum implantation data (IDATA) set containing no more than
six items:
• Date of implantation
• ID number of medical facility where implantation occurred (per
HSS FDA 92-4247)
• Coded ID number creating link to patient hospital records (may
included previous two items)
• Device identity (per HSS FDA 91-4246)
• Indication for use (per DRG)
• Device serial/model number (when available) and/or coded ID
number creating link to manufacturing records

22.5.2 Organizational Elements

The proposed NIDRA structure designed to generate, collate, and analyze
these data sets has four principal elements:

* HSS FDA 91-4246: Classification names for medical devices and in vitro diagnostic products.
August 1991.
** These correspond, respectively, to DRA classes (section 22.2): RG1: class 3, RG2: class 2, RG3:
class 1 (possibly class 2), RG4: class 2.
Clinical Performance of Biomaterials 453

• Clinical institutions
• Study centers
• Data analysis and device management center (DADMC)
• Steering committee

The proposed responsibilities of each element and their manner of interac-

tion are briefly outlined next.

22.5.2.1 Clinical Institutions

Any participating clinical institution would be expected to modify its
(informed) “consent for treatment” procedure to enable off-site analysis of
devices, to operate a recovery system, and to enroll each device by informing
the DADMC (see later section) by FAX transmission of a minimum EDATA
set on the day of explantation of each device. The clinical institution would
then hold the device for a set period (2 to 10 days; recommendation: 3
working days) and subsequently:

• If informed affirmatively (during the holding period) of a need for

the device for an approved DRA study in another center and, upon
receipt of a prepaid shipping container, ship the device to a specified
study center, or
• If not informed affirmatively, dispose of the device in accordance
with recovery system routine practice

The costs for a clinical institution to participate in this program would be

minimal and deviation from conventional recovery practice would in many
cases not be necessary beyond NIDRA case enrollment (FAX preparation
and transmission).

22.5.2.2 Study Centers

The principal public need is for an accessible flow of uniform, high-quality
data concerning properties and clinical performance of implants. It is pro-
posed to support present analysis programs and, if needed, to encourage
the establishment of new ones in academic, medical, or industrial settings,
by defining and funding a set of prospective studies. Support would be
provided from traditional public funding sources as well as from a fund to
be established in relation to the DADMC.
Briefly, experimental questions would be proposed by the NIDRA Steering
Committee and advertised through traditional RFP/RFC channels for
response by interested groups. Examples of such areas of study are in vivo
degradation of silicone elastomers and fatigue processes in spinal fixation
devices. A study center with a funded NIDRA study would have real-time
access to the device enrollment data flowing into the DADMC so that devices
454 Biological Performance of Materials: Fundamentals of Biocompatibility

meeting the criteria defined for each experimental program could be

retrieved rapidly in large numbers. The study center would, as part of its
responsibilities, design and fabricate appropriate shipping containers. It
would identify devices that met the criteria of its study, cause DADMC to
make an affirmative selection of these devices on its behalf, and dispatch
appropriate prepaid shipping containers to the clinical institutions that
enrolled the selected devices. The study center would also bear the respon-
sibility (and cost) of directly contacting the clinical institutions from which
it received selected devices to obtain supplementary data, laboratory test
results, etc. as needed for the specific study. As the study progresses, the
study center would transmit results in real time to the DADMC.
Although many highly capable interdisciplinary research groups are cur-
rently active in the analysis of device materials and performance, as men-
tioned previously, it may sometimes prove necessary to establish dedicated
study centers focusing on a particular perceived device related clinical prob-
lem, on a class of devices in a particular medical/surgical field, or on a
particular biomaterial class. The need for such centers would be defined by
the NIDRA Steering Committee and conventional center support funds
would be sought from NIH, NSF, and other public sources as well as funds
provided from the central fund.

22.5.2.3 Data Analysis and Device Management Center (DADMC)

The DADMC would be the only new permanent federal element of the
proposed NIDRA data management system. Its principal responsibility
would be the creation of the software and hardware to host a publicly
accessible relational database to house the minimum EDATA sets and the
integrated results of focused studies by the study centers, on a grouped basis
as well as through generation of an expanded EDATA set for each specific
device studied. This relational database would be provided with functional
linkages to present databases, such as the MDR system, and proposed bases
such as the proposed FDA-based Biomaterials Compendium.* The
DADMC’s secondary roles would include management and oversight of
clinical institutions’ recovery and enrollment systems, recruitment of new
clinical institutions to the program, performance of statistical analyses on
the database, and design and provision of products for electronic access and
hard-copy publication.

22.5.2.4 Steering Committee

It is intended that the NIDRA Data Management System be coordinated and
directed by a nationally organized steering committee. This committee
would be responsible for further design of NIDRA, for field test and

* At the time of the original NIDRA study (1992), the FDA (CDRH) was compiling data base of
clinically used biomaterials, including properties and relevant standards, from various forms of
pre-approval applications. This effort has apparently been abandoned.
Clinical Performance of Biomaterials 455

implementation, and for scientific oversight and management in the steady

state. Initially, this committee should be staffed by invitation; however, it
might be reasonable for a definitive committee to have identified seats to be
filled by representatives from various professional, scientific, and industrial
organizations. As of today, NIDRA remains a proposal, although limited
efforts are under way in the U.S. and elsewhere to develop and test various
elements of such a national system.

22.6 Autopsy Retrieval Studies

Conventional DRA programs, however well conceived and executed, will
continue to be studies of “failure”; that is, they focus on the few devices for
which the outcome has been unexpectedly less than desired and/or frank
damage is noted on retrieved implants. For a long time, researchers have
recognized this problem and have attempted to compensate for this by in
situ studies and recovery of successful implants after death.
In situ studies have so far been largely limited to the use of conventional
imaging techniques, such as x-ray, CT, and MRI, and occasional sampling
and analysis of fluids and tissues. However, ethical and practical consider-
ations have severely limited the scope and utility of these studies. This is a
still primitive field of effort when compared to the in situ study of natural
systems, but it can be expected to develop in the future.
A number of investigators, such as Sir John Charnley, have recognized the
need to study success and have asked patients prospectively to “return”
their devices when they are no longer needed — that is, after death. The
success of the studies of such devices and the important insights that they
have provided have encouraged a wider approach to autopsy retrievals of
successfully functioning devices and associated tissues, as well as tissues
and fluids from systemic and remote locations.
Involvement in one such program since 1990 (Jacobs et al. 1999) has high-
lighted the benefits and the inherent difficulties of its operation. Several
comments can be made about such programs in general:

• Patients and their close relations are generally interested in their

medical condition and have an inherent willingness to take part in
studies, if they have a fairly low impact on their day-to-day lives,
in order to benefit others. However, individuals have social, reli-
gious, and ethical standards and principles that must always be
honored. Thus, programs must be flexible enough to accommodate
a wide variety of individual concerns.
• Permitting study of one’s body after death for scientific purposes is
a personal decision, much like agreeing to donate organs. As a result,
456 Biological Performance of Materials: Fundamentals of Biocompatibility

successful autopsy retrieval programs depend to a great extent on

continuing, repeated contact with the prospective subjects and their
families by caring, concerned personnel with minimum interference
after death.
• Death rarely arrives on schedule or at a convenient time and the
window of opportunity for satisfactory, uncontaminated retrieval of
device components and tissue specimens is usually quite brief.
Therefore, successful programs require well defined and established
protocols, previously prepared instrument and sample recovery kits,
trained one- to three-person retrieval teams on 24-hour standby
(with adequate coverage for sick leave, vacations, holidays, etc.), and
a reliable communication system to alert all parties as quickly as
possible after death occurs.

In general, autopsy retrieval programs function much as more conven-

tional DRA studies do. However, as the previous comments suggest, costs
are considerably higher. On balance, the scientific results from the few in
operation today have more than repaid the effort required. and they repre-
sent one of the frontiers of biomaterials research.

22.7 Concluding Remarks

It should be a truism that one can only really learn about the clinical per-
formance of biomaterials by actually examining that clinical performance. A
vast human experiment is under way; significant numbers of patients now
have had chronic devices in situ for more than 20 years. The time when
important new discoveries about the biological performance of biomaterials
can be made in the laboratory or in limited animal studies without primary
reference to this clinical experience has probably passed. Device retrieval
and analysis studies, national data systems, and autopsy retrieval programs
will come to play important roles in obtaining data and insight to benefit
future generations of patients.
It is clear that the failure of such systems to emerge, for whatever reasons,
has profound impacts on the quality and cost of health care. In 2000, a
U.S.–NIH-sponsored national consensus development conference* con-
cluded in part that:

• “Implant retrieval and analysis is of critical importance in the pro-

cess of improving care of patients…”

* Improving medical implant performance through retrieval information: challenges and oppor-
tunities, Bethesda, MD, January 10–12, 2000.
Clinical Performance of Biomaterials 457

• “The [continuing] failure to appreciate the value of…retrieval and

analysis is a serious impediment to medical device research…
[R]etrieval will lead to the acquisition of information necessary to
improve the quality of future devices.”

In other words, better understanding of biological performance of devices

(and their materials of construction) in actual clinical settings is and remains
important and necessary.

References
Black, J. and Fielder, J.H., Ethical aspects in device retrieval, Proc. Implant Retrieval
Symposium, Society for Biomaterials, St. Charles, LA, 9/17–20/92, 14–1.
Jacobs, J.J. et al., Postmortem retrieval of total joint replacement components, J.
Biomed. Mater. Res. (Appl. Biomat.), 48(3), 385, 1999.
Moss, A.J., Advance Data from Vital and Health Statistics, No. 191, National Center for
Health Statistics, 1991, 1.

Bibliography
Anderson, J.M., Procedures in the retrieval and evaluation of vascular grafts, in
Kambic, H.E., Kantrowitz, A. and Sung, P. (Eds.), Vascular Graft Update: Safety
and Performance, STP 898, American Society for Testing and Materials, Phila-
delphia, 1986, 156.
Anderson, J.M., Cardiovascular device retrieval and evaluation, Cardiovasc. Pathol.,
2(3)(suppl.), 199S, 1993.
Black, J., An overview of goals and perspectives of implant retrieval, Int. J. Risk Safety
Med., 8, 99, 1996.
Brooks, C.R. and Choudury, S.A., Metallurgical Failure Analysis, McGraw–Hill, New
York, 1992.
Collins, J.A., Failure of Materials in Mechanical Design, 2nd ed., John Wiley & Sons,
New York, 1993.
Das, A.K., Metallurgy of Failure Analysis, McGraw–Hill, New York, 1997.
Engel, L. et al., An Atlas of Polymer Damage: Surface Examination by Scanning Electron
Microscope, Prentice Hall, Englewood Cliffs, NJ, 1981.
Fraker, A.C. and Griffin, C.D. (Eds.), Corrosion and Degradation of Implant Materials:
Second Symposium, STP 859, American Society for Testing and Materials, Phil-
adelphia, 1985.
Scheirs, J., Compositional and Failure Analysis of Polymers: A Practical Approach, John
Wiley & Sons, New York, 2000.
Syrett, B.C. and Acharya, A. (Eds.), Corrosion and Degradation of Implant Materials, STP
684. American Society for Testing and Materials, Philadelphia, 1979.
Weinstein, A., Horowitz, E. and Ruff, A.W. (Eds.), Retrieval and Analysis of Orthopaedic
Implants, NBS Special Publication 472, U.S. Government Printing Office, Wash-
ington, D.C., 1977.
Glossary

G.1 Introduction
From its beginning, the intellectual field of biomaterials science and engi-
neering has been hampered by having grown up from a group of supporting
basic and applied endeavors (Chapter 1). As a result, its vocabulary has been
drawn from a number of varied historical sources. The practice in the disci-
pline has been, in some cases, to give new meanings to old terms. In addition,
practitioners in the field have had to invent or adopt terminology to describe
their insights. The resulting vocabulary is so far a piecemeal assemblage and,
except for efforts by various authors, does not appear in any one place.
Perhaps the most noteworthy attempt to solve this problem has been The
Williams Dictionary of Biomaterials (1999). An earlier effort by Szycher (1992)
is of little practical use because it is simply a compilation of U.S. legal medical
device definitions as required under the Medical Device Amendments (1976)
(Chapter 20) and later legislation, previously published in Title 21 of the
Code of Federal Regulations (CFR) (FDA91-4246).
At international consensus conferences in 1987 (Williams 1987) and 1991
(Williams et al. 1992), attempts were made to develop a standard core nomen-
clature for biomaterials. All of the definitions considered at these two meet-
ings are included in the following two sections. Adopted ones are identified
as follows:

* = 1987 conference consensus definition

*p = 1987 conference provisional definition
** = 1991 conference consensus definition

Beyond these specialized terms, the vocabulary of biomaterials is still

drawn from a broad base in the engineering and scientific disciplines. Many
of the terms used have common language meanings, so a popular dictionary
can be used. For medical terms, a more specialized source is required. Many
medical dictionaries are available; Dorland’s Illustrated Medical Dictionary,
30th edition (2003) in the pocket edition is recommended. Its small size
makes it convenient to keep on the desk and to carry to the library, leaving
no excuse for misunderstanding.

459
460 Biological Performance of Materials: Fundamentals of Biocompatibility

G.2 Glossary
Note: When two definitions are provided, the first is the more common usage.

Acute Duration of less than 30 days; however, durations associated with

clinical treatment (such as use of instruments, dialysis equipment,
etc.) are usually termed short term or intraoperative. See also:
chronic.
Adaptation The ability of tissues to adapt to local requirements, includ-
ing reaction to the chemical, physical, or electrical properties of
implants.
Allograft See: graft, allo-.
Artificial organ* A medical device that replaces, in part or in whole, the
function of one of the organs of the body.
Autograft See: graft, auto-.
Bioactive The ability of a biomaterial surface or coating to adhere di-
rectly to soft or hard tissue without an intermediate layer of
modified tissue.
Bioactive material
1.** A biomaterial designed to elicit or modulate biological
activity.
2.* One designed to induce specific biological activity.
Bioadhesion* The adhesion of cells and/or tissue to the surface of a
material.
Bioattachment* The fastening of cells and/or tissue to the surface of a
material, including mechanical interlocking (see: ingrowth; on-
growth).
Bioceramic Strictly, any ceramic biomaterial. Usually used as equiv-
alent to bioactive material, although bioactive polymers also
exist.
Biocompatible material One having acceptable host and material re-
sponse in a specific application (see: host response; material re-
sponse; first definition of biocompatibility).
Biocompatibility
1.* The ability of a material to perform with an appropriate host
response in a specific application.
2. Biological performance in a specific application that is judged
suitable to that situation. Note: the implication that biocom-
patibility implies little or only beneficial host response is to
be avoided by context.
Glossary 461

Biodegradation
1.** The breakdown of a material mediated by a biological
system.
2.*p The gradual breakdown of a material mediated by specific
biological activity.
Biological environment See: environment, biological.
Biological performance The interaction between materials and living
systems (see: host response; material response).
Biomaterial
1.** A material intended to interface with biological systems to
evaluate, treat, augment, or replace any tissue, organ, or func-
tion of the body (most general; see below; see also bioactive
material).
2.* A nonviable material used in a medical device, intended to
interact with biological systems.
3.
A material of natural or manmade origin that is used to
direct, supplement, or replace the functions of living tissues.
Also compound forms may be used: ceramic biomaterial, com-
posite biomaterial, metallic biomaterial, polymeric biomaterial.
Biomaterial, inert One that elicits little or no host response. Also
termed: type 1 biomaterial.
Biomaterial, interactive One designed to elicit, promote, or modulate a
specific host response, such as hard tissue adhesion. Also termed:
type 2 biomaterial.
Biomaterial, manmade (or manufactured) One that is significantly pro-
cessed from raw materials of inorganic or organic origin.
Biomaterial, native (or natural) One obtained from natural organic
sources and implanted essentially unprocessed.
Biomaterial, replant One consisting of live native cells or tissue, cul-
tured in vitro from cells obtained previously from specific patients.
Also termed: type 4 biomaterial.
Biomaterial, viable One that incorporates host tissue and/or live cells
and/or active DNA plasmids and is capable of being remodeled
and/or is resorbable and/or bioresorbable. Also termed: type 3
biomaterial.
Biomaterials The organized study of the materials properties of the
tissues and organs of living organisms; the development and
characterization of pharmacologically inert materials to measure,
restore, and improve function in such organisms; and the inter-
action between viable and nonviable materials.
462 Biological Performance of Materials: Fundamentals of Biocompatibility

Biomaterials engineering The application of the principles of biomate-

rials science and its foundation sciences to the solution of practical
problems of human health, disability, and disease.
Biomaterials science The study and knowledge of the interaction be-
tween living and nonliving materials.
Biomaterials science and engineering Compound form; modern de-
scriptor for the intellectual, academic, and industrial field previ-
ously referred to as biomaterials (see: biomaterials engineering;
biomaterials science).
Biophysiological environment See: environment, biophysiological.
Bioprosthesis* An implantable prosthesis that consists totally or sub-
stantially of nonviable, treated donor tissue.
Bioresorbable The ability of a biomaterial to be digested by or as a
consequence of cellular activity and thus disappear in part or in
whole after implantation. Should be used to imply specific action
of cells or tissues (see: resorbable).
Bioresorption*p The process of removal by cellular activity and/or dis-
solution of a material in a biological environment.
Bone bonding** The establishment by physicochemical processes of
continuity between implant and bone matrix.
Calor Local tissue temperature rise; one of the four classic signs of
inflammation (see also: dolor; rubor; tumor).
Cancer A disease of multicellular organisms characterized by uncon-
trolled multiplication and spread of abnormal forms of host cells.
Capsule Tissue surrounding an implant produced by local host response
(see also: incapsulization; host response, local).
Carcinogen An agent capable of causing cancer.
Carcinogenesis Malignant, inheritable change in mammalian cells.
Carcinogenesis, chemical (attribute of an implant) Carcinogenesis in-
duced by the chemical composition of an implant or its degrada-
tion products.
Carcinogenesis, foreign body (attribute of an implant) Carcinogene-
sis induced by the physical form of an implant, independent of
its chemical composition.
Chelation A type of interaction between an organic compound (having
two or more points at which it may coordinate with a metal) and
the metal to form a ring-type structure.
Chemotaxis Orientation or movement of cells towards a chemical
source.
Chronic Duration of 30 days or longer (see also: acute).
Glossary 463

Coagulation Sequential process in blood leading to thrombus formation

(see also: thrombus).
Colony-forming unit The minimum number of bacteria required to
grow a cell cluster or colony on a suitable solid culture medium.
Abbreviated: cfu.
Control material See: reference material.
Coordination The joining of an ion or molecule to a metal ion by a
nonionic valence bond to form a complex ion or molecule.
Cytokine Chemical species used for intercellular signaling.
Cytotoxic Having a deleterious or adverse effect on cells. Note: does not
necessarily imply cell death.
Device matching Selecting an implant suited to the expected implant
life history of a particular patient. (See also implant life history.)
Device, medical* An instrument, apparatus, implement, machine, con-
trivance, in vitro reagent, or other similar or related article, includ-
ing any component, part, or accessory, that is intended for use in
the diagnosis of disease or other conditions, or in the cure, miti-
gation, treatment, or prevention of disease in man. (Note: in mod-
ern usage, for “man” read “humans.” Such a definition may be
equally well used in veterinary medicine; however, there the pre-
ferred term would be veterinary medical device.)
Device, percutaneous*p A medical device that passes through the skin,
remaining in position for a significant length of time (may be
equivalent to implant [acute; chronic], permucosal).
Device, permucosal*p A medical device that passes through a mucosal
layer and remains in position for a significant length of time (may
be equivalent to implant [acute; chronic], percutaneous).
Diapedesis The outward passage of blood cells through intact vessel
(arterial or venous) walls.
Dilantant Property of a lubricant; increasing shear viscosity with in-
creasing shear rate (see also: thixotropic).
Dolor Local pain; one of the four classic signs of inflammation (see also:
calor; rubor; tumor).
Environment, biological Conditions encountered within an animal or
human body.
Environment, biophysiological Controlled chemical (inorganic) and
thermal conditions, with addition of appropriate cell products,
simulating a portion of a biological or pericellular environment.
Environment, pericellular Conditions encountered immediately adja-
cent to living cells in vitro or, more generally, within an animal or
human body.
464 Biological Performance of Materials: Fundamentals of Biocompatibility

Environment, physiological Controlled chemical (inorganic) and ther-

mal conditions simulating a portion of a biological, biophysiolog-
ical, or pericellular environment.
Extrusion Resolution in which implants in contact with epithelial tissue
(skin and the lining of natural internal body cavities) are sur-
rounded by a down-growing extension of such tissue, directed
towards extruding the implant from the body. This is termed
marsupialization, due to the resemblance of the newly formed
tissue to a kangaroo’s pouch.
Factor XII Initial factor in intrinsic pathway for blood coagulation; also
called Hageman factor.
Foreign body reaction A variation in normal tissue behavior caused by
the presence of a foreign material (see also: host response, local).
Glycocalyx A protective enveloping film formed on the surface of im-
plants by some types of bacteria.
Graft* A piece of viable tissue or collection of viable cells transferred
p

from a donor site to a recipient site for the purpose of reconstruc-

tion of the recipient site (most general; see following).
Graft, allo-*p A graft taken from another individual of the same species
as the recipient. (Note: all humans are members of a single
species.)
Graft, auto-*p A graft taken from a source in the individual who receives
it; that is, the donor and the recipient are the same person (see
also: replant).
Graft, xeno-*p A graft taken from an individual of a different species
from the recipient’s. (Note: the source is explicitly nonhuman.)
Granuloma Actively growing provision soft tissue that precedes remod-
eling phase of inflammatory response; may become chronic in the
absence of resolution.
Hageman factor See factor XII.
Hemolysis Release of hemoglobin due to damage to red blood cells.
Heterograft Old term for autograft (see: graft, auto-).
Homeostasis The maintenance of conditions necessary for mammalian
life.
Homograft Old term for allograft (see: graft, allo-).
Host response
1. The local and systemic response, other than the intended
therapeutic response, of living systems to the material; a
component of biological performance (most general, see
below).
2.* The reaction of a living system to the presence of a material.
Glossary 465

Host response, level of The nature of the host response in a standard

test with respect to the response obtained with a reference ma-
terial.
Host response, local The response, other than the intended therapeutic
response, of tissue and organs contacting a biomaterial.
Host response, remote The response, other than the intended therapeu-
tic response, of remote tissue and organs in an individual with
one or more implants.
Host response, systemic The distributed or disseminated response, oth-
er than the intended therapeutic response, of tissue and organs
in an individual with one or more implants.
Hybrid artificial organ* An artificial organ that is a combination of vi-
able cells and one or more biomaterials (see: biomaterial, viable;
hybrid device).
Hybrid device One utilizing cells from patient or donor sources cul-
tured in vitro and combined with resorbable or metabolizable
supports and matrices (may be equivalent to hybrid artificial
organ; see also: biomaterial, viable).
Implant
1. A device placed within an animal or human body by the act
of implantation (see: implantation).
2.* A medical device made from one or more biomaterials that
is intentionally placed within the body, totally or partially
buried beneath an epithelial surface (most general, see
following).
Implant, acute A device that remains in situ for less than 30 days.
Implant, chronic (or permanent) A device that remains in situ for 30 or
more days.
Implant, intraoperative A device removed within hours to days at the
termination of the surgical or therapeutic procedure.
Implant, percutaneous A device that, after placement, penetrates the
skin (see also: device, percutaneous).
Implant, permucosal A device that, after placement, penetrates the mu-
cosa (see also: device, permucosal).
Implant life history Lifetime performance requirements for an implant.
Implantation Placement of a device or material within the body of an
animal or human by a medical or surgical professional, in such a
way as to breach one or more epithelial layers and to leave ma-
terials and/or components in place after the initial procedure is
completed.
Incapsulization Resolution in which the implant is surrounded
and walled off from normal tissue by a collagenous, relatively
466 Biological Performance of Materials: Fundamentals of Biocompatibility

acellular tissue termed capsule that much resembles scar tissue.

In a bony location, the capsule may be mineralized and is called
a sequestrum.
Inflammation See: inflammatory response.
Ingrowth Formation of tissue within pores, etc. in the body of an im-
plant (see also: ongrowth).
Integration Resolution for a very limited number of materials, such as
“bioactive” glasses of selected compositions and some metals,
such as pure titanium for which direct “bonding” or apparent
adhesion to normal tissue may take place.
Iontopheresis Facilitation of diffusion by imposing an electrical gradi-
ent on a charged diffusing species.
Ligand Any ion or molecule that, by donating one or more pairs of
electrons to a central metal ion, is coordinated with it to form a
complex ion or molecule.
Marsupialization See: extrusion.
Material response The response of a material to living systems; a com-
ponent of biological performance.
Material response, level of The nature of the material response in a
standard test with respect to the response obtained with a refer-
ence material.
Mutagenesis Induction of a permanent (inheritable) genetic change.
Ongrowth Formation of tissue directly on the surface of an implant (see
also: ingrowth). Does not imply adhesion.
Oppenheimer effect Induction of primarily subcutaneous tumors in ro-
dents through a foreign body mechanism; named after its discov-
erers, E. and B.S. Oppenheimer.
Opsonization Coating of bacteria or biomaterial particles with native
proteins, such as complement factors, rendering them dectable as
“foreign” by phagocytic cells.
Orthosis A device applied externally to the body to provide stability
and to control motion. May or may not replace a portion of a limb
(see also: prosthesis).
Osseointegration
1. Clinical stability of an implant anchored in bone; often taken
to refer to implants with bioactive coatings.
2.** A description of clinical performance of devices; not appli-
cable to the description of biomaterial–bone interactions.
(Note: sometimes spelled osteointegration; however, ossein-
tegration is preferred.)
Glossary 467

Osteoconductive Property of a biomaterial that encourages bone, al-

ready being formed, to lie close to or adhere to its surface.
Osteogenic Property of a biomaterial that stimulates bone growth in the
implant site.
Osteoinductive Property of a biomaterial that encourages bone to form
close to or adhering to its surface. Note: the term should be used
to apply to the material (matrix) in the absence of specific osteoin-
ductive signaling molecules or ligands.
Osteointegration See osseointegration.
Osteolysis Cellularly mediated bone loss (also called small particle dis-
ease) secondary to debris production and/or release by implants
in or near to bone. Notes: previously, also incorrectly called “ce-
ment disease.” Do not confuse with stress shielding.
Pericellular environment See: environment, pericellular.
Phagocytosis The process of internalizing small particles by mammalian
cells.
Phagocytosis, frustrated The failure of mammalian cells to phagocy-
tose particles due primarily to their size, resulting in release of
cytokines.
Physiological elements Elements (calcium, phosphorus, potassium,
sulfur, sodium, chlorine, and iron) — other than oxygen, hy-
drogen, nitrogen, and carbon — required for mammalian
homeostasis.
Physiological environment See: environment, physiological.
Prosthesis* A device that replaces a limb, organ, or tissue of the body.
Note: externally worn prostheses, especially ones not permanent-
ly attached to the body, are more properly termed orthoses; see:
orthosis.
Pseudointima Tissue consisting of a firm fibrin clot, with occasional
islands of endothelial cells, formed by resolution on interior
(blood-contacting surfaces) of cardiovascular implants.
Pseudoneointima Pseudointima in which cells form a continuous layer.
Pyrogen A substance producing fever (heat) in vivo.
Reference material A material that, by standard test, has been deter-
mined to elicit a reproducible, quantifiable host or material re-
sponse.
Replant An autograft produced in vitro from DNA, cells and/or tissues
obtained from the donor, utilizing one or more techniques of
tissue engineering.
Resolution The stable end state of the inflammation or inflammatory
response associated with an implant.
468 Biological Performance of Materials: Fundamentals of Biocompatibility

Resorbable The ability of a biomaterial to be dissolved or digested and

thus disappear after implantation. Note: does not imply specific
action of cells or tissues; see: bioresorbable.
Resorption Resolution associated with resorbable or bioresorbable im-
plants in which the tissue site condenses to a collapsed scar or,
in the case of bone, completely remodels to normal (regenerated)
tissue.
Response, acute Host response or material response in less than 30 days
after implantation.
Response, chronic Host response or material response in 30 days or
more after implantation.
Response, host See host response.
Response, immune Host response involving humoral or cellular specif-
ic immune mechanisms.
Response, inflammatory The cell-mediated local and regional response
directed towards stabilizing injured tissue, restoring physiologi-
cal status quo ante, removing dead or damaged tissue elements
and foreign material, and correcting the structural and functional
loss due to the initial insult. The four classical signs of inflamma-
tion are: redness (rubor), swelling (tumor), pain (dolor), and heat
(calor).
Response, material See: material response.
Rubor Local tissue reddening; one of the four classic signs of inflamma-
tion (see also: calor; dolor; tumor).
Sequestrum Mineralized capsule, generally in or on bone. In the absence
of bone, more properly referred to as ectopic calcification.
Small particle disease See osteolysis.
Sonopheresis Facilitation of diffusion, particularly transdermally, by in-
ducing reversible ultrasonic microcavitation.
Standard test A well-defined, repeatable test of host or material re-
sponse, generally involving the use of one or more reference
materials.
Stress shielding Effect resulting in a decreased density of bone as a
consequence of load sharing between an implant and tissue.
Thixotropic Property of a lubricant; decreasing shear viscosity with in-
creasing shear rate (see also: dilantant).
Three-phase junction Referring to percutaneous implants; the point at
which tissue, implant, and air meet.
Thrombogenicity* The property of a material that induces and/or pro-
motes the formation of a thrombus (most general, see below).
Glossary 469

Thrombogenicity, general*p Thrombogenicity of a blood–material

system.
Thrombogenicity, inherent
1.** Thrombus formation controlled by the material surface.
2.*p Reaction-controlled thrombogenicity at the surface of a ma-
terial.
Thrombogenicity, non-*p The characteristic of a material that leads to
minimal thrombogenicity.
Thrombus A solid mass formed from the molecular and cellular con-
stituent of blood (see also: thrombogenicity).
Tissue engineering (provisional definition) Elaboration of cells and
tissues outside a living organism, intended for use as components
of a viable biomaterial or replant by use of engineering methods
and techniques.
Tumor Local tissue swelling; one of the four classic signs of inflamma-
tion (see also: calor; dolor; rubor).
Vroman effect The temporal succession of molecular species adherent
to surfaces of implants, named after its discoverer, Leo Vroman.
Wolff’s law “The form being given, tissue adapts to best fulfill its me-
chanical function” (after Wolff 1892 in Maquet and Furlong 1986).
Xenograft See graft, xeno-.

G.3 Deprecated Terms

Williams (1987) also lists the following terms, which the attendees to the first
consensus conference considered redundant or inappropriate under certain
conditions; it is suggested that these should be deprecated:

Antithrombogenic
Bioceramic To be consistent with the advice on biopolymer and biometal
(see following), this term should also be deprecated. However,
see ceramic biomaterial and bioceramic (Section G.2), notwith-
standing the advice of the 1991 conference to deprecate the latter.
Biocompatible When used as an adjective.
Bioinert
Biological performance Preferred term is biocompatibility.
Biometal Preferred term is metallic biomaterial.
Biopolymer Term has an agreed meaning in molecular biology; the
preferred term is polymeric biomaterial.
470 Biological Performance of Materials: Fundamentals of Biocompatibility

Biostability
Blood compatibility
Material rejection
Material response
Thromboresistant The preferred term is nonthrombogenic (see: non-
thrombogenicity, Section G.2).
Tissue response Preferred term is host response, local.

References
FDA 91-4246: Classification Names for Medical Devices and in Vitro Diagnostic Products.
U.S. Government Printing Office, Washington, D.C., 1991.
Maquet, P. and Furlong, R., The Law of Bone Remodeling, Wolff, H. (1892), Springer–Ver-
lag, Berlin, (transl.), 1986.
Newman Dorland, W.A., Dorland’s Illustrated Medical Dictionary, 30th ed., W.B. Saun-
ders, Philadelphia, 2003.
Szycher, M., Szycher’s Dictionary of Biomaterials and Medical Devices, Technomic Pub-
lishing Co., Lancaster, PA, 1992.
Williams, D.F. (Ed.), The Williams Dictionary of Biomaterials, Liverpool University Press,
Liverpool, U.K., 1999.
Williams, D.F. (Ed.), Definitions in Biomaterials: Proceedings of a Consensus Conference
of the European Society for Biomaterials, Chester, England, March 3–5, 1986. Else-
vier, Amsterdam, 1987.
Williams, D.F., Black, J. and Doherty, P.J., Definitions in biomaterials, Second Con-
sensus Conference on Definitions in Biomaterials, in Doherty, P.J., Williams,
R.L., Williams, D.F. and Lee, A.J.C. (Eds.), Biomaterial–Tissue Interfaces. Advances
in Biomaterials Vol. 10, Elsevier, Amsterdam, 1992, 525.
Index

A surface/interface interactions, 77, 78–79

Age, and inflammatory response, 148
Ablative surfaces, thromboresistant materials Aggregation, surface/interface interactions,
development, 170 74
Abrasion, 116, 117 Albumin, surface/interface interactions, 77,
Absorbable materials, fracture strength, 101 81
Absorption, 36–38 Alkaline/basic conditions, corrosion, 50
drug release devices, 46 Allergic foreign body response, 225–241
mechanical properties clinical performance of biomaterials, 331
elastic modulus, 92 hypersensitivity reactions, classes of, 230
environment effects, 91 hypersensitivity reactions, implant-
fracture strength, 103 associated, 230–240
mineral metabolism metals, 231–240
chromium, 283–295 polymers, 230–231
iron, 276–277 immune response, mechanisms of,
undesirable, examples of, 38–41 226–229
and wear debris, 121 specific versus nonspecific response, 225,
Acellular fibrous capsule, 146 226
Acid treatment, passivation, 54 systemic/remote site response
Activation, surface/interface interactions, mechanisms, 323
80–81 Allo-, defined, 210
Acute response, inflammation, 148, 149 Allobiopsy cells, in vitro tissue growth and
Adaptation, 183–199 replantation, 211, 214
examples, implant applications, 186–197 Allografts
bone response, remodeling near cell sources, 211, 212
implants, 191–196 vascular, 174
bone response, to electrified implants, Alloys
196–197 carcinogenicity, 253, 255
tendon replacement, 189–191 corrosion, 65–66
vascular prosthesis, 186–189 leaching, 62
factors in local host response, 198–199 pitting, 61
interdependency of effects, 198–199 uniform attack, 59
systemic/remote site response corrosion product distribution, 299
mechanisms, 322 electrochemical series, 54, 55, 56
tissue growth strategies, 183–185 inflammation, phagocytosis, 156–157
utility of active tissue response, 197–198 systemic distribution and excretion
Adhesion dissolved species, distribution and
adaptation, physical induction factors, excretion of, 303
199 multicompartment distribution
surface/interface interactions, 78–86 models, 306–309
Adhesive wear, size of debris particles, 119 one compartment distribution models,
Adsorption 304, 305
hypersensitivity reactions, implant- test methods, in vivo, 369
associated, 230–231 wear, size of debris particles, 119, 120

471
472 Biological Performance of Materials: Fundamentals of Biocompatibility

Alumina, 320 Associative bonding, surface/interface

corrosion, 68 interactions, 74
fracture strength, 103 ASTM F-562, leaching, 62–63
implant material properties, 131 Asymmetry, implant life history, 26
Aluminum Atomic bonds, and mechanical properties,
corrosion 90, 91
leaching, 62 Atrophy, adaptation, 199
products of, 297 Auto-, defined, 210
mineral metabolism, 273, 274 Autobiopsy cells, in vitro tissue growth and
systemic distribution and excretion, 303 replantation, 211, 214
Amalgam, dental, 320 Autografts, cell sources, 211
American Dental Association, 406–407 Autoimmune hemolytic anemia, 231–232
American National Standards Institute Autoimmunity, 226
(ANSI), 126–129 dental amalgam, 320
American Society for Testing and Materials hypersensitivity reactions, 231
(ASTM) International Standards, Autoreplant, 210–211
125–129, 407–414
device retrieval and analysis, 446
preclinical tests, 388, 389, 390, 391
in vivo test protocol, 358, 374–378 B
American Type Culture Collection (ATCC),
340 Bacteria
Ames carcinogenicity test, 346–347 Ames carcinogenicity test, 346–347
Amperes per square centimeter (corrosion infection, 150–155
rate), 57 iron and, 280–282
Analysis of devices, see Clinical performance Bacterial toxins, inflammation, 140
of biomaterials Barrier membrane, drug release devices, 45
Anaphylaxis, 230 Basophils, blood composition, 24
Angiogenesis Benign, defined, 245
inflammation, 145 Berg limit, 173
small particle disease, 158 Bertrand curve, drug release, 44
Animal models, see also Test methods, in vivo Beryllium, 319
implant models Biliary excretion, iron, 301–302
animal welfare issues, 356–358 Bioactive materials
biological performance, 8 classification of responses, 8
Anode/anodic corrosion definitions, 7
crevice, 60 Bioavailability, mineral metabolism, 288
electrochemical series, 55 Biocompatibility definitions and issues, 3–14
potential-current relationship, 57 biological performance, 5–6
rate of, 56, 57 consensus definitions, 6–7
Anodic polarization, passivation, 54 definitions, 5, 6
Anomalous wear, 121 paradigm shift, 12–14
Antagonistic interactions, inflammation, 148 professional and disciplinary issues,
Antibiotic-impregnated implants, 151–152 10–11
Antibody, immune response mechanisms, reactogenicity and, 8
226–227 terminology, 3–4
Antigen-antibody complex formation, Biodegradation
immune response mechanisms, definitions, 7
226–228 paradigm shift, 12
Aortic valve, mechanical conditions, 22 Bioglass™
Archard's constant, 115 bone bonding and adhesion, 195
Arginine-glycine-aspartic acid (RGD), corrosion, 68
208–209, 213 Bioinert (classification of responses), 8
Arsenic, 319, 320 Biological differences, and implant life
Articular cartilage, remodeling, 146 history, 22–23
Index 473

Biological environment, 17–29 Bond strength, metals and ceramics, 91

classes of exposure environments, 19 Bond structure
comparison of internal and external and mechanical properties, 90
conditions, 17–18 Bone
corrosion, 67 adaptation
elements of, 20–22, 23, 24 adhesion, 194–195
implant life history, 22–23, 25–27 to electrified implants, 196–197
preimplantation handling, 28–29 fracture healing, 184–185
problems in definition of, 18–20 ingrowth into porous biomaterials,
Biological models, paradigm shift, 12 193–194
Biological molecule-biomaterial surface physical induction factors, 199
interactions, see Surface remodeling near implants, 191–196
interactions, biological molecules stress shielding, 191–193
and biomaterial surfaces clinical performance of biomaterials, 330,
Biological performance, definitions and 331
issues, 5–6 materials testing sites, 355, 356
Biological transducer, Wolff's law, 197 mechanical conditions, 22
Biomaterials, definitions, 6, 7 remodeling, 146
Biomaterials science, 10–11 small particle disease, 159–160
Biophysiological environments, 19 wear, anomalous, 121
Biotolerant (classification of responses), 8 Bone bonding, definitions, 7
Blood, see also Coagulation and hemolysis Bone cements, polymer material properties,
absorption of materials from, 39–40 130
composition of, 24 Bone endoprostheses, systemic distribution
and corrosion, 65 and excretion of monomers,
host-implant interactions, 19 296–297
leaching, 42 Bone particulates/fragments, wear
paradigm shift, 12 mechanisms, 118
Pourbaix diagram, 52 Bony plate formation, 196
Blood cells, see also Coagulation and Boundary conditions, leaching, 42
hemolysis Boundary lubrication, 109, 111, 112
blood composition, 24 Brain, materials testing sites, 355
blood contact tests, 347–349 Brave New World (Huxley), 218
coagulation cascade, 166 British Standards Institute (BSI), 125–129, 359
inflammation, 141–145 Brittle fracture, mechanical properties,
leukemia and lymphoma, defined, 245 99–100, 101
test methods, in vivo, 379 Brittle materials
Blood contact tests, performance testing, fracture strength, 98–99
347–350 mechanical properties, fracture strength,
dynamic, 349–350 101
static, 348–349
Blood vessels, see Endothelium/endothelial
cells; Vascular response
Blood viscosity, 368 C
Body compartments, corrosion product
distribution, 300–301 Calcification, thromboresistant materials
Body fluids/body water, see Fluids, body development, 174
Bonding Calcium
adaptation, bone adhesion, 195 coagulation cascade, 166, 167
definitions, 7 dietary metal intake, 286
and mechanical properties, 91 metabolic disorders, 320
mechanical properties, yield strength, 96 Calcium hydroxyapatite, corrosion, 68
surface/interface interactions, 74 Callus, bone fracture healing, 184
Bonding energy, organometallic compounds, Cancellization, 196, 199
75 Cancellous bone
474 Biological Performance of Materials: Fundamentals of Biocompatibility

adaptive remodeling, 196 Cathodic potentials, electrochemical series,

mechanical conditions, 22 55
Cancer, definitions, 245; see also Cell, paradigm shift, 13, 14
Carcinogenesis Cell adhesion molecules, 207–208
Canine vena cava test, 367, 368 Cell biology, 207
Capillary vasodilation, see Vascular response Cell culture spectrum (CCS), 344
Capsular tissue, local host response in Cell culture/tissue culture/in vitro studies,
humans, 380–381 328
Capsule formation biological performance, 8
adaptation, bone adhesion, 194–195 carcinogenesis, foreign-body, 261
inflammation, 146–148 host-implant interactions, 20
test methods, materials performance, paradigm shift, 14
361–362, 379 performance testing
Carbon dioxide levels, physicochemical carcinogenicity, 345–347
conditions, 22 cell types and observation, 340–341
Carbon fibers delayed hypersensitivity tests, 345
implant material properties, 133 examples of use in materials testing,
tendon prosthesis, 190 341–342
Carcinogenesis, 245–268; see also Neoplastic generic methods, 338
response host response screening, 342–344
chemical, 246–257 problems with, 339–340
carcinogens, 249–252 types of, 338
common knowledge, folklore and surface/interface interactions, 81
facts, 246–249 Cell debris, inflammation, 145
latent period, 256, 257 Cell environment, 19
metals, 252–256 Cell growth/proliferation, adaptation, 183,
types of carcinogens, 249 184
clinical performance of biomaterials, 331 Cell-mediated immunity, immune response,
definitions, 245–246 mechanisms of, 227
foreign body, 257–262 Cell-receptor paradigm, in vitro tissue growth
additional studies, 260–262 and replantation, 206–210
early observations of, 257–258 Cells, blood, blood composition, 24
mechanisms, 258–260 Cell size, adaptation, 184
implant-related, evidence for, 263–268 Cell sources
local host response, 148, 149 types of tissue engineering, 205
nonspecific, 262–263 in vitro tissue growth and replantation,
systemic/remote site response 210–214
mechanisms, 322, 323 Cellular attack, particle size and, 69
test methods, materials performance Cellular invasion
in vitro, 345–347 implant, see Tissue ingrowth
in vivo, 366–367 inflammation, 140, 141–145
Carcinogens, chemical, 249–252 Cell wall (plasma membrane), 206–207
Carcinoma, defined, 245 Cement disease, 158–159
Cardiovascular function tests, in vivo, Ceramics
367–369 bone adaptive remodeling
Carticel™, 216–217 adhesion, 195
Cartilage ingrowth into porous biomaterials,
remodeling, 146 193–194
in vitro tissue growth and replantation, corrosion, 68
213, 216–217 dissolution of, 68–69
Cast metals, intergranular corrosion, 61–62 implant material properties/standards,
Cathode corrosion 131–132
crevice, 60 infection hazards, 154–155
potential-current relationship, 57 lubrication, transfer film, 117
rate of, 56 mechanical properties
Index 475

elastic modulus, 90, 91, 95 Chromium

fracture strength, 99–100, 103 carcinogenicity, 253, 254, 255, 267, 268
organ function, effects of degradation corrosion
products on, 320–321 in biological environment, 67
Cerclage wires intergranular, 62
clinical performance of biomaterials, corrosion product distribution, 298, 299
330 hypersensitivity reactions, 232, 233,
corrosion, 65 234–235, 237, 239
Cerebral cortex, materials testing sites, 355 inflammation
Cervital®, 195 chemotaxis inhibition, 158
Chain scission, mechanical properties phagocytosis, 157
elastic modulus, 93, 94 metabolic disorders, 319, 320
environment effects, 91 mineral metabolism, 273, 274, 283–285
yield strength, 96 dietary intake, 286, 287
Charge, see also Electrical potentials organometallic compounds, 75–76
corrosion, in biological environment, 67 Pourbaix diagram
interfacial interactions, 82–83 reactions in presence of chloride ion,
surface/interface interactions, 82 52–54
Chelation, 75 reactions in pure water, 51–52
Chemical activity, phagocytes, 144–145 systemic distribution and excretion, 303
Chemical bonds equilibrium models, 310–311
and mechanical properties, 90, 91 multicompartment distribution
mechanical properties models, 308
fracture strength, 103 one compartment distribution models,
yield strength, 96 304, 305
surface/interface interactions, 74 Chromium alloys
Chemical carcinogenesis, 246–257 corrosion and leaching, 62–63, 68
adaptation, physical induction factors, fracture strength, 102
199 Chromium release assays, 349–350
carcinogens, 249–252 Chronic response, inflammation, 148, 149
clinical performance of biomaterials, 331 Circulation, host-implant interactions, 19
common knowledge, folklore and facts, Clasis
246–249 clinical performance of biomaterials, 331
latent period, 256, 257 physical induction factors, 199
metals, 252–256 Classification of biomaterials, 8–10
systemic/remote site response Clinical performance of biomaterials,
mechanisms, 322 441–457
types of carcinogens, 249 autopsy retrieval studies, 455–456
Chemical environment, 21 common concerns about device retrieval
Chemical factors, adaptation, 198, 199 and analysis, 447–448
Chemical mediation, inflammation, see also cost, 447–448
Inflammatory mediators ownership of device, 448
capsule formation, 147 patient confidentiality, 449
development of local host response, 148 recovery system, 449
Chemical reaction elements of
corrosion, 50 data sets, 451–452
mechanical properties, elastic modulus, organizational, 452–455
94 historical perspective, 441–443
surface/interface interactions, 73 implants, 327–332
Chemotaxis epidemiology of biomaterials, 328–329
inflammation, 141 human physiology of biomaterials,
metals and, 157–158 329
Chick embryo studies, 355 total hip replacement, 330–332
Chloride, chromium reactions in presence of, national implant device retrieval and
52–54 analysis program (NIDRA)
476 Biological Performance of Materials: Fundamentals of Biocompatibility

elements of, 451–455 inflammation

proposed system, 450–451 chemotaxis inhibition, 158
procedures for device retrieval and cytotoxicity, 157
analysis, 443–447 metabolic disorders, 319, 320
goals, 444 mineral metabolism, 273, 274
methodology, 445–446 organometallic compounds, 76
reporting, 446–447 systemic distribution and excretion, 303
responsibility, 445 multicompartment distribution
Clinical testing, 383–400 models, 308
conclusions from, 392–394 one compartment distribution models,
decision for general clinical use, 394–398 304, 305
design of, 384–392 testing, in vivo, 369
clinical trials, 391–392 Cobalt alloys
preclinical tests, 387–391 ASTM F-562 leaching, 62–63
goal of, 383–384 carcinogenicity, 253, 255
Cloning, ethical and social concerns, 218–219 corrosion, 62–63, 68
Clubbing, capsule, 147 hypersensitivity reactions, implant-
Coagulation and hemolysis associated, 231–232
adaptation, vascular prosthesis, 186–187 implant material properties/standards,
blood contact tests, 347–349 127
coagulation cascade, 165–169 superalloys, 126–127
extrinsic, 166–167 Cocarcinogens, 249
implant-induced, 167–169 Coexisting conditions, and inflammatory
intrinsic, 165–166 response, 148
hemolysis, 176–179 Cold solution sterilization, 28, 29
flow velocity, experimental relation to, Collagen
177–179 inflammation, 145
general description, 176 tendon replacement, 189–190
paradigm shift, 12 vascular grafts, 174
surface/interface interactions, 81–82 in vitro tissue growth and replantation,
systemic/remote site response 213
mechanisms, 322 Complement factor five (C5a), 142
thromboresistant materials development, Complement system
170–176 antigen-antibody complex formation, 227
critical surface tension, 173, 174 coagulation, implant-induced, 168
mechanisms of thromboresistance, immune response, mechanisms of,
170–171 228–229
natural surfaces, 174–175 inflammation, 142
negative surface charge, 171–172 Complexation/complex ion formation
overview, 175–176 corrosion, 50, 67
Coagulation factors organometallic compounds, 75–76
coagulation cascade, 165–166 surface/interface interactions, 75, 80
extrinsic, 167 Complications, clinical testing, 393–394
intrinsic, 165, 166 Composite materials
inflammation, 139–140 implant material properties/standards,
thromboresistant materials development, 132–133
173 mechanical properties, 91
Coagulation factor XII (Hageman factor), 139, fracture strength, 100–101, 103
165, 166, 173 yield strength, 96–97
Coating, see Film formation; Passivation Composition of implant, and corrosion, 66
Cobalt Compound alloys, corrosion, 68
carcinogenicity, 253, 254, 267, 268 Conductivity, corrosion, 58
corrosion products, 297 Contact lenses, 42
hypersensitivity reactions, 231–232, 233, Continuous fiber carbon reinforced
234–235, 237, 239 polysulfone, 133
Index 477

Controlled delivery systems, antibiotic- systemic distribution and excretion,

impregnated implants, 151–153 297–301
Control (reference) materials, definitions, 5 body water distribution, 300–301
Control systems local effects, 297–300
adaptation, 184 one compartment distribution models,
homeostasis, 19 304, 305
Wolff's law, 197 wear
Coordination compounds, surface/interface anomalous, 121
interactions, 75 mechanisms of, 117
Copper Cortical bone
corrosion, 62, 65 adaptive remodeling, 196
metabolic disorders, 319 mechanical conditions, 22
mineral metabolism, 273 Covalent bonds
organometallic compounds, 76 mechanical properties
Copper-aluminum alloys, leaching, 62 elastic modulus, 93
Cornea, materials testing sites, 355 yield strength, 96
Corrosion and dissolution polymers, 90–91
adaptation, bone response to electrified surface/interface interactions, 82
implants, 196–197 Cracks
ceramics, dissolution of, 68–69 corrosion
chemistry of corrosion, 49–50 crevice, 60
classification of reactions, 50–51
intergranular, 61–62
corrosion rate, 56–57
stress and fatigue, 63–64
engineering variables affecting, 66
fracture strength, 98–104
electrochemical series, 54–56
swelling/leaching and, 46
ideal, 54–55
Cranial plates, corrosion, 65
practical, 55–56
Crazing, swelling/leaching and, 46
factors peculiar to biological
Creatinine, 304
environments, 67
Creep
forms of corrosion, 58–64
clinical performance of biomaterials, 330
crevice, 60–61
environment effects, 96–97
erosion, 63
galvanic, 59–60 and wear debris, 121
intergranular, 61–62 Crevice corrosion
leaching, 62–63 forms of corrosion, 60–61
pitting, 61 implants, 64
stress and fatigue, 63–64 Critical interfacial tension, hemolysis, 178
uniform attack, 59 Critical surface tension, thromboresistant
in implant applications, 64–66 materials development, 170, 173,
and inflammatory response 174
capsule formation, 147 Cross-links
infection, 155 and mechanical properties, 91
mechanical properties mechanical properties
elastic modulus, 93 elastic modulus, 93, 94–95
fracture strength, 101, 102–103 environment effects, 91
metal carcinogens, 252 Cross sensitivity, metals, 233
paradigm shift, 12 Crystal fields, organometallic compounds, 76
polymers, dissolution of, 70 Crystal field stabilization energy (CFSE), 75
potential-current relationship, 57–58 Crystalline ceramics, corrosion, 69
Pourbaix diagram, 51–54 Crystallinity
reactions of chromium in presence of and carcinogenicity, 347
chloride ion, 52–54 and polymer mechanical properties, 94
reactions of chromium in pure water, Crystals, intergranular corrosion, 61–62
51–52 Culture, microbial, 151
478 Biological Performance of Materials: Fundamentals of Biocompatibility

Cultured cells, vascular prosthesis seeding protein, host immune response, 81

with cultured endothelial cells, Density
187 ceramic implant material properties, 131
Cumulative toxicity index (CTI), 344 and infection hazards, 154–155
Curing, polymer material properties, 129 Dental amalgam, 320
Cyclic loading/deformation Depolymerization, mechanical properties
with corrosion elastic modulus, 92, 94
crevice, 61 environment effects, 91
stress and fatigue corrosion and, 64 Dermatitis, metal implant-associated,
fracture strength, 102–103 232–235
Cytokines Design and selection of implant materials,
small particle disease, 158, 159 427–439
systemic/remote site response design process, 429–438
mechanisms, 322 value of prospective design, 438–439
Cytotoxicity Desorption, organometallic compounds, 76
copper, 319 Device retrieval and analysis (DRA), see
inflammation, phagocytosis, 156–157 Clinical performance of
inflammatory response, 147 biomaterials
test methods, materials peformance Dialysis equipment, 318–319
in vitro, 341, 345 Diapedesis, inflammation, 141
in vivo, 355, 361–362 Dielectric constant, surface/interface
interactions, 83
Dietary metal intake, 285–288
chromium, 284
D iron, 276
Diffusion, see also Swelling/leaching,
Dead space, infection hazards, 154 materials
Debris, cell drug release devices, 46
fracture fixation devices, 191–192 Fick's laws of, 35–36
inflammation, 145 host-implant interactions, 19
Debris, implant, see Particulates; Wear debris leaching, 42
Deep infections, 150 Diffusion constants, concentration-
Definitions dependent, 42
biological performance, 5–6 Diffusion-controlled delivery, antibiotic-
carcinogenicity, ?? impregnated implants, 151–153
consensus, 6–7 Dilatant lubricants, 112, 113
deprecated terms, 469–470 Dipole interactions, surface/interface
glossary, 460–469 interactions, 82
terminology, 3–4 Discontinuous fiber reinforcing phases,
Deformations properties, 133
mechanical properties, yield strength, Disorder effects, and mechanical properties,
96–97 90
mechanical testing, 89–90 Disorder zones, elastic modulus, 94
recoverable, 89; see also Elastic modulus Dissociation, surface/interface interactions,
and wear debris, 121 73
Degradation Dissolution
biomaterials paradigms and paradigm ceramics, 68–69
shift, 12, 13 implant materials, systemic distribution
corrosion, in biological environment, 67 and excretion, see Systemic
Delayed hypersensitivity distribution and excretion
classes of hypersensitivity reactions, 230 metal debris, 121
test methods, in vitro, 345 polymers, 70
Delayed union, bone fracture healing, 185 Distribution models
Denaturation multi-compartment, 306–309
interfacial interactions, 74 one-compartment, 304–306
Index 479

Dominance, Pourbaix diagram, 54 Electrochemical stimulation, and

Dorn-Weertman equation, 97 inflammatory response, 147
Dosage Electrolyte solution, corrosion, 58
drug release dose response curve, 44 Electron beam irradiation, polymer material
in vitro systems limitations, 340 properties, 129
Drug release devices, 44–46 Electron transfer, corrosion rate, 56
Dry heat sterilization, 28, 29 Electrophoresis, 83
Ductile fracture, mechanical properties, 101 Electroplating, 55
Ductility Electrostatic charge, thromboresistant
absorption and, 41 materials development, 170,
polymer material properties, 129 171–172
Dynamic remodeling, wear versus, 121 Elimination reactions, 94
Dynamic stress, stress and fatigue corrosion Elution, leaching, 42
and, 64 Embryonic stem cells, 212
Encapsulation
adaptation, physical induction factors,
199
E biological performance, 8
clinical performance of biomaterials, 331
Edema (swelling), inflammation, 139, 140 local host response, resolution phase, 149
Effective local potentials, Pourbaix diagram, paradigm shift, 12
52, 53 Endoprostheses, corrosion, 65
Effects, of swelling/leaching, materials, 46 Endothelium/endothelial cells
Elasticity coagulation cascade, 165–166
absorption and, 41 vascular prosthesis seeding with cultured
swelling/leaching and, 46 cells, 187
yield strength, 96–97 Endotoxin, inflammation, 140
Elastic modulus, 89 Endurance limits, stress and fatigue
environment effects, 91–95 corrosion and, 64
fracture strength, 99 Enough: Staying Human in an Engineered Age
fundamentals, 90–91 (McKibben), 218–219
Elastohydrodynamic lubrication, 110, 111 Environments
Electrical factors, adaptation, 198, 199 biological, 17–29; see also Biological
Electrical mediation, inflammation environment
capsule formation, 147 classes of, 19
development of local host response, 148 Enzyme activity, surface/interface
Electrical potentials interactions, 80
corrosion, 50 Eosinophils, 24, 142
galvanic, 59–60 Epidemiology of biomaterials, 328–329
potential-current relationship, 57–58 Epoxides, carcinogens, 250
Pourbaix diagram, 52, 53 Epoxy composites, implant material
homeostasis, 19 properties, 133
and inflammatory response, capsule Equilibrium conditions
formation, 147 corrosion, potential-current relationship,
surface/interface interactions, 82–83 58
Electrified implants, bone response to, surface/interface interactions, 78–79
196–197 Equilibrium models, systemic distribution
Electrochemical charge, corrosion in and excretion, 309–311
biological environment, 67 Erosion
Electrochemical equilibria, surface/interface forms of corrosion, 63
interactions, 83 implants, 65
Electrochemical series, 54–56 Erythrocytes
corrosion, galvanic, 59–60 blood composition, 24
ideal, 54–55 coagulation cascade, 166
practical, 55–56 hemolysis, 176–179
480 Biological Performance of Materials: Fundamentals of Biocompatibility

Essential trace elements, see Trace elements paradigm shift, 12

Etching tendon prosthesis as scaffold for
erosion, 63 regenerating cells, 190
intergranular, 62 test methods, materials performance, 361
Ethical and social concerns Fibrous capsule, see Capsule formation
animal tests, 355, 356–358 Fick's laws of diffusion, 35–36
clinical trials, 386 Film, fluid film lubrication, 110, 111
in vitro tissue growth and replantation, Film formation
214–220 adaptation, vascular prosthesis, 187
Excretion, see also Systemic distribution and corrosion, 67, 68
excretion surface/interface interactions, 78–79,
iron, 279 81–82
mineral metabolism, chromium, 285 Finger joint prostheses, absorption problems,
paradigm shift, 12 40–41
Extracellular matrix Fixation, fracture, see Fracture fixation
cell-receptor paradigm, 207–208, 209 devices
in vitro systems limitations, 339 Flow conditions
Extrinsic coagulation cascade, 166–167 blood contact tests, 348
Extrusion, local host response, resolution erosion, 63
phase, 148, 149 hemolysis, 177–179
leaching, 42
organometallic compounds, 76–77
Fluid film lubrication, 111
F Fluids, body
corrosion, Pourbaix diagram, 53
Fabrication, polymer material properties, 129 corrosion product distribution, 300–301
Factors, coagulation, see Coagulation factors homeostasis, 19
Failure in tension, 87–88 implant life history, 23
Faradic stimulation, and inflammatory Fluoride ions, corrosion, 62
response, 147 Foreign body carcinogenesis, 257–262
Fatigue additional studies, 260–262
clinical performance of biomaterials, 330 early observations of, 257–258
forms of corrosion, 63–64 mechanisms, 258–260
metal corrosion, 297 systemic/remote site response
polymer material properties, 130 mechanisms, 323
swelling/leaching and, 46 Foreign body giant cells
Fatigue fracture, fracture strength, 101–102 capsule formation, 146
Fatigue wear, 118, 120 inflammation, 143–144
Feedback control Foreign body response
adaptation, 184 allergic, see Allergic foreign body
Wolff's law, 197 response
Fetal rodent models, 355 coagulation, 167
Fetal stem cells, 212 Fracture, device/materials
Fever, pyrogens and, 29 absorption and, 41
Fibrin clinical performance of biomaterials, 330
adaptation, vascular prosthesis, 187 paradigm shift, 12
coagulation cascade, 166, 167, 168 swelling/leaching and, 46
Fibroblasts, 143, 145 wear mechanisms, 118
Fibrocartilage, remodeling, 146 work of, 89
Fibronectin, 209 Fracture (bone) fixation devices
Fibrosis, see also Encapsulation adaptive remodeling, 191
clinical performance of biomaterials, 331 carcinogenicity, 263–264
coagulation cascade, 166 corrosion
corrosion and, 66 crevice, 61
inflammation, 145 distribution of products of, 298
Index 481

hypersensitivity reactions, 238–239 Geometric factors

systemic distribution and excretion and carcinogenesis, foreign-body, 258–259
corrosion products, 298 inflammatory response, 152–155
large particles, 292 in vitro systems limitations, 340
test methods, in vivo, 363 wear and, 114
Fracture (bone) healing, stages of, 184–185 Glassy materials, see Ceramics
Fracture strength, materials, 97–104 Glassy phases, corrosion, 68
environment effects, 101–104 Glossary, 460–469
fundamentals, 97–101 Glycocalyx, 151, 153
Free radical reactions, 94 Gold
Friction, wear, lubrication, 107–122 electrochemical series, 55, 56
absorption and, 41 hypersensitivity reactions, 232
friction, 107–109 Gott test, 367, 368
lubrication Grain boundaries, metals, 90
boundary, 111 Grain size, ceramic implant material
elastohydrodynamic, 110 properties, 131
hydrodynamic, 110 Granulation tissue
mixed, 111 adaptation, vascular prosthesis, 187
squeeze film, 111 inflammation, 145
types of behavior in response to shear, small particle disease, 158
111–114 Granules, intergranular corrosion, 61–62
wear, 114–121 Granulomas, resolution, 149
anomalous, 121 Graphite-reinforced polytetrafluoroethylene
debris size, 119–121 (Proplast™), 189–191
other mechanisms, 117–118 GRAS (generally regarded as safe) standard,
transfer film, 114–117 4
in vivo evidence of, 118–119 Gravity, and friction, 108
Functional side chains Griffith cracks, 98–100
and mechanical properties, 91 Growth, adaptation, physical induction
mechanical properties, elastic modulus, factors, 199
94 Growth factors
Functional tests, in vivo, 362–367 inflammation, see Inflammatory
mediators
systemic/remote site response
mechanisms, 322
Gut sutures, 101
G
Galvanic corrosion
forms of corrosion, 59–60
implants, 64–65 H
surface/interface interactions, charged
interfaces and ions, 83 Hageman factor (coagulation factor XII), 139,
Gamma radiation 165, 166
elastic modulus, 94–95 Handling
polymer material properties, 129 and corrosion, 66
sterilization with, 28 preimplantation, 28–29
Gas sterilization, 28, 29 Hard tissue, tendon replacement, 189
Gels, osmotic equilibrium, 42 Healing, physical induction factors, 199
Generally regarded as safe (GRAS) standard, Heart valves
4 absorption in, problems with, 38–40
Genetic engineering, 213 mechanical conditions, 22
Gene transfer, in vitro tissue growth and wear mechanisms, 118–119
replantation, 214 Heat, tissue, 139, 140
Genotypic differences, and implant life Heat sterilization, 28, 29
history, 22–23 Heat treatment
482 Biological Performance of Materials: Fundamentals of Biocompatibility

corrosion, intergranular, 62 Hydrolysis, mechanical properties, elastic

polymer material properties, 129 modulus, 94
Hemangioendothelioma, 263 Hydrophilic gels, osmotic equilibrium, 42
Hematoma, bone fracture healing, 184 Hydrophilic polymer corrosion, 69–70
Hemochromatosis, 319 Hydrophobic polymers corrosion, 69–70
Hemodialysis, 167 Hydrostatic pressure, homeostasis, 19
Hemoglobin Hydroxides, corrosion, 297
degradation products, 301 crevice, 61
hemolysis, 176–179 passivation, 52, 54
Hemolysis, 176–179 Pourbaix diagram, 53
chromium release assays, 349–350 Hydroxyapatite, 68
flow velocity, experimental relation to, Hydroxylation, corrosion, 50
177–179 Hypersensitivity
general description, 176 adaptation, physical induction factors,
Hemosiderin, 297, 319 199
Hip replacement, see Joint prostheses classes of reactions, 230
Histology, capsule, 147 clinical performance of biomaterials, 331
Homeostasis, 19 defined, 225
immune response, mechanisms of, 226
Host response
implant-associated, 230–240
and corrosion, 66
test methods, in vitro, 345
definitions and issues
Hypertrophy, adaptation, 184
biological performance, 8
biomaterials classification, 9
definitions, 5, 6
discipline of biomaterials engineering
and science, 11 I
paradigm shift, 12, 13
Immune response, host
implant, grading, 146
clinical performance of biomaterials, 331
inflammation, see Inflammation
inflammation, development of local host
sterilization and, 29
response, 148, 149
test methods, materials performance
iron and, 282
in vitro, 337, 342–344
mechanisms of, 226–229
in vivo, 380–381
antigen-antibody complex formation,
Human body
226–228
inorganic composition, 23 complement system activation,
physicochemical and mechanical 228–229
conditions, 22 specific versus nonspecific, 225, 226
Human tests, in vivo, 369–370, 380–381 surface/interface interactions, 81
Humidity, and elastic modulus, 92, 93 systemic/remote site response
Humoral immune response, 226 mechanisms, 321
Huxley, Aldous, 218 Immunity, material
Hyaline cartilage remodeling, 146 corrosion, 56, 68
Hydrated calcium phosphate, 68 electrochemical series, 54
Hydrated ceramics, corrosion, 69 Implant-related carcinogenesis, 263–268
Hydrated layers, thromboresistant materials Implants
development, 170 corrosion, 61, 64–66
Hydrodynamic lubrication, 110, 111 material properties
Hydrogen ceramics, 131–132
gaseous, reaction with water, 52 composites, 132–133
potential-current relationship, 57 metals, 126–129
Hydrogen bonding, surface/interface polymers, 129–131
interactions, 74 stainless steel, 126
Hydrogen ions, 53; see also pH mechanical properties
Hydrogen line, Pourbaix diagram, 52 elastic modulus, 92, 93
Index 483

fracture strength, 101 coagulation cascade, 165–166

pitting, 61 macrophage and, 143
Implant site inflammation, hypersensitivity small particle disease, 158–159
reactions, 235–239 systemic/remote site response
Impurities, and metal mechanical properties, mechanisms, 322
90 Ingrowth, see Tissue ingrowth
Inert, host response concepts, 9 Inherent thrombogenicity, definitions, 7
Infection Initiation, carcinogenesis, 249
animal tests, 365–366 Integration
and corrosion rates, 298 bone, adaptive remodeling, 194–195
and inflammatory response, 148 local host response, 149
inflammatory response Integrins, 208, 209
antibiotic-impregnated implants, Interactions of biomaterials and host
151–152 host response concepts, 9
geometric factors, 152–155 paradigm shift, 12, 13
types, 150–151 Interatomic bonds
iron and, 279–282 and mechanical properties, 90, 91
systemic/remote site response surface/interface interactions, 74
mechanisms, 323 Interdisciplinary science
xenografts and, 214
biomaterials as, 11
Inflammation, 139–160
regenerative medicine as, 205–206
adaptation, physical induction factors,
Interfaces
199
crevice corrosion, 60
antigen-antibody complex formation,
critical surface tension, thromboresistant
227–228
materials development, 173, 174
clinical performance of biomaterials, 331
fracture strength, 98–99
hypersensitivity reactions, metal implant-
Interfacial interactions, biological molecules
associated
and biomaterial surfaces, 73–84
dermatitis, 232–235
adhesion, results of, 80–86
implant site, 235–239
charged interfaces and ions, 82–83
implant degradation products and,
155–160 denaturation, 74
effects of phagocytosis, 158–160 mechanical aspects of interfaces, 77–80
effects on phagocytosis, 155–158 organometallic compounds, 74–77
infection Intergranular corrosion
antibiotic-impregnated implants, forms of corrosion, 61–62
151–152 implants, 65
geometric factors, 152–155 stress and fatigue corrosion and, 63–64
types, 150–151 Interindividual variability, implant life
mechanisms of inflammatory response, history, 23, 25
139–149 Interleukins
capsule formation, 146–148 inflammation, see Inflammatory
cellular invasion, 140, 141–145 mediators
initial events, 139–141 small particle disease, 159
remodeling, 145–146 Internal dipoles, surface/interface
resolution, 148–149 interactions, 82
paradigm shift, 12 International Standards Organization (ISO),
reactogenicity and, 8 126–129
signs and symptoms, 165 clinical testing, 388, 390–391
systemic/remote site response device retrieval and analysis, 446
mechanisms, 321 standards, 414–415, 416–417
test methods, in vivo, 379, 380 Interspecies variations
wear and, 114 biological performance, 8
Inflammatory mediators, 140 test methods, in vivo, 363
capsule formation, 147 Intramedullary rods, corrosion, 65
484 Biological Performance of Materials: Fundamentals of Biocompatibility

Intramolecular bonds, surface/interface dissolved species, 301–302

interactions, 74 equilibrium models, 309–311
Intrauterine device, 65 one compartment distribution models,
Intrinsic coagulation cascade, 165–166, 168 304, 305
Intrinsic parameters, stress-strain curve, 88, Irradiation
89 and elastic modulus, 94–95
In vitro studies, see Cell culture/tissue and polymer material properties, 129
culture/in vitro studies; Test sterilization with, 28, 29
methods, in vitro
Isotropic materials, diffusion in (Fick's laws),
In vitro tissue growth and replantation, 35–36
203–222
cell-receptor paradigm, 206–210
early ideas, 206–207
membrane receptor, 207–208
receptor-matrix interaction, 208–210
J
classes of biomaterials, 203–204
Joint prostheses
ethical and social concerns, 214–220
absorption problems, 40–41
matrices and cell sources, 210–214
carcinogenicity, 253, 264–266
tissue engineering, defined, 204–206
clinical performance of biomaterials,
In vivo evidence of wear, 118–119
330–332
In vivo testing, see Test methods, in vivo
hypersensitivity reactions, 235–237,
implant models
238–239
Ion exchange and complex formation
infection, antibiotic-impregnated
corrosion, 50
implants, 151–153
organometallic compounds, 76
local host response in humans, 380–381
Ionic bonds, polymers, 91
systemic/remote site response
Ionic conditions
mechanisms, 321
corrosion, potential-current relationship,
test methods, in vivo, 363
58
wear
drug release devices, 46
mechanisms of, 117–118
host-implant interactions, 19
size of debris particles, 119, 120
Ionization, corrosion, 49
Joints, lubrication, 113
Iontopheresis, drug release devices, 46
Iron
carcinogenicity, 253
and infection, 155
inflammation, 157 K
metabolic disorders, 319
Kallikrein, 140
mineral metabolism, 273, 274–283
Karyotyping, 211
absorption, 276–277
Kevlar, 133
dietary intake, 286
implants, role in TIBC saturation, Kidneys
282–283 iron excretion, 302–305
metabolic compartments, 275 metal ion excretion, 301–302
overload, 278–279 systemic distribution and excretion,
storage and excretion, 278 equilibrium models, 309–311
and susceptibility to infection, 279–282 Kinetics
transportation, 277–278 corrosion, 54
utilization, 278 surface/interface interactions, 77–78
systemic distribution and excretion Kininogen, 165, 166
corrosion product distribution, 297, Knife-edge attack, 62
298, 299, 300 Kusserow test, 367–368
Index 485

L homeostasis, 19
in humans, 380–381
Laminin, 207 inflammation, see Inflammation
Late infections, 151 metal corrosion products, 297–300
Latent period, chemical carcinogenesis, 256, physical induction factors, 198–199
257 Local potentials, Pourbaix diagram, 52, 53
Leaching, 43 Lymphatics
clinical performance of biomaterials, 331 debris particles in, 121
drug release, 44–46 host-implant interactions, 19
forms of corrosion, 62–63 inflammation, 145
implants, 65 particle accumulation in, 145
and inflammation, capsule formation, 147 Lymphocytes
mechanical properties blood composition, 24
elastic modulus, 92, 93 immune response, mechanisms of,
environment effects, 91 226–228
yield strength, 96 implant site, 146
planned, controlled drug release devices, inflammation, 143, 145
44–46 Lymphoma
systemic distribution and excretion of clinical performance of biomaterials, 331
monomers, 296–297 defined, 245
Lead, 319, 320 metal carcinogens, 254
Legislation, biomedical device standards, Lysis, erythrocyte (hemolysis), 176–179
415, 418–420
Leukemia, defined, 245
Leukocytes, see also specific cell types
blood composition, 24 M
coagulation cascade, 166
Level of response, definitions, 5 Macromolecular scale, surface/interface
Life expectancy, implant life history, 26, 27 interactions, 73
Ligament, mechanical conditions, 22 Macrophages
Ligament replacement, implant life history, and corrosion, 67
25–26 inflammation, 145, 155–157
Ligands small particle disease, 159
membrane receptors and, 208–209 surface denaturation of proteins and, 81
surface/interface interactions, 75 Magnetic fields, surface/interface
Lipid absorption interactions, 83
heart valve, 39–40 Malignant, defined, 245
joint prostheses, 40–41 Manganese
Liver inflammation, 157
iron excretion, 301–302 mineral metabolism, 273
particle accumulation in, 145 Manufacturing, and corrosion, 66
Load elongation curve, 88 Marine environment, pericellular
Load/loading environment and, 67
crevice corrosion, 61 Mass loss per unit time, corrosion, 69
fracture strength, 103 Mass transfer, absorption problems, 40
implant life history, 26 Material composition of implant, and capsule
stress and fatigue corrosion and, 64 formation, 147
tests, 87–88 Material properties
Local host response, see also Host response biological performance, 8
biological performance, 8 implant materials
biomaterials paradigms and paradigm ceramics, 131–132
shift, 12, 13 composites, 132–133
and corrosion, 66 metals, 126–129
in biological environment, 67 polymers, 129–131
Pourbaix diagram, 53 stainless steel, 126
486 Biological Performance of Materials: Fundamentals of Biocompatibility

sterilization and, 28–29 Medical science, biomaterials as, 11

Material response Membrane attack complex (MAC), 228, 229
definitions, 5 Membrane receptor, cell-receptor paradigm,
test methods, materials performance 207–208
in vitro, 337 Mercury, 320
in vivo cardiovascular functional tests, Metabolism
367–368 bacterial, 155
Materials, choice of, 4 metals, essential trace elements, 319
Materials science, biomaterials as, 11 Metallic bonds, and mechanical properties,
Matrix, biomaterial 90, 91
drug release devices, 45 Metallosis, 297
and mechanical properties, 91 Metals
mechanical properties biological performance, 8
elastic modulus, 92, 95 bone adaptive remodeling
yield strength, 97 fracture fixation devices, 191–192
paradigm shift, 13, 14 ingrowth into porous biomaterials,
in vitro tissue growth and replantation, 193–194
210–214 carcinogens, 252–256, 267–268
Matrix, tissue, in vitro tissue growth and clinical performance of biomaterials, 330,
replantation, 204, 213–214 331
McKibben, B., 218–219 corrosion, see also Corrosion and
Mechanical conditions, human, 22 dissolution
Mechanical deformation, stress and fatigue
in biological environment, 67
corrosion and, 64
chemistry of, 49–50
Mechanical effects
electrochemical series, ideal, 54, 55
interfacial interactions, 77–80
galvanic, 59–60
surface/interface interactions, 73
and infection, 155
Mechanical environment, 21
intergranular, 61–62
Mechanical factors
Pourbaix diagram, 51–54
adaptation, 198, 199
hypersensitivity reactions, implant-
and inflammatory response, capsule
associated, 231–240
formation, 147
Mechanical mediation, inflammation dermatitis, 232–235
capsule formation, 147 implant site inflammation, 235–239
development of local host response, 148 implant material properties/standards,
Mechanical performance, bone adaptive 126–129
remodeling, 193–194 inflammation
Mechanical properties, biomaterials, 87–104 capsule formation, 147
corrosion and phagocytosis, 155–157
crevice corrosion, 61 mechanical properties, elastic modulus,
stress and fatigue corrosion, 64 90, 91
elastic modulus, 90–95 metabolic disorders, 319–320
environment effects, 91–95 mineral metabolism
fundamentals, 90–91 chromium, 283–285
fracture strength, 97–104 dietary metal intake, 285–288
environment effects, 101–104 iron, 274–283
fundamentals, 97–101 organ function, effects of degradation
mechanics of materials, 87–90 products on, 319–320
swelling/leaching and, 46 surface/interface interactions, stability, 75
absorption and, 41 systemic distribution and excretion
osmotic equilibrium and, 42 corrosion products, 297–301
tendon replacement, 189–191 dissolved species, distribution and
yield strength, 96–97 excretion of, 301–304
Mechanical properties, fibrocartilage, 146 multicompartment distribution
Mechanical properties, tissue, 206 models, 306–309
Index 487

one compartment distribution models, inflammation, 157

304–306 mineral metabolism, 273
test methods, materials performance systemic distribution and excretion
in vitro, 340–341 corrosion product distribution, 298,
in vivo, 369, 380 299
thromboresistant materials development, corrosion products, 297
171–172 one compartment distribution models,
wear 304, 305
anomalous, 121 Monocytes
mechanisms of, 117 blood composition, 24
size of debris particles, 119 inflammation, 143
Metastasis, defined, 245 Monomers
Methyl methacrylate dissolved species, transport of, 296–297
clinical performance of biomaterials, 331 and mechanical properties, 93
hypersensitivity reactions, 237 Mononuclear phagocyte (MNP),
Microcracks/crazing, swelling/leaching inflammation, 142–143
and, 46 Morphology, capsule, 147
Microenvionment, corrosion, 53 Movement, see Friction, wear, lubrication
Micromotion, bone adaptive remodeling, 193 Moving charges, surface/interface
Microstructure, and mechanical properties, interactions, 83
91 Multicompartment metal distribution
Mineralization, bone, adaptive remodeling, models, 306–309
194 Multimolecular structures, surface/interface
Mineral metabolism, 273–288 interactions, 74
chromium, 283–285 Multinuclear foreign body giant cell (FBGC),
dietary metal intake, 285–288 143–144, 146
iron, 274–283 Multiphase alloys, corrosion, 62
absorption, 276–277 Multiphase mixtures, corrosion, 68
implants, role in TIBC saturation, Muscle, mechanical conditions, 22
282–283
overload, 278–279
storage and excretion, 278
and susceptibility to infection, 279–282 N
transportation, 277–278
utilization, 278 Nanoparticles, phagocytes and, 144
Mirra scale, 370 National Formulary, 405, 406
Mitral valve, 22 National implant device retrieval and
Mixed corrosion potential, bone response to analysis program (NIDRA)
electrified implants, 196–197 elements of, 451–455
Mixed lubrication, 111 proposed system, 450–451
Modulus, mechanical properties, Native biomaterials, 10
environment effects, 91 Natural materials
Moist heat sterilization, 28, 29 thromboresistant materials development,
Molecular level interactions, surface/ 170
interface, 73 in vitro tissue growth and replantation,
Molecular orbitals, organometallic 213, 214
compounds, 75–76 Natural surfaces, thromboresistant materials
Molecular shape, surface/interface development, 174–175
interactions, 81 Necrosis
Molecular weight adaptation, physical induction factors,
polymer material properties, 129 199
and yield strength, 96 complement system activation and, 229
Molybdenum small particle disease, 160
ASTM F-562 leaching, 62–63 Negative feedback control
carcinogenicity, 255 adaptation, 184
488 Biological Performance of Materials: Fundamentals of Biocompatibility

Wolff's law, 197 O

Negative surface charge, thromboresistant
materials development, 170, Octacalcium phosphate, corrosion, 68
171–172 One compartment distribution models,
Neointima 304–306
adaptation, vascular prosthesis, 187 Operating rooms, bacteria in, 150
test methods, materials performance, 362 Opsonization/opsonins, inflammation,
Neoplasm, defined, 245 141–142
Neoplastic response, see also Carcinogenesis Organic substrates, surface/interface
adaptation, physical induction factors, interactions, 80
199 Organometallic compounds
clinical performance of biomaterials, 331 carcinogens, 252
inflammation, development of local host corrosion, 67
response, 148, 149 interfacial interactions, 74–77
Neotendon, tendon prosthesis, 190–191 surface/interface interactions, 74–77
Neovascularization definitions, 74–75
inflammation, 145 production of, 76–77
small particle disease, 158 stability, 75–76
Net acceptability index, vascular prosthesis Organ transplantation
adaptation, 187–188 ethical and social concerns, 214–220
Neutrophils sources of organs, 211–212
blood composition, 24 Osmotic equilibrium, swelling/leaching, 42
inflammation, 145 Osmotic pressure
cellular invasion, 141–142 drug release devices, 46
implant degradation product effects, homeostasis, 19
157–158 Osseointegration, 195
Newtonian lubricants, 113 Osseous transducer, Wolff's law, 197
Nickel/nickel alloys Osteoblasts
ASTM F-562 leaching, 62–63 inflammatory mediators, 143
carcinogenicity, 253, 254, 267, 268 small particle disease, 159, 160
electrochemical series, 55, 56 Osteoclasts, 143, 159
hypersensitivity reactions, 232, 233, 234, Overload, iron, 278–279, 319
237, 239 Oxidation
inflammation corrosion, 49, 50
chemotaxis inhibition, 158 in biological environment, 67
cytotoxicity, 157 rate of, 56
mineral metabolism, 273, 274 and elastic modulus, 94
systemic distribution and excretion inflammatory cells, 144
corrosion products, 298, 299 polymer material properties, 130
multicompartment distribution Oxides
models, 306–308 crevice corrosion, 61
one compartment distribution models, passivation, 52
304, 305 Oxygen, reaction with water, 52
urinary excretion, 303 Oxygen line
Noble/cathodic potential Pourbaix diagram, 52
corrosion, galvanic, 59–60 Oxygen partial pressures
electrochemical series corrosion
ideal, 55 ceramics, 68
practical, 56 crevice, 60
Nonspecific carcinogenesis, 262–263 galvanic, 59–60
Nonspecific foreign body response, 225, 226 pitting, 61
Nonunions potential-current relationship, 58
bone fracture healing, 185 Pourbaix diagram, 53
clinical performance of biomaterials, 330 and corrosion, local host response, 66
Nutritional disorders, 319–320 homeostasis, 19
Index 489

and inflammatory response, 147 electrical currents, and inflammatory

physicochemical conditions, 22 response, 147
Pourbaix diagram, 52 electrochemical series, ideal, 54–55
homeostasis, 19
mechanical properties, fracture strength,
101
physicochemical conditions, 22
P
Pourbaix diagram, 51, 53
Pain, inflammation, 139, 140 surface/interface interactions
Paradigm shift, biocompatibility, 12–14 charged interfaces and ions, 83
Particulates denaturation, 74
clinical performance of biomaterials, 330 Phagocytes
corrosion and capsule formation, 146
ceramics, 69 and corrosion, in biological environment,
polymers, 69–70 67
inflammation debris particle transport to lymph nodes,
capsule formation, 146 121
phagocytosis, 155–157 inflammation, 142–145
Phagocytosis
phagocytosis, effects of, 158–160
clinical performance of biomaterials,
systemic/remote site response
331
mechanisms, 321
complement system activation and, 229
phagocytes and, 144
implant degradation product effects,
systemic distribution and excretion
155–160
large particles, 291–292
inflammation, 141–142
movement of solid bodies, 291–296
membrane receptors and, 208
phagocytic transport, 293–296
physical induction factors, 199
test methods, materials performance, 361
receptor-ligand complex and, 209
in vitro, 340–341
test methods, in vivo, 379, 380
in vivo, 362, 380, 381
Phase transformations, surface roughness,
wear
117
debris, size of, 119–121
Phenotypic differences, and implant life
mechanisms of, 118 history, 22–23
transfer film, 115 Phosphates, crevice corrosion, 61
Passivation, 54, 68 Phthalates, 318
in biological environment, 67 Physical induction factors, local host
chromium, 52, 53 response, 198–199
crevice, 60 Physicochemical conditions, human, 22
electrochemical series, 55, 56 Physiological environments, 19
uniform attack, 59 Physiology
Patency, vascular implants, 172 biological performance, 8
Peak load, implant life history, 26 clinical performance of biomaterials,
PEEK, implant material properties, 131 329
Pellethane™, 327 systemic distribution and excretion, see
Pericellular (circumcellular) environments Systemic distribution and
biological environment, 19, 21 excretion
and corrosion, 53, 67 Pins, corrosion, 65
Peritoneal cavity, materials testing sites, 355 Pitting corrosion
pH erosion corrosion versus, 63
corrosion, 50 forms of corrosion, 61
and corrosion, 66 implants, 64
corrosion Plasma membrane, cell-receptor paradigm,
in biological environment, 67 206–210
ceramics, 68 Plasma sterilization, 28, 29
galvanic, 59–60 Plastic deformation, yield strength, 96
490 Biological Performance of Materials: Fundamentals of Biocompatibility

Plasticizers Polygalactic acid sutures, fracture strength,

and mechanical properties, yield strength, 101
96, 97 Polylactic acid polymers
systemic effects, 318 implant material properties, 131
Plasticizing, mechanical properties tendon prosthesis, 190
elastic modulus, 92, 93 Polymerization, and yield strength, 97
environment effects, 91 Polymers
Plastics, see Polymers bone adaptive remodeling, ingrowth into
Plate formation, bone, 196 porous biomaterials, 193–194
Platelets carcinogens, 250
blood composition, 24 dissolution of, 70
coagulation hypersensitivity reactions, implant-
coagulation cascade, 166 associated, 230–231
implant-induced, 167 implant material properties/standards,
thromboresistant materials development, 129–131
172, 173 inflammatory response, capsule
Plates formation, 147
adaptive remodeling, 192 mechanical properties
corrosion elastic modulus, 90–95
crevice, 61 fracture strength, 99–100, 101, 103
galvanic, 64–65 organ function, effects of degradation
Platinum products on, 318–319
electrochemical series, 55, 56 organometallic compounds, 76–77
hypersensitivity reactions, 232 sterilization and, 28–29
PMMA (polymethylmethacrylate) surface/interface interactions, host
bone cement immune response, 81
hypersensitivity reactions, implant- systemic distribution and excretion of
associated, 231 monomers, 296–297
inflammation, 158–159
test methods, materials performance
systemic distribution and excretion,
in vitro, 342
296–297
in vivo, 380, 381
systemic effects, 318
wear
systemic/remote site response
size of debris particles, 120, 121
mechanisms, 321
transfer film, 114–115
cements, antibiotic-impregnated
Polymorphonuclear leukocytes
implants, 151–153
blood composition, 24
clinical performance of biomaterials, 330
elastic modulus, 92–95 inflammation, 141–142
implant material properties, 130, 131 Polysulfone polymers, implant material
composites, 133 properties, 131
wear mechanisms, 118 Polyurethane biomaterials, 327
Poisson ratios, 95 Poppets, heart valve
Polarization, anodic absorption problems, 38–40
electrochemical series, 55 wear mechanisms, 118–119
passivation, 54 Porosity, biomaterials
Polarization resistance, biological adaptation
performance, 8 bone, 193–194, 196
Polishing, transfer film, 117 vascular prosthesis, 187–188
Polycrystalline ceramics, corrosion, 69 and carcinogenesis, foreign-body, 258
Polyethylene, see also UHMWPE (ultrahigh and infection hazards, 154–155
molecular weight polyethylene) Porous materials
implant material properties, 131 ceramics, 69
infection hazards, 154–155 mechanical properties, 95, 99
organometallic compounds, 76–77 Porphyrins, 301
Polyethylene disease, 158–159 Positioning of implant, and corrosion, 66
Index 491

Potential-current relationship, corrosion, Reduction, corrosion rate, 56

57–58 Reference (control) materials
Pourbaix diagram, 51–54 blood contact tests, 348
galvanic, 60 definitions, 5
reactions of chromium in presence of preclinical tests, 387
chloride ion, 52–54 Regeneration, inflammation, 146
reactions of chromium in pure water, Regenerative medicine, types of tissue
51–52 engineering, 205–206
uniform attack, 59 Regional lymph nodes, particle accumulation
Precipitation, corrosion, 61, 299 in, 145
Preimplantation handling, 28–29 Regulation of biomaterials, see also Standards,
Prekallikrein, 165, 166 standardization, and regulation
Pressure, lubrication, 112 of implant materials
Pretreatment, vascular prosthesis seeding clinical trials, 386–387
with cultured endothelial cells, Reinforcing materials
187 implant material properties, 132
Primary neoplasm, defined, 245 and mechanical properties, 91
Procarcinogens, 250, 251, 252 mechanical properties, elastic modulus,
Promotion, carcinogenesis, 249 95
Properties, paradigm shift, 13 tendon replacement, 189
Proplast™ Rejection, implant, 226
bone tissue ingrowth into pores, 194 Remodeling
tendon replacement, 189–191 adaptation, 199
Prostaglandins, small particle disease, 159 bone fracture healing, 184
Proteins coagulation cascade, 166
adaptation, vascular prosthesis, 187 inflammation, 145–146
blood composition, 24 systemic/remote site response
coagulation, 165, 166, 167 mechanisms, 322
iron-binding, 231, 275, 276, 277, 279, 281, Remote host response
283, 301 organ function, effects of degradation
metal complexation, 231–232 products on, 317–325
organometallic compounds, 76, 77 ceramics, 320–321
surface/interface interactions, 74, 81–82 metals, 319–320
thromboresistant materials development, polymers, 318–319
173 systems involved in host response,
Pseudoarthrosis, bone fracture healing, 185 321–323
Pseudointima, thromboresistant materials paradigm shift, 12, 13
development, 174 particle accumulation in organs, 145
Pus, inflammation, 145 systemic distribution and excretion, see
Pyrogens, sterilization and, 29 Systemic distribution and
excretion
Renal embolus test, 367–368
Renal excretion
Q iron, 302–305
systemic distribution and excretion,
Quaternary structure, denaturation, 74 equilibrium models, 309–311
Renewal of tissue, adaptation, 184
Repetitive motion/strain, implant life
history, 26
R Replantation, see also In vitro tissue growth
and replantation
Reactogenicity, 8; see also Inflammation cell sources, 210–212
Receptor-ligand binding model, 207–208, 209 host response concepts, 9
Receptors, cell-receptor paradigm, 206–210 terminology, 203
Redness (erythema), inflammation, 139, 140 types of tissue engineering, 205
492 Biological Performance of Materials: Fundamentals of Biocompatibility

Reporting, DRA studies, 446–447 Secondary structure, surface/interface

Reservoir drug release devices, 45 interactions, 74
Resins Self-polarization, electrochemical series, 55
fracture strength, 103 Serum
implant material properties, 129, 130, 131 lubrication, 109, 110, 116, 117
Resolution, inflammation, 148–149 surface/interface interactions, host
Resorbable materials, in vitro tissue growth immune response, 81
and replantation, 214 Shape changes, see Geometric factors
Resorption Shear
adaptation, physical induction factors, and hemolysis, 176, 177–179
199 types of behavior in response to, 111–114
local host response, resolution phase, 149 Shear rate, lubrication, 111, 113
vascular graft matrix, 204 Shock, anaphylactic, 230
Retrieval of devices, see Clinical performance Side chains, and mechanical properties, 91
of biomaterials Signal, paradigm shift, 13, 14
Revascularization, bone adaptive Silicone
remodeling, 193 carcinogenicity, 266
RGD (arginine-glycine-aspartic acid), hypersensitivity reactions, 231
208–209, 213 systemic effects, 319
Rigidity, bone adaptive remodeling, 192, 193 test methods, in vitro, 342
Rods Silver, corrosion products, 297
corrosion, 65 Skin, remodeling, 146
fracture fixation devices, adaptive Skin bacteria, 150
remodeling, 191–192 Skin reactions, metal hypersensitivity,
Roughness, surface, see Surface properties 232–235
Rupture strength, tendon replacement, Small particle disease, 158–159
189–191 Soft tissue
Rwanda test, 219–220 bone, adaptive remodeling, 194
capsule formation, 146
tendon replacement, 189
test methods, materials performance, 355
S wear, anomalous, 121
Spatial arrangements, surface/interface
Safety standards, 4 interactions, 74
Saline solutions, elastic modulus, 93 Species specificity, inflammatory response,
Sarcoma 148
defined, 246 Specific foreign body response, 225, 226
fracture fixation devices and, 263 Spherical pores, 99
metal carcinogens, 254 Spleen, particle accumulation in, 145
Scaffold function Spring clips, corrosion, 65–66
tendon prosthesis, 190 Spring constant, 89; see also Elastic modulus
in vitro tissue growth and replantation, Squeeze film lubrication, 111
204 Stainless steel
Scarring carcinogenicity, 253
capsule formation, 146 corrosion, 65–66, 67, 68
corrosion and, 66 electrochemical series, 55, 56
inflammation, 145 hypersensitivity reactions, 239
Scission reactions, 93, 94 implant material properties/standards,
Scoring, and corrosion, 64 126, 127
Screws and iron load
clinical performance of biomaterials, 330 corrosion products, 282
corrosion debris, 280–281
crevice, 61 systemic distribution and excretion,
galvanic, 64–65 309–311
pitting, 64 Standard man/human, 20
Index 493

Standards, 358–359 Strength, material

mechanical properties, 125–129 mechanical properties, 91
in vivo testing, 374–378 mechanical testing, 89–90
Standards, standardization, and regulation of Strength, tissue, adaptive remodeling of
implant materials, 403–423 bone, 192, 193
biomaterials standardization activities, Stress, wear mechanisms, 118
406–415, 416–417 Stress corrosion
American Dental Association, 406–407 forms of corrosion, 63–64
ASTM, 407–414 fracture strength, 101
ISO, 414–415, 416–417 implants, 65
biomaterials supply crisis, 422–423 Stress-enhanced attack, 101
clinical trials, 386–387 Stresses, mechanical conditions, 22
clinical use, approval for, 394–395 Stress shielding, bone implants, 191
drug standardization activities, 404–406 Stress-strain curve, 88–90
historical perspective, 403–404 Stress-versus-number of cycles (S-N) curve,
preclinical tests, 388 102–103
regulation of materials for implants, Strontium, 320
419–422 Structure
U.S. federal regulation of medical devices paradigm shift, 13
and biomaterials, 415, 418–419 surface/interface interactions, 74, 82
Standard test Subchondral condensation, physical
biological performance, 7–8 induction factors, 199
definitions, 5 Subcutaneous air pouches, materials testing
Staphylococcus aureus, 150, 151, 153, 154 sites, 355
Staphylococcus epidermidis, 150, 151 Submicroscopic particles
Starr-Edwards poppet-type heart valve, phagocytes and, 144
38–40 wear, 120
Static (nonflow) conditions, corrosion, 61 Substrates, surface/interface interactions, 80
Static fatigue Sulfur-containing molecules, and corrosion,
fracture strength, 103 65
swelling/leaching and, 46 Superalloys, 126–127
Statutory regulations, biomedical device corrosion, 68
standards, 415, 418–420 implant material properties/standards,
Steady loads, fracture strength, 103 127
Steam sterilization, 28, 29 Surface active materials, 195
Steel alloys, see also Stainless steel Surface energy, organometallic compounds,
corrosion, 65–66, 298 77
mechanical properties, fracture strength, Surface films
102 and corrosion, 68
Stem cells wear, transfer film, 114–115
adaptation, 184 Surface fracture, wear mechanisms, 118
in vitro tissue growth and replantation, Surface interactions, biological molecules
212 and biomaterial surfaces, 73–84
Sterilization, preimplantation handling, adhesion, results of, 80–86
28–29 charged interfaces and ions, 83–84
Stiffness, mechanical testing, 89–90 denaturation, 74
Storage, biomaterials mechanical aspects of interfaces, 77–80
and mechanical properties, 92 organometallic compounds, 74–77
paradigm shift, 12 Surface potentials, surface/interface
polymer material properties, 130 interactions, 83
Storage and excretion, mineral metabolism Surface properties
chromium, 285 bone adaptive remodeling, adhesion, 195
iron, 278 and carcinogenesis, foreign-body, 258
Strain rate, polymer material properties, corrosion and, polymers, 69
129 fracture strength, 98–99
494 Biological Performance of Materials: Fundamentals of Biocompatibility

lubrication, transfer film, 117 ceramics, 320–321

organometallic compounds, formation of, metals, 319–320
76–77 polymers, 318–319
passivation, see Passivation systems involved in host response,
preimplantation handling, 28 321–323
thromboresistant materials development, solid bodies, movement of, 291–296
170 large particles, 291–292
critical surface tension, 173, 174 phagocytic transport, 293–296
natural surfaces, 174–175 Systemic host response
negative surface charge, 171–172 homeostasis, 19
Surface recession, uniform attack, 59 paradigm shift, 12, 13
Surface separation, lubrication, 111
Surface tension
fracture strength, 99
surface/interface interactions, 79 T
thromboresistant materials development,
170, 173, 174 Tanning, vascular grafts, 174
Surfactants, lubrication, 109 Tantalum implants
Surgical materials corrosion products, 297
corrosion, 65–66 material properties, 129
fracture strength, 101 Teflonoma, 321
Sustained reservoir drug release curve, 45 Temperature
Sutures, fracture strength, 101 homeostasis, 19
Swelling, materials, 35–46 mechanical properties
Swelling, tissue (edema), inflammation, 139, elastic modulus, 91, 93
140 fracture strength, 99–100
Swelling/leaching, materials yield strength, 96–97
absorption, 36–38 physicochemical conditions, 22
undesirable, examples of, 38–41 polymer material properties, 129
effects of, 46 Temperature environment, 21
and elastic modulus, 92 Tendon, 22
Fick's laws of diffusion, 35–36 Tendon prostheses
leaching, 43 adaptation, 189–191
planned, controlled drug release systemic distribution and excretion of
devices, 44–46 large particles, 292
osmotic equilibrium, 42 Tensile strength
Synergistic interactions, inflammation, 148 fracture strength, 97–104
Synovial fluid, lubrication, 109, 110 suture materials, 101, 102
Synovial tissue, local host response in Tension, failure in, 87–88
humans, 380–381 Terminology
Systemic distribution and excretion, 291–312 carcinogenicity, ??
debris particles, 121 consensus definitions, 6–7
dissolved species, distribution and definitions and issues, 3–4
excretion of, 301–311 deprecated terms, 469–470
distribution models, multi- glossary, 460–469
compartment, 306–309 Test methods, clinical trials, see Clinical
distribution models, one- testing
compartment, 304–306 Test methods, in vitro, 337–351
equilibrium models, 309–311 blood contact tests, 347–350
metal ions, 301–304 dynamic, 349–350
dissolved species, transport of, 296–301 static, 348–349
metal corrosion, 297–301 test strategies, 337–338
monomer leaching, 296–297 tissue culture tests
remote organ effects of degradation carcinogenicity, 345–347
products, 317–325 cell types and observation, 340–341
Index 495

delayed hypersensitivity tests, 345 Thrombus formation, coagulation cascade,

examples of use in materials testing, 165
341–342 TIBC (total iron binding capacity), see Total
generic methods, 338 iron binding capacity (TIBC)
host response screening, 342–344 Time, polymer material properties, 129
problems with, 339–340 Tissue analysis, implant life history, 23
types of, 338 Tissue composition, homeostasis, 19
Test methods, in vivo implant models, Tissue engineering, see also In vitro tissue
355–373, 374–381 growth and replantation
animal welfare, 356–358 defined, 204–206
approaches to, 355–356 Tissue growth
ASTM standard procedures, 374–378 adaptation, 183–185
evaluation of human local host response in vitro, see In vitro tissue growth and
in synovial and capsular tissue, replantation
380–381 Tissue ingrowth, 12
rating system for animal implant sites, 379 bone, porous biomaterials and, 193–194
test types, 358–370 paradigm shift, 12
cardiovascular functional tests, physical induction factors, 199
367–369 test methods, in vivo, 361–362, 364, 369
functional, 362–367 in vitro tissue growth and replantation,
human tests, 369–370 204
nonfunctional, 358–362 Tissue necrosis
Tissue necrosis, small particle disease, 160
Tests
Tissue types, biological environment, 21
biological performance, 7–8
Titania, 68, 320
definitions, 5
Titanium/titanium alloys
mechanical, 87–90
carcinogenicity, 253, 255
Thermoplastic resins
corrosion, 68
implant material properties, 129, 130, 131
electrochemical series, 55, 56
mechanical properties, fracture strength,
fracture fixation devices, adaptive
103
remodeling, 191–192
Thermosets, implant material properties, 129,
hypersensitivity reactions, 239
130
implant material properties/standards,
Thioneins, 231 128
Thixotropic lubricants, 112, 113 material properties, 129
Three-body wear, 116 mineral metabolism, 273, 274
Three-dimensional structures, paradigm systemic distribution and excretion, 303
shift, 13 test methods, in vivo, 369
Thrombogenicity To Live Forever (Vance), 217–218
blood contact tests, 347–349 Total hip replacement, see Joint prostheses
coagulation, see Coagulation and Total iron binding capacity (TIBC)
hemolysis implants and, 282–283
definitions, 7 iron overload and, 280
hemolysis and, 176 systemic distribution and excretion,
surface/interface interactions, 83 equilibrium models, 309–311
systemic/remote site response transport, 277–278
mechanisms, 322 Toxic complex syndrome, classes of
Thromboplastin, coagulation cascade, 167 hypersensitivity reactions, 230
Thromboresistant materials development Toxicity, chromium, 285
critical surface tension, 173, 174 Toxins, inflammation, 140
mechanisms of thromboresistance, Trace elements
170–171 blood composition, 24
natural surfaces, 174–175 human body, composition of, 23
negative surface charge, 171–172 inflammation, 156
overview, 175–176 metabolic disorders, 319
496 Biological Performance of Materials: Fundamentals of Biocompatibility

Trade associations, standards developments, van't Hoff's law, 42

415 Vascular grafts/prostheses
Transducer, biological, 197 adaptation, 186–189
Transfer film wear, 114–117 thromboresistant materials development,
Transferrin, 231, 275, 281, 283, 299 171–172, 174
Transitional tests, cardiovascular functional in vitro tissue growth and replantation,
tests, 368 204
Transplantation, sources of materials, Vascularization
210–212 bone, adaptive remodeling, 193
Transport, paradigm shift, 12 inflammation, 145
Transportation, iron, 277–278 small particle disease, 158
Tricalcium phosphate, 68, 320 Vascular response
Tumor, defined, 246 coagulation
Tumorigenic, defined, 245 coagulation cascade, 165, 166
Tumor necrosis factor-alpha, 159 implant-induced, 168
Tungsten implants endothelial, see Endothelium/endothelial
inflammation, 157 cells
material properties, 129 inflammation, 139–140
Turbulent flow test methods, materials performance, 362
blood contact tests, 348 Velocity, lubrication, 112, 113
hemolysis, 176 Vena cava test, canine, 367, 368
Viable, host response concepts, 9
Viscoelastic materials, material properties,
129
U Viscosity
absorption and, 41
UHMWPE (ultrahigh molecular weight blood, 368
polyethylene) lubrication, 111, 112
clinical performance of biomaterials, 330 Volume
elastic modulus, 92–95 osmotic equilibrium, 42
implant material properties, 131 swelling/leaching and, 46
lubrication, 116 Vroman effect, 81, 82
wear, size of debris particles, 120
wear mechanisms, 117–118
Ultimate stress, 89, 90, 97–104
Ultrastructural properties, capsule, 147 W
Uniform attack, 59
Union, bone fracture healing, 185 Water
United States Pharmacopeia, 404–405, 406 chromium reactions in, 51–52
Urinary excretion host-implant interactions, 19
iron, 302–305 lubrication, transfer film, 116
metal ions, 301–302 mechanical properties
systemic distribution and excretion, elastic modulus, 93
equilibrium models, 309–311 fracture strength, 103
metal corrosion products, systemic
distribution and excretion,
300–301
V vascular prosthesis, porosity, 187–188
Weak-associative bonding, surface/interface
Valence, corrosion rate, 56, 57 interactions, 74
Vanadium, 273 Wear, 114–121
Vance, Jack, 217–218 absorption and, 41
Van der Waals bonds anomalous, 121
cell adhesion, 194 clinical performance of biomaterials, 330,
surface/interface interactions, 74 331
Index 497

corrosion Xenobiopsy cells, in vitro tissue growth and

in biological environment, 67 replantation, 214
product distribution, 299 Xenografts
fracture fixation devices, 191 cell sources, 212
other mechanisms, 117–118 vascular, 174
paradigm shift, 12 X-irradiation, and elastic modulus, 94–95
transfer film, 114–117
in vivo evidence of, 118–119
Wear debris
clinical performance of biomaterials, 330,
331 Y
corrosion product distribution, 299
hypersensitivity reactions, 237 Yield stress/strain, 89, 90
iron, 280–281 environment effects, 96–97
size of particles, 119–121 mechanical properties
test methods, materials performance, 361 environment effects, 91
Welding, intergranular corrosion, 62 fracture strength, 99–100
Wetting, surface/interface interactions, 79 Young-Dupree equation, 78–79, 173
Whisker fiber reinforcing materials, 133 Young's modulus, 89
Wilson's disease, 319
Wires, corrosion, 65
Wolff's law, 198
bone response to electrified implants, 196
feedback control system, 197
Z
fracture healing, 185, 186 Zinc, 62, 300
remodeling, 194
Zirconia, 320
Work of fracture, 89
corrosion, 68
Woven materials, vascular prosthesis,
187–188 implant material properties, 131
Zisman plot, 79, 80
Zones of disorder, metals, 90
Zr-2.5, material properties, 129
Zwitterions, surface/interface interactions,
X 82
Xeno-, defined, 210
3959 Cover 11/22/05 4:36 PM Page 1
C M Y CM MY CY CMY K

Biomedical Engineering

of MATERIALS Fundamentals of Biocompatibility

BIOLOGICAL PERFORMANCE
FOURTH EDITION
FOURTH EDITION
BIOLOGICAL PERFORMANCE
of MATERIALS
Fundamentals of Biocompatibility
Bioengineers need a thorough grounding in biocompatibility — the
BIOLOGICAL
biological performance of materials. Until now, there were no publications
suitable for a neophyte in the field; other publications are either not
comprehensive or focused on rather narrow interests. Drawing on the
author’s 35 years of experience as a teacher, researcher, and consultant
PERFORMANCE
in biomaterials science and engineering (BSE), Biological Performance
of Materials: Fundamentals of Biocompatibility, Fourth Edition
focuses primarily on principles of biological performance at a relatively
fundamental level, analyzing interactions between living organisms and
of
nonliving materials used in medical devices — the subject that sets BSE
apart as a distinct field of investigation.

Following an introductory section, this fully revised and updated fourth

edition consists of three sections: the material response to biological
MATERIALS
systems, host response to biomaterials, and test methods for determining
biological response in vitro as well as in animal models and clinical
Fundamentals of
settings.

Supplemental “Interparts” summarize the properties of implant materials

Biocompatibility
and examine the clinical performance of implanted biomaterials.

FEATURES
• Focuses on principles of biological performance at a relatively
fundamental level
• Examines the concept of biocompatibility and the arguments
for the broader concept of biological performance
• Offers references, test methods, and approaches for establishing
the biological performances of materials
• Tabulates generic materials properties Black
• Explores approaches for detecting clinical issues associated
with biomaterials in animal models and in patients
• Covers the design, qualification, standardization, and regulation FOURTH
of implant materials
3959
EDITION
Jonathan Black
6000 Broken Sound Parkway, NW
Suite 300, Boca Raton, FL 33487
270 Madison Avenue
New York, NY 10016
2 Park Square, Milton Park
www.taylorandfrancisgroup.com Abingdon, Oxon OX14 4RN, UK