Massive Brain Hemorrhage: A Review of 144 Cases

and an Examination of Their Causes

BY WILLIAM F. McCORMICK, M.D.,»
AND DAVID B. ROSENFIELD, M.D.f

Abstract: • A detailed clinicopathological study of the causes and locations of massive non-
Massive traumatic brain hemorrhage in 144 patients is reported. A cause of the hemorrhage,
Brain such as an aneurysm, angioma, arteritis, neoplasm or a blood dyscrasia (leukemia,
Hemorrhage:
A Review of hemophilia), was proved in two-thirds (95) of these patients. Twelve normotensive
144 Cases patients had no cause found to explain their hemorrhage. Systemic hypertension, gen-
and an erally mild, defined as a pre-ictal pressure of > 140/90 or by excessive heart weight,
Examination was present in 58 of the 144 patients. Twenty-one of these 58 hypertensive patients
of Their
Causes had a clear discernible cause for their brain hemorrhage (i.e., leukemia, metastatic
carcinoma, angioma, aneurysm), whereas no satisfactory morphological cause could
be found in only 37. Thus, in only about one-fourth of our patients could any serious
claim be made that hypertension was the cause of the hemorrhage. Our data would
indicate the need for a critical study of the causes of intracranial hemorrhages, and
re-evaluation of the true relationship of systemic hypertension to such strokes. The
widespread dogma that hypertension is the outstanding cause of nontraumatic brain
hemorrhage no longer seems warranted.

Additional Key Words saccular aneurysms neoplasms angiomas

Charcot-Bouchard aneurysm leukemia arteritis
venous thrombosis
Downloaded from http://ahajournals.org by on February 20, 2023

• "When an erroneous hypothesis becomes en- increases both the severity and the rapidity of
trenched and generally accepted, it is transformed development of complicated atherosclerosis in the
into a kind of tenet that no one is allowed to brain. The alleged relation between hypertension and
question and investigate; and it then becomes an evil massive brain hemorrhage is less clearly established,
which endures for centuries." with the majority of reports so poorly "controlled"
Goethe and documented as to be of little scientific validity.
Such statements as "hypertension is the most com-
Introduction mon cause, from 10—20 times more frequent than
Historically, hypertension and nontraumatic brain all other causes [of cerebral hemorrhage]," can be
hemorrhage have been so closely correlated that the found in a current textbook of pathology1 or
assumption that hypertension causes most massive ".. . overwhelming frequent cause of single, sizeable
brain hemorrhage has become virtual dogma. This intracerebral hemorrhage is hypertension,"2 would
reputed association has retarded the critical evalua- convey the notion that hypertension is proved to be
tion of the pathogenesis of the lesion. Numerous the major cause of massive brain hemorrhage and
well-controlled studies have revealed a clear associa- that its frequency is greater than that of all other
tion between hypertension and cerebral (and causes combined. We do not believe that these
systemic) occlusive vascular disease, and there conclusions are warranted, based on the data
appears to be little doubt that systemic hypertension currently available.
In denning "hypertensive" brain hemorrhage, it
is necessary to use both positive and "negative"
•Professor of Pathology and Neurology, and Chief, Division
of Neuropathology, University of Texas Medical Branch, data. Documentation of pre-ictal blood pressure
Galveston, Texas 77550. Formerly, Chief, Division of recordings of a level generally accepted as hyper-
Neuropathology, University of Iowa, Iowa City, Iowa. tensive, evidence of left ventricular thickness,
tChief Resident in Neurology, Duke University, increased heart weight (in the absence of cardiac
Durham, North Carolina. Formerly, Junior Resident in valvular lesions) and other anatomical concomitants
Neurology, University of Iowa, Iowa City, Iowa.
Reprint requests to Dr. McCormick, Galveston, of hypertension should be present. Moreover, other
Texas. independent causes of brain hemorrhage must be
946 Stroke, Vol. 4, November-December 1973
 MASSIVE BRAIN HEMORRHAGE

considered. Hypertension is so frequent in Western TABLE 2

civilization adults that any group of adult patients "Confro/" Autopsy Population — 250 Patients*
would be expected to contain a large number with Age by No. with clinical No. with enlarged Total no. of
hypertension. Recently, Langfeld3 has reported that decade hypertension hearts (400 am) patients
4 1 % of patients over age 35 were hypertensive if a 20-29 0 1 19
pressure of 140/90 was used. Kurtzke 4 has noted 30-39 3 4 9
that at least 20% of patients with cerebrovascular 40-49 4 5 28
disease would be expected to have prior cardio- 50-59 14 33 66
vascular disease even if the two entities were totally 60-69 10 42 72
unrelated. Our data suggest that even these estimates 70-79 6 17 40
may be too low. 80-89 0 10 16
Total 37 + 112 250
The mere documentation of pre-ictal hyper-
tension is not, in our opinion, sufficient to establish *Patients: 134 male and 116 female.
hypertension as the cause of a brain hemorrhage. tThere were nine patients with elevated blood pressures
Arteriolar necrosis and fibrin-platelet thrombosis of but with normal weight ( < 400 gm) hearts.
arterioles occur virtually constantly in the immediate
region of a hemorrhage. Small arterial and arteriolar
were 144 patients with massive brain hemorrhage
disease due to hypertension may be demonstrated,
coining to autopsy during the study period.
however, throughout the nervous system. Even when At autopsy, the brains and spinal cords of all
both proved hypertension (and/or increase in patients were removed and fixed in 10% buffered
cardiac size) and the characteristic small arterial formalin for a period of 8 to 14 days. Following
and arteriolar lesions are present, they do not prove fixation, detailed dissection was carried out by the same
that hypertension is the cause of a given case of neuropathologist. Photographs were made of the uncut
massive brain hemorrhage. and cut brain during varying stages of dissection. An
It is our intention to document the many causes average of nine color photographs of the gross brain
of massive cerebral hemorrhage encountered in the were available for review on each case. Detailed
autopsy population of this medical center. examination of the vessels of the circle of Willis were
made in situ and following the removal from the brain
in the manner similar to that previously described.5 The
Methods majority of the brains were sectioned in the coronal
Downloaded from http://ahajournals.org by on February 20, 2023

In August, 1964, a prospective study of massive brain plane at 1-cm intervals; the others were sectioned in
hemorrhage was begun at the University of Iowa either the horizontal or sagittal planes. The hematoma
Hospitals. Detailed review was made of the clinical and its walls were examined meticulously with a
records of all patients with brain hemorrhage coming to minimum of eight blocks taken from the central
autopsy. When possible, telephone or written communi- hematoma and its adjacent tissue. These blocks were
cation with the referring family physician was made in
order to utilize any clinical information obtained prior
to the patients' hospitalization at this institution. There TABLE 3

Known Causes of Massive Nontraumatic Brain Hemorrhage

— University of Iowa (144 Total Patients)
TABLE 1 Leukemia 21 (4)
Aneurysm 20 (4)
Massive Brain Hemorrhages Angioma 15* (4)
No. of patients with Neoplasm 13 (3)
Age massive hemorrhage
(year) {"hypertensive" only) Total autopsy population Cortical vein/dural sinus thrombosis 6 (1)
Liver failure with bleeding 4 (1)
0- 1 3* 589 Sepsis with vasculitis (including bacterial
1- 9 8 180 endocarditis) 5 (1)
10-19 15(1) 163 Collagen-vascular disease
20-29 8 156
4t (2)
Aplastic anemia with thrombocytopenia 4 (1)
30-39 12 (2) 150 Hemophilia A 1
40-49 24(9) 276 Anticoagulant therapy (poorly controlled) 1
50-59 36 (12) 475 Disseminated intravascular coagulation 1
60-69 26(7) 577 Total 95 (21)
70-79 7(4) 299
80-89 5 (2) 135 Number in parenthesis ( ) pertains to patients with hyper-
90 + 0 10 tension.
Total 144 (37) 3,010 *One patient on anticoagulants.
fTwo of these patients on anticoagulants but not included
"Cases of classical subependymal matrix hemorrhages in the under anticoagulant therapy heading. The coincidence may
neonate as excluded from this study. have been important in causing the bleed, however.
Stroke, Vol. 4, November-December 1973 947
Downloaded from http://ahajournals.org by on February 20, 2023
 MASSIVE BRAIN HEMORRHAGE

stained with hematoxylin and eosin, trichrome and

Movat stains. In addition, a minimum of 12 "routine"
blocks were taken from the cerebrum, brain stem and
cerebellum in areas remote from the hematoma.
We have, somewhat arbitrarily, elected to use a
pre-ictal blood pressure recording of 140/90 or above as
evidences of hypertension and/or a heart weight (in the
adult) of 400 gm (without evidences of valvular
disease). Since patients with increased intracranial
| pressure, as from a massive hemorrhage, will commonly
have high blood pressure, recordings made after the
clinical ictus were not accepted as evidence of pre-
existing hypertension. A "massive hemorrhage" is
defined as one in which the maximal dimension is 3 cm
or greater in the cerebral hemisphere and 2 cm or
3 greater in the brain stem or cerebellum. Strict adherence
O to these definitions have been maintained in this
study.

in to DEFINITION OF STUDY POPULATION

CM — The University of Iowa Medical Center is a 1,100-bed
referral center with an average of approximately 31,000
patient admissions per year. During the period covered
s by this study (August 1, 1964, to April 1, 1973) there
c
were 6,622 deaths in the population for which the
• — •> a Department of Pathology of the University of Iowa
§ "> P Hospitals is responsible. Of these, 4,600 patients (70%)
*J oo g
8
were autopsied and 3,372 had examination of their
brains; the remainder were limited autopsies. Thus, only
about 50% of all patients dying in this medical center
had examination of their brains. Of total brains, 3,010
0 of 3,372, including all of those with massive hemor-
_O rhage (table 1), were examined by the senior author.
Downloaded from http://ahajournals.org by on February 20, 2023

The male:female ratios for this population are 1.4:1 for

I 2 du" both deaths and autopsies.
a
£ 'u c In order to determine, as nearly as possible, the
u percentage of our adult autopsy population that would
4)
U
o
o
I
o meet our criteria for hypertension, the autopsy
B. protocols of 250 patients over the age of 19 were chosen
D
by selecting 50 protocols from each of the years 1967 to
a 1971 inclusive. The 50 cases were consecutive except
for discarding all patients less than 19 years of age and
all cases with severe cardiac valvular disease. The time
periods were chosen so that all months were represented
in the sample (i.e., January and February, 1967, March
.2 o and April, 1968, etc.). The results are summarized in
81 table 2 and indicate the prevalence of "hypertension" in
our autopsy population to be 45%.

^8 Results
S9 o a> The pathological processes considered to be the
o -i:
cause for the brain hemorrhages are given in table 3.
E a!
U A satisfactory cause for the hemorrhage could be
identified in 95 of the 144 patients with massive
< u brain hemorrhage. Leukemia, generally associated
with severe thrombocytopenia, was the single most
common disease (table 4) in our series. Rupture of
CO 00
a saccular aneurysm (table 5) or of an angioma
.2 II (table 6) constituted the next most common
00
00 recognizable causes.
Tables 4 through 7 list specific data for those
Q patients whose hemorrhages were attributed to
Stroke, Vol. 4, November-December 7973 949
 McCORMICK, ROSENFIELD

TABLE 5

Ruptured Saccular Aneurysms (20)

Autopty Location of
no. Age/sex Hypertension hematoma

A64-385 56 F L. mid. cerebral L. temp.

A64-433 37 M + L. vertebral-post, L. cerebellar
inf. cerebellar hemisphere
A64-491 49 F _ L. mid. cerebral L. striate
A65-139 51 M L. ant. cerebral-ant, comm. L. frontal
A65-441 54 M _ R. ant. cerebral-ant, comm. L. frontal
A65-561 26 M L. mid. cerebral L. striate
A65-616 41 M R. mid. cerebral R. striate
A66-483 38 F _ R. mid. cerebral R. striate
A66-499 60 M — R. int. carotid-post, comm. R. temp.
A66-500 45 F — L. mid. cerebral L. striate
A67-216 31 F + R. mid. cerebral R. frontal
A67-308 52 M L. mid. cerebral L. frontal
A68-455 53 F _ R. ant. cerebral-ant, comm. L. frontal
A69-272 56 F + L. ant. cerebral-ant, comm. R. frontal
A70-24 54 F — L. int. carotid-post, comm. L. temp.
A70-148 51 M + L. int. carotid-post, comm. L. striate
A71-533 38 M _ L. int. carotid-ophthalmic L. frontal
A72-461 47 F _ R. mid. cerebral R. temp.
A73-23 60 M _ R. mid. cerebral R. striate
A73-110 52 F - R. mid. cerebral R. frontal

leukemia, ruptured saccular aneurysms, angiomas confined to the right striatum in the remaining three.
and either primary or metastatic brain neoplasms. Only one of these patients, the oldest in the series,
Table 8 itemizes those patients with severe systemic had significant hyaline arteriolar sclerosis within the
bleeding disorders other than leukemia, and table 9 brain. These cases are summarized in table 10.
lists those patients with cortical vein thrombosis. Among 58 patients in our material who had
Downloaded from http://ahajournals.org by on February 20, 2023

There were 12 normotensive patients, ranging in acceptable evidence of systemic hypertension, 21

age from 8 to 75 years, for whom no cause of the had other causes for the hemorrhage independent of
massive brain hemorrhage could be determined. Five the hypertension. Thus, only 37 of our 144 patients
were less than 15 years of age. In eight of these had "hypertension" as the only identified cause of
patients, the massive hemorrhage was lobar in their hemorrhage. Nearly half of these patients also
distribution, right cerebellar hemisphere in one, and had significant vascular disease outside of the central

TABLE 6

Ruptured Angiomas (15)

No. Age/»ex Hypertension Location Type

A65-297 17 M — Thalamus and hypothalamus Telangiectasis

A66-380 15 M — L. striate Arteriovenous
A67-171 29 F — Thalamus and hypothalamus Arteriovenous
A68-136 12 F — Cerebellar vermis and L. Arteriovenous
hemisphere
A68-206 55 M + L. striatum Arteriovenous
A68-253 48 F — L. striatum Arteriovenous
A69-47 61 M + L. cerebellar hemisphere Arteriovenous
A69-356 64 M + L. temporal Arteriovenous
A69-439 15 M — L. frontal Arteriovenous
A70-147 1 da M — L. cerebellar hemisphere Arteriovenous
A71-38 37 M — L. cerebellar hemisphere Arteriovenous
A71-409 31 F — L. cerebellar hemisphere Arteriovenous
A71-491 48 F + R. temporal Arteriovenous
A71-525 65 F — Pons and medulla Arteriovenous
A73-36 61 M - L. parietal Arteriovenous

950 Sfroke, Vol. 4, NoYember-Dactmber 1973

 MASSIVE BRAIN HEMORRHAGE

TABLE 7

Massive Hemorrhage into Neoplasms (13 Cases)

Case no. Age/sex Hypertension Type and location of primary neoplasm

1. Primary brain neoplasms (5):

A66-182 12 F~ Medulloblastoma; cerebellum*
A66-493 52 F Glioblastoma; right frontal
A68-183 11 M Malignant astrocytoma; right cerebral,
massive
A70-176 29 F Astrocytoma; right thalamus
A73-59 9 mo. M Cerebral "neuroblastoma"; L. cerebral
subependymal matrix
2. Metastatic brain neoplasms (8):
A65-207 69 F Melanoma; left arm
A67-168 59 M Melanoma; chest wall
A68-188 44 F Adenocarcinoma; lung
A68-288 79 M Sq. cell; primary site unknown
A71-235 50 F Adenocarcinoma; lung
A72-312 65 M Large cell carcinoma; lung
A72-338 58 F Giant cell carcinoma; lung
A73-92 27 F Chorionepithelioma

•Published in / Neurosurg 26:78-81, 1967.

nervous system and six had severe coronary artery The locations of the hematomas in all 144
disease with myocardial infarction. The youngest patients were cerebral in 107, cerebellar in 15, brain
patient in this group was 19 years old (with chronic stem in six, and multiple, often in cerebrum, brain
glomerulonephritis). stem and cerebellum, in 16. The basal ganglia
Hypertension, generally mild, as determined ("striatum" and thalamus) were primarily involved
either by pre-ictal recordings of blood pressure in 54 of the single cerebral hemorrhages and one of
and/or cardiac hypertrophy also was present in 21 the lobes (i.e., "lobar") in 53 (tables 10 and 11).
Downloaded from http://ahajournals.org by on February 20, 2023

of the 95 patients (table 3) in whom a satisfactory

cause of the hemorrhage, independent of hyper- Discussion
tension, was found. Four (of 20) patients with The location of a hemorrhage would seem to be of
ruptured saccular aneurysms and four (of 15) of the considerable importance when considering etiology.
patients with ruptured angiomas were hypertensive. Multiple hemorrhages were seen most commonly in
Three of the patients with massive hemorrhage into patients with severe bleeding disorders, such as
a neoplasm were mildly hypertensive—all of these in leukemia, liver failure, etc. No examples of multiple
patients with metastatic neoplasm. Four of the hemorrhages were seen in our patients with
patients with leukemia were hypertensive but in all "hypertension" as the only recognized factor for the
of these patients the hypertension was mild and hemorrhage.
asymptomatic. The youngest patient with hyperten- The so-called "striate-hemorrhage" is the most
sion and leukemia was 58 years of age. common pattern seen in our "hypertensives," but

TABLE 8

Severe Bleeding Disorders — Nonleukemic

Disorder No. Age/sex Hypertension Location of hemorrhage(s)

Hemophilia A71-351 11 M — L. frontal

Alcoholic cirrhosis with severe A65-433 62 M + L. cerebellar hemisphere
hepatic failure A67-54 54 M — Multiple cerebral and cerebellar
A70-305 48 M — L. parietal
A71-41 49 M — R. temporal
Aplastic anemia with severe A69-75 18 M — R. parietal
thrombocytopenia A69-385 8 F — Intracallosal and bifrontal
A69-267 58 F - R. striate
A69-297 70 M + Multiple cerebral
Disseminated intravascular A70-58 47 F — Multiple cerebral and cerebellar
coagulation

Stroke, Vol. 4, November-December 1973 951

 McCORMICK, ROSENFIELD

TABLE 9

Cortical Vein and/or Dural Sinus Thrombosis

No. Age/tax Hypertension Involved veins and other significant findings

A67-299 33 F Thrombosis of right and left veins of Trolard; positive

contrast myelography 12 hours prior to onset of signs
and symptoms of brain lesions
A68-211* 25 F Thrombosis of right frontal and parietal veins; systemic
venous thrombosis with pulmonary emboli
A68-483 41 M Superior sagittal sinus; severe systemic arterial and
venous thrombosis
A72-67 36 F Cortical veins and superior sagittal sinus; metastatic
carcinoma from breast
A72-260 6 wk. M Vein of Galen and internal cerebral veins
A73-41 51 M Septic thrombosis of left frontal veins; subdural empyema
and purulent leptomeningitis
•Patient on oral contraceptives.

bleeding in this location is by no means limited to with known causes. There were 12 additional patients
hypertension. Nineteen patients with a known with hypertension plus a cause for the hemorrhage
nonhypertensive etiology and three normotensive (saccular aneurysm in four, leukemia in three,
patients with no known cause for their hemorrhage angioma in two, etc.) who had their only bleed in
bled massively into the striatum. Five patients with one lobe of the cerebrum.
other causes (two with metastatic carcinoma, one Of the 15 patients with a cerebellar hemor-
with an aneurysm, one with an angioma, and one rhage, seven were in "hypertensive" patients; four
with collagen-vascular disease) who were also
hypertensive (all clinically asymptomatic) had their were in patients with only hypertension and three in
hemorrhage in the striatum. patients with another satisfactory cause for the bleed
A cerebral "lobar" hematoma is apparently (one with a ruptured aneurysm, one with an
angioma, and one with a severe, nonleukemic
Downloaded from http://ahajournals.org by on February 20, 2023

much less likely to be related to hypertension, and

was found in only four of the 37 patients with bleeding disorder). Eight occurred in normotensive
"hypertension (table 12). Lobar hemorrhages were patients, four in patients with ruptured angiomas,
present in eight normotensive patients with no known three in patients with brain tumors, and one in a
cause of their bleed and in 30 normotensive patients patient in whom no cause could be found.

TABLE 10

Massive Brain Hemorrhage (No Known Cause — Normotensive)

Moderate or severe
arteriolar disease
No. Age/sex Location of hemorrhage in CNS Miscellaneous data
A64-407 44 F L. parietal lobar Sudden onset of severe headache while resting
A64-426 8 F L. parieto-occipital lobar Awakened by "feeling of fear" and headache
A65-447 58 F R. frontal lobar No known ictus, "gradual" decline, carcinomatosis
with icterus
A66-104 7 M R. parietal lobar Onset of headache while resting and watching
television
A66-129 52 M L. frontoparietal lobar R. focal seizure — R. hemiplegia
A66-435 14 F R. striate Recovering from "flu," sudden loss of consciousness
A67-506 62 M R. striate Sudden onset of dizziness — fell — L. hemiplegia
— severe coronary artery disease and aortic
atherosclerosis, heart weight 320 gm
A68-72 8 M R. striate Sudden onset of severe headache while at play —
rapid coma
A68-438 61 F R. frontoparietal lobar Slow onset of L. hemiparesis
A68-481 11 M R. frontolobar "Suddenly fell unconscious" while in school
A71-47 60 M R. temporal lobar Onset with headache and L. hemiparesis, death
in three days
A71-490 75 M R. cerebellar Moderate cerebral atherosclerosis, lacunar state

952 Stroke, Vol. 4, November-December 1973

 MASSIVE BRAIN HEMORRHAGE

TABLE 11
Location of Hematomas — 95 Patients With Known Cause
Single cerebral
Cause of hemorrhage basal ganglia* lobar Cerebellar Brain stem Multiple

Leukemia (21) 1 10 0 l 9
Aneurysm (20)f 7 12 1 0 0
Angiomas (15) 5 4 5 I 0
Neoplasms (13) 4 4 3 0 2
Bleeding disorders — nonleukemic (10) 1 5 1 0 3
Vasculitis (bacterial endocarditis, sepsis
and collagen disease) (10) 5 2 0 1 2
Cortical vein/dural sinus thrombosis (6) 1 5 0 0 0
Total (95) 24 42 10 3 16
*Basal ganglia = striatum and/or thalamus.
tThe hematoma was always directly contiguous with the ruptured aneurysm at one point.

Three of the six patients with brain stem hospital, the preciseness of the definition of the
hemorrhage were hypertensive. Leukemia, a rup- terms used in the study and, in the case of a
tured angioma, and vasculitis accounted for the potentially highly lethal disorder, the number of
remaining three. cases reaching the hospital for study. Other factors
The relative prevalences of the hemorrhages by to be considered are the special interests of the
major sites (e.g., cerebral, cerebellar and brain different clinicians in the referral institution, and, of
stem) do not differ significantly from those widely great importance, the manner in which the individual
reported in the literature.6 cases are studied by the pathologist. There is no
Nontraumatic cortical vein and/or dural sinus possible mechanism by which all possible biases can
thrombosis was the cause of a massive brain be eliminated from such a study. The best that can
hemorrhage in six patients (table 9 ) . One was be hoped for is that these biases can be recognized,
clearly due to a septic process surrounding the right identified clearly, and their possible significance
frontal veins. One patient was a young woman on examined. To do this, precision of definition and
clear delineation of the population under study are
Downloaded from http://ahajournals.org by on February 20, 2023

oral contraceptives who also had severe systemic

venous thrombosis and died from a massive imperative.
pulmonary embolus. The remaining four patients There is a marked racial selection in this study
had no clear cause for their intracranial venous population, in that virtually all patients have been
thrombosis. One patient had widely disseminated white—only one was black. (This marked difference
breast carcinoma with many brain metastases. No reflects the regional racial difference and does not
tumor thromboemboli were found and no marantic appear to reflect an exceptional referral pattern.) It
endocarditis was present. Another patient had the is often stated that hypertension is significantly more
onset of her neurological signs and symptoms common and has a higher average level in many
approximately 12 hours after positive contrast black populations, but this cannot be evaluated from
myelography for vague sciatic pain. No known our material.
allergies were present in this patient. While initially With a highly lethal disease such as massive brain
having an occlusive vascular disease, all of these hemorrhage, it is quite possible that a significant
patients had massive bleeding into the areas of number of patients will not live long enough to reach
ischemia producing hematomas larger than 3 cm. any hospital, this one in particular due to its referral
There are a number of possible explanations for nature. However, this statement should operate in
the striking discrepancies in this hemorrhage popula- many, if not all, cases of massive hemorrhage and
tion and in most of those previously reported. not just those "due" to "hypertension." Moreover, as
Inherent in any such study are certain problems as to the great majority of patients with a massive brain
choice of population, method of referral to the hemorrhage will survive for at least several hours

TABLE 12
Location of Hemorrhage — Hypertensive and Unknown — 49 Patients
Cerebral
Basal ganglia Labor Cerebellar Brain stem

Hypertensive (37) 26 4 4 3
Normotensive, no known cause (12) 3 8 1 0

Stroke, Vol. 4, November-December 1973 953

 McCORMICK, ROSENFIELD

following their ictus, patients from a significant part 2). Moreover, 21 of these 57 patients had a recog-
of our referral area would, and do, have time to nizable cause for their hemorrhage, regardless of
reach the hospital. the blood pressure levels (table 3). The argument
Note must be made of the special referral for hypertension as the significant—indeed the over-
nature of this University Hospital. It is the central whelming—cause of massive nontraumatic brain
registry for the Cooperative Aneurysm Study and hemorrhage is not supported by these prevalence
many patients with proved or suspected ruptured figures.
aneurysms are referred to this institution from a There is a widespread belief that hypertension
large part of the Midwest. Twenty of our 144 is the significant etiological factor in causing
patients had a ruptured aneurysm as the cause of spontaneous massive cerebral hemorrhage. Many
their hemorrhage. However, several of these patients years ago DeVries stated that cerebral hemorrhage
did not have their aneurysm(s) proved during their occurs in people with hypertension: "Bleeding into
life. There is an active hematologic-oncologic group the brain . . . is an affliction of the older hypertensive
in this institution with special interests in the therapy patient, more frequent in males than in females."7
of the blood dyscrasias, particularly leukemias. More recently Ransohoff et al.2 were to claim that
Blood dyscrasias accounted for the single largest the " . . . overwhelming frequent cause of single,
group of our patients. Finally, there are relatively sizeable [size not mentioned] intracerebral hemor-
large and active neurology and neurosurgery units in rhage is hypertension. When other etiologic factors
this institution and a relative paucity of private are excluded, hypertension is present in over 90% of
practitioners of these specialties in this general cases." Many well-known current pathology and
geographic area. neurology texts adopt this view, as do the recent
All three of these factors may well operate in reviews by Ziilch,8 Stehbens,6 Schwartz,9 and
causing a bias in both the number and the type of Luyendijk.10 Most authors fail to adequately de-
brain hemorrhages seen in this center. scribe the many known causes of massive intra-
Further altering the usual reported ratios cerebral hemorrhage in the various locations that
between "hypertensive" hemorrhages and hemor- such hemorrhages can occur.
rhages attributable to recognizable structural abnor-
malities of the brain is the uniform and meticulous Acknowledgments
method of examining the specimens adhered to in Appreciation is given to Dr. Sydney S. Schochet. Jr.. for his
this institution. At the onset, this study was set up as advice and suggestions during this study.
Downloaded from http://ahajournals.org by on February 20, 2023

a prospective one in which all brains were to be

examined, photographed, and described by a single References
observer (WFMcC) and appropriate blocks taken 1. Robbins SL: Pathology. Philadelphia, WB Sounders,
p 1400, 1967
and examined by the same observer. A minimum
2. Ransohoff J, Derby B, Kricheff I: Spontaneous
number of sections (12) were taken from unin-
intracerebral hemorrhage. Clin Neurosurg 18: 247-
volved areas of cerebrum, cerebellum, brain stem 266, 1971
and spinal cord as well as a minimum of eight blocks 3. Langfeld SB: Hypertension: Deficient case of the
from the wall of the hemorrhage and the central medically served. Ann Int Med 78: 19-23 (Jan)
hematoma. In many cases, a much larger number of 1973
blocks were taken and subsequently sectioned and 4. Kurtzke JF: Epidemiology of Cerebrovascular Dis-
examined. ease. New York, Springer-Verlag, p 81-89, 1969
Undoubtedly, other biases not recognized by 5. McCormick WF: Problems and pathogenesis of
the authors could be in operation. It is obvious that intracranial arterial aneurysms. In Toole JF, Moossy
some of the above-mentioned biases will increase the J, Janeway R (eds) : Cerebral Vascular Diseases.
accuracy of a study of this nature, whereas others, New York, Grune and Stratton, p 219-231, 1971
such as the basically all-white population and the 6. Stehbens WE: Pathology of the Cerebral Blood
referral nature of this hospital, will decrease the Vessels. St Louis, CV Mosby, p 284-350, 1972
application of our findings to other institutions and 7. DeVries E- Apoplexy; some facts and recent reviews.
to other regions of the country. Chinese Med J 4 6 : 679-715, 1932
8. Ziilch KJ: Hemorrhage, thrombosis, embolism. In
When we consider only those patients 20
Minckler J (ed) : Pathology of the Nervous System.
years of age and over, thus removing from considera- New York, McGraw-Hill, 2: 1499-1536, 1971
tion 25 without and one with hypertension, there are 9. Schwartz P: Apoplectic lesions of the brain in adults.
118 "adults" with massive brain hemorrhage. Fifty- In Vinken PJ, Bruyn GW (eds) : Handbook of
seven (49%) of these patients had "hypertension," Clinical Neurology. New York, American Elsevier
as we have defined it. This is not significantly differ- Publishing Co, 1 1 : 578-659, 1972
ent from the expected, based on a prevalence of 10. Luyendijk W : Intracerebral haematoma. Ibid, p 660-
45% derived from our "control" populaton (table 719

954 Stroke, Vol. 4, November-December 1973