Basic pharmacokinetics

Frédérique Servin
APHP – hôpital Bichat
Paris, FRANCE
Conflict of interest declaration
WHAT DECLARATION
Grants/research support/P.I. none
Employee none
Consultation fees none
Honoraria Baxter
Speakers bureau none
Company sponsored Fisher & Paykel
Stock shareholder none
Spouse/partner none
Scientific Advisory Board none
Other none

www.esahq.org
DOSE CONCENTRATION EFFECT

Pharmacokinetics Pharmacodynamics
What the body does to What the drug does to
the drug the body

• Transfer to the effect site: PK or PD?

• Estimation of the transfer to the effect site requires a
measure of effect
Pharmacokinetic steps

• Absorption
• Digestive
• Trans-mucosal (nasal; sub-lingual; rectal…)
• Distribution
• Elimination / transformation
• Metabolism
• Excretion in bile or urine

www.esahq.org
Absorption: first pass effect

www.esahq.org
Distribution
• Describes the diffusion (active or passive) of a drug in
various tissues. The penetration of the drug varies
with different tissues.

• If the drug is protein bound, only the free

(active) fraction will be distributed
• Acid drugs are bound to plasma albumin
• Basic drugs are bound to a1-acid-glycoprotein
• BUT if the affinity for the metabolic enzyme is superior to
that for the binding protein, the binding will not limit the
clearance (non restrictive clearance)

www.esahq.org
Protein binding of anaesthetic agents

Albumin a1 AGP binding <50%

Thiopental Fentanyl Ketamine
Propofol Alfentanil Pancuronium
Etomidate Sufentanil Vecuronium
Midazolam Remifentanil Rocuronium
Flumazenil Atracurium
Morphine Cisatracurium

www.esahq.org
Hepatic drug elimination
• The liver renders xenobiotics accessible to
excretion in urine or bile  water-soluble
• Biliary excretion
• Metabolism
• Phase 1 reactions
• Hydrolysis
• Oxidation (CYP 450)
• Reduction
• Phase 2 reactions
• Conjugation +++
• Methylation, acetylation ….

www.esahq.org
Hepatic clearance

ClH = QH  E = QH  [ [Cli] / [QH + Cli] ]

• If E is close to 1, ClH ~ QH The clearance is flow rate
dependant
• If E is small (low extraction), Cli is small compared
to QH, thus [Cli] / [QH + Cli] is close to Cli / QH , and
ClH ~ QH  [Cli / QH ] ~ Cli The clearance is enzyme
dependent, and insensitive to changes in blood
flow

www.esahq.org
Hepatic clearance of anaesthetic
agents
Rate dependant Intermediate Enzyme
dependant
Propofol Midazolam Thiopental
Etomidate Alfentanil Diazepam
Ketamine Vecuronium Lorazepam
Flumazenil Bupivacaïne Flunitrazepam
Morphine
Fentanyl
Sufentanil
lidocaïne

www.esahq.org
Hepatic metabolism of anaesthetic
agents
Phase 1 CP450 Phase 1 other Phase 2
Thiopental Etomidate Propofol
Ketamine (Atracurium) Morphine
Midazolam
Flumazenil
Fentanyl
Sufentanil
Alfentanil
(Vecuronium)
(Rocuronium)

www.esahq.org
Application of pharmacokinetics to
clinical anaesthesia

www.esahq.org
 Pharmacokinetic analysis
C1 = A * e -a*t
Concentration

100
C2 = B * e -b*t
C3 = C * e -g*t
(µg/ml)

This three exponential

10
analysis can be
modified (MATHS !!!)
1 in other forms
« compartment
analysis »
0.1
0 10 20 30 40 50 60

Time (min)

www.esahq.org
Pharmacokinetic model

V3
V2 V1
CL2 CL3

CL1
Cl1 = V1*k10
Cl2 = V1*k12 = V2*k21
1 - V1, V2, V3, Cl1, Cl2, Cl3 V2 = V1* (k12/k21)
2 – V1, k10, k12, k21, k13, k31

www.esahq.org
 Compartment models
• Simple ++ (max 3 cpts)
• Actually not much to do with reality
• the drug concentration drug is assumed to be uniformly
equal within the compartment, with instant homogeneous
distribution…
• Mammillary model (the peripheral compartments are only
connected through the central one)
• Elimination usually occurs only from the central
compartment
• Advantage: they work! At least at the level of complexity
adequate for clinical practice

www.esahq.org
 Bring the drug concentration in the
3.5 « therapeutic » zone and keep it
conc
there as long as necessary
3

2.5

1.5

0.5

0
0 5 10 15 20 25 30 35 40 45 50 55 60

time (min)

www.esahq.org
 V3
The BET
V2 V1
 Bolus CL2 CL3
 The target concentration is obtained
 by filling up the central compartment CL1
 Elimination
 A constant rate infusion
 Compensates metabolism and / or excretion
 Transfer
 A gradually decreasing rate infusion
 Compensates the distribution to the peripheral
compartments

Lauven et al A microprocessor controlled infusion scheme for midazolam to

achieve constant plasma levels, Anaesthesist 31: 15-20, 1982)

www.esahq.org
 BET equations
• Bolus:
Ct X V1
• Elimination:
Ct X Cl = Ct X V1 X k10
• Transfer ...
Ct X V1(k12 e(-k21t)+ k13 e(-k31t))
Maintenance rate infusion is therefore:
Dose = Ct X V1 (k10 +k12 e(-k21t)+ k13 e(-k31t))

www.esahq.org
Building a PK model

• Patients or healthy volunteers

• Bolus dose or infusion
• Many blood samples (critical phases)
• Two stages analysis
• Population kinetics

www.esahq.org
 Population kinetics
• Blood samples from all the patients are studied
together (pooled approach)
• Significant covariates improve the statistical likelihood
of the model (ex : weight, age, sex …)
• The influence of the covariates is considered for every
parameter.
• Bayesian adaptation allows the estimation of individual
parameters (posthoc values)
• Some populations (children) require specific models

www.esahq.org
Induction

www.esahq.org
Effect site concentration, fentanyl

Scott, Anesthesiology 1985

www.esahq.org
Time course of action : alfentanil

Scott J, Anesthesiology, 1985

www.esahq.org
 Effect hysteresis
effet

conc

www.esahq.org
Cl1 = V1 x k10
Cl2 = V1 x k12 = V2 x k21
Cl4 = V1 x k14 = VE x ke0;
VE = e, so k14 = e, and V1
is proportional to Ke0

V3
V2 V1

CL2 CL3

ke0 CL1

Effect site
 Time to peak effect is a physiological variable
independent from the model
• The ke0 constant associated to the pharmacokinetic model is dependent
on the PK model and the time to peak effect

Female, 75 y.o., 62 kg, 59 cm

propofol 70 mg, no opioid
• Time to peak = 2.2 min
100.00
• Marsh model
BIS
• V1 = 14.1 L
80.00
• Ke0 = 1.11 (0.267)
60.00
• Schnider’s model
• V1 = 4.27 L
40.00

20.00 • Ke0 = 0.408 (0.456)

0.00
0.0 2.0 4.0

www.esahq.org
 Opioids time to peak effect.

sufentanil
100

80
fentanyl
% of EEG maximum effect

60

40 alfentanil

20

remifentanil
0

0 2 4 6 8 10
Minto et al, Anesthesiology 1997
www.esahq.org
 Onset time  time to peak effect
Conc (ng/ml)

16
14
12
10 Minimum concentration recommanded
with remifentanil (3 ng/ml)
8
6 Lower concentration limit
for endotracheal intubation
4
2
0
time (min)
0 10 20 30
Alfentanil Rémifentanil Sufentanil Fentanyl
www.esahq.org
Effect site concentration, therapeutic
window
Propofol bolus, 2.5 mg/kg

Propofol plasma
conc ( µ g/ml) Effect site
12

10

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13
time (min)

www.esahq.org
Instantaneous mixing is an approximation

Avram, Anesthesiology 2003

www.esahq.org
 Pharmacokinetic models currently
used for propofol TCI
• Marsh (1991) • Schnider (1998) • Allometric approach
• V1 = 0.228*TBW • V1 = 4.27 L Eleveld (2014, 2018)
• K10 = 0.119 min-1 • V2 = 18.9 – 0.391*(age-53) L
• K12 = 0.112 min-1 • V3 = 238 L
• K13 = 0.0419 min-1 • Cl1 = 1.89 + 0.0456*(TBW-77)
• K21 = 0.055 min-1 – 0.0681*(LBM-59) +
• K31 = 0.0033 min-1 0.0264*(height-177) L/min-1
• Ke0 Diprifusor • Cl2 = 1.29 – 0.024*(age – 53)
0.267 L/min-1
• Ke0 Base Primea • Cl3 = 0.836 L/min-1
1.21 • Ke0 = 0.456
• Azena PK Tpeak
1.6min

www.esahq.org
Maintenance

www.esahq.org
Concentration / effect relationship

100 probability of effect

90

80

70

60 E Emax Cy
50
g
EC50Cg
40

30
efficiency range
20

10
EC95
0 EC50
20 30 40 50 60 70 80 90 100

concentration
www.esahq.org
 Hypnotics

Opioids

Stimuli
Recovery

www.esahq.org
From maintenance to recovery …

Vuyk, Anesthesiology 1997

www.esahq.org
Alfentanil

Vuyk, Anesthesiology Dec. 1997

www.esahq.org
Remifentanil

Vuyk, Anesthesiology Dec. 1997 www.esahq.org

7 Propofol conc.
(µg/ml)
6

4
Context sensitive half life
3

2 Decrement time

0
time (min)
The recovery time depends on the ratio
between the maintenance concentration and
the C50 for recovery
6
End of infusion
5

1 4’15 22’13

0
50 55 60 65 70 75 80 85 90
Time (min)
www.esahq.org
Context Sensitive Half Time

Shafer SL, ASA Refresher Course, Chapter 19, 1996

Are drug models predictive?

• Mathematical models of drug behaviour

incorporating effect site concentrations and drug
interactions predict anaesthetic drug effect (e.g.,
loss of response to stimulation) as well as:
• Measured concentrations
• BIS
• AAI

• I am not aware of any counter examples.

Slide S. Shafer

www.esahq.org
Remember
• Thinking in concentration is mandatory to
understand a drug time course of effect and
interactions
• Mathematics are necessary to predict drug
concentration over time (thus effect)
• BUT pharmacokinetics is primarily proteins,
enzymes, liver, kidney, blood, fluids and cells!
• If you understand pharmacokinetics, you will be
able to understand what happens to your drug
even in patients widely different from the standard
population

www.esahq.org