PKPD: PAST, PRESENT AND FUTURE - AFMPS · 2018-11-13 · PHARMACOKINETICS Pharmacokinetics is the...

PKPD: PAST, PRESENT AND

FUTUREApplied to antibiotics - Prof. Dr. An Vermeulen - 23 Nov 2017

DEPARTMENT OF BIOANALYSIS

LABORATORY OF MEDICAL BIOCHEMISTRY AND CLINICAL ANALYSIS

2

0 20 40 60 80 100

0.0

0.2

0.4

0.6

0.8

1.0

Efficacy

Toxicity

Dose (exposure)

Effect

DOSE SELECTION MADE EASY

3

0 20 40 60 80 100

0.0

0.2

0.4

0.6

0.8

1.0

Dose (exposure)

Effect

IN REALITY …

4

0 20 40 60 80 100

0.0

0.2

0.4

0.6

0.8

1.0 Efficacy

Toxicity

Dose (exposure)

Effect

… KNOWLEDGE IS IMPRECISE

Uncertainty

5

0 20 40 60 80 100

0.0

0.2

0.4

0.6

0.8

1.0 Efficacy

Toxicity

Dose (exposure)

Effect

ALSO, ALL PATIENTS ARE DIFFERENT

Inherent

Biological

Variability

6

− We try to establish relationships between exposure (mostly plasma

concentrations) and effects

− In a way, if we focus on PK, it is because we believe it is a biomarker for

effects, and that understanding the PK profile helps to understand the PD

(effects) profile

− Most of the times, we can not measure the concentrations at the effect site,

and we sample plasma concentrations instead (easy matrix to sample and

to measure concentrations in)

− The relationship between plasma and effect site concentrations is then

again linked mathematically

− Once a model is developed, it allows simulations of scenarios we have

not studied/for which we do not have information

− Time points w/o sampling

− Additional doses/dose regimen

− Repeated dosing (using single dose data)

− At population & individual level

− For different covariate subgroups (pediatrics, CrCL, BWT, ..)

PKPD MODELING

7

PKPD MODEL APPLICATIONS

• Conceptualization and quantification

• In vitro In vivo extrapolation

• Species scaling

• Clinical trial design

• Design of new dosage forms

• Comparative efficacy/toxicity

• Patent enhancement

• Labeling advice

8

PHARMACOKINETICS

9

PHARMACOKINETICSPharmacokinetics is the study and mathematical characterization of the time

course of drug absorption, distribution, metabolism, and excretion (ADME)

processes that determine the time-course of drug action.

Adapted from Atkinson, Principles of Clinical Pharmacology (2001)

PRESCRIBED DOSE

DRUG PRODUCT

ABSORPTION

PLASMA

FREEPROTEINBOUND

ELIMINATION

METABOLISM RENALEXCRETION

MOST TISSUES

NONSPECIFIC

BINDING

BIOPHASE

RECEPTORBINDING

DISTRIBUTION

EFFECT

BIOPHARMACEUTICS COMPLIANCE

10

11

NONCOMPARTMENTAL ANALYSIS

12

NONCOMPARTMENTAL ANALYSIS

PRIMARY PHARMACOKINETIC PROPERTIES

van de Waterbeemd and Gifford. Nature Rev Drug Disc. 2:192-204 (2003)13

COMPARTMENTAL ANALYSIS

14

COMPARTMENTAL ANALYSIS

0 5 10 15 20

0.1

0.5

1.0

# 4 50001

Time, h

R2

091

30

, ng

/ml

15

16

THE ASPECT VARIABILITY

− Modelling is essentially similar at the population level !

− But ..− At the population level, the aspect variability comes into play

− In a population analysis, the focus is on

− estimating the magnitude of this variability

− trying to identify patient characteristics that can explain (part of) this variability

− dose-adjustments in subsets? LABELLING impact !

− dose-individualisation in individuals? (avoid TDM - therapeutic drug monitoring)

− So ..− Three levels of variability have to be considered:

− IIV or BSV: inter-individual or between-subject variability

− IOV: inter- or between-occasion variability

− Intra-subject, within-subject or residual (unexplained) variability16

THE ASPECT VARIABILITY

0 5 10 15 20

0.1

0.5

1.0

# 4 50001

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l0 5 10 15 20

0.5

1.0

# 4 50002

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l0 5 10 15 20

0.5

1.0

# 4 50003

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l0 5 10 15 20

0.2

0.3

0.5

# 4 50004

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

0 5 10 15 20

0.5

1.0

# 4 50005

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

0 5 10 15 20

0.5

1.0

# 4 50006

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

0 5 10 15 20

0.1

0.5

1.0

# 4 50007

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

0 5 10 15 20

0.2

0.4

0.7

# 4 50008

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

0 5 10 15 20

0.5

1.0

# 4 50009

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

0 5 10 15 20

0.5

1.0

# 4 50010

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

0 5 10 15 20

0.1

0.5

1.0

# 4 50011

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

Time, h

R2

09

13

0, n

g/m

l

17

− Nonlinear Mixed Effects Modelling (NONMEM)

− Mixed effects− Fixed effects ()

− Random effects (, )

− Fixed effects− Estimate mean PK(PD) parameters in the population (= typical values,

central tendency, population average, ..)

− Identify factors that influence the PK (PD) parameters (demographics, lab values, PGx, smoking habits etc…)

− Random effects− Estimate unexplained variability

=> IIV, IOV and residual variability

POPULATION PK PRINCIPLES

18

PHARMACODYNAMICS

19

PK-PD RELATIONSHIPS: TYPE OF EFFECTS

Drug effect

Reversible

Irreversible

Direct

Indirect

20

REVERSIBLE EFFECTS – DIRECT CONCENTRATION-EFFECT MODELS

)(

)ln(

50

max

50

max

p

p

p

p

p

p

p

CEC

CEEffect

CEC

CEEffect

CEffect

CEffect

CEffect

Linear

Log-linear

Power

Emax

Sigmoid Emax

Effect directly

related to plasma

concentration

EC50/IC50 values: the potency of the drug

Emax/Imax values: the efficacy of the drug

γ : sigmoidicity factor 21

22

23

REVERSIBLE EFFECTS – INDIRECTMODELING A BIOPHASE

Segre, Il Pharmaco (1968)

Sheiner et al., CPT (1979)

eCpCeokdt

edC

eC50EC

eCmaxEE

24

25

REVERSIBLE EFFECTS – INDIRECT

26

Chemotherapeutic effects

Simple cell killing

Killing and regrowth

Cell cycle effects

Clinical therapy

Irreversible enzyme inhibition

Reactive drug metabolites

Simple drug toxicity

Carcinogenicity models

IRREVERSIBLE EFFECTS

27

SIMPLE CELL KILLING

ln SF = -k * D0/CL

J. Pharm. Sci. 60:892 (1971)

Drecker functionN = R

28

SIMPLE CELL KILLING

Spleen cells

Osteosarcoma cells

Cyclophosphamide - mice

29

CELL KILLING – CELL GROWTH

30

CELL KILLING – CELL GROWTH

ANTIBACTERIAL EFFECTS

31

32

PKPD: PAST, PRESENT AND FUTURE …

33

MIC

34

MIC

kg = ks

35

SUMMARY PK MEASURES

36

ANTIBACTERIAL EFFECTS

37

ANTIBACTERIAL EFFECTS

38

Antibiotics

Thigh Lung

39

MIC DISTRIBUTION

40

TARGET ATTAINMENT

41

TARGET ATTAINMENT

42

CLINICAL AND MICROBIOLOGICAL CURE

43

CLINICAL CURE

44

CLINICAL BREAKPOINTPKPD BREAKPOINT

45

MODELING TIME COURSE OF ANTIBIOTIC EFFECT

• Better to model time course

of bacterial growth & killing

• Differentiate between the

system (bacteria) and drug

effects

• Can address resistance development

• .. and the effects of the host’s immune

system

=> Use all available data: in vitro, animal, clinical

46

SYSTEMS PHARMACOLOGY: PBPK

47

… COMBINED WITH PD MODEL AT INFECTION SITE

48

… ALLOWS TO DERIVE THE TISSUES WITH HIGHEST EFFECT

An VermeulenPharmacokinetics and Pharmacodynamics/PBPK/Pharmacometrics

DEPARTMENT OF BIOANALYSIS

E AnMC.Vermeulen@ugent.be

T +32 9 264 81 12

M +32 474 55 08 65

www.ugent.be

Ghent University

@ugent

Ghent University