Population PK/PD modeling for Individualized drug therapy ... · Population PK/PD modeling for Individualized drug therapy: immunosuppressive agents . ... Case Study 1 9 of 42. ...

Zheng Jiao, PhDHuashan Hospital, Fudan University, ChinaEmail: [email protected]

Population PK/PD modeling for Individualized drug therapy: immunosuppressive agents

Two types of drugs

Drugs with high therapeutic index

Drugs with narrowtherapeutic index and

patients at risk

It is not necessary to adjust regimen on the basis of the knowledge of individual PK parameters

The dose must be adjusted before each drug administration from the knowledge of individual PK parameters

Introduction

1 of 40

•Drug administration (test dose)• Extensive blood sampling (min 3 per phase)• Data modeling of the conc-time curve using classical approach

•compartmental or non-compartmental approach

BUTVery long and expensive method

Difficult to carry out

Ethical Problems (elderly, pediatric …)

How to determine individual PK parameters?

2 of 42

Population parameters (average of each PK parameter

+ variability)

Data from the patient (drug concentration [1, 2 or 3]

+covariates)

Bayesian estimation

Estimation of individual PK parameters

Use of population approach

＋

3 of 42

Variability in PK parameters

Inter-individual Variability

Specificity of PK parameters for each patient

For a specific patient, PK parameters can vary with time (i.e. during treatment)

Intra-individual Variability

must be taken into account before adjusting regimen

Variability

4 of 42

Definition of Population PK

FDA Guidance:“Population pharmacokinetics seeks to identify the measurable pathophysiologic factors that cause changes in the dose-concentration relationship and the extent of these changes so that, if such changes are associated with clinically significant shifts in the therapeutic index, dosage can be appropriately modified.”

5 of 42

The difference between Population PK & Classical PK

healthy subjects or highly selected patientsrich sampling average PK profile

one variable at a time

minimized through complex study design and/or restrictive inclusion/exclusion criteria

Subjects:

PK data: Focus of interest:

Endpoint

Variability

Specific settings

study target population

sparse and rich samplingaverage and individual PK profilemultiple variables, allow to study interactions among variablesNot controlled

ICU, pediatrics and elderly

6 of 42

China—kidney & liver transplants in past decade

Huang JF et al. Lancet 2008 372(9654) 1937

7 of 42

Population analysis for immunosuppressive agents

• Optimal dose is important to – Prevent acute rejection– Try and prolong graft (and patient) survival

• Case studies by population analysis– Mycophenolic acid (MPA)– Tacrolimus (TAC)– Sirolimus (SRL)

8 of 42

Population PK of mycophenolic acid in Chinese

Case Study 1

9 of 42

Background

Mycophenolic acid (MPA)

– Enterohepatic recirculation

– High protein binding ~97%

One dose fits all?

Clinical application shows

– Large inter-patient variability

– Narrow therapeutic range

– AUC 30~60 mg/L/h (Caucasian)Require lower dose for Chinese?

10 of 42

Objective

To characterize the PK of MPA in Chinese

– typical value, variability & covariates

– healthy subjects & transplant patients

To design individualized medication in Chinese renal

transplant patients

11 of 42

Study design

Classical PK study of MPA & MPAG

in Healthy Subjects

Classical PK study of MPA & MPAG

in RT patients（intensive sampling）

Pop PK study ofMPA in RT patients(intensive + sparse

sampling)

Individualized drug therapy

EnterohepaticCirculation Model

Limited SamplingStrategy

Multiple regression

Bayesian Method

Pop parameter

Individual parameter

Bioassay for MPA and MPAG

UGT1A9Genotyping

12 of 42

HPLC assay by online post-column alkalization

MPAG MPA

•Validated method based on FDA's guidance•Equivalent to LC/MS/MS in terms of sensitivity and accuracy, but more cost-effective

Jiao Z, et al. Chromatographia, 2005,62(7/8):363Shen J, et al. J Chromatogr B, 2005,817(2):207

13 of 42

UGT1A9 Genotyping

ID GC% Tm Sequence

p1f 25 49.5 TTT TCA ATT GTT CAT TGC TA

p1r 45 48.9 CTA CTC AAT GGA GGA CAA TC

p2f 50 51.8 AGC AGA CTG AGA GAG ACA AG

p2r 45 49.7 CAG ACA CAC ACA TAG AGG AA

Primer

PCR Reaction

98℃ 2min; 96℃ 30S, 60℃ 30S, 72℃ 45S; 94℃ 30S, 55℃30S, 72℃ 45S; 72℃ 7min; 4 ∞

10cycles (↓0.5℃/cycle) 35 cycles

Girard H et al. Pharmacogenetics. 2004; 14(8): 501

14 of 42

UGT1A9 Polymorphism

MutationSNP

Caucasian Chinese

-665 42% 0

-440 70% 4.8%

-331 70% 4.8%

-2152 5% 0-275 7% 0

Ref. Girard H. Pharmacogenetics. 2004; 14(8):501

Kypers DR. Clin Pharmacol Ther. 2005,78(4):351

Jiao Z, et al. Br J Clin Pharmacol, 2008,65(6):89315 of 42

Enterohepatic Circulation Model

Model Structure

2 recycling process within first 24hr was modeled

Jiao Z, et al. Br J Clin Pharmacol, 2008,65:89316 of 42

Mean time-concentration of MPA & MPAG profile (Observation v.s. prediction)

0.1

1

10

100

0 5 10 15 20 25

MPAG

MPA

Circle: observation; Line: prediction

Time (hr)

Con

cent

ratio

n (m

g/L)

Jiao Z, et al. Br J Clin Pharmacol, 2008,65:89317 of 42

Results

Present Study(Chinese)

Cremer S et al.(Caucasians)

CLMPA (L/hr/kg) 0.156 0.158

CLMPAG (L/hr/kg) 0.0211 (48%↑) 0.0144

EHC fraction (%) 29 25

Jiao Z, et al. Br J Clin Pharmacol, 2008,65(6):893

Cremer S, et al. Br J Clin Pharmacol 2005;60(3):249

18 of 42

0.001

0.010

0.100

1.000

10.000

100.000

1000.000

0 2 4 6 8 10 12Time (hr)

Classical PK study in RT patients (n=12)

AUC Present Study Korean Caucasian African American

Total MPA 27.0 ± 8.72 24.6 ± 5.67 33.1 ± 11.7 31.7 ± 19.6

Free MPA 0.98 ± 0.7 Not Available 1.13 ± 0.4 1.43 ± 0.6

Total MPAG 875 ± 198 Not Available 1732 ± 272 Not Available

Free MPAG 101* Not Available 177.2* Not Available

* Median

free MPA

Total MPA

Total MPAG

Free MPAG

Jiao Z, et al. EJCP. 2007, 63(1):2719 of 42

Limited sampling strategy

Algorithm– Multiple regression to estimate AUCAUC0-12＝ Intercept ＋M1×Ct1＋M2×Ct2＋……+Mi×Cti

Total MPAAUC0-12 ＝ 2.135+1.186·C1+1.217·C2+4.191·C4

Free MPAAUC0-12 ＝ 0.194+1.149·C1+0.819·C2+2.759·C4

Dose adjustment

Jiao Z, et al. EJCP. 2007, 63(1):2720 of 42

Population analysis

Intensive Sampling

– 18 patients (10, 30, 90 d post-operation)

– 24 Full PK profiles (11 sampling points)

Sparse Sampling

– 34 patients

– Randomly selected from C0, C1, C2 and C4

Jiao Z, et al. Chinese Pharmaceutical Journal. 2007, 63(1):2721 of 42

Final PPK model

⎟⎠⎞

⎜⎝⎛⋅−⎟

⎠⎞

⎜⎝⎛⋅−

⋅⋅⎟⎠⎞

⎜⎝⎛⋅= 100

253.032

858.0

3008.80)/(/

CLcrALB

eeCsAhLFCL

⎟⎠⎞

⎜⎝⎛⋅−

⋅= 3211.3

2 1240)(/ALB

eLFV

Jiao Z, et al. Chinese Pharmaceutical Journal. 2007, 63(1):2722 of 42

Covariates

20

30

40

50

50

100

150

200

20

30

40

20

30

40

50

20

30

40

50

50

100

150

200

10

15

20

20

30

40

50ALB ALB

CLcr CLcr

CL/FCL/F

CsA=150mg CsA=300mg

CsA dose↓ ，ALB ↑ ，CLcr ↑ CL/F ↓

23 of 42

Model Evaluation

Prediction Error Basic Model Final Model

MPE: Mean ± SD (95% CI) 0.357 ± 28.7 (-6.94~7.65) -0.184 ± 1.23 (-0.496 ~ 0.129)

MAE: Mean ± SD (95% CI) 2.24 ± 2.93 (1.50~2.99) 0.888 ± 0.863 (0.669~1.11)

Pred

ictio

n Er

ror (

mg/

L)

22 patients on 21d post-transplantation (1st Shanghai People’s Hospital)

24 of 42

Results

Mean (SE%) Shum B, et al (1) Van HR, et al (2) Present Study

Subjects 22 adults 140 adults 61 adults

Period (Mean) < 1 mo 0 -140 d 3-122 d (M: 19d)

CL/F (L/h)CL/F (L/h) 27.1 (20%)27.1 (20%) 33 (31%)33 (31%) 26.7 (43%)26.7 (43%)

V1/F (L) 98 (56%) 91 (91%) 50.2 (70%)V2/F (L) 206 (151%) 237 (102%) 187Q/F (L/h) 25.7 35 25ka (1/h) 2.27 4.1 1.33

Lag time (h) 0.145 0.21 0.268

1. Br J Clin Pharmacol 2003,56:1882. Clin Pharmacokinet. 2005, 44, 1083 25 of 42

Limited Sampling Strategy

2 Sampling：C2-C4

3 Sampling：not improved a lot

Combined with Bayesian method

DOSE = AUC × CL/F

26 of 42

Clinical Application

ALB(g/L)

CLcr(ml/min)

CsADose

(mg/d)

CL/F(L/h)

MMF Dose(mg bid)

Early phase 25 50 320 38.8 ~1000

Stable phase 32↑ 100↑ 300↓ 26.5↓ ~750↓

Target AUC: 30 mg·L/h

Jiao Z, et al. Chinese Pharmaceutical Journal. 2007, 63(1):27

27 of 42

Conclusion

• There was no obvious PK difference in MPA (but MPAG) between Chinese and Caucasians

• 1000 mg bid in the early phase– <~ 1 month

• 750 mg bid in the stable phase– >~ 1 month

• Limited sampling strategy: C2-C4• The findings of this study & clinical outcomes resulted in

the revision of clinical pathway of the hospital

28 of 42

Association of ABCB1, CYP3A4*18Band CYP3A5*3 Genotypes on the PK of Tacrolimus in Healthy Chinese: a Population PK Analysis

Case Study 2

J Clin Pharm Ther (in press)

29 of 42

Result:Non-compartmental approach

J Clin Pharm Ther (in press)

ANOVA after Logarithm transformation of AUC

30 of 42

Result:Population analysis by NONMEM

Carriers/non-carriersGenotypes

Non-CMT NONMEM

Contribution of totalinter-subject variability (%)

CYP3A4*18B 1.9 2 18.4

CYP3A5*3/*3 2 2 22.2

CYP3A4-CYP3A5 *1/*1-*3/*3 2.1 2.1 28.4

•Confirmed the results by non-compartmental analysis

•Estimated the contribution of the covariates of the total ISV

J Clin Pharm Ther (in press)31 of 42

Monte Carlo Simulation (n=2200)

0

10

20

30

40

200 400 600 800 1000 1400

CYP3A4*1/*1 CYP3A5*3/*3

200 400 600 800 1000 1400

CYP3A4-CYP3A5 *1/*1-*3/*3

CYP3A4*18B

200 400 600 800 1000 1400

CYP3A5*1/*3

0

10

20

30

40CYP3A4-CYP3A5 non-*1/*1-*3/*3

AUC

Per

cent

of T

otal

1200

12001200

inf

J Clin Pharm Thera (in press)32 of 42

Population PK-PD Modeling of Sirolimus and Its Impact on Lipids in Chinese Renal Transplant Patients

Case study 3

Jiao Z et al. BJCP. 2009 68:4733 of 42

Study Design and Patients

Non-randomized, open-label, multi-center, sequential 2 phases study

To evaluate the safety and efficacy

SRL：6～12 ng/mL

CsA：125～250 ng/mL

Phase I CsAPhase II SRL M

50～100 ng/mLonth 4~6 12～20 ng/mL; Month 7-12 10～20 ng/mL

0 0.5 1 2 3 6 9 12 Month

Phase I SRL + CsA↓+ Pred (n = 71)

Phase II: CsA ↓25% per week SRL + Pred in Month 4 (n = 41)

SRL + CsA + Pred(n=112)

Pre-dose Conc.

Jiao Z et al. BJCP. 2009 68:47 Study on sirolimus 34 of 42

Final Population PK Model

( ) ( )[ ]479.0

0 2661.065.010400417.066.5662.01.10/ ⎟

⎠⎞

⎜⎝⎛⋅⋅⋅−⋅−−⋅−=DDSCsACTCFCL GLZSLM

( )10427.73670/ 0 −⋅−= CsACFV

TC: total cholesterol levels (mmol/L);CsA C0: CsA trough concentration (ng/mL); DDS: SRL daily dose (mg): SLM=1 when a patient with hepatic impairment is receiving silymarin co-herapy, otherwise SLM = 0;GLZ＝1：1 when a patient with hepatic impairment is receiving glycyrrhizin co-therapy, otherwise GLZ = 0;


Silymarin → ~30%↓in CL

• Silymarin– hepatoprotection drug widely use around the world – a flavonolignan from the ‘milk thistle’– Added when patients have abnormal liver function

indices• Cotherapy with Silymarin → CL ↓

– Silymarin itself– Abnormal liver function– Silymarin + Abnormal liver function


Silymarin + liver impairment → CL ↓

• CYP3A4 & P-gp substrate– In vitro: a strong inhibitor but has high IC50 – In vivo: no effect

• Liver Impairment– ALT and other hepatoprotection medications did

not have an impact• Silymarin + Liver Impairment

– Liver Impairment → [silymarin] ↑→ SRL CL ↓

Schrieber SJ, et al. Drug Metab Dispos 2008; 36:1909Hawke RL, et al. J Clin Pharmacol. 2010 50(4):434Study on sirolimus

Total cholesterol(R)

Production (Kin) Outflow (Kout)

Indirect Response Model

Kin and Kout is daily production or outflow

Jiao Z et al. ASCPT. 2009

TCCCCsAC

dtdTC

SRL

SRLGEN ⋅+⋅

⋅⋅−−⋅=11.6311.0)

10000749.00926.0(917.0373.0 0

CsA deduction/elimination may be a good strategy for lowering total cholesterol

Study on sirolimus 38 of 42

Clinical application for initial dose

0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

0 2 4 6 8 10 12

Total cholesterol (mmol/L)

Sirolimus Daily Dose (mg)

CsA C0=100 ng/mL and not co-therapy with SRL

Css=10ng/ml

Css=15ng/ml

39 of 42

Summary

• To estimate relevant pharmacokinetic or PK/PD parameters •in the target patient population

• To identify and measure potential sources of variability•demographic & patho-physiological•environmental & genetic

• To quantify the unexplained variability•variability within an individual•measurement error, model misspecification, etc

40 of 42

Acknowledgement

Transplantation Center, Huashan Hospital– Dr. Ming Zhang, Dr. Jian-yongZhong & Dr.

Fu-ming Lu– All nurses and staffs

Clinical Pharmacy Laboratory– Prof. Ming-kang Zhong; Prof Xiao-jin Shi; – Dr Xian-yan Qiu

41 of 42

Thank you for your attention!

Population PK/PD modeling for Individualized drug therapy ... · Population PK/PD modeling for Individualized drug therapy: immunosuppressive agents . ... Case Study 1 9 of 42. ...

Documents