I ntroduction to population PKPD modelling modelling I ntroduction - PowerPoint PPT Presentation

I ntroduction to population PKPD modelling modelling I ntroduction to population PKPD in paediatric paediatric clinical pharmacology clinical pharmacology in Catherijne Knibbe, Oscar , Oscar Della Pasqua Della Pasqua, , Meindert Danhof Meindert Danhof Catherijne Knibbe Leiden/ Amsterdam Center for Drug Research Division of Pharmacology 1

What is the problem? What is the problem? • Drugs dosing in children is largely empirical • Frequent under-and overdosing problems • Efficacy and safety of drugs, in particular in (premature) newborns is largely unknown Body weight is used for dose adjustment Body weight is used for dose adjustment instead of the PKPD relationships instead of the PKPD relationships 2

PKPD MODELLI NG: What is it? PKPD MODELLI NG: What is it? 3

Clinical Questions Clinical Questions • How to identify a safe and effective dosing regimen in children in different age groups? – First time in kids (early drug development) – Change in indication or age group, including neonates (clinical practice) • Which factor(s) should be used to adjust the dose for the individual child in different age groups? – dosing recommendation in the label 4

Paediatric Research I ssues Research I ssues Paediatric Unbalanced Unbalanced vs vs balanced balanced designs: designs: – 100 observations for subject A 100 observations for subject A – 1 observation for subject B 1 observation for subject B Sparse vs. serial Sparse vs. serial data: data: – 2 measurements per subject 2 measurements per subject 5

Population approach Population approach Simultaneous analysis of all available data: PK and/or PD parameters are simultaneously estimated taking into account differences between patients 120 1. POPULATION PK and/or PD 100 parameters (fixed effects) 80 2. Inter-individual variability 60 40 3. Residual error 20 0 0 6 12 18 24

Population PKPD modelling modelling Population PKPD ID=1 (pred) 120 120 Concentration (mg/L) ID=1 (obs) Predicted 100 100 ID=2 (pred) Observed 80 ID=2 (obs) 80 ID=3 (pred) 60 Residual error 60 ID=3 (obs) 40 40 20 Inter- 0 20 individual 0 6 12 18 24 0 variability TIme (hr) 0 6 12 18 24 7

Population PK/ PD modelling modelling Population PK/ PD • Applicable to sparse and unbalanced data sets (neonates, children, etc) • Scientific basis for study/trial simulations, • Scientific basis for study/trial simulations, Scientific basis for study/trial simulations, • dose adjustment or labeling extensions in dose adjustment or labeling extensions in dose adjustment or labeling extensions in other populations other populations (intra and interspecies) other populations (intra and interspecies) (intra and interspecies) • Covariate analysis for identification of • Covariate analysis for identification of Covariate analysis for identification of • predictors of variability in PK and PD predictors of variability in PK and PD predictors of variability in PK and PD (genetics, body weight, age, interactions etc) (genetics, body weight, age, interactions etc) (genetics, body weight, age, interactions etc) 8

Ventilated children (1 (1- -5 5 yrs yrs) ) following following Ventilated children cardiac surgery in the I CU in the I CU cardiac surgery Best 6 samples of 250 ul ul per child per child 6 samples of 250 2 2 6 children 6 children 1 1 Children Children Adults Adults 0 0 0 200 400 600 Cl (l/min) Cl (l/min) 0 200 400 600 2.3 2.3 Propofol concentration (mg/l) Median (ml/kg/min) (ml/kg/min) 35* 35* 28* 28* 2 2 V1 (l) V1 (l) 12 12 21 21 (l/kg) (l/kg) 0.78* 0.78* 0.26* 0.26* 1 1 Q (l/min) Q (l/min) 0.35 0.35 1.4 1.4 0 0 0 200 400 600 0 (l/kg/min) (l/kg/min) 200 400 600 23 23 18 18 Worst V2 (l) V2 (l) 24 24 139 139 2 2 (l/kg) (l/kg) 1.54 1.54 1.88 1.88 1 1 0 0 0 200 400 600 0 200 400 600 9 Knibbe et al., Br J Clin Pharmacol 2002

Population PKPD modelling modelling Population PKPD • Applicable to sparse and unbalanced data sets (neonates, children, etc) • Scientific basis for study/trial simulations, dose adjustment or labeling extensions in other populations (intra and interspecies) • Covariate analysis for identification of • Covariate analysis for identification of Covariate analysis for identification of • predictors of variability in PK and PD predictors of variability in PK and PD predictors of variability in PK and PD (genetics, body weight, age, interactions etc) (genetics, body weight, age, interactions etc) (genetics, body weight, age, interactions etc) 10 10

Propofol in non in non- -ventilated children ventilated children Propofol 2.0 Knibbe observed conc Propofol concentration (mg/l) Rigby-Jones Schuttler 1.5 current study 1.0 0.5 0.0 1200 time (min) 0 200 400 600 800 1000 11 Peeters MYM et al., Anesthesiology 2006 ; 104(3):466-474

Propofol in in nonventilated nonventilated children children Propofol 2.0 Knibbe observed conc Propofol concentration (mg/l) Rigby-Jones Schuttler 1.5 current study Peeters MYM et al., Anesthesiology 2006 Mar; 1.0 104(3):466-474 0.5 0.0 1200 time (min) 0 200 400 600 800 1000 12

COMFORT-B 6 behaviour items • alertness • Calmness/agitation • Respiratory response / crying • Physical movement • Muscle tone • Facial tension 13

Non- -agitated children agitated children Non non-agitated, median performance 30 Propofol concentration (mg/l) 1.0 24 0.8 20.00 h 07.00 h COMFORT-B 0.6 18 0.4 12 0.2 no propofol 6 0 0 200 400 600 800 1000 1200 1400 time (min) 14 Peeters et al., Anesthesiology, March 2006

Agitated children Agitated children B) agitated, median performance 30 Propofol concentration (mg/l) 1.0 24 0.8 COMFORT-B 0.6 18 0.4 24 mg/h 12 8.6 kg 0.2 start propofol 18 mg/h 6 0 0 200 400 600 800 1000 1200 1400 time (min) propofol propofol 15 Peeters et al., Anesthesiology, March 2006

Model based advised propofol propofol dose dose Model based advised 30 mg/h for a postoperative child of 10 kg 30 mg/h for a postoperative child of 10 kg Propofol concentration (mg/l) 1.0 22 0.8 COMFORT-B 18 0.6 0.4 14 0.2 10 0.0 6 0 200 400 600 800 1000 1200 1400 time (min) 16 Peeters et al., Anesthesiology, March 2006

Population PKPD modelling modelling Population PKPD • Applicable to sparse and unbalanced data sets (neonates, children, etc) • Scientific basis for study/trial simulations, dose adjustment or labeling extensions in other populations (intra and interspecies) • Covariate analysis for identification of predictors of variability in PK and PD (genetics, body weight, age, interactions etc) 17 17

Body weight or age? Body weight or age? 15000 10000 BWS 5000 0 0 400 800 1200 PCA age age 18

Covariate analysis Covariate analysis Identification of potential covariates Identification of potential covariates (body weight, gender, age, renal function, PGx etc). Graphical evaluation of each covariate versus • The individual post-hoc PK or PD parameter estimate • the weighted residuals Statistical evaluation using standard techniques Statistical evaluation using standard techniques 1. Change in objective function 2. Standard error of the additional parameter 3. Improvement of individual fits 4. Diagnostics: B) observed versus model-predicted Krekels et al, Expert Opin. Pharmacother. (2007) 8 (12):1787-1800 19 19 Peeters MY et al., Anesthesiology, March 2006 and Dec 2006, CP&T March 2008

Covariate analysis Covariate analysis When more than one significant covariate for the simple model is found, the covariate-adjusted model with the largest decrease in objection function is chosen as a basis to explore the influence of additional covariates sequentially with the use of the same criteria Forward inclusion and backward deletion Krekels et al, Expert Opin. Pharmacother. (2007) 8 (12):1787-1800 20 20 Peeters MY et al., Anesthesiology, March 2006 and Dec 2006, CP&T March 2008

Covariate analysis Covariate analysis Nature of the influen Nature of the influence of t ce of the covariat he covariate – preferably non-empirical (mechanism/physiologically based) – Consider the possibility of potential extrapolation or interpolation Validation Validation confirms the influence of the covariates Krekels et al, Expert Opin. Pharmacother. (2007) 8 (12):1787-1800 21 21 Peeters MY et al., Anesthesiology, March 2006 and Dec 2006, CP&T March 2008

Morphine PK in PK in Children Children Morphine • 250 children: – 70 premature neonates, 70 premature neonates, 1.2 – 60 neonates, 60 neonates, Clearance – 60 < ½ 60 < ½ yr, r, 0.8 – 30 < 1 yr, 30 < 1 yr, CL – 30 < 3 yr 30 < 3 yr 0.4 • 1-4 samples/24 h/pt • BW median 2.8 kg 0.0 500 3000 5500 8000 10500 13000 15500 18000 Body weight BWS Supported by a grant of the Sophia Stichting voor 22 Wetenschappelijk Onderzoek

I nfluence of post natal age > / < 10 d I nfluence of post natal age > / < 10 d 0.5 0.0 ET1 -0.5 Independent of gestational -1.0 time or body weight at birth 10 -1.0 2 3 4 5 6 10 0.0 2 3 4 5 6 10 1.0 2 3 4 5 6 10 2.0 2 3 4 5 6 10 3.0 2 3 23 PNA