Pavan Vaddady, Pratik Bhagunde, Ming Xu, Alok Maniar, Luke F. Chen, Amanda Paschke, Matthew Rizk
Merck & Co., Inc., Kenilworth, NJ, USA
Objectives:
A fixed-dose combination of imipenem/cilastatin (IMI) and relebactam (REL), with a 2:1 IMI:REL ratio, is currently developed for pediatric populations to combat severe Gram-negative bacterial infections. The goal of this work is to develop a pediatric population pharmacokinetic (Pop PK) model based on sparse imipenem and REL pediatric pharmacokinetic (PK) data in different age cohorts while leveraging prior adult information to identify dosing regimens that optimize imipenem and REL pharmacokinetic/pharmacodynamic (PK/PD) and safety target attainment for continued investigation.
Methods:
PK data from three different age cohorts (Cohort 1: 12 to < 18 yr, Cohort 2: 6 to < 12 yr and Cohort 3: 2 to < 6 yr) comprising of six patients per cohort was available. For each patient, three PK samples were drawn at predefined time intervals following intravenous administration of single dose of study drug. Leveraging a Bayesian framework, a pediatric Pop PK model comprising of both imipenem and REL components was developed using adult Pop PK parameters and respective covariate relationships as priors. A multivariate normal prior was used for fixed effect parameters and an inverse-Wishart prior was used for both inter-individual variability and residual variability parameters. Allometric scaling was used to describe the differences in clearance and volumes from adult population across pediatric cohorts. Standard allometric exponent of 0.75 was added to body weight (WT) on CL and inter-compartmental clearance (Q), while allometric exponent of 1 was added to WT on central volume (V1) and peripheral volume (V2) respectively.
Virtual pediatric populations were created using 2011 to 2016 National Health and Nutrition Examination Survey (NHANES) data[1], and published serum creatinine distributions[2] comprising of age, gender, weight, height, and serum creatinine measures. Simulations were run using these virtual populations (N=2000 per age cohort) to evaluate the probability of target attainment (PTA) for PK/PD targets (imipenem: %fT>MIC ≥ 30%, REL: AUC0-24h ≥ 38.5 μM.hr) and safety targets (imipenem: AUC0-inf ≤ 216.5 μM.hr, Cmax ≤ 161 μM, REL: AUC0-inf ≤ 190 μM.hr, Cmax ≤ 118 μM). Different dosing regimens, with imipenem dose capped at 500 mg and REL dose capped at 250 mg, were evaluated and a dosing regimen where the predefined targets are jointly achieved that have at least 90% PTA for each age cohort were chosen.
Results:
A two compartment pediatric Pop PK model comprising both imipenem and REL components leveraging priors from adult Pop PK described the observed pediatric data adequately. Priors for key PK parameters CL and V1 were uninformative and priors on Q were informative. Based on the adult Pop PK model, both creatinine clearance (CRCL) and WT on CL, and WT on V1 were chosen as covariates for imipenem; CRCL on CL and WT on V1 were chosen as covariates for REL. Based on the PTA simulations and the aforementioned criteria, a fixed dose of 250 mg REL as an 1 hr infusion was chosen for cohort 1 and a weight-based dose of 7.5 mg/kg REL as 1 hr infusion was chosen for cohorts 2 and 3.
Conclusions:
A comprehensive modeling and simulation based approach was successfully developed to recommend final doses in cohorts 1, 2 and 3 for the upcoming efficacy study. Leveraging Bayesian approach with prior information from the adults aided in the pediatric model development despite sparse PK data in the pediatric study.
References:
[1] Centers for Disease Control and Prevention (CDC). National Center for Health Statistics (NCHS). National Health and Nutrition Examination Survey Data. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention [https://wwwn.cdc.gov/nchs/nhanes/Default.aspx].
[2] Savory DJ. Ann Clin Biochem (1990) 27, 99-101.
Reference: PAGE 28 (2019) Abstr 8876 [www.page-meeting.org/?abstract=8876]
Poster: Drug/Disease Modelling - Paediatrics