Model-based approaches to prospectively power pediatric pharmacokinetic trials with limited sample size

• Rare disease studies in pediatric subjects are challenging due to small sample sizes. Pharmacokinetic (PK) information in pediatric subjects is important and often used for matching strategy towards adults informing pediatric development program. Prior to studying PK in children, it is important to optimize the sparse sampling schedule and show that the study is designed to estimate key PK parameters with sufficient certainty. In this work, the sampling schedule in children was optimized for marzeptacog alfa activated (MarzAA) and dalcinonacog alfa (DalcA), two drugs in development for treatment of hemophilia. Subsequently, evaluation of different model-based approaches to calculate the power to estimate clearance (CL) and volume of distribution (V) using a fixed sample size (n=24) was performed. Usage of Bayesian priors (up to 2x inflation of the adult priors) performed well (power   80 %), but with lower power with decreasing informativeness (5x and 10x inflation of the adult priors), in particular for DalcA. Reusing the full adult model or a simplified model for standalone analysis of the pediatric data did not perform well (<80% power). Fixing the adult PK parameters except for CL and V performed well when pooling adult and pediatric data (power 100 %). In general, the power to estimate V alone or CL together with V was lower than for CL, indicating that the sampling schedules were more informative for CL. Although Bayesian prior approaches were shown to perform well without need of pooling data, other approaches that require less technical expertise and no need for simplification of the adult model were found to be good alternatives when pooling of data is possible. 

Ämnesord

MEDICIN OCH HÄLSOVETENSKAP  -- Klinisk medicin -- Pediatrik (hsv//swe)

MEDICAL AND HEALTH SCIENCES  -- Clinical Medicine -- Pediatrics (hsv//eng)

Publikations- och innehållstyp

vet (ämneskategori)

ovr (ämneskategori)