Background: Bridging study is a concept for extrapolating information gathered from clinical study in an original region, e.g. an adult patient population, to a new region, e.g. a paediatric patient population. Such studies generally include additional pharmacokinetic information to provide knowledge of the time course of exposure in the new region. Since systematic differences are possible between the original and new region then optimally designed studies based solely on the original region will be suboptimal when applied to the new region.
Aim: To develop an optimal adaptive design method for bridging studies.
Methods: The adaptive bridging study was designed and assessed using simulated data. Two hundred adult patients and twenty five paediatric patients were simulated in this research. The paediatric patients were divided into five batches with five patients in each batch. Data were simulated to follow a Bateman pharmacokinetic model. For adult patients: dose = 100mg, CL = 4Lh-1, V = 20L, Ka = 1h-1 and for paediatric patients: dose = 28.57mg, CL = 1.56Lh-1, V = 5.71L, Ka = 1h-1. The variance of the log-normal between subject variability was 0.1 for both populations. A combined residual error model was assumed. The two hundred adult patients each provided 6 blood samples following an empirical sampling schedule. The adult patient data was fitted using NONMEM and the estimated parameter values obtained was used to design the optimal sampling schedule for the first batch of five paediatric patients. The data from forty adult patients was then replaced by the data from five paediatric patients and a combined estimation was performed using NONMEM. The estimated parameter values were then used to design the optimal sampling time for the second batch of five paediatric patients. The process of estimation and design is repeated till the last batch of paediatric patients. Using this method the population Fisher information matrix for the optimal design (MF) is a combination of population Fisher information matrix from adult patient (AF) and population Fisher information matrix from paediatric patient (PF). MF = alpha(AF) + PF where alpha is the proportion of information from the adult patients used in the optimal design. We have considered fixed reduction of alpha at the current stage of this research.
Results: Two hundred patients with 6 samples per patient provided precise parameter estimates for the adult data set. The adaptive design with fixed reduction of alpha (20% per iteration) provided precise parameter estimates for the paediatric population at the 5th (final) iteration.
Conclusion: Optimal adaptive designs for bridging studies are a useful method for learning about new populations.