Model-Based Elaboration of a Limited Sampling Strategy in the Bioequivalence Assessment of Highly Variable Dabigatran

Background: The bioequivalence (BE) assessment of generic (Test) and brand name (Reference) formulations of drugs with steep exposure-response relationships exhibiting high pharmacokinetic (PK) variability such as dabigatran represent an expensive challenge for pharmaceutical companies. Supported by the population pharmacokinetics (pop-PK) approach, the present article investigates modelling potential to assess BE using a reduced number of blood samples. Methods: Pop-PK models for the Reference and Test formulations were developed retrospectively using standard modeling techniques for a BE study of dabigatran. Reduced sampling scenarios were selected and the developed pop- PK models were refitted on each dataset for the respective formulations. These models were simulated to generate virtual PK profiles to be tested with the standard BE criteria, in order to identify the scenarios maintaining the original BE conclusions with the least samples required. Results: The BE study original data was best described as a pop-PK model presenting two compartments with first order elimination and absorption, as well as an absorption lag time. Sex was identified as a significant covariate with impact on bioavailability. Using a rational sampling selection procedure under the framework of modeling and simulation, the results proved that the BE verdict could be maintained with only five of the 20 original blood samples using the current regulatory BE standards and criteria. Conclusion: We conclude that the pop-PK model-based BE assessment can be an efficient tool for aiding the BE assessment of dabigatran by significantly reducing the number of samples required, and consequently lower trial costs and increase benefits for enrolled participants.