Operating Characteristics of the Simulated Healthy Participant Approach in Impaired Clearance Studies

Minimizing harm is a cornerstone of ethical research practices. A drug that has undergone extensive clinical pharmacological testing in healthy participants (HPs) and a diverse selection of patients can be described with a sufficiently predictive population pharmacokinetic (PopPK) model. In impaired clearance trials, recruitment is minimized and underpowered for all but major exposure differences. Virtual HP arms have been reported to support similar conclusions to conventional impaired clearance studies, and further minimize potential harm of drug exposure without medical benefit by eliminating an arm of the study. However, the extent to which the conventional analysis of impairment studies compare to the simulation approach is unknown. Here we assess the operating characteristics of the virtual cohort approach along with the conventional approach through controlled simulations. These simulations included a simple, widely accessible PopPK model and several internal models that have been used in a previous meta-analysis of the virtual cohort approach. In the pairwise comparisons assessed, the virtual cohort simulation approach had greater power per sample size than the conventional approach and the same power under the null hypothesis. Given key methodological differences, it is recommended that the simulation and conventional approaches be treated as having approximately the same power under equivalent conditions. These results provide a strong justification for the use of the virtual cohort approach when an adequate PopPK model is available, minimizing unnecessary exposure to study drugs that will not benefit healthy study participants.