PhD Candidate University of Minnesota Minneapolis, Minnesota, United States
Background: Clinical studies that assess the impact of renal and hepatic impairment on pharmacokinetics typically recruit participants with different degrees of organ impairment and a control group with no impairment. The control group should be similar to patients with organ impairment in age, weight, gender and other factors that can potentially affect pharmacokinetics. However, it can be challenging to recruit healthy participants who match desired characteristics to all participants with organ impairment. This work aimed to assess alternative matching approaches for control groups in organ impairment studies which leverage available Phase 1 data. Methods: Two approaches were assessed using upadacitinib and elagolix as model drugs. Approach 1: Control groups were generated in R using 3 different statistical matching algorithms: propensity score matching, Mahalanobis distance matching, and genetic matching that utilizes Mahalanobis distance, to match organ impairment patients based on age, sex, weight, and race. Approach 2: A population pharmacokinetics model was developed in NONMEM for each drug and was used to simulate pharmacokinetics in a virtual control group that matches the characteristics of the organ impairment patients. The ratios of Cmax and AUCinf for each drug in organ impairment patients relative to the control group were calculated for each approach and compared to the in-study results. Results: The virtual control group generated using Approach 1 provided a closer match to organ impairment patients than the in-study control groups for all key characteristics, except for age. The difference between virtual controls and in-study controls in estimating exposure change in organ impaired patients ranged from -28% to 40% for Approach 1 and -30% to 41% for Approach 2 across both drugs. Both approaches are expected to result in the same dosing recommendations as the use of in-study controls for both drugs. Conclusion: Existing data from healthy participant studies can potentially be leveraged to complement or replace control groups in organ impairment studies. Relaxing the maximum age limit for healthy participants in Phase 1 studies can increase the ability to accurately match organ impairment participants using virtual controls.
