Generating Control Groups for Organ Impairment Studies: A Case-Study Comparing Statistical and Population...

Area of Science:

• Pharmacokinetics
• Clinical Pharmacology
• Drug Development

Background:

• Phase 1 studies assessing drug pharmacokinetics in organ impairment face challenges in recruiting demographically matched control groups.
• This difficulty can hinder the accurate evaluation of drug exposure and safety profiles in patient populations with hepatic or renal impairment.

Purpose of the Study:

• To evaluate alternative methods for generating control groups in phase 1 organ impairment studies.
• To assess the feasibility of statistical matching and population pharmacokinetic modeling approaches for creating virtual control groups.

Main Methods:

• Retrospective analysis of phase 1 data from upadacitinib and elagolix clinical programs.
• Evaluation of a statistical matching approach (genetic matching with Mahalanobis distance, 3:1 k-match) and a population pharmacokinetic (PopPK) model-based approach.
• Comparison of drug exposure (Cmax, AUCinf) between organ impairment groups and matched/virtual control groups using geometric mean ratios.

Main Results:

• The statistical matching approach achieved adequate demographic balance, except for age, and minimized prediction error.
• The PopPK model-based approach yielded results generally comparable to in-study controls.
• Differences in geometric mean ratios between alternative approaches and in-study controls were within acceptable ranges for both case studies.

Conclusions:

• Both statistical matching and population pharmacokinetic modeling are demonstrated as viable alternative strategies for generating control groups in phase 1 organ impairment studies.
• These methods can potentially overcome recruitment challenges and facilitate the assessment of drug pharmacokinetics in diverse patient populations.
• Further validation may support the broader adoption of these approaches in clinical drug development.