Exposure-response analysis for time-to-event data in the presence of adaptive dosing: efficient approaches and pitfalls

MIDD

Trial design and analysis

Methodology

CPT Pharmacometrics Syst Pharmacol. 2025 Nov 20. Online ahead of print.

Lavalley-Morelle A, Le Louedec F, Anziano RJ, Mentré F, Bergstrand M

A critical challenge in pharmacometrics is how to adequately and confidently analyze exposure-response (E-R) relationships for time-to-event data, especially in complex studies involving adaptive dosing, such as down-titration for improved tolerability, a practice prominent in oncology and commonly applied in other therapeutic areas.

By using simulations to compare time-static vs. time-varying metrics across scenarios (adaptive vs. fixed dosing, drug accumulation, event type, and onset, the authors demonstrate a significant pitfall: no time-static metric produces consistently reliable results across all conditions.

The authors provide guidance for confident assessment, recommending the use of time-varying exposure for adequate statistical power and to control type-I error – a methodology demonstrating robust performance across all scenarios.

Research of this nature validates and refines the very tools used to measure a drug's actual effect - a critical link in translating complex data into confident decisions, enabling more efficient clinical trial designs and reducing the risk of late-stage failures and, most importantly, better patient outcomes in the long run.