Adaptive choice of patient subgroup for comparing two treatments

Abstract

This paper is motivated by a randomized controlled trial to compare an endovascular procedure with conventional medical treatment for stroke patients, in which the endovascular procedure may be effective only in a subgroup of patients. Since the subgroup is not known at the design stage but can be learned statistically from the data collected during the course of the trial, we develop a novel group sequential design that incorporates adaptive choice of the patient subgroup among several possibilities which include the entire patient population as a choice. We define the type I and type II errors of a test in this design and show how a prescribed type I error can be maintained by using the closed testing principle in multiple testing. We also show how asymptotically optimal tests can be constructed by using generalized likelihood ratio statistics for parametric problems and analogous standardized or Studentized statistics for nonparametric tests such as Wilcoxon’s rank sum test commonly used for treatment comparison in stroke patients.

A multicenter randomized controlled trial of endovascular therapy following imaging evaluation for ischemic stroke (DEFUSE 3)

“…18 The adaptive design, developed specifically for DEFUSE 3, is based on closed testing theory and the group sequential methods for the generalized likelihood ratio (GLR) statistic developed by Lai and Shih. 19 …”

Power of an Adaptive Trial Design for Endovascular Stroke Studies

“…6, 7 Briefly, in the case that the trial adapts its patient selection criteria midway, subsequent recruitment and the primary efficacy analysis are limited to patients who are part of the selected subgroup. Subjects who have already been enrolled but are not part of this subgroup are excluded from the primary efficacy analysis.…”

“…4, 5 Our group developed a novel adaptive trial design that restricts selection criteria to the subgroup with the best chance of showing a treatment effect if the results of an interim or final analysis demonstrate futility in the overall sample . 6, 7 In the event that the enrollment criteria are restricted (adapted), both subsequent recruitment and the population in which the primary hypothesis is tested are limited to the selected subgroup. (Figures 1 and 2)…”

Design of clinical trials with failure-time endpoints and interim analyses: An update after fifteen years

“…In fact, Section 4.3 of [37] and the recent paper [55] have developed such designs for comparing a new method against standard medical care for stroke patients. The endpoint of that trial is the Rankin score, which is much easier to handle than the censored failure-time endpoint in survival trials.…”

“…The endpoint of that trial is the Rankin score, which is much easier to handle than the censored failure-time endpoint in survival trials. Lai, Liao and Tsang have recently extended the approach in [55] to survival outcomes, and the method and results will be presented elsewhere. An important innovation of their work is that unlike [51] and [54] which rely heavily on the logrank statistic, a more flexible and powerful statistic of the type Eq.…”