Current Statistical Considerations and Regulatory Perspectives on the Planning of Confirmatory Basket, Umbrella, and Platform Trials

Abstract

Master protocols have received a growing interest during the last years. By assigning patients to specific substudies, they aim at targeting and accelerating clinical development. Given their complexity, basket, umbrella, and platform designs have raised challenging regulatory and statistical questions, especially the control of multiplicity in confirmatory trials. In basket trials, regulatory assessment of the benefit/risk in pooled populations and choice of the treatment indication is challenging. We provide here our perspectives on these topics. In master protocols, as long as the statistical hypotheses tested between the different substudies are independent, no supplementary adjustment for multiplicity over the different substudies should be required. Moreover, sharing a control arm within an umbrella or a platform trial investigating different drugs would not require a correction for the type I error rate, whereas the chance of multiple false positive regulatory decisions should be recognized. In basket trials, pooling across substudies requires a rationale supporting the intended indication and should be preplanned. Assessment of the benefit/risk in pooled target populations can be complicated by differences in design or in efficacy/safety signals between the substudies. While trials governed by a master protocol can offer logistic and financial advantages, more experience is needed to gain a deeper insight into this novel framework.

Potential Statistical Issues Between Designers and Regulators in Confirmatory Basket, Umbrella, and Platform Trials

“…I appreciate and agree with Collignon et al, 1 on the description of the type I error allocated to each arm that enters a platform trial as being an appropriate metric and not the probability of a false positive in the larger platform trial (Master Protocol Wide Error Rate (MPWER)). I think it is appropriate that if three experimental arms enter a platform trial, they each have separate type I error calculations and the probability of at least one type I error of the three (MPWER) is not relevant.…”

“…Collignon et al 1 comment on a need for blinding across therapies-the idea that a patient randomized to one experimental arm would take the blinding agent for another experimental arm. As the authors point out, there would be many challenges with this-to the extent I do not think it is feasible.…”

“…Collignon et al 1 recommend an arm only use concurrent controls in a platform trial: "When investigating the effect of a particular treatment, one solid way to account for possible changes to the patient population over time is to restrict the control group to concurrent patients." I agree there are some settings in which this may be appropriate, but there are many trials in which there is a great deal of information about the control arm from previous and maybe even future control patients enrolled, while the particular arm was not enrolling.…”

“…I highlight some of these new issues and discuss the regulatory challenges emerging from these trials. I highlight some critical points made by Collignon et al, 1 while providing some conflicting views on these statistical challenges.…”

“…Collignon et al, 1 present an interesting, well-written, and clear paper on the emerging innovations of platform and basket trials that will help move the science forward. Their paper represents their personal views, from their perspectives, as the issues regulators face in the review of these innovative trial designs and structure.…”