Article
When and how to adjust for multiplicity in platform trials?
Search Medline for
Authors
-
Franz Koenig
- Medical University of Vienna, Vienna, Austria
-
Frank Bretz
- Novartis Pharma AG, Basel, Switzerland
-
Elias Meyer
- Medical University of Vienna, Vienna, Austria
-
Martin Posch
- Medical University of Vienna, Vienna, Austria
| Published: | February 26, 2021 |
|---|
Outline
Text
Background: In recent years there has been the rise of more flexible clinical trial designs allowing the investigation of several sub-studies within a single clinical trial platform. These types of designs are labelled with the term “master protocol”. Such trials can have both exploratory and confirmatory elements (e.g., see [1]). Especially for the latter false positive decisions should be controlled at a certain level. Due to the complex trial design with various sub-studies allowing for multiple treatment arms, subgroups and adaptations, it is not clear when and how to adjust for multiplicity. It is not clear which type I error rates should be controlled in the first place (e.g. the family-wise error rate (FWER), the per-comparison error rate, ...)? For example Wason et al. [2] reviewed papers published in 2012 and showed there was no consensus on whether multiplicity corrections are needed in a multi-arm trials with different treatments.
Results: First, we will give an overview on which multiplicity adjustment strategies and methods have been proposed for master protocols (see e.g., [3], [4]). We will identify the main sources for multiplicity issues arising in platform trials such as multiple treatment arms, multiple (overlapping) subgroups and adaptations (sample size re-assessment, dropping/adding of treatments and/or subgroups, modification of endpoints).
Especially for confirmatory parts the strict control of the type I error rate in pre-defined families of hypotheses is paramount regardless which adaptations are eventually performed, e.g., see Bauer et al. [5]. We will discuss error rate definitions beyond classical ones such as the “population-wise error rate” by Hillner et al. [6] in the context of integrated platform trials. Another complexity is how to incorporate non-concurrent data in the analysis, because if data are naively used, this can lead to a substantial inflation of the type I error rate (see e.g., Eichler et al. [7] ) and other biases (see e.g. Schmidli et al. [8]).
The authors declare that they have no competing interests.
The authors declare that an ethics committee vote is not required.
References
- 1.
- Meyer EL, Mesenbrink P, Dunger-Baldauf C, Fülle HJ, Glimm E, Li Y, Posch M, König F. The Evolution of Master Protocol Clinical Trial Designs: A Systematic Literature Review. Clin Ther. 2020 Jul;42(7):1330-1360. DOI: 10.1016/j.clinthera.2020.05.010
- 2.
- Wason JM, Stecher L, Mander AP. Correcting for multiple-testing in multi-arm trials: is it necessary and is it done? Trials. 2014 Sep;15:364. DOI: 10.1186/1745-6215-15-364
- 3.
- Stallard N, Todd S, Parashar D, Kimani PK, Renfro LA. On the need to adjust for multiplicity in confirmatory clinical trials with master protocols. Ann Oncol. 2019 04;30(4):506-509. DOI: 10.1093/annonc/mdz038
- 4.
- Collignon O, Gartner C, Haidich AB, James Hemmings R, Hofner B, Pétavy F, Posch M, Rantell K, Roes K, Schiel A. Current Statistical Considerations and Regulatory Perspectives on the Planning of Confirmatory Basket, Umbrella, and Platform Trials. Clin Pharmacol Ther. 2020 May;107(5):1059-1067. DOI: 10.1002/cpt.1804
- 5.
- Bauer P, Bretz F, Dragalin V, König F, Wassmer G. Twenty-five years of confirmatory adaptive designs: opportunities and pitfalls. Stat Med. 2016 Feb;35(3):325-47. DOI: 10.1002/sim.6472
- 6.
- Hillner C, Brannath W, Rohmeyer K. Group sequential and adaptive designs with control of the “population-wise error rate”. In: Workshop Adaptive Designs. Münster; 2019.
- 7.
- Eichler HG, Bloechl-Daum B, Bauer P, Bretz F, Brown J, Hampson LV, Honig P, Krams M, Leufkens H, Lim R, Lumpkin MM, Murphy MJ, Pignatti F, Posch M, Schneeweiss S, Trusheim M, Koenig F. “Threshold-crossing”: A Useful Way to Establish the Counterfactual in Clinical Trials? Clin Pharmacol Ther. 2016 Dec;100(6):699-712. DOI: 10.1002/cpt.515
- 8.
- Schmidli H, Häring DA, Thomas M, Cassidy A, Weber S, Bretz F. Beyond Randomized Clinical Trials: Use of External Controls. Clin Pharmacol Ther. 2020 Apr;107(4):806-816. DOI: 10.1002/cpt.1723
