Statistical methodologies for absolute and relative efficacy assessment based on single-arm trials: a scoping review

Regulatory decision-making on the marketing authorisation of Advanced Therapy Medicinal Products (ATMP) is challenging, in particular since the evidence that is provided in the marketing authorization application is frequently not from randomized controlled clinical trials, but based on single-arm trials (SAT). Additionally, if a conditional marketing authorisation is applied, cross-trial comparisons are often necessary against other authorised treatments in the same indication. While various biostatistical solutions to these problems have been proposed recently, a comprehensive overview of the available methods is missing. Therefore, our aim is to provide a general overview of statistical methods that have been proposed to support efficacy claims in marketing authorization applications mainly based on non-randomised evidence. We carefully developed a systematic search strategy, which initially yielded 63,671 results. Finally, following predefined in- and exclusion criteria, the methodologies from 120 papers were summarized in this review, followed by a discussion of their potential relevance for regulatory decision making, as well as promising future directions of biostatistical research. There is indeed a broad range of different methodological approaches available, but hardly any systematic empirical comparisons of these methods exist. Therefore, biostatisticians should be encouraged to systematically generate such comparative evidence, in order to allow for subsequently formulating recommendations regarding which methods are appropriate for supporting efficacy claims in the approval of medicines for ATMPs.