Regulators have usually required at least two significant independent pivotal trials as substantial evidence of the effectiveness of a new drug. This standard requirement is known as the two-trial paradigm and has remained the conventional approach for decades.

However, to adhere to this standard rule, sponsors commonly design and conduct two identical trials. As an alternative approach, one might combine the data from the two trials into a single trial (one-trial paradigm) and obtain a higher power. It can be shown that this method would ensure the same level of type I error protection as the two-trial paradigm only under a specific scenario, but there is little investigation on the type I error protection over the whole null region.

In this talk, we compare the two-trial paradigm to the one-trial paradigm to better understand the regulatory decision-making in the assessment of drugs’ effectiveness, specifically what statistical errors the regulators are trying to protect against. We consider scenarios in which the two trials are conducted in identical or different populations as well as with equal or unequal size. With identical populations, the results show that a single trial provides better type I error protection and higher power. Conversely, with different populations, although the one-trial rule is more powerful in some cases, it does not always protect against the type I error.

Reference

Zhan, SJ, Kunz, CU, Stallard, N. Should the two-trial paradigm still be the gold standard in drug assessment? Pharmaceutical Statistics. 2023; 22(1): 96-111.

Regulatory guidance suggests controlling the family-wise error rate (FWER) in confirmatory clinical trials. The two-trial paradigm represents a further requirement to demonstrate efficacy in a clinical submission: A statistically significant outcome in at least two adequate and well-controlled clinical trials. Within each trial, different endpoints may require different sample size to achieve the adequacy of power. Sometimes the sample size driven by one endpoint could be twice as large as that required by other endpoints. These unbalanced requirements of resources in a single trial are amplified under the two-trial convention and may lead to financial and logistical challenges for the trial sponsor. It is, therefore, often of interest to combine the data from the two trials for an endpoint to make a confirmatory claim without doubling the sample size. However, it remains unclear what approaches could be used to manage multiplicity adjustments for the combined analysis using data from two identically designed trials. In this talk, we provide principles of controlling the submission-wise error rate (SWER) with combined analyses when success claims for endpoints tested in individual trials should be based on significance in both trials. We also discuss examples of SWER under other requirements where success claims could be based on significance from a single trial.

References

Bretz, F., Maurer, W., & Xi, D. (2019). Replicability, reproducibility, and multiplicity in drug development. Chance, 32(4), 4-11.

Bretz, F., & Xi, D. (2019). Commentary on “Statistics at FDA: Reflections on the Past Six Years”. Statistics in Biopharmaceutical Research, 11(1), 20-25.

In this talk, we present case studies that generated pivotal evidence using combined data from multiple pivotal trials in an overarching formal testing hierarchy. This goes beyond the traditional two-trials convention which requires independent pivotal evidence from (at least) two trials, separately. For each case study, we discuss the situation and rationale, the approach taken with its advantages and caveats, any experiences with health authorities, and the final outcomes such as resulting label claims.