Performance of Control Arms in Trials Leading to FDA Cancer Drug Approvals

Background: Selecting an appropriate control arm is essential when evaluating the efficacy of new anticancer therapies. Methods: We analyzed the progression-free survival (PFS) performance of control arms in trials leading to FDA drug approvals between January 2014 and December 2023 for the treatment of advanced solid cancers. For each phase III or double-arm phase II trial, we extracted (1) the assumed median PFS (amPFS) value for the control arm, as specified in the study protocol for sample-size calculation, and (2) the observed median PFS (omPFS) and its confidence interval (CI) for that arm. Control arms were categorized as underperforming when the amPFS was greater than the upper bound of the omPFS CI, and as outperforming when the amPFS was lower than the lower bound of the omPFS CI. All remaining cases were classified as “within the range.” Results: We identified 72 trials that led to new cancer drug approvals. Overall, control arms underperformed in 24 of 72 (33.3%) trials and outperformed in 8 of 72 (11.1%), while the remaining 40 (55.5%) fell “within the range.” Notably, control arms underperformed in 12 of 25 (48.0%) trials leading to approvals of immune checkpoint inhibitors and in 7 of 25 (18.9%) trials leading to approvals of targeted therapies. By tumor type, underperformance occurred in 6 of 14 (42.9%) lung cancer trials, 5 of 16 (31.3%) breast cancer trials, and 5 of 16 (31.3%) genitourinary cancer trials. In 18 of the 24 (75.0%) underperforming control arms, the discrepancy between amPFS and omPFS exceeded 20%. Among the 18 trials with underperforming control arms that received expedited approval, a final overall survival advantage was not demonstrated in 7 (38.9%) cases. Conclusions: The PFS benefit observed in several trials leading to FDA approval of new anticancer drugs may have been influenced by shorter-than-expected PFS in the control arms. Accurate analysis of control-arm outcomes in clinical trials is therefore essential to ensure an accurate and comprehensive assessment of the efficacy of experimental therapies.