Representation of women in clinical trials supporting FDA‐approval of contemporary cancer therapies

Abstract

AbstractContemporary anticancer therapies frequently have different efficacy and side effects in men and women. Yet, whether women are well‐represented in pivotal trials supporting contemporary anticancer drugs is unknown. Leveraging the Drugs@FDA database, clinicaltrials.gov, MEDLINE, and publicly available FDA‐drug‐reviews, we identified all pivotal (phase II and III) non‐sex specific trials supporting FDA‐approval of anticancer drugs (1998–2018). Observed‐enrollment‐rates were compared to expected‐population‐rates derived from concurrent US‐National‐Cancer‐Institute's Surveillance‐Epidemiology‐and‐End‐Results (SEER) reported rates and US‐Census databases. Primary outcome was the proportional representation of women across trials, evaluated by a participation‐to‐prevalence ratio (PPR), according to cancer type. Secondary outcome was the report of any sex‐specific analysis of efficacy and/or safety, irrespective of treatment‐arm. Overall, there were 148 trials, enrolling 60,216 participants (60.5 ± 4.0 years, 40.7% female, 79.1% biologic, targeted, or immune‐based therapies) evaluating 99 drugs. Sex was reported in 146 (98.6%) trials, wherein 40.7% (24,538) were women, compared to 59.3% (35,678) men (p < .01). Altogether, women were under‐represented in 66.9% trials compared to the proportional incidence of cancers by respective disease type; weight‐average PPR of 0.91 (relative difference: ‐9.1%, p < .01). Women were most under‐represented in gastric (PPR = 0.63), liver (PPR = 0.71), and lung (PPR = .81) cancer trials. Sex‐based safety data was reported in 4.0% trials. There was no association between adequate female enrollment and drug efficacy (HR: 0.616 vs. 0.613, p = .96). Over time, there was no difference in the percentage of women recruited into clinical trials. Among pivotal clinical trials supporting contemporary FDA‐approved cancer drugs, women were frequently under‐represented and sex‐specific‐efficacy and safety‐outcomes were commonly not reported.