Impact of broadening trial eligibility criteria on the inclusion of patients with brain metastases in cancer clinical trials.

Abstract

1505 Background: On October 2, 2017, the American Society of Clinical Oncology, Friends of Cancer Research, and the U.S. Food and Drug Administration (ASCO/FoCR/FDA) issued a joint research statement recommending that certain eligibility criteria in cancer clinical trials, including those related to brain metastases, be broadened to make trials more inclusive. We examined whether patterns of exclusions regarding patients with brain metastases changed over time in relation to these recommendations. Methods: We used data from ClinicalTrials.gov to evaluate patterns of trial eligibility criteria in phase I-III U.S.-based interventional clinical trials for patients with advanced breast, colorectal, or lung cancers from January 2013 through October 2021. Trial inclusion and exclusion criteria were abstracted; to enhance validity, reviewers were blinded to the year of trial activation. For each trial, we determined whether patients with brain metastases were not excluded, conditionally excluded (i.e., excluded in some circumstances), or wholly excluded. Trial registrations between October 2, 2017-December 31, 2018 were excluded to allow 1 year for newly conceived trials to adopt the recommendations, plus 3 months to account for the lag time between loading trial records and trial activation dates; thus, the independent exposure variable was January 1, 2019. An interrupted time series analysis with multinomial logistic regression was used to assess whether the ASCO/FoCR/FDA recommendations were associated with changes in brain metastases eligibility criteria. Results: We evaluated N = 1998 trials. Patients with brain metastases were not excluded in 307 trials (15.4%); conditionally excluded in 1459 trials (74.8%); and wholly excluded in 196 trials (9.8%). In the post-recommendation period, we found a 92% increase in the odds of trials with brain metastases not excluded compared to conditionally excluded (OR = 1.92, 95% CI, 1.08-3.45, p =.03). The estimated proportion of trials in which patients with brain metastases were not excluded increased from 9.2% had the recommendations not been made to 15.6% (p =.04). Conversely, the proportion of trials in which patients with brain metastases were conditionally excluded (76.9%) was lower than expected (85.3%, p =.02). We found no difference in the proportion of trials in which patients with brain metastases were wholly excluded (7.5% vs. 5.4%, p =.28). Conclusions: The ASCO/FoCR/FDA recommendations were associated with a shift in patterns of brain metastases exclusion criteria from conditionally excluded to not excluded. To our knowledge, this is the first evidence that cancer clinical trials have become more inclusive of a broader set of patients in response to the ASCO/FoCR/FDA recommendations. More inclusive eligibility improves trial access and representativeness, increasing trial validity and the pace at which trials enroll.