Adherence of published randomized phase 3 cancer trials to principles proposed by common-sense oncology.

Abstract

11019 Background: Randomized clinical trials (RCTs) remain the gold standard for evaluating the efficacy and safety of novel cancer therapies. Some RCTs are well-designed and show meaningful improvements in patient outcomes while others are confounded by various types of bias, or do not reflect outcomes that matter to patients. Published RCTs should be designed, analyzed and reported to provide optimal, unbiased information for clinicians to enhance treatment decision-making Common-Sense Oncology (CSO) is an initiative of clinicians, patient advocates, researchers, and policymakers with the mission of ensuring that cancer care and research are focused on outcomes that matter to patients. CSO has published a checklist for the design, analysis and reporting of RCTs evaluating systemic treatments for cancer. In the present study, we have applied the checklists to a cohort of cancer drug trials to assess the extent to which CSO principles were incorporated in reports of RCTs published in 2023 in high-impact journals. Methods: We reviewed retrospectively phase 3 RCTs evaluating systemic therapies for adult solid tumors published in 2023 in The New England Journal of Medicine, Lancet, Lancet Oncology, JAMA, JAMA Oncology, Journal of Clinical Oncology, and Annals of Oncology . These journals were selected based on their high impact. For each trial we evaluated the trial design in the methodology, how the results were reported and the discussion section using the CSO RCT Checklist. Results: 50 RCTs evaluating systemic therapies for solid tumors were published in 2023. The most common tumor types were lung, liver, and prostate cancer. Progression-free survival and overall survival were the primary endpoints in 44% (22) and 42% (21) of trials, respectively. Only 36/50 trials justified the control arm, 25/50 justified the primary endpoint, and 18 included Quality-of-Life as a secondary endpoint. Only two trials addressed strategies to limit censoring and dropout; numbers of censored patients (with numbers at risk) were shown under Kaplan-Meier curves in only 21/50 trials, and sensitivity analysis to determine the potential effects of censoring was done in only 5 trials. Chronic toxicities were reported in only one trial, only 7 trials included patient-reported outcomes and only 3/50 trials mentioned cost of the drug. Conclusions: Our findings underscore the need for standardized methodologies, comprehensive design, and reporting in oncology RCTs. By identifying gaps in RCT design and reporting, CSO aims to improve the quality and consistency of future trials.