Statistical aspects in adjuvant and neoadjuvant trials for gastrointestinal cancer in 2020: focus on time-to-event endpoints

Abstract

Purpose of review Clinical-trial design, analysis, and interpretation entails the use of efficient and reliable endpoints. Statistical issues related to endpoints warrant continued attention, as they may have a substantial impact on the conduct of clinical trials and on interpretation of their results. Recent findings We review concepts and discuss recent developments related to the use of time-to-event endpoints in studies on adjuvant and neoadjuvant therapy for colon, pancreatic, and gastric adenocarcinomas. The definition of endpoints has varied to a considerable extent in these settings. Although these variations are relevant in interpreting results from individual trials, they probably have a small impact when considered in aggregate. In terms of surrogacy, most published reports so far have used aggregated data. A few studies based on the preferred method of a metaanalysis of individual-patient data have shown that disease-free survival (DFS) is a surrogate for overall survival in the adjuvant therapy of stage III colon cancer and in gastric cancer, whereas DFS with a landmark of six months is a surrogate for overall survival in the neoadjuvant therapy of adenocarcinoma of the esophagus, gastroesophageal junction, or stomach. Testing novel agents in gastrointestinal cancer requires continued attention to statistical issues related to endpoints.