Disease-free survival as a surrogate for overall survival in HR+/HER2– early breast cancer: A correlation analysis

Abstract

BackgroundOverall survival (OS) is a universally accepted measure of clinical benefit; however, prolonged follow-up is needed to observe sufficient events. Disease-free survival (DFS) has been widely adopted as a primary endpoint for early breast cancer (EBC) trials, as follow-up is comparatively shorter. Here, we present an analysis evaluating DFS as a surrogate for OS for adjuvant treatment of hormone receptor–positive/human epidermal growth factor receptor 2–negative (HR+/HER2–) EBC. MethodsA systematic literature review which included randomized controlled trials (RCTs) with ≥80% of adult patients with HR+/HER2– EBC was conducted. The RCTs evaluated various systemic therapeutic categories; key inclusion criteria included reporting of DFS and OS hazard ratios (HRs) and mature OS data. Spearman rank correlation and weighted linear regression analyses evaluated DFS and OS HR correlation. A scenario analysis tested base-case analysis robustness, and a parallel analysis using patient-level data was conducted. ResultsThe base case (N = 14 RCTs) showed an unweighted Spearman coefficient of 0.81 between OS and DFS (weighted: 0.81), with 84% of the variability in OS explained by DFS differences (R2 from weighted regression). The surrogate threshold effect (Burzykowski T, Buyse M. Pharm Stat. 2006;5:173–186) was 0.82 for DFS/OS HR. Scenario analysis (n = 9 RCTs), which excluded chemotherapy trials, and patient-level analysis using FACE trial data were consistent with the base-case analysis. ConclusionsThese analyses support DFS as a reliable surrogate endpoint for OS in adjuvant HR+/HER2− EBC trials. Using DFS as a surrogate measure will permit timelier access to novel treatments for patients with HR+/HER2− EBC.