Meta-analysis of recurrence-free survival or disease-free survival as a potential surrogate endpoint for overall survival in esophageal cancer trials

Abstract Background Cancer trials increasingly use surrogate endpoints, but it is unclear how well recurrence-free (RFS) or disease-free survival (DFS), specifically, predict overall survival (OS) in resectable esophageal cancer (EC).Methods A systematic literature review identified trials with RFS/DFS and OS endpoints. A meta-analysis assessed RFS/DFS as surrogates for OS, estimating pooled hazard ratios (HRs) from trial HRs. Forest plots and heterogeneity tests showed effect sizes and pooled estimates. Unweighted linear regression and weighted sensitivity analysis estimated the correlation between OS and RFS/DFS, producing a regression plot.Results Of 975 articles identified, nine met the criteria, with two more from references. The pooled HR for OS and RFS/DFS were 0.90 and 0.87, respectively. The primary analysis showed a strong Pearson correlation between RFS/DFS and OS (ρ = 0.89, p Conclusion Subject to known methodological limits, RFS/DFS was demonstrated to be a potentially-suitable surrogate endpoint for OS in resectable EC.