The surrogacy of disease-free survival for overall survival in locally advanced esophageal squamous cell carcinoma after neoadjuvant therapy.

Abstract

Introduction: Disease-free survival (DFS) is increasingly proposed as a surrogate endpoint for overall survival (OS) in oncology trials, yet its validity in esophageal squamous cell carcinoma (ESCC) remains unconfirmed. This study evaluated the surrogacy of DFS for OS in patients with locally advanced ESCC receiving neoadjuvant therapy.

Methods: Individual patient data from a prospectively maintained cohort of patients who underwent neoadjuvant therapy followed by radical esophagectomy (N = 578) were analyzed. Individual-level surrogacy was assessed by comparing long-term OS with expected survival in a demographically matched general population using standardized mortality ratios (SMR) and landmark analysis. Trial-level surrogacy was evaluated via weighted linear regression of treatment effects on DFS and OS across published clinical trials.

Results: Patients remaining recurrence-free at 3 years achieved a 5-year OS rate of 93.0%, comparable to the matched general population (SMR = 0.935; 95% CI: 0.491-1.380). Conversely, early recurrence was associated with substantially worse outcomes, with a 51.3% absolute risk difference in 5-year OS between groups in landmark analysis. At the trial level, short-term DFS demonstrated a strong correlation with OS (R² = 0.776).

Conclusions: Short-term DFS is robustly correlated with long-term OS and may serve as a potential surrogate endpoint in locally advanced ESCC following neoadjuvant therapy, supported by both individual- and trial-level evidence. Patients achieving 3-year recurrence-free status exhibit long-term survival indistinguishable from the general population, reinforcing the clinical utility of DFS as an early efficacy marker in neoadjuvant trials.