PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
April 12, 2026BMC Cancer0 citationsOpen Access

Event-free survival as a surrogate for overall survival in locally advanced esophageal cancer: a correlation analysis of trials assessing definitive chemoradiation therapies

WXWeiguang XueInstitute of Group AnalysisAVAdriana ValderramaMerck & Co., Inc., Rahway, NJ, USA (United States)KRKarthik RamakrishnanMerck & Co., Inc., Rahway, NJ, USA (United States)

Key Points

  • This research aims to assess the validity of event-free survival as a surrogate for overall survival in patients with unresectable locally advanced esophageal cancer receiving definitive chemoradiotherapy.
  • Conducted a systematic literature review of randomized controlled trials (RCTs) evaluating definitive chemoradiotherapy (dCRT) for unresectable locally advanced esophageal cancer.
  • Included studies that reported hazard ratios for both event-free survival (EFS) and overall survival (OS).
  • Utilized weighted linear regression to evaluate the correlation between log hazard ratios of EFS and OS.
  • Performed leave-one-out cross-validation and sensitivity analyses to confirm the robustness of correlation findings.
  • Identified a strong correlation between EFS and OS, with an estimated slope coefficient of 0.91 (P<0.001).
  • The coefficient of determination (R2) = 0.80, indicating good surrogacy between EFS and OS.
  • Median EFS was reported as 17.5 months, while median OS was 25.1 months.
  • The leave-one-out cross-validation showed that 71.4% of observed OS hazard ratios fell within the predicted confidence intervals.

Abstract

Treatments for esophageal cancer are evolving rapidly, highlighting the need for surrogate endpoints to expedite patient access to therapy. Overall survival (OS) remains the gold standard endpoint in oncology clinical trials, but its assessment often requires extended follow-up, delaying treatment evaluations. Event-free survival (EFS) has emerged as a promising surrogate for OS in resectable locally advanced esophageal cancer. However, its validity in the context of definitive chemoradiotherapy (dCRT), the standard treatment for unresectable disease, has not been established. We assessed the trial-level correlation between EFS and OS in this setting. A systematic literature review identified randomized controlled trials (RCTs) evaluating dCRT for unresectable locally advanced esophageal cancer. Studies reporting hazard ratios (HRs) for both EFS (or similarly defined early endpoints) and OS were included. Weighted linear regression (WLR) was used to assess the correlation between log(HR) for EFS and log(HR) for OS. The strength of the correlation was evaluated using the coefficient of determination (R2), with thresholds of ≥ 0.65 or ≥ 0.75 indicating good surrogacy. The robustness of the WLR model and correlation analyses was confirmed by leave-one-out cross-validations and sensitivity analyses, respectively. Fourteen between-treatment comparisons from 11 RCTs (N = 2,812) in unresectable locally advanced esophageal cancer were included. Eight studies were conducted in China and three in Europe, with sample size and follow-up ranging from 86 to 434 patients and 24 to 60 months, respectively. Reported median EFS was 17.5 months, whereas median OS was 25.1 months. Most patients had advanced-stage squamous cell carcinoma. The correlation analysis identified a strong EFS–OS correlation (estimated slope coefficient: 0.91; 95% confidence interval CI, 0.62–1.20, P<0.001; R2=0.80; 95% CI, 0.38–0.96), indicating that EFS was a significant predictor of OS. The leave-one-out cross-validation confirmed this association, with 71.4% of observed OS HRs falling within the predicted 95% CIs. Sensitivity analyses further supported the robustness of the correlation. We identified a strong, statistically significant, correlation between EFS and OS, suggesting that EFS may serve as a valid surrogate endpoint for OS in patients with unresectable locally advanced esophageal cancer receiving dCRT, with potential to streamline clinical trials and regulatory assessments.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Xue et al. (2026) studied this question.

synapsesocial.com/papers/69db37774fe01fead37c5841https://doi.org/10.1186/s12885-026-15914-6
Ask AI
Helpful
Bookmark
Share
View Full Paper