Survival outcomes of intrathoracic vs. cervical anastomosis post-esophagectomy in middle and lower thoracic esophageal squamous cell carcinoma: a retrospective propensity score matching analysis

TL;DR

This study compares long-term survival outcomes of two surgical techniques for esophageal cancer and finds that initial survival benefits of one technique disappear after adjusting for other factors.

Contribution

The novel contribution is the use of propensity score matching to assess the true impact of anastomotic technique on survival in esophageal cancer patients.

Findings

Cervical anastomosis initially showed better survival outcomes compared to intrathoracic anastomosis.

After adjusting for confounding factors, the survival advantage of cervical anastomosis disappeared.

Tumor stage and lymphovascular invasion were stronger predictors of survival than anastomotic technique.

Abstract

This study aimed to compare long-term survival outcomes between cervical anastomosis (CA) and intrathoracic anastomosis (IA) in patients with middle and lower thoracic esophageal squamous cell carcinoma (ESCC).

A retrospective cohort analysis was conducted on 571 patients who underwent esophagectomy at a single institution. Patients were stratified into CA and IA groups based on anastomotic technique. Propensity score matching (PSM, 1:1) was applied to balance baseline covariates. Overall survival (OS) and disease-free survival (DFS) were evaluated using Kaplan-Meier analysis and Cox regression. Secondary outcomes included postoperative complications.

In the unmatched cohort, CA demonstrated superior OS (median: 51.17 vs. 34.50 months; HR: 1.368, 95% CI: 1.062–1.763; p=0.015) and DFS (median: 45.07 vs. 28.87 months; HR: 1.289, 95% CI: 1.013–1.641; p=0.039) compared to IA. However, after PSM, the survival advantage attenuated (OS: HR = 1.303, 95% CI: 0.953–1.780, p=0.097; DFS: HR = 1.295, 95% CI: 0.962–1.744, p=0.089). Multivariate analysis identified pathological T3/T4 stages (OS: p=0.002–0.009; DFS: p<0.001) and lymphovascular invasion (DFS: p=0.023) as dominant prognostic factors, overshadowing anastomotic technique. The CA group exhibited more extensive lymph node dissection (>7 stations, p<0.001), but short-term mortality (30-/90-day) did not differ between groups (p≥0.382).

In conclusion, our study suggests that there may be a potential survival advantage of CA over IA in patients undergoing esophagectomy for ESCC. However, the initial survival benefits associated with CA diminished after adjusting for confounding factors.

Figures