The impact of perioperative outcomes on long-term survival after esophageal cancer resection

Abstract

Objective

Current metrics for quality of care after esophagectomy rely on short-term perioperative outcomes. We aimed to evaluate the relationship of perioperative outcomes and long-term survival after esophagectomy for esophageal cancer.

Methods

Institutions in the National Cancer Database performing ≥20 esophagectomies (2010-2019) were ranked by a weighted composite score of major perioperative events—length of stay >20 days, unplanned readmissions, and 30-day mortality—and stratified into quintiles. Survival between centers with the least perioperative morbidity (quintile 1) and all other centers (quintiles 2-5) was compared using Kaplan-Meier, Cox proportional hazard, and landmark analyses.

Results

Overall, 11,036 patients with esophageal cancer who underwent esophagectomy at 209 institutions were included. Of those, 1825 (17%) patients were treated at 42 quintile 1 centers and 9211 (83%) were treated at 167 quintiles 2-5 centers. Prolonged length of stay (15% vs 7.9%, P < .001), unplanned readmissions (7.1% vs 4.1%, P < .001) and 30-day mortality (3.1% vs 0.5%) were all significantly greater at the quintile 2-5 centers. Five-year overall survival was better at quintile 1 centers in both univariable (51.9% vs 47.7%, P < .001), and in multivariable analysis (hazard ratio 0.89, P = .004). Significantly better survival continued to be observed at quintile 1 centers on landmark analysis that excluded patients who died within 6 months (55.8% vs 47.7%, P = .036).

Conclusions

Our study demonstrates that patients undergoing esophagectomy for esophageal cancer at centers with lower rates of major perioperative events had improved long-term survival, suggesting that short-term metrics can serve as indicators of long-term performance and oncologic efficacy.