Background: The use of epidural analgesia has been proposed to improve the prognosis of esophageal cancer by attenuating the stress response and being less immunosuppressive than opioids. This study aims to evaluate the association, if any, between non-epidural pain management compared to epidural analgesia during minimally invasive or open esophagectomy and esophageal cancer prognosis.

Materials and methods: This was a population-based nationwide retrospective cohort study in Finland, using the Finnish National Esophago-Gastric Cancer Cohort (FINEGO). Esophagectomy patients with epidural and no epidural analgesia were compared. Multivariable Cox regression provided hazard ratios (HR) with 95% confidence intervals (CI) non-epidural pain management compared to epidural analgesia, adjusted for the calendar period of surgery, sex, age, comorbidity (Charlson Comorbidity Index), tumor stage, tumor histology, neoadjuvant therapy, type of surgery, and esophageal cancer surgery volume.

Results: After exclusions, there were 1381 patients available with information on epidural analgesia. Of these, 969 (70.2%) were men and 832 (60.2%) had esophageal adenocarcinoma. After adjustment for confounding factors, non-epidural pain management was not associated with higher 90-day mortality (HR 1.022 95% CI 0.582-1.794), overall mortality up to 5 years (HR 1.156 95% CI 0.909-1.470), nor with 5 years cancer-specific mortality (HR 1.134 95% CI 0.884-1.456) compared to epidural analgesia.

Conclusion: Although the point estimates may hint at a potentially improved prognosis associated with epidural use, this population-based nationwide study suggests no statistically significant association between epidural analgesia during esophagectomy and esophageal cancer prognosis.

Editorial comment: This large esophagectomy (cancer) cohort in Finland was used to compare those who received epidural analgesia with those who did not for associations with late mortality in a retrospective analysis and where anesthesia and analgesia treatments were not controlled. The findings showed that when other recognized risks for mortality were taken into account, there was not a meaningful difference in relative risk for late mortality related to the presence or absence of epidural analgesia, though the analgesia treatments were not randomly allocated. These results do not rule out associations of analgesia choice with other outcomes that might be important to patients.