Post-treatment mortality after definitive chemoradiotherapy versus resection for esophageal cancer

In efforts to better characterize incidence and predictors of 30- and 90-day mortality following operative versus nonoperative approaches for locally advanced esophageal cancer (EC), we conducted a novel investigation of a large, contemporary US database. The National Cancer Database was queried for newly-diagnosed T1-3N0-1 squamous cell or adenocarcinoma receiving surgical-based therapy (esophagectomy alone or preceded by chemotherapy and/or radiotherapy) versus definitive chemoradiotherapy (dCRT). Statistics included graphing cumulative incidences of mortality before and following propensity score matching (PSM), based on age-based intervals. Cox regression determined factors independently predictive of 30- and 90-day mortality. Of 15,585 patients, 9,278 (59.5%) received surgical-based therapy and 6,307 (40.5%) underwent dCRT. In the unadjusted population, despite nonsignificant differences at 30 days (3.3% dCRT, 3.6% surgical-based), the dCRT cohort experienced higher 90-day mortality (11.0% vs. 7.5%, P < 0.001). Following PSM, however, dCRT patients experienced significantly lower 30-day mortality (P < 0.001), with nonsignificant differences at 90 days (P = 0.092). Surgical-based management yielded similar (or better) mortality as dCRT in ≤70-year-old patients; however, dCRT was associated with reduced mortality in subjects > 70 years old. In addition to the intervention group, factors predictive for 30- and 90-day mortality included age, gender, insurance status, facility type, comorbidity index, tumor location, histology, and T/N classification. In summary, surgical-based therapy for EC is associated with higher 30-day mortality, which becomes statistically similar to dCRT by 90 days. Differences between surgery and dCRT were most pronounced in patients > 70 years of age. These data may better inform shared decision-making between multidisciplinary providers and patients.