Effect of extended postoperative recovery on long-term oncological outcomes.

Abstract

765 Background: Curative cancer surgery remains the cornerstone of treatment of localized malignancies. Extended post operative recovery (eLOS) is associated with debility and immunosuppression. We hypothesized that eLOS is associated with worse long term oncological outcome. Methods: The NCDB participant user file from 2004-13 was utilized to identify patients with localized gastrointestinal malignancies undergoing curative cancer resections. Quartiles (Qx) of length of stay (LOS) were calculated based on histology and site specific indicators. Survival modeling was performed using cox-proportional hazards and conditional models were created. Alpha of 0.05 was set for statistical significance. Results: Of 1,616,215 patients in the cohort, 577,860 patients were included who met inclusion criteria. Majority of patients were ≥65 years (56%) with a predominance of colon cancer (70%). The median length of stay (LOS) varied from 5 days (partial hepatectomy) to 10 days (esophagectomy). After adjusting for age, gender, stage, and site of disease, increasing length of stay was significantly associated with worse oncological survival with a dose response effect (Q1 HR 1.0, Q2 1.12 (1.10-1.14) p<0.001, Q3 1.42 (1.41-1.44) p<0.001 and Q4 1.82(1.80-1.85), p<0.001). Conditional overall survival for patients who survived 3, 6 and 12 months after surgery, was worse with eLOS (p<0.001). Conclusions: Longer than expected length of stay after an index curative cancer operation leads to worse overall oncological outcomes. This effect persists even after excluding patients who suffer delayed post-operative mortality within 12 months after the operation. More research needs to be done to determine if this outcome is related to perioperative immunosuppression.