Influence of post-resection time to chemotherapy in pancreatic adenocarcinoma on overall survival.

Abstract

e15791 Background: Pancreatic adenocarcinoma is an aggressive malignancy with a high propensity for distant spread. To date, surgery remains the only potentially curative treatment option; which has traditionally been followed by adjuvant chemotherapy. Past research from the pancreatic literature as well as other disease sites has shown conflicting outcomes as it relates to timely delivery of adjuvant chemotherapy. We thus used the National Cancer Database (NCDB) to evaluate time to initiation of chemotherapy following pancreatic resection and if there was any correlate with outcome. Methods: We identified patients diagnosed with stage 1-3 pancreatic adenocarcinoma in the NCDB, excluding those with stage IV disease, those treated nonoperatively, with adjuvant or neoadjuvant chemotherapy or radiotherapy and those with unknown or inadequate ( < 3 months) follow-up. Receiver operator curve analysis identified an interval of 66 days as associated with outcome. Multivariable logistic regression analysis identified variables associated with increased time to chemotherapy postoperatively ( > 66 days). Propensity matching was done to account for indication bias. Overall survival (OS) was compared between both propensity-matched groups receiving earlier and postponed chemotherapy post-resection, with an additional breakdown by stage. Results: In total, 6,873 and 3,348 patients received chemotherapy before and after the 66 day cutoff, respectively. Predictors of expedited chemotherapy included lower comorbidity, treatment outside a community program in an urban location, having insurance, Caucasian race and treatment after 2009. Median overall survival was 21.8 months for all patients and of these, 6462 were stage 1. 5-year OS was 20% in patients receiving chemotherapy within 66 days and 18% in those not (p = 0.0266). In stage 1 patients, 5-year OS was 23% versus 21% (p = 0.0116) in favor of expedited chemotherapy. In stage 2 patients, 5-year OS was 16% in both arms (p = 0.7231). Higher stage and comorbidity score, treatment at a community cancer program, lower salary, increased time from surgical resection to chemotherapy, higher grade, positive margins, and more distant year of treatment were associated with worse prognosis. Conclusions: The present propensity-matched analysis showed a significant benefit to earlier delivery of chemotherapy in the adjuvant setting, with the largest benefit seen in stage 1 patients.