Adjuvant chemotherapy with or without sequential radiation in resected pancreatic adenocarcinoma: Who will win RTOG 0848?

Abstract

334 Background: The current role of adjuvant radiation in the management of resected pancreatic adenocarcinoma (PC) is controversial. A prospective trial, RTOG 0848, is currently in process to determine optimal adjuvant therapy, comparing adjuvant chemotherapy (CT) with or without radiation (RT). We determined outcomes in this population using the National Cancer Database (NCDB). Methods: Patients with non-metastatic PC diagnosed from 2004 to 2015 were identified from the NCDB. Patients who underwent surgery and adjuvant chemotherapy were included. Patients who received RT were only included if RT began at least 4 months after CT, to emulate RTOG 0848 schema. A 6 month landmark analysis was performed. Overall survival (OS) was compared using log-rank Kaplan-Meier for univariate analysis (UVA), and Cox proportional hazards for multivariate analysis (MVA). Logistic regression was used to determine predictors of RT utilization and generate propensity scores. Results: We identified 16075 patients. Median follow-up was 30 months, 37 months for survivors. Patients received sequential RT in 9.5% of cases. On UVA, sequential RT demonstrated OS benefit compared to adjuvant chemotherapy alone (3-year OS 44.6% vs 35.3%, p < 0.001). RT utilization was more likely with diagnosis after 2010, age ≤ 65 years, private insurance, higher income, ≤ 10 miles to treatment facility, treatment in the northeast, node positive disease, positive margins, and head of pancreas primary. On MVA, benefit to RT on OS was maintained (HR 0.68, 95% CI 0.63-0.73, p < 0.001). Other factors that improved OS included diagnosis after 2010, no comorbidity, non-government insurance, higher income, and treatment at academic facility. Factors that predicted for worse OS included increasing pathologic tumor and nodal staging, positive margins, and CA 19-9 above a cut-off of 90. After propensity-adjustment, OS benefit was maintained for RT (HR 0.68, p < 0.001). Conclusions: In a large national database analysis, sequential RT improves survival in patients who receive at least 4 months of adjuvant chemotherapy following upfront resection for pancreatic adenocarcinoma. The results of RTOG 0848 are awaited to confirm these findings.