Adjuvant Chemotherapy ± Chemoradiotherapy for Adenocarcinoma of the Pancreatic Head: Results of the Radiotherapy Random Assignment of NRG Oncology/RTOG 0848

PURPOSE To assess whether adding fluoropyrimidine sensitized radiotherapy (CXRT) to adjuvant chemotherapy improves overall survival (OS) after curative intent resection of the pancreatic head. METHODS This was a multicenter, randomized phase III, two step trial. Step 1: gemcitabine versus gemcitabine + erlotinib (previously reported). Step 2: random assignment to sixth chemotherapy cycle ± CXRT after five cycles of step 1 chemotherapy without progression. Outcomes of step 2 random assignment are reported here. Assuming 17 months median OS (chemotherapy alone), the sample size was 354 patients (hazard ratio [HR], 0.76, 80% power, one-sided α = .05, 316 OS events). OS/disease-free survival (DFS) were estimated by Kaplan-Meier and arms compared using the log-rank test. RESULTS A total of 354 patients (median age 63, 55% male, 56% performance status, 1) were randomly assigned to chemotherapy (174) or chemotherapy + CXRT (180). Univariable median and 5-year OS (90% CIs) were 2.6 years (2.1-3.1) and 23.1% (17.7-28.6) for chemotherapy alone, and 2.3 years (2.0-2.6) and 27.9% (22.2-33.6) for chemotherapy + CXRT. The OS primary end point was not met (HR, 0.96 [90% CI, 0.79 to 1.18]; one-sided P = .38, two-sided P = .77). Chemotherapy + CXRT was associated with a trend for improved DFS (univariably; HR, 0.82 [95% CI, 0.65 to 1.03]; P = .089), without increase in grade 4/5 toxicities. However, grade 3 toxicity increased (38% v 19%, P < .001). Significantly, treatment by nodal status interactions showed that CXRT improved OS ( P = .0063) and DFS ( P = .014) in node-negative patients. CONCLUSION Overall, the addition of adjuvant CXRT to adjuvant gemcitabine did not statistically significantly improve OS, DFS, or increase grade 4/5 toxicity. For node-negative patients, CXRT improved OS and DFS. These results, if confirmed in studies with current or future systemic therapies, would support the use of adjuvant/neoadjuvant CXRT for node-negative patients.