Benefit of Chemoradiotherapy Versus Chemotherapy After Induction Therapy for Conversion of Unresectable Into Resectable Pancreatic Cancer: The Randomized CONKO-007 Trial

PURPOSE To determine the benefit, measured as complete removal of a tumor so that no tumor cells are detectable during histopathologic examination of the resection margin (R0 resection rate), of induction chemotherapy plus chemoradiotherapy (CRT) compared with chemotherapy alone for unresectable pancreatic tumors. PATIENTS AND METHODS CONKO-007, an investigator-initiated open-label, multicentric, phase III randomized clinical trial, enrolled 525 patients with unresectable tumors, and 495 patients received induction chemotherapy (402 with fluorouracil, irinotecan, and oxaliplatin [FOLFIRINOX] and 93 with gemcitabine). Patients without progression after 3 months of induction chemotherapy (n = 336) were randomly assigned for continuation of the same chemotherapy (n = 167) or CRT (n = 169; 50.4Gy concurrently with gemcitabine). Resectability was centrally reassessed by a panel of surgeons. Surgery was recommended if possible. After an interim analysis, the primary end point was changed from overall survival (OS) to overall R0 resection rate because of slow recruitment. The median follow-up was 76 months. Important planned secondary end points were R0 resection rate in the surgically treated population and OS. RESULTS The primary end point (overall R0 resection rate) was not significantly different between treatment arms with 25% (43 of 169) in the CRT arm versus 18% in the chemotherapy arm (30 of 167; P = .113). Secondary end point analysis showed that surgery was performed equally often ( P = .91); R0 resection rate in patients who underwent surgery was higher after CRT, 69.4% (43 of 62) compared with chemotherapy alone: 50.0% (30 of 60 patients, P = .04). Other parameters of resection (ratio of R0/R1/R2/no resection) also favored CRT ( P = .02). No difference in OS was seen between treatment arms (hazard ratio [HR], 0.937 [95% CI, 0.747 to 1.174]; P = .57; randomly assigned intention-to-treat patients). Surgery was associated with longer OS ( P < .001, HR, 0.525 [95% CI, 0.408 to 0.676]). CONCLUSION Although not improving overall R0 resection rate or survival, CRT enables a R0 resection in surgically treated patients more often than chemotherapy alone.