External and intraoperative radiotherapy for resectable and unresectable pancreatic cancer: Analysis of survival rates and complications

Purpose: Clinical results of intraoperative radiotherapy (IORT) and/or external beam radiotherapy (EBRT) for both resectable and unresectable pancreatic cancer were analyzed. Methods and Materials: Between 1980 and 1995, 332 patients with pancreatic cancer were treated with surgery and/or radiation therapy (RT). Of the 332 patients, 157 patients were treated,vith surgical resection of pancreatic tumor, and the remaining 175 patients had unresectable pancreatic tumors. Among the 157 patients with resected pancreatic cancer, 62 patients were not treated with RT, while 40 patients were treated with EBRT alone (mean RT dose; 46.3 Gy) and 55 patients with IORT (25.2 Gy) +/- EBRT (44.0 Gy). On the other hand, among the 175 patients with unresectable pancreatic cancer, 58 patients were not treated with RT, 46 patients were treated with EBRT alone (39.2 Gy), and the remaining 71 patients with IORT (29.3 Gy) +/- EBRT (41.2 Gy). Results: For 87 patients with curative resection, the median survival times (MSTs) of the no-RT, the EBRT, and the IORT +/- EBRT groups were 10.4, 13.0, and 15.5 months, respectively, without significant difference. For 70 patients with noncurative resection, the MSTs of the no-RT, the EBRT, and the IORT +/- EBRT groups were 5.3, 8.7, and 6.5 months, respectively. When the EBRT and the IORT +/- EBRT groups were combined, the survival rate was significantly higher than that of the no RT group for noncuratively resected pancreatic cancers (log rank test; p = 0.028). The 2-year survival probability of the IORT +/- EBRT group (16%) was higher than that of the EBRT group (0%). For unresectable pancreatic cancer, the MSTs of 52 patients without distant metastases were 6.7 months for palliative surgery alone, 7.6 months for EBRT alone, and 8.2 months for IORT +/- EBRT. The survival curve of the IORT +/- EBRT group was significantly better than that of the no-RT group (p<0.05), and the difference between the IORT +/- EBRT and the EBRT alone groups was marginally significant (p = 0.056). In addition, the 2-year survival probability for the IORT +/- EBRT group was 14%, while no 2-year survival was observed in the no RT or the EBRT groups. Multivariate analysis using the Cox proportional hazards model revealed that tumor size, stage (Stages 1, 2 vs. Stages 3, 4), and curability of resection were significant variables for resectable pancreatic cancer, while distant metastases and performance of IORT were significant variables for unresectable pancreatic cancer. The dose of EBRT was a marginally significant factor for both resectable and unresectable tumors (both p = 0.06). In terms of complications, ulcers of gastrointestinal tract were noted in 14% of the 126 patients treated with IORT. Conclusion: Although prolongation of the MST by IORT was not remarkable, long survivals (>2 years) were obtained by IORT +/- EBRT for noncuratively resected and unresectable pancreatic cancer. IORT combined with EBRT is indicated for noncurative resected or unresectable pancreatic cancer without distant metastases. (C) 1997 Elsevier Science Inc.