Neoadjuvant/preoperative Gemcitabine for patients with localized pancreatic cancer : a meta-analysis of prospective studies

BACKGROUND: Long-term prognosis for localized pancreatic cancer remains poor. We sought to assess the benefit of neoadjuvant/preoperative chemotherapy with or without radiotherapy. METHODS: Prospective studies where gemcitabine with or without radiotherapy was provided before surgery in patients with initially resectable or unresectable disease were reviewed by meta-analysis. Primary outcome was survival, and secondary outcomes were tumor response after therapy, toxicity, surgical exploration, and resection rates. RESULTS: Twenty independent studies with 707 participants were included, 366 with resectable lesions and 341 with unresectable lesions. Seven studies were phase I/II trials, 10 phase II, and 3 prospective cohort studies. Estimated 1- and 2-year survival probabilities after resection were 91.7% (95% confidence interval [CI] 75-100) and 67.2% (95% CI 38-87) for initially resectable patients, and 86.3% (95% CI 78-100) and 54.2% (95% CI 25-100) for initially unresectable patients. The complete/partial response rate was 12% (95% CI 4-23) and 27% (95% CI 18-38) in resectable and unresectable lesions, respectively. The rate of treatment-related grade 3-4 toxicity was 31% (95% CI 21-42). Of resectable patients evaluable after restaging, 91% (95% CI 83-97) underwent surgery, and 82% (95% CI 65-95) of explored patients underwent resection. R0 resections amounted to 89% (95% CI 83-94). Of unresectable patients evaluable after restaging, 39% (95% CI 28-50) underwent surgery, and 68% (95% CI 53-82) of explored patients were resected, with 60% (95% CI 50-71) R0 resections. CONCLUSIONS: Current analysis provides marginal support to the assumed benefits of neoadjuvant therapies for patients with resectable cancer, and indicates a potential advantage only for a minority of those with unresectable lesions.