Neoadjuvant therapy for pancreatic cancer: Systematic review of postoperative morbidity, mortality, and complications

The purpose of this review was to assess whether neoadjuvant chemotherapy and chemoradiotherapy (CRT) result in differential postoperative morbidity and mortality as compared with pancreatic tumor resection surgery alone. Using PRISMA guidelines and the PubMed search engine, we reviewed all prospective phase II trials of neoadjuvant chemotherapy and CRT for pancreatic cancer that examined postoperative morbidities and mortalities. A total of 30 articles were identified, collated, and analyzed. Risks of postoperative complications vary based on trial. With surgery alone, the most common postoperative complications included delayed gastric emptying (DGE) (17% to 24%), pancreatic fistula (10% to 20%), anastomotic leaks (0% to 15%), postoperative bleeding (2% to 13%), and infections/sepsis (17% to 20%). With surgery alone, the mortality was <5%. Neoadjuvant chemotherapy showed comparable fistula rates (3% to 4%), leaks (3% to 11%), infection (3% to 7%), with mortality 0% to 4% in all but 1 study. CRT for resectable/borderline resectable patients also showed comparable complication rates: DGE (6% to 15%), fistulas (2% to 3%), leaks (3% to 7%), bleeding/hemorrhage (2% to 13%), infections/sepsis (3% to 19%), with 9/13 studies showing a mortality of ≤4%. As compared with initially borderline/resectable tumors, CRT for initially unresectable tumors (despite less data) showed higher complication rates: DGE (13% to 33%), fistulas (3% to 25%), infections/sepsis (3% to 16%). However, the confounding factor of the potentially higher tumor burden as an associative agent remains. The only parameters slightly higher than historical surgery-only complication rates were leaks and bleeding/hemorrhage (13% to 20%). Mortality rates in these patients were consistently 0%, with 2 outliers. Hence, neoadjuvant chemotherapy/CRT is safe from a postoperative complication standpoint, without significant increases in complication rates compared with surgery alone. Resectable and borderline resect able patients have fewer complications as compared with unresectable patients, although data for the latter are lacking.