Abstract
The primary treatment for operable pancreatic cancer (PC) involves surgery followed by adjuvant therapy. Nevertheless, perioperative or neoadjuvant chemotherapy (CT) may be used to mitigate the likelihood of recurrence and mortality. This network meta-analysis (NMA) assesses the comparative efficacy of various treatment approaches for resectable PC. A thorough search was carried out on January 31, 2023, encompassing PubMed/MEDLINE, Cochrane Library, and Embase databases. We incorporated randomized clinical trials (RCTs) that compared surgical interventions with or without (neo)adjuvant or perioperative therapies for operable PC. We conducted a fixed-effects Bayesian NMA. We presented the effect sizes in terms of hazard ratios (HRs) for overall survival (OS) along with 95% credible intervals (95% CrIs). The treatment was deemed statistically superior when the 95% credible interval (CrI) did not encompass a null value (hazard ratio < 1). Treatment rankings were established based on the surface under the cumulative ranking curve (SUCRA). A total of 24 studies were incorporated, comparing 21 treatments with surgery in isolation. Eleven treatments showed superior efficacy compared to surgery alone, with HRs ranging from 0.38 for perioperative treatments to 0.73 for adjuvant 5-fluorouracil. After the exclusion of studies conducted in Asia, it was found that the perioperative regimen of gemcitabine combined with nab-paclitaxel was the most effective regimen (SUCRA, p = 0.99). The findings endorse the utilization of perioperative CT, especially multi-agent CT, as the favored intervention for operable PC in Western nations.