Pancreaticojejunostomy Versus Pancreaticogastrostomy for the Prevention of Postoperative Pancreatic Fistula Following Pancreaticoduodenectomy: a Comprehensive Systematic Review and Meta-Analysis of Randomized Controlled Trials.

Background: Pancreaticoduodenectomy (PD) is the primary curative procedure for malignancies of the pancreatic head and periampullary region. The critical step of reconstructing the pancreatic stump is most commonly performed via pancreaticojejunostomy (PJ) or pancreaticogastrostomy (PG). Postoperative pancreatic fistula (POPF) remains the most formidable complication, which drives significant morbidity and mortality. The ongoing debate regarding the optimal reconstruction technique lacks a definitive evidence-based consensus. Objective: To conduct a rigorous systematic review and meta-analysis comparing the efficacy and safety of PG versus PJ in preventing POPF and other critical postoperative outcomes in patients undergoing PD. Methods: We systematically searched major electronic databases (CENTRAL, MEDLINE, Embase, CINAHL) and clinical trial registries up to September 2016 for all relevant randomized controlled trials (RCTs). Our primary outcomes were the incidence of POPF (any grade, A-C), clinically significant POPF (grades B & C) and 90-day postoperative mortality. Secondary outcomes encompassed length of hospital stay, re-intervention rates, overall surgical complications, postoperative bleeding, intra-abdominal abscess, quality of life and cost-analysis. Methodological quality was appraised using the Cochrane Risk of Bias tool and the certainty of evidence was graded using the GRADE framework. Results: Our analysis incorporated 10 RCTs with a pooled cohort of 1,629 participants. The meta-analysis demonstrated no statistically significant difference in the overall incidence of POPF between PJ and PG (24.3% vs. 21.4%; RR 1.19, 95% CI 0.88 to 1.62; low-quality evidence). The evidence for clinically significant POPF was highly uncertain (19.3% vs. 12.8%; RR 1.51, 95% CI 0.92 to 2.47; very low-quality evidence). A comparable postoperative mortality between groups was found (3.9% vs. 4.8%; RR 0.84, 95% CI 0.53 to 1.34; moderate-quality evidence). A notable trade-off was observed: PJ was associated with a significantly lower risk of postoperative bleeding (9.3% vs. 13.8%; RR 0.69, 95% CI 0.51 to 0.93) but a significantly higher risk of intra-abdominal abscess (14.7% vs. 8.0%; RR 1.77, 95% CI 1.11 to 2.81). No significant differences for other secondary outcomes were detected. Conclusion: The current body of evidence does not establish the superiority of either PJ or PG for pancreatic reconstruction following PD. The choice of technique should be individualized, relying on surgeon expertise and specific patient- and pancreas-related factors, while weighing the distinct risk profiles of each procedure (bleeding vs. abscess).