Meta-analysis of pancreaticogastrostomy versus pancreaticojejunostomy after pancreaticoduodenectomy.

Abstract

Surgical reconstruction following pancreaticoduodenectomy (PD) is associated with significant morbidity and mortality. Because of great variability in definitions of specific complications, it remains unclear whether there is a difference in complication rates following the two commonest types of reconstruction, pancreaticogastrostomy (PG) and pancreaticojejunostomy (PJ). Published consensus definitions for postoperative pancreatic fistula (POPF) have led to a series of randomized clinical trials (RCTs) uniquely placed to address this question. A literature search was carried out to identify all RCTs comparing postoperative complications of PG versus PJ reconstruction following PD published between January 1995 and December 2013. Pooled odds ratios (ORs) with 95 percent confidence intervals (c.i.) were calculated using fixed-effect or random-effects models. In total, seven RCTs with 1121 patients were included. Four of these trials applied definitions as published by the International Study Group on Pancreatic Fistula (ISGPF). Using ISGPF definitions, the incidence of POPF was lower in patients undergoing PG than in those having PJ (OR 0·50, 95 per cent c.i. 0·34 to 0·73; P < 0·001). Using definitions applied by each individual study, PG was associated with significantly lower rates of POPF (OR 0·51, 0·36 to 0·71; P < 0·001), intra-abdominal fluid collection (OR 0·50, 0·34 to 0·74; P < 0·001) and biliary fistula (OR 0·42, 0·18 to 0·93; P = 0·03) than PJ. Meta-analysis of four RCTs based on ISGPF criteria, and seven RCTs using non-standard criteria, revealed that PG reduced the incidence of POPF after PD compared with PJ.