Roux-en-Y versus Billroth I reconstruction after distal gastrectomy for gastric cancer: a meta-analysis

Abstract

Aim: To conduct a meta-analysis to compare Roux-en-Y (R-Y) gastrojejunostomy with gastroduodenal Billroth I (B-I) anastomosis after distal gastrectomy (DG) for gastric cancer.

Methods: A literature search was performed to identify studies comparing R-Y with B-I after DG for gastric cancer from January 1990 to November 2012 in Medline, Embase, Science Citation Index Expanded and the Cochrane Central Register of Controlled Trials in The Cochrane Library. Pooled odds ratios (OR) or weighted mean differences (WMD) with 95%CI were calculated using either fixed or random effects model. Operative outcomes such as operation time, intraoperative blood loss and postoperative outcomes such as anastomotic leakage and stricture, bile reflux, remnant gastritis, reflux esophagitis, dumping symptoms, delayed gastric emptying and hospital stay were the main outcomes assessed. Meta-analyses were performed using RevMan 5.0 software (Cochrane library).

Results: Four randomized controlled trials (RCTs) and 9 non-randomized observational clinical studies (OCS) involving 478 and 1402 patients respectively were included. Meta-analysis of RCTs revealed that R-Y reconstruction was associated with a reduced bile reflux (OR 0.04, 95%CI: 0.01, 0.14; P < 0.00 001) and remnant gastritis (OR 0.43, 95%CI: 0.28, 0.66; P = 0.0001), however needing a longer operation time (WMD 40.02, 95%CI: 13.93, 66.11; P = 0.003). Meta-analysis of OCS also revealed R-Y reconstruction had a lower incidence of bile reflux (OR 0.21, 95%CI: 0.08, 0.54; P = 0.001), remnant gastritis (OR 0.18, 95%CI: 0.11, 0.29; P < 0.00 001) and reflux esophagitis (OR 0.48, 95%CI: 0.26, 0.89; P = 0.02). However, this reconstruction method was found to be associated with a longer operation time (WMD 31.30, 95%CI: 12.99, 49.60; P = 0.0008).

Conclusion: This systematic review point towards some clinical advantages that are rendered by R-Y compared to B-I reconstruction post DG. However there is a need for further adequately powered, well-designed RCTs comparing the same.

Keywords: Billroth I; Distal gastrectomy; Gastric cancer; Meta-analysis; Reconstruction; Roux-en-Y.

Figure 1

Flow diagram depicting the process of identification and inclusion of selected studies.

Figure 2

Roux-en-Y versus Billroth I-randomized controlled trials comparison. A: Operation time; B: Intraoperative blood loss; C: Hospital stay; D: Anastomotic leakage; E: Anastomotic stricture; F: Bile reflux; G: Remnant gastritis; H: Reflux esophagitis; I: Delayed gastric emptying. Pooled weighted mean difference (WMD) or odds ratio (OR) with 95%CI was calculated using the fixed-or random effects model.

Figure 3

Roux-en-Y versus Billroth I-observational non-randomized clinical studie comparison. A: Operation time; B: Intraoperative blood loss; C: Hospital stay; D: Anastomotic leakage; E: Anastomotic stricture; F: Bile reflux; G: Remnant gastritis; H: Reflux esophagitis; I: Dumping symptoms; J: Delayed gastric emptying. Pooled weighted mean difference (WMD) or odds ratio (OR) with 95%CI was calculated using the fixed-or random effects model.

Figure 4

Funnel plot. A: Operation time-randomized controlled trial; B: Remnant gastritis-observational clinical studies. None of the studies lay outside the limits of the 95%CIs, and there was no evidence of publication bias.