Effect of intraoperative botulinum toxin injection on delayed gastric emptying and need for endoscopic pyloric intervention following esophagectomy: a systematic review, meta-analysis, and meta-regression analysis.

Abstract

The aim of this study was to evaluate the effect of intraoperative botulinum toxin (BT) injection on delayed gastric emptying (DGE) and need for endoscopic pyloric intervention (NEPI) following esophagectomy. In compliance with Preferred Reporting Items for Systematic reviews and Meta-Analyses statement standards, a systematic review of studies reporting the outcomes of intraoperative BT injection in patients undergoing esophagectomy for esophageal cancer was conducted. Proportion meta-analysis model was constructed to quantify the risk of the outcomes and direct comparison meta-analysis model was constructed to compare the outcomes between BT injection and no BT injection or surgical pyloroplasty. Meta-regression was modeled to evaluate the effect of variations in different covariates among the individual studies on overall summary proportions. Nine studies enrolling 1070 patients were included. Pooled analyses showed that the risks of DGE and NEPI following intraoperative BT injection were 13.3% (95% confidence interval [CI]: 7.9-18.6%) and 15.2% (95% CI: 7.9-22.5%), respectively. There was no difference between BT injection and no BT injection in terms of DGE (odds ratio [OR]: 0.57, 95% CI: 0.20-1.61, P = 0.29) and NEPI (OR: 1.73, 95% CI: 0.42-7.12, P = 0.45). Moreover, BT injection was comparable to pyloroplasty in terms of DGE (OR: 0.85, 95% CI: 0.35-2.08, P = 0.73) and NEPI (OR: 8.20, 95% CI: 0.63-105.90, P = 0.11). Meta-regression suggested that male gender was negatively associated with the risk of DGE (coefficient: -0.007, P = 0.003). In conclusion, level 2 evidence suggests that intraoperative BT injection may not improve the risk of DGE and NEPI in patients undergoing esophagectomy. The risk of DGE seems to be higher in females and in early postoperative period. High quality randomized controlled trials with robust statistical power are required for definite conclusions. The results of the current study can be used for hypothesis synthesis and power analysis in future prospective trials.