Abstract
OBJECTIVE: The objective of this article is to assess the rate of anastomotic leak and other perioperative outcomes in patients undergoing esophagectomy with either thoracic or cervical anastomosis. METHODS: This meta-analysis was conducted by searching relevant literature studies in Web of Science, Cochrane Library, PubMed, and Embase databases. Articles that included patients undergoing esophagectomy and compared perioperative outcomes of McKeown with Ivor Lewis procedures were included. The primary outcome parameter was anastomotic leak, and secondary outcome parameters were grade ≥2 anastomotic leak, chylothorax, recurrent laryngeal nerve injury, hospital length of stay, intensive care unit (ICU) length of stay, postoperative mortality rate, operative time, blood loss, R0 resection rate, and lymph nodes examined. RESULTS: A total of eight studies, with 3,291 patients (1,857 Ivor Lewis procedure and 1,434 McKeown procedure) were eligible for analysis. Meta-analysis showed that Ivor Lewis procedure was associated with lower rate of anastomosis leak of all grades [risk ratio (RR), 0.67; 95% confidence interval (CI), 0.55-0.82; P = 0.0001], lower rate of recurrent laryngeal nerve injury (RR, 0.14; 95% CI, 0.08-0.25), and shorter length of hospital stay (weighted mean difference, 0.13; 95% CI, 0.04-0.22). Grade ≥2 anastomotic leak, chylothorax, ICU length of stay, postoperative mortality rate, operative time, blood loss, R0 resection rate, and lymph nodes examined were similar between the two groups. CONCLUSIONS: Although all grades of anastomotic leak and recurrent laryngeal nerve injury are higher in the McKeown procedure, this meta-analysis supports similar short-term outcomes and oncological efficacy between Ivor Lewis and McKeown esophagectomy.