Comparison of the clinical outcomes after esophagectomy between intrathoracic anastomosis and cervical anastomosis: a systematic review and meta-analysis

Abstract

Objectives: Esophageal cancer is a high-mortality disease. Esophagectomy is the most effective method to treat esophageal cancer, accompanied with a high incidence of post-operation complications. The anastomosis has a close connection to many severe post-operation complications. However, it remains controversial about the choice of intrathoracic anastomosis (IA) or cervical anastomosis (CA). The study was conducted to compare the clinical outcomes between the two approaches. Methods: We searched databases for both randomized controlled trials (RCTs) and cohort studies comparing post-operation outcomes between IA and CA. Primary outcomes were the incidences of anastomotic leakage and mortality. Secondary outcomes were the incidences of anastomotic stenosis, pneumonia and re-operation. Results: Twenty studies with a total of 7,479 patients (CA group: n = 3,183; IA group: n = 4296) were included. The results indicated that CA group had a higher incidence of anastomotic leakage than IA group (odds ratio [OR] = 2.05, 95% confidence intervals [CI] = 1.61–2.60, I2 = 53.31%, P < 0.01). Subgroup analyses showed that CA group had higher incidences of type I (OR = 2.19, 95%CI = 1.05–4.57, I2 = 0.00%, P = 0.04) and type II (OR = 2.75, 95%CI = 1.95–3.88, I2 = 1.80%, P < 0.01) anastomotic leakage than IA group. No difference was found in type III anastomotic leakage (OR = 1.23, 95%CI = 0.82–1.86, I2 = 20.92%, P = 0.31). The 90-day mortality (OR = 1.66, 95%CI = 1.11–2.47, I2 = 0.0%, P = 0.01) in IA group were lower than that in CA group. No difference was found in in-hospital mortality (OR = 1.31, 95%CI = 0.91–1.88, I2 = 0.00%, P = 0.15) and 30-day mortality (OR = 1.08, 95%CI = 0.69–1.70, I2 = 0.00%, P = 0.74). Conclusions: IA might be a better anastomotic approach than CA, with a lower incidence of anastomosis leakage and no increase in short-term mortality. Significant heterogeneity and publication bias might limit the reliability of the results. More high-quality studies are needed to verify and update our findings.