Semimechanical anastomosis during oesophagectomy reduces leaks and stenosis: a propensity score matched analysis.

OBJECTIVES: Several techniques for oesophageal anastomosis during oesophagectomy have been described, all of which are associated with variable leakage and stricture rates. Given the notable morbidity of oesophageal fistula, reducing its incidence is of paramount importance. We report our single-centre experience with the semimechanical (SM) technique as compared to a totally manual (TM) technique. METHODS: Three hundred and twelve partial and subtotal oesophagectomies performed between January 1998 and April 2018 were analysed. The series was split into a training period (January 1998-September 2015), when both TM and SM techniques were used, and a validation period (October 2015-April 2018), during which SM technique became standard practice. Propensity score matching was used to reduce confounding. RESULTS: The training period included 212 oesophagectomies (90 TM, 122 SM); SM technique was initially used in the neck and afterwards also in the thorax, mainly with gastric conduits (92%), whereas the TM group contained a prevalence of jejunal loops (48%). SM anastomosis was associated with a significant reduction in both leak (0.8% vs 12%; P < 0.001) and stricture rate (0% vs 7%; P = 0.005). After propensity score matching, the difference in leak (0% vs 14%; P = 0.013) and stricture rate (0% vs 10%; P = 0.022) was confirmed. During the validation period, which included 100 oesophagectomies performed with SM anastomosis, 1 leak (1%) and 1 stricture (1%) were observed. CONCLUSIONS: SM technique for oesophageal anastomosis outperforms TM technique and allows achieving very low complication rates.