There seems to be a decreased anastomotic leak rate and a late stricture formation after linear-stapled (LS) cervical oesophagogastric anastomosis compared with hand-sewn (HS) technique. The aim of our study was to compare the surgical outcomes of intrathoracic side-to-side LS and end-to-end HS anastomosis after transthoracic oesophagectomy.
We conducted a retrospective review of all patients undergoing Ivor Lewis oesophagectomy with LS or HS anastomosis for neoplasia at our institution from 2005 to 2012. Anastomotic leak was radiologically and clinically graded as minor or major.
End-points included overall and major leak rate, morbidity, mortality, length of hospital stay and endoscopically identified late anastomotic stricture. A propensity score-matched analysis was done to compensate for the differences in baseline characteristics between HS and LS groups.
Multivariable analyses of the associations of anastomotic technique and other preoperative and pathological variables with anastomotic leak and stricture were performed. There were 415 patients, 134 with HS and 281 with LS anastomoses.
Anastomotic leak occurred in 56 patients (13.5%), significantly more after HS than LS technique (20.9 vs 10.0%; P = 0.002). Major leak rate was not significantly different (9.0 vs 5.7%; P = 0.216, respectively).
Overall morbidity (54.7%), in-hospital mortality (3.9%) and length of hospital stay (median 12 days) were not affected by the anastomotic technique. A follow-up endoscopic evaluation was available in 248 patients (59.8%).
An anastomotic stricture was detected in 24 patients (9.7%), significantly more after HS than LS technique (20.3 vs 6.3%; P = 0.002). The propensity score-matched analysis of 105 patient pairs confirmed a significantly decreased overall leak rate (11.4 vs 22.9%; P = 0.045) and stricture formation (7.5 vs 18.2%; P = 0.041) in LS technique compared with HS technique.