Implications of anastomotic leakage after total gastrectomy for gastric carcinoma.

Background/objectives: Oesophagojejunal anastomotic leakage is a serious complication following total gastrectomy, and is reported to be decreasing in frequency. This study was an audit of the radiological and clinical frequency of such leakage and its consequences.
Methods: A retrospective (1983-2000) cohort study was undertaken in a tertiary institution where 175 patients had undergone a total gastrectomy for gastric carcinoma with oesophagojejunal reconstruction using a stapling device and a 50 cm J-loop. Anastomotic leakage was sought 3-9 days postoperatively using a Gastrograffin (diatrizoate meglumine) swallow.
Results: Leakage was demonstrated in 7 patients (4%), being subclinical in 2, minor in 4, and fatal in 1. There was no correlation between leakage and patient factors (age, medical risk, haemoglobin, albumin), surgical factors (surgical seniority, approach, reconstruction, splenectomy, lymph node dissection) or tumour factors (stage, nodes examined, and margin positivity). However, intraoperative difficulties or mishaps were recorded in most cases of leakage. Subclinical leakage was marked by an uneventful postoperative course, and low-volume enterocutaneous fistulas were self-limiting. One patient developed a subphrenic abscess that required drainage. One patient suffered an intrathoracic leak which proved fatal.
Conclusions: Anastomotic leakage was an infrequent complication of total gastrectomy when using a stapling device and a 50 cm J-loop. It was related to intraoperative surgical difficulty and mishap rather than conventional patient and tumour factors. It was subclinical or self-limiting, if occurring in the abdomen, but fatal if in the chest.