OBJECTIVE. Gastroesophageal anastomotic leak after cancer resection has a m
ortality rate of up to 60% and significant morbidity, whatever the mode of
treatment. We assessed the efficacy of esophageal stenting as a therapeutic
option to reduce the mortality and morbidity associated with symptomatic i
ntrathoracic anastomotic leakage.
SUBJECTS AND METHODS. During a 52-month period, 14 patients had placement o
f stents for clinically significant postoperative leaks: 10 patients had an
esophagogastrectomy and four patients had a total gastrectomy with esophag
ojejunal anastomosis. Thirteen of 14 patients had tumors that were histolog
ically staged as T3 N1 MO or worse. Significant anastomotic leaks were reve
aled by a contrast-enhanced study at 3-28 days after surgery. Stents were i
nserted in patients in whom the leakage was debilitating or initial conserv
ative treatment had failed. Stenting outcome in terms of clinical and radio
logic healing, hospital stay, survival, and complications was assessed.
RESULTS. No procedural morbidity or 30-day mortality occurred. Immediate po
stprocedural leak occlusion was obtained in all patients. Clinical healing
of the leak occurred in 13 (92.8%) of 14 patients, with a median healing ti
me of 6 days. Of the 13 patients, healing occurred within 10 days in 10 pat
ients (76.9%). Eight of these 10 early closures received a knitted nitinol
stent (p = 0.02). One patient (7%) died as a consequence of leakage at 135
days. Median survival for all 14 patients was Il months (Kaplan-Meier metho
d). Complications included five episodes of food blockages in three patient
s, which required endoscopic clearance, and one case of stent-related upper
gastrointestinal hemorrhage. No patients developed anastomotic structure o
r occlusive epithelial hyperplasia.
CONCLUSION. Covered esophageal stenting appears to reduce the mortality and
morbidity of symptomatic anastomotic leakage after surgery for gastroesoph
ageal cancer. Knitted nitinol stents may be best suited to this purpose.