Introduction Resectional surgery for oesophageal cancer is associated with significant risk of anastomotic dehiscence. In the era of enhanced recovery after surgery (ERAS), some anastomotic leaks may not present until after the patient is discharged home. This study aims to identify markers which may predict likelihood of anastomotic dehiscence.

Methods All patients undergoing oesophageal cancer resection from January 2001 to November 2011 were identified from the Upper GI database where data is recorded prospectively. Retrospective review of patient demographics, operation type, radiology, blood results, histology, length of stay and mortality was performed. Blood results of patients with clinical evidence of anastomotic leak and 50 matched controls were collected for the first 10 post-operative days.

Results 309 patients (median age 66 years) underwent oesophageal resection during this period. In-hospital mortality occurred in 22 patients (7.1%). 216 patients (69.9%) underwent routine contrast studies and 12 radiological leaks were identified. Two patients with radiological leaks had clinical findings suggestive of anastomotic leak, all were managed conservatively. Anastomotic leakage occurred in a further 20 patients. 13 of these should have undergone contrast swallow at day 7 according to the protocol at the time. Two had contrast swallows at day 7, which were normal, but subsequently leaked. 16 patients demonstrated clinical deterioration at days 2–7 which prompted either imaging with CT, endoscopy or surgical exploration. Two patients were never fit enough to undergo contrast study. The overall mortality in patients with leak was 5/32 (15.6%). Results of mean white cell count on days 1 to 9 are given in the Abstract PWE-017 table 1 below. There was a significant difference in white cell count from days 6 to 9 between the control group and leak group. There was a difference in mean CRP in the control group (163.91 mg/l) compared to the leak group (267.13 mg/l), p=0.00021. There was no significant difference in mean albumin between groups.

Conclusion Routine contrast swallows are of limited value following oesophageal cancer resection. If there is suspicion of anastomotic leakage, radiology and endoscopy can be utilised. The finding of raised inflammatory markers in the absence of other causes is associated with anastomotic leakage and should be further investigated. In the clinically well patient with normal blood results, ERAS can be safely implemented after oesophageal cancer resection.

Competing interests None declared.