Anastomotic leak (AL), which is when the join between two segments of bowel fails in the immediate post­operative period, is the most feared major complication of colorectal surgery. This surgery is performed as a primary treatment for colorectal cancer and in 70% of patients with inflammatory bowel disease. AL rates are between 3% and 10% of surgeries, and mortality rates up to 39%. Few improvements have been made in the rate of leak in the past 20 years, despite the best efforts of surgeons around the globe.

Recent work in models suggests a specific role for the microbiome in AL. The research suggests that in patients with AL the microbiome plays a role in inhibiting the wound healing process after anastomosis, with particular bacteria associated with a higher incidence of leaks when present in abundance at the site of anastomosis. 

We have previously shown that the microbiome can respond to tissue damage and inflammation, upregulating genes that aid bacterial survival while also potentially damaging the host. In particular, bacterial quorum sensing which drives bacterial accumulation at the site of injury and bacterial proteases which damage healing tissue and may impair the healing process leading to anastomotic leak.
We hypothesise that the microbiome profile at the site of anastomosis will be predictive for AL risk, and that the presence of specific bacteria and bacterial genes will be associated with AL.

This study proposes to collect samples of the bacteria present at the site of anastomosis during colorectal procedures at the John Hunter Hospital and Newcastle Private Hospital. We will collect samples from 200 patients and the profile of bacteria present will be compared to the clinical outcomes of the patients, to identify associations between bacteria and leak in these patients.

Samples collected will be analysed in two ways. Firstly, we will examine the genetic profile of all bacteria in the sample to assess whether the patient’s bacterial community (microbiome) as a whole, is associated with the presence of bacteria capable of breaking down collagen. Secondly, we will culture (grow) bacteria to assess what species are present, and whether they express genes and proteins that impair wound healing. In addition, we will perform simple laboratory assays to examine how efficient the bacteria, isolated from each patient, are at degrading collagen a major protein involved in wound healing. 

Identification of bacteria associated with AL will pave the way for rapid translation of diagnostic testing to allow identification and prophylactic treatment of patients at high risk for post-surgery AL. This would have a profound effect on the management and survival of these patients.



A/Prof Simon Keely, Dr Peter Pockney, A/Prof Stephen Smith, A/Prof Ian Grainge, Dr Andrea Mathe

Project type 
Project Grant
Year of funding