Anastomotic leak is a subject close to the heart of colorectal surgeons, particularly as it can have devastating effects on patients. We know that early recognition and early source control are key to ensuring the best outcome after anastomotic leak.

A multicentre cohort study of serum and peritoneal biomarkers to predict anastomotic leakage after rectal cancer resection
Sparreboom et al Colorectal Dis 2020 22: 36-45

This study is a multicentre cohort which looks at a range of biomarkers to detect anastomotic leak following rectal resection by day 3. The biomarkers were taken from blood (CRP) or from drain fluid (MMP, and interleukins). The authors undertook logistic regression and found that Matrixmetalloprotease 2 (MMP2) and CRP rises were associated with anastomotic leak.

This fits with the rest of the literature. CRP (or the change in CRP) is constantly revisited as a detector of anastomotic leak, with varying degrees of success. MMP is an enzyme which is known to degrade extracellular matrices, leading to anastomotic breakdown. The presence of this protein may be linked to the microbiome (i.e. bacterial product).

A few caveats with this paper. With 38 events, a regression model including 12 explanatory variables may be too many to detect all effects (rule of thumb is 5-10 events per variable). This means that some markers may have been erroneously discounted. The calibration of the model also seems to have been performed on the dataset it was derived from. This will always fit reasonably well, so external validation is required. There is also the question of how to implement this into practice. Most UK surgeons will not use drains following an anterior resection, so testing for MMP by this route is not an option.