Prediction and early detection of the leaking colorectal anastomosis – do inflammatory pathways hold the key?

Anastomotic leak remains a significant complication following colorectal surgery. Strategies for prevention and early detection remain a subject of considerable attention and investigation given the morbidity this causes our patients, not to mention both the health-economic and oncological costs. This month, BJS Open published two papers exploring the potential for utilising inflammatory mediators as predictive biomarkers for anastomotic complications.

Shi et al

Shi et al undertook a prospective study of 119 patients undergoing restorative surgery or APER for (predominantly left sided) colorectal cancer¹. Drain fluid assays of known inflammatory mediators were undertaken and several prediction models evaluated, demonstrating correlation with both anastomotic leak and pelvic collection.

Widespread adoption of this would be challenging given that enhanced recovery after surgery (ERAS) programmes have led to a significant reduction in the use of drains for many patients. What is not clear from the present study is whether the inflammatory mediators assayed are truly predictive of a leak i.e. representative of a pro-inflammatory environment which is identifiable prior to the leak occurring, or a means of early detection once the leak has occurred. This is particularly prudent given the AScore reported by the authors had the greatest AUC (0.88) on the first post-operative day. If this assay is truly predictive, moving it into the intra-operative arena for peritoneal [rather than drain] fluid may provide helpful information for real-time decision making in high risk or complex cases whilst eliminating the requirement for drainage.

Holmgren et al

The second study by Holmgren et al also investigated for a pro-inflammatory state by retrospectively analysing the serum of 41 patients with anastomotic complications and their matched controls using mass spectrometry for a pre-determined panel of inflammatory mediators and ELISA for a post hoc sensitivity analysis. A number of known pro-inflammatory peptides were statistically significantly raised in the cohort of patients who suffered from anastomotic leakage. Interestingly, when tumour levels of these peptides were evaluated by immunohistochemistry no difference was seen. This complements the findings of Shi et al and suggests a systemic pro-inflammatory environment. It will be of interest to learn the results of a validation dataset derived from a prospective cohort.

Take home

These papers show promise for further investigation into the identification, modelling and manipulation of dysregulation in the native inflammatory pathway to help understand and address this age old and elusive complication.