Biomarkers and anastomotic leakage in colorectal surgery: C-reactive protein trajectory is the gold standard.

Abstract Anastomotic leakage is a feared complication following colorectal surgery. Early prediction results in improved clinical outcome, but accurate predictive factors remain elusive. Many biomarkers have been studied with respect to diagnosis of anastomotic leakage but the concept of trajectory testing, using biomarkers, has not been assessed with regards to early diagnosis of anastomotic leak. C-reactive protein (CRP), procalcitonin (PCT), white cell count (WCC) and gamma-glutamyl transferase were assessed for predictive utility in diagnosing anastomotic leakage with emphasis on identifying an association with change in their levels or trajectory. Levels were collected preoperatively and daily for the first 5 post-operative days on patients undergoing elective colorectal surgery, involving an anastomosis. Anastomotic leakage was defined clinically by operative or radiological intervention. Comparison was made between biomarkers and clinical anastomotic leakage, using receiver operator characteristic curves for logistic models, based on trajectory of the four biomarkers. A total of 197 consecutive patients were analysed. Eleven patients developed clinical anastomotic leakage. An association of biomarker trajectory with anastomotic leakage was observed for WCC, PCT and CRP, but not for gamma-glutamyl transferase. CRP was the superior biomarker based on trajectory, with area under the receiver operator curve of 0.961. This study identifies change in CRP, WCC and PCT as potential markers of anastomotic leakage following colorectal surgery and in particular highlights CRP trajectory as extremely accurate in diagnosing anastomotic leakage requiring intervention. External validation should be sought before incorporating this into routine clinical practice, given the numbers in this study.