Early diagnosis of anastomotic leakage after colorectal cancer surgery using an inflammatory factors-based score system.

BACKGROUND

Anastomotic leakage (AL) is a severe complication after colorectal surgery. This study aimed to investigate a method for the early diagnosis of AL after surgical resection by analysing inflammatory factors (IFs) in peritoneal drainage fluid.

METHODS

Abdominal drainage fluid of patients with colorectal cancer who underwent resection between April 2017 and April 2018, were prospectively collected in the postoperative interval. Six IFs, including interleukin (IL)-1β, IL-6, IL-10, tumour necrosis factor (TNF)-α, matrix metalloproteinase (MMP)2, and MMP9, in drainage were determined by multiplex immunoassay to investigate AL (in patients undergoing resection and anastomosis) and pelvic collection (in patients undergoing abdominoperineal resection). Sparreboom and colleagues' prediction model was first evaluated for AL/pelvic collection, followed by a new IF-based score system (AScore) that was developed by a least absolute shrinkage and selection operator (LASSO) regression, for the same outcomes. The model performance was tested for the area under the curve (AUC), sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV).

RESULTS

Out of 123 patients eligible, 119 patients were selected, including 12 patients with AL/pelvic collection. Sparreboom and colleagues' prediction model was documented with the best diagnostic efficacy on postoperative day 3 (POD3), with an AUC of 0.77. After optimization, AScore on POD3 increased the AUC to 0.83 and on POD1 showed the best diagnostic efficiency, with an AUC of 0.88. Based on the Youden index, the cut-off value of AScore on POD1 was set as -2.46 to stratify patients into low-risk and high-risk groups for AL/pelvic collection. The model showed 90.0 per cent sensitivity, 69.7 per cent specificity, 98.4 per cent NPV, and 25.0 per cent PPV.

CONCLUSIONS

The early determination of IFs in abdominal drainage fluid of patients undergoing colorectal surgery could be useful to predict AL or pelvic collection.