Khirurgiia (Mosk). 2020;(8):82-87. doi: 10.17116/hirurgia202008182.
OBJECTIVE: Systematic review and meta-analysis of data on C-reactive protein (CRP) as a predictor of anastomotic leakage (AL) after surgery for colorectal cancer.
MATERIAL AND METHODS: Literature searching was performed in Medline, Elibrary, Scopus, Web of Science databases. Literature request consisted of keywords «CRP», «colorectal surgery», «anastomotic leakage» for the period 2008-2018. Meta-analysis included 2 manuscripts for the second postoperative day, 7 articles for the third postoperative day and 6 articles for the fourth postoperative day. ROC-analysis was made to determine optimal prognostic values.
RESULTS: ROC-curve for the second postoperative day - AUC 0.758; optimal CRP value - 154 mg/l (sensitivity 70.1%, specificity 55.6%), 95% confidence interval 0.698-0.819. ROC-curve for the third postoperative day - AUC 0.715; optimal CRP value - 144.5 mg/l (sensitivity 79.1% specificity 60.3%), 95% confidence interval 0.68-0.75. ROC-curve for the fourth postoperative day - AUC 0.767; optimal CRP value - 122.91 mg/l (sensitivity 72.3% specificity 60%), 95% confidence interval 0.73-0.804.
CONCLUSION: Increased CRP is an early predictor of AL after surgery for colorectal cancer. CRP level ≥144.5 mg/l on the third postoperative day can predict AL (sensitivity 79%, specificity 60%).