Systematic Review Of Serum And Tissue Biomarkers For Predicting Anastomotic Leak After Elective Colorectal Cancer Surgery

Abstract

Background: Anastomotic leak (AL) remains one of the most serious complications following elective colorectal cancer (CRC) surgery, contributing to major morbidity, prolonged hospitalization, and adverse oncologic outcomes. Early identification of patients at risk remains challenging because clinical deterioration often occurs after the leak has progressed. Serum biomarkers reflecting postoperative inflammation and infection have been widely investigated as adjunctive tools for early risk stratification; however, the evidence is heterogeneous and lacks standardized, clinically validated thresholds. Objective: To systematically evaluate and synthesize the diagnostic accuracy of serum biomarkers for predicting anastomotic leak after elective colorectal cancer resection. Methods: A PRISMA-compliant systematic review was conducted. PubMed/MEDLINE, Embase, Scopus, Web of Science, and CENTRAL were searched for studies published between January 2014 and April 2024. Eligible studies were observational cohorts or randomized studies evaluating preoperative or postoperative serum biomarkers for AL prediction in adults undergoing elective CRC surgery. Diagnostic accuracy outcomes included area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and/or extractable contingency data. Methodological quality was assessed using QUADAS-2. Due to heterogeneity in biomarkers, sampling schedules, leak definitions, and cutoffs, a qualitative synthesis was performed. Results: Eight observational cohort studies comprising 2,414 patients were included, with 177 AL events (7.3%). C-reactive protein (CRP) was the most frequently evaluated biomarker and demonstrated moderate-to-high discrimination when measured on postoperative days (POD) 3–5 (AUC range: 0.76–0.91), with peak performance commonly observed on POD 4. Procalcitonin showed promising accuracy in two cohorts (AUC approximately 0.83–0.85 on POD 3–4). Evidence for interleukin-6 and presepsin was limited to single-cohort evaluations but suggested potential utility (AUC ~0.80–0.89). QUADAS-2 assessment indicated an overall moderate risk of bias, most commonly related to patient selection, symptom-triggered verification, and derivation of optimal thresholds within study cohorts. Conclusion: Serum biomarkers—particularly CRP and procalcitonin measured on POD 3–5—show clinically meaningful potential as adjunctive tools for early postoperative risk stratification of anastomotic leak after elective colorectal cancer surgery. However, substantial heterogeneity and lack of prospectively validated cutoff values limit their standalone diagnostic use. Future multicenter studies should focus on external validation of standardized thresholds and evaluation of biomarker-guided management pathways