Postoperative C-Reactive Protein-to-Albumin Ratio Predicts Anastomotic Leakage After Elective Right Hemicolectomy for Colon Cancer.

AIM: Anastomotic leakage (AL) is a life-threatening complication of colorectal surgery carrying significant risks of postoperative morbidity and mortality. Timely identification of patients at elevated risk for AL is critical for optimizing surgical outcomes. This study aimed to evaluate factors associated with AL and to assess the predictive utility of the postoperative day 3 C-reactive protein-to-albumin ratio (CAR) in patients undergoing elective right hemicolectomy for colon cancer. METHODS: This retrospective observational cohort study included patients who underwent elective right hemicolectomy with primary ileocolic anastomosis for colon cancer at Sancaktepe Şehit Prof. Dr. İlhan Varank Training and Research Hospital (January 2020 to January 2025). Patients undergoing emergency surgery, multivisceral resection, diverting stoma formation, or with incomplete data were excluded. ROC curve analysis and limited univariable logistic regression were used to evaluate the predictive performance of postoperative day 3 CAR. RESULTS: A total of 194 patients were included, and AL developed in 14 patients (7.2%). No significant differences were observed between groups regarding age, body mass index, operative duration, surgical technique, or comorbidities, confirming the homogeneity of the study population. Early enteral feeding was initiated significantly later in patients with AL (3.36 ± 0.74 vs. 2.83 ± 0.76 days, p = 0.021). Postoperative day 3 C-reactive protein (CRP) levels were significantly higher in the AL group (451.3 [398.7-534.2] mg/L vs. 98.4 [71.2-138.6] mg/L, p < 0.001), whereas serum albumin levels were significantly lower (2.30 ± 0.16 g/dL vs. 3.09 ± 0.34 g/dL, p < 0.001). The postoperative day 3 CAR was markedly elevated in patients with AL compared with those without leakage (196.4 [168.2-248.7] vs. 32.1 [22.4-45.8], p < 0.001). Receiver operating characteristic (ROC) curve analysis demonstrated strong predictive performance for CAR, with an area under the curve (AUC) of 0.91 (95% CI: 0.84-0.98, p < 0.001). The optimal cutoff value was 148.7, yielding a sensitivity of 85.7% and specificity of 88.3%. In limited univariable logistic regression analysis, postoperative day 3 CAR remained significantly associated with AL (odds ratio [OR]: 6.84, 95% CI: 3.11-11.52, p < 0.001). CONCLUSIONS: On postoperative day 3, CAR demonstrated strong discriminatory performance for anastomotic leakage after elective right hemicolectomy in this retrospective cohort. The high negative predictive value (98.8%) suggests potential utility as a rule-out tool for early postoperative risk stratification; however, these findings require prospective multicenter validation before any clinical implementation can be considered.