TITLE:

POSTOPERATIVE CRP, PROCALCITONIN, ESR, AND CLINICAL PARAMETERS AS PREDICTORS OF ANASTOMOTIC LEAK AFTER COLON RESECTION: A PROSPECTIVE COHORT STUDY

 

Introduction

Anastomotic leak (AL) is a serious complication in colorectal surgery, often diagnosed late. Biomarkers may allow earlier detection, but their combined use with bedside parameters remains unclear. This study evaluated the predictive value of C-reactive protein (CRP), procalcitonin (PCT), erythrocyte sedimentation rate (ESR), heart rate (HR), blood pressure (BP), and temperature (Temp) for early AL diagnosis.

 

Methods and Procedures

Seventy consecutive patients undergoing colon resection with primary anastomosis (2021–2025) were prospectively enrolled. Laboratory and clinical data were collected on postoperative days (POD) 0, 3, and 5. The primary endpoint was AL within 30 days (n=4; 5.7%). Univariate analyses compared leak vs no-leak groups. ROC curves (AUC, 95% CI) assessed discrimination and cutoffs. A penalized logistic regression explored independent predictors.

 

Results

Patients with AL had higher CRP and PCT levels, especially on POD3 and POD5. CRP reached median values of 144 mg/L in leaks versus 89 in non-leaks on POD3 (p=0.04, AUC 0.80) and 106 versus 51 on POD5 (AUC 0.79). PCT showed a similar trend, with leak patients at 2.2 versus 0.9 ng/mL on POD3 (AUC 0.82) and 2.3 versus 0.6 on POD5 (AUC 0.84). Optimal thresholds were approximately 100 mg/L for CRP on POD3, approximately 90 mg/L on POD5, and 1.0–1.2 ng/mL for PCT. ESR values were higher in leaks but had modest predictive accuracy (AUC approximately 0.65–0.70).

 

Clinical parameters also differed. HR was elevated in leaks on POD3 (median 104 vs 86 bpm, p=0.05, AUC 0.73), and temperature was significantly higher (38.1°C vs 37.2°C, p=0.04, AUC 0.72). Both markers provided sensitivity and specificity at cutoffs of 100 bpm and 37.8°C. BP trended lower in leaks (118 vs 128 mmHg, p=0.09) but was not discriminatory (AUC approximately 0.60). In multivariable analysis, CRP-POD3 and PCT-POD3 were the strongest predictors, while HR and Temp modestly improved discrimination.

 

Conclusions

CRP and PCT on POD3–POD5 are strong predictors of AL with actionable cutoffs (approximately 100 mg/L and approximately 1 ng/mL). ESR added limited value but contributes to the inflammatory profile. HR and temperature improved prediction and may enhance biomarker-based algorithms. BP was less informative but may contribute in combined scores. Together, routine laboratory markers and vital signs offer an effective surveillance strategy to improve early recognition and patient safety after colon resection.