Abstract
<title>Abstract</title> <p> Background This study aimed to identify clinicopathologic and serum-based inflammatory variables associated with tumor recurrence in patients with rectal cancer treated with neoadjuvant chemoradiotherapy (CRT) followed by curative-intent resection, and to develop a practical, recurrence-oriented scoring system. Methods A cohort of 129 patients with rectal adenocarcinoma was included in the study. Univariate and multivariate logistic regression analyses were used to evaluate associations with tumor recurrence. Kaplan-Meier and ROC analyses were applied to assess the predictive accuracy of the proposed scoring model. Results Multivariate analysis identified elevated preoperative neutrophil-to-lymphocyte ratio (NLR) (NLR > 3.0 HR:2.884; <italic>p</italic> = 0.006) and systemic-inflammatory-index (SII) (SII > 750 HR:4.557; <italic>p</italic> = 0.001), presence of extra-nodal extension (ENE) (HR: 6.831; <italic>p</italic> = 0.002), CEA level > 5.0 (HR:5.975; <italic>p</italic> = 0.018), and pT3 (HR:7.349; <italic>p</italic> = 0.012) and pT4 (HR:15.320; <italic>p</italic> = 0.003) as independent predictors of recurrence. Patients were categorized into low-, intermediate-, and high-risk groups using the REACT score, a three-tiered risk model. In Kaplan-Meier analysis, the REACT score was substantially correlated with disease-free survival (DFS) (p < 0.001) and showed high prognostic discrimination (AUC: 0.812). 7.9% of low-risk, 32.1% of intermediate-risk, and 100% of high-risk patients experienced a recurrence. Conclusion The REACT score provides a low-cost, easily applicable, and clinically relevant method for perioperative recurrence risk stratification in rectal cancer. </p>