Prognostic value of routine blood biomarkers in 3-year survival of resectable colorectal cancer patients: a prognostic nomogram for clinical practice.

Abstract

Purpose: This study aimed to develop a prognostic model for colorectal cancer (CRC) patients using biomarkers from routine preoperative peripheral blood examinations combined with clinical factors.

Methods: This observational study comprised CRC patients (stages I-III) who underwent curative surgery between January 2011 and December 2019. Study variables included patient demographics, tumour characteristics, and immune/inflammatory markers from preoperative blood tests. Cut-off thresholds for continuous variables were determined using maximally selected rank statistics. Univariate and multivariate analyses identified variables associated with 3-year cancer-specific survival (CSS) and disease-free survival (DFS). Cox regression models were developed and validated using a random split-sample approach. Nomograms based on these models were constructed, and receiver operating characteristic (ROC) curves were generated for 12, 24 and 36 months.

Results: A total of 764 patients were included. Independent factors for 3-year DFS included laparoscopic surgery, prognostic nutritional index (PNI), neutrophil count, lymphocyte count, and Charlson comorbidity index. The DFS prediction model showed AUC values of 66.6%, 64.8%, and 69% for years 1, 2, and 3, respectively. For CSS, independent factors included age, systemic immune-inflammation index (SII), serum albumin, and platelet count, with AUC values of 89.2%, 76.8%, and 71% for years 1, 2, and 3. The most significant contributors to the CSS model were SII and platelet cut-off values.

Conclusion: Inflammatory biomarkers combined with clinical parameters robustly predict 3-year survival outcomes in CRC patients undergoing curative resection. These findings highlight the importance of systemic inflammation in CRC prognosis and support its inclusion in preoperative risk stratification.