A predictive model for colorectal cancer complicated with intestinal obstruction based on specific inflammation score.

ABSTRACT

PURPOSE: Inflammatory factors play an important role in the onset and progression of colorectal cancer (CRC). This study aimed to develop and validate a novel scoring system that utilizes specific inflammatory factor indicators to predict intestinal obstruction in CRC patients. METHODS: This study conducted a retrospective analysis of 1,470 CRC patients who underwent surgical resection between January 2013 and July 2018. These patients were randomly allocated to the training group (n = 1060) and the validation group (n = 410). Univariate and multivariate logistic regression analyses were performed to identify independent predictive factors for intestinal obstruction. The CRC peculiar inflammation score (CPIS), comprising lymphocyte-to-monocyte ratio (LMR), prognostic nutrition index (PNI), and alanine transaminase-to-lymphocyte ratio index (ALRI) scores, was significantly associated with the occurrence of intestinal obstruction. A nomogram combining CPIS with other clinical features was developed to predict this occurrence. Model accuracy was assessed by determining the area under the receiver operating characteristic (ROC) curve (AUC). RESULTS: The CPIS generated by multi-factor logistic regression was as follows - 1.576 × LMR - 0.067 × PNI + 0.018 × ALRI. Using CPIS cutoff values of 50% (- 7.188) and 85% (- 6.144), three predictive groups were established. Patients with a high CPIS had a significantly higher risk of intestinal obstruction than those with a low CPIS (odds ratio [OR] 10.0, confidence interval [CI] 5.85-17.08, P < 0.001). The predictive nomogram demonstrated good calibration and discrimination abilities. The AUC of the ROC curve for the obstruction nomogram was 0.813 (95% CI 0.777-0.850) in the training set and 0.806 (95% CI 0.752-0.860) in the validation set. The calibration curve exhibited neither bias nor high credibility. Decision curve analysis indicated the utility of this predictive model. CONCLUSION: CRC-associated intestinal obstruction is closely linked to inflammatory markers in patients. CPIS is a CRC-specific inflammatory predictive score based on a combination of inflammatory-related indicators. A high CPIS serves as a strong indicator of intestinal obstruction. Its integration with other clinical factors and preoperative inflammatory-specific indicators significantly enhances the diagnosis and treatment of CRC patients with intestinal obstruction.