Development and validation of a nomogram for predicting immune-mediated colitis in lung cancer patients treated with immune checkpoint inhibitors: a retrospective cohort study in China.

BACKGROUND

The increasing utilization of immune checkpoint inhibitors (ICIs) has led to a concomitant rise in the incidence of immune-related adverse events (irAEs), notably immune-mediated colitis (IMC). This study aimed to identify the clinical risk factors associated with IMC development in patients with lung cancer and to develop a risk prediction model to facilitate personalized treatment and care strategies.

METHODS

The data collected included 21 variables, including sociodemographic characteristics, cancer-related factors, and routine blood markers. The dataset was randomly partitioned into a training set (70%) and a validation set (30%). Univariate and multivariate logistic regression analyses were conducted to identify independent predictors of IMC development. On the basis of the results of the multivariate analysis, a nomogram prediction model was developed. Model performance was assessed via the area under the receiver operating characteristic curve (AUC), calibration curve analysis, decision curve analysis (DCA), and clinical impact curve (CIC).

RESULTS

Among the 2103 patients, 66 (3.14%) developed IMCs. Multivariate logistic regression analysis revealed female sex, small cell lung cancer (SCLC), elevated β2 microglobulin (β2-MG) and globulin (GLB) levels, and an increased neutrophil-lymphocyte ratio (NLR) as independent predictors of IMC development (all < 0.05). Conversely, a higher white blood cell (WBC) count, CD4/CD8 ratio, and platelet-lymphocyte ratio (PLR) were identified as factors associated with a reduced risk of IMC development (all < 0.05). The nomogram prediction model demonstrated good discrimination, achieving an AUC of 0.830 (95% CI: 0.774-0.887) in the training set and 0.827 (95% CI: 0.709-0.944) in the validation set. Analysis of the calibration curve, DCA, and CIC indicated good predictive accuracy and clinical utility of the developed model.

CONCLUSION

This study identified eight independent predictors of IMC development in patients with lung cancer and subsequently developed a nomogram-based prediction model to assess IMC risk. Utilization of this model has the potential to assist clinicians in implementing appropriate preventive and therapeutic strategies, ultimately contributing to a reduction in the incidence of IMC among this patient population.