Related Experiment Video
Updated: Mar 8, 2026
Predictive Immune Modeling of Solid Tumors
Published on: February 25, 2020
Construction and Validation of a Risk Prediction Model for Immune-Related Adverse Events in Lung Cancer Patients
Yitao Chen1,2,3, Xiaoyu Xiang3, Lichun Wang3
1Department of Oncology, Inner Mongolia Medical University, Hohhot, Inner Mongolia, China.
Objective:
To examine the factors influencing immune-related adverse events (irAEs) in lung cancer (LC) patients undergoing treatment with immune checkpoint inhibitors (ICIs) and to develop and validate a prediction model.
Study Design:
An observational study. Place and Duration of the Study: Department of Oncology, the Second People's Hospital of Datong, Shanxi, China, from January 2020 to July 2024.
Methodology:
The demographic characteristics and blood biomarkers of 108 LC patients treated with ICIs were evaluated retrospectively within one week before and after at least one course of therapy treatment. To compare categorical variables, χ2 or Fisher's exact test was used. Univariate and multifactorial logistic regression analyses were conducted using the forward stepwise method to determine independent predictors of irAEs. Brain metastasis (BrMs), ECOG-PS, TNM stage, IL-1 (after), IL-4 (after), IL-6 (after), IL-8 (after), IL-10 (after), WBC (after), and NLR (after) were statistically significant in the univariate logistic analysis and were then included in the multifactorial analysis.
Results:
Univariate and multivariate logistic regression models were employed to estimate the odds ratios (OR) and their 95% confidence intervals (CI). The analysis revealed that BrMs (OR = 0.053, 95% CI: 0.009 to 0.305, p = 0.001), interleukin-10 (IL-10, OR = 1.866, 95% CI: 1.347 to 2.584, p <0.001), and white blood cells (WBC, OR = 1.310, 95% CI: 1.060 to 1.620, p = 0.013) were identified as independent factors influencing the occurrence of irAEs (p <0.05). The joint prediction model achieved a C-index of 0.937 (95% CI: 0.889-0.981). The ROC curve for the subjects achieved an AUC of 0.966 (95% CI: 0.935-0.997). The calibration curve closely approximated the ideal curve, suggesting that the model exhibited strong discrimination and calibration. The decision curve analysis demonstrated that the predictive model provided a substantial net clinical benefit.
Conclusion:
The model developed using post-dose IL-10, WBC levels, and BrMs indicators demonstrates significant diagnostic value for irAEs.
Key Words:
Immune checkpoint inhibitors, Lung cancer, Immune-related adverse events, White blood cells, Interleukin-10, Brain metastases.