Constructing a nomogram model for close-range radiotherapy using radioactive 125I particles to treat radiation-induced pneumonia in advanced lung cancer

Abstract

Purpose: The most common and potentially fatal side effect of postoperative close-range radiotherapy using radioactive 125I particles in the chest is radiation-induced pneumonia (RP). This study aims to develop a dynamic nomogram to accurately predict radiation-induced pneumonia in lung cancer patients following close-range radiotherapy with radioactive 125I particles. Methods: A retrospective analysis was conducted on data from 436 advanced lung cancer patients who underwent close-range radiotherapy using radioactive 125I particles at the Northern Theater General Hospital of the Chinese People's Liberation Army from January 2016 to December 2023. Risk factors for radiation-induced pneumonia were identified through LASSO logistic regression and multivariable logistic regression analysis, which were then used to construct a dynamic nomogram. The predictive performance of the nomogram was validated using ROC curves, calibration plots, and decision curve analysis (DCA). Lastly, the grading of radiation-induced pneumonia and Kaplan-Meier analysis were performed. Results: Preoperative N and M staging, maximum dose, and whether chemotherapy was administered were identified as important factors for accurately predicting radiation-induced pneumonia. A dynamic nomogram for radiation-induced pneumonia was developed using these risk factors. The area under the ROC curve was 0.878 (95% CI 0.814-0.942) for the training cohort and 0.828 (95% CI 0.787-0.870) for the validation cohort. The nomogram demonstrated good calibration. The maximum dose parameter in both the training and validation cohorts achieved the most ideal clinical benefits, showing promising clinical utility. Patients staged as T1 and T3 preoperatively were more likely to develop radiation-induced pneumonia compared to those staged as T2 (P<0.001). Patients staged as M1 preoperatively, those receiving a maximum dose above the mean, and those who had received chemotherapy had a higher probability of developing radiation-induced pneumonia (P<0.001). Conclusion: The developed dynamic nomogram provides a precise and convenient tool for clinical application to predict the risk of radiation-induced pneumonia in lung cancer patients undergoing close-range radiotherapy with radioactive 125I particles.