Association of image-defined risk factors with clinical features in thoracic neuroblastoma and the development of a prognostic prediction model

Background: Image-defined risk factors (IDRFs) were critical in managing and predicting outcomes for neuroblastoma. This study systematically evaluated baseline clinical features, IDRFs prevalence, and prognosis in thoracic neuroblastoma, explored the association between IDRFs and these features, and developed a nomogram to predict event-free survival (EFS). Methods: Clinical and prognostic data were collected retrospectively from pediatric patients diagnosed with thoracic neuroblastoma at Children's Hospital of Nanjing Medical University. Logistic regression was used to identify factors associated with IDRFs presence. Kaplan-Meier analysis assessed the effect of IDRFs on survival. Cox regression identified independent predictors of EFS, with model selection based on the Akaike Information Criterion (AIC). A nomogram was developed from these predictors and evaluated using calibration curves, time-dependent AUC curves, and decision curve analysis (DCA). Results: < 0.05). Cox regression identified infiltrative IDRFs, vascular IDRFs, LDH ≥273.00 U/L, and TP ≥ 70.00 g/L as independent risk factors for poor EFS. Nomogram based on these variables showed favorable discrimination with C-index = 0.77, good calibration, and clinical utility. Conclusions: Thoracic neuroblastoma had unique demographic and clinical features. TP levels and tumor size were associated with the presence of IDRFs, which significantly affected EFS. The nomogram accurately predicted EFS and held potential for clinical utility.