Abstract
OBJECTIVE: This study aimed to delineate the long-term overall survival (OS) landscape and identify factors associated with OS in a large, population-based cohort of patients diagnosed with vestibular schwannoma (VS). MATERIALS AND METHODS: Clinical, demographic, and treatment data for VS patients from the SEER database (2000-2019) were extracted and randomly divided into training and validation cohorts. Cohort comparability was assessed using Chi-square, Fisher's exact, and Mann-Whitney U tests. Univariate and multivariate Cox proportional hazards models were used to identify factors associated with OS. Nomograms for predicting 3-, 5-, and 10-year OS were constructed. Model performance was evaluated using the concordance index (C-index), receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA). RESULTS: From the initial pool of 28,693 VS patients in the SEER database, 20,235 were included in the analysis. Among these, 14,167 patients (70.01%) were allocated to the training cohort, and 6,068 patients (29.99%) to the validation cohort. Multivariate Cox regression identified age (hazard ratio [HR] = 12.895; 95% confidence interval [CI] 8.347-19.809; p < 0.001), sex (HR = 0.761; 95% CI 0.683-0.849; p < 0.001), race (HR = 0.214; 95% CI 0.089-0.515; p < 0.001), tumor size (HR = 3.024; 95% CI 2.464-3.712; p < 0.001), primary site surgery (HR = 0.409; 95% CI 0.300-0.557; p < 0.001), and radiation therapy (HR = 0.812; 95% CI 0.710-0.929; p = 0.002) as independent prognostic factors. Nomograms based on these variables demonstrated robust predictive capability, with a C-index of 0.737 ± 0.065 for the training cohort and an AUC of 0.759, 0.766, and 0.790 for predicting 3-, 5-, and 10-year OS, respectively. In the validation cohort, the model yielded a C-index of 0.763 ± 0.053 and AUCs of 0.757, 0.765, and 0.760 for the corresponding timeframes. Other factors, such as year of diagnosis, reporting source, surgical-radiation sequence, laterality, and chemotherapy, did not exhibit significant associations with OS. CONCLUSION: This analysis of over 20,000 VS patients provides a comprehensive overview of long-term survival in this patient population. We identified age, sex, race, tumor size, and treatment modality (surgery and radiotherapy) as factors associated with OS. The developed nomograms offer a tool for visualizing the combined impact of these demographic and clinical factors on long-term survival, which may contribute to comprehensive patient assessment and counseling at the time of diagnosis.