Nomograms for Predicting the Overall Survival and Cancer-specific Survival of Patients with Cerebral Gliomas: An Analysis of the Surveillance, Epidemiology and End Results (SEER) Database

Introduction: Cerebral gliomas pose a therapeutic challenge due to their heterogeneity. Given the need for a predictive model assessing the overall survival (OS) and cancer-specific survival (CSS), we aim to construct and validate a nomogram for cerebral glioma patients.

Methods: Data of patients diagnosed with cerebral gliomas between 1975 and 2019 were extracted from SEER database. We identified and randomized 4329 cerebral glioma patients into a training (n=3030) and validation group (n=1299) at a ratio of 7:3. Independent OS and CSS prognostic factors used to build both nomograms were determined using multivariate Cox regression performed on the training set. Discrimination and consistency for both nomograms at 6-, 12-, 24-, and 36-months were assessed using C-indices, area under the curve (AUC), and calibration curves.

Results: Overall, patients had a median age of 61 years, and 58% were male. The median OS and CSS were 16.0 months (95% CI: 15.1-16.9) and 17.0 months (95% CI: 16.1-17.9), respectively. Multivariate Cox regression identified age, race, chemotherapy, radiotherapy, surgery, histological subtype, and WHO grade as independent prognostic factors affecting OS and CSS. In the training cohort, the C-indices for the OS nomogram and the CSS nomogram were 0.805 (95% CI: 0.793-0.817) and 0.804 (95% CI: 0.792-0.816), respectively. In the validation cohort, the c-indices for the OS nomogram and CSS nomogram were 0.798 (95% CI: 0.764-0.832) and 0.795 (95% CI: 0.759-0.831), respectively. The AUC values at 6-, 12-, 24-, and 36-months were 0.870, 0.853, 0.852, and 0.874 for OS and 0.873, 0.854, 0.851, and 0.871 for CSS, respectively. Calibration curves exhibited great consistency between the predicted and actual survival probabilities for both OS and CSS.

Conclusion : Our nomogram shows great consistency in predicting patients’ survival based on a comprehensive list of demographic and clinical features. This nomogram may help in guiding individualized treatment in cerebral glioma patients.