From Data to Decision: Machine Learning Models for Personalized Survival Predictions in Spinal Cord Gliomas

Introduction: Numerous factors have been associated with the survival outcomes in patients with spinal cord gliomas (SCG). Recognizing these specific determinants is crucial, yet it is also vital to establish a reliable and precise prognostic model for estimating individual survival outcomes. The objectives of this study are to create an array of interpretable machine learning (ML) models developed for predicting survival outcomes among SCG patients; and to integrate these models into an easily navigable online calculator to showcase their prospective clinical applicability.

Methods: The National Cancer Database (NCDB) was utilized to identify adults aged ≥ 18 who were diagnosed with histologically confirmed SCGs between 2010 and 2019. We employed five ML algorithms—TabPFN, CatBoost, XGBoost, LightGBM, and Random Forest—along with the Optuna library for hyperparameter optimization. The models that yielded the highest AUROC values were chosen for integration into the online calculator. To enhance the explicability of our models, we utilized SHapley Additive exPlanations (SHAP) for assessing the relative significance of predictor variables and incorporated partial dependence plots (PDPs) to delineate the influence of singular variables on the predictions made by the top-performing models.

Results: For the 1-year survival analysis, 4,913 patients [5.6% with 1-year mortality]; for the 3-year survival analysis, 4,027 patients (11.5% with 3-year mortality]; and for the 5-year survival analysis, 2,854 patients (20.4% with 5-year mortality) were included. The top models achieved AUROCs of 0.938 for 1-year mortality (TabPFN), 0.907 for 3-year mortality (LightGBM), and 0.902 for 5-year mortality (Random Forest).

Conclusion : This study demonstrates ML techniques can develop personalized prognostic models for SCG patients with excellent discriminatory ability. The interactive online calculator provides individualized risk assessments to be assessed by physicians. Local interpretability informs prediction influences for a given patient. This interpretable methodology aligns with precision medicine, establishing a foundation for continued research leveraging ML's predictive power to enhance patient care.