LUNAR: A Deep Learning Model to Predict Glioma Recurrence Using Integrated Genomic and Clinical Data

Gliomas account for approximately 25.5% of all primary brain and central nervous system tumors, with a striking 80.8% of these being malignant. The prognosis varies significantly; low-grade gliomas (LGGs) can exhibit 5-year survival rates of up to 80%, while higher-grade gliomas (HGGs) often see rates below 5%. Recurrence is a common challenge, occurring in 52%-62% of LGGs and 90% of HGGs, complicating clinical management and treatment planning. Currently, no widely available models exist for predicting glioma recurrence, which is critical for optimizing patient outcomes. Machine learning (ML) and deep learning (DL) techniques have shown promise in predicting recurrence for various cancers, most using Electronic Health Records (EHR). This study introduces g L ioma rec U rre N ce A ttention-based classifie R (LUNAR), a DL-based model to predict early versus late glioma recurrence by integrating clinical, genomic, and mRNA expression data from patients with primary grade II-IV gliomas. The data was obtained from The Cancer Genome Atlas and the Glioma Longitudinal Analysis Consortium (GLASS).