Rosenblum-Mahaley Clinical Research Award - Glucagon-Like Peptide-1 Receptor Agonist Use Significantly Reduces Primary Malignant Brain Tumor Risk: A Propensity-Matched Analysis of 415,548 Patients with Type 2 Diabetes

Introduction: Glucagon-like peptide-1 receptor agonists (GLP-1RAs) effectively treat type 2 diabetes (T2D) and are associated with reduced obesity-associated cancer (OAC) risk. Notably, GLP-1 receptors are widely expressed in both normal and neoplastic brain tissue, and preclinical studies suggest that GLP-1RAs may exhibit antitumorigenic effects against gliomas; however, clinical evidence is lacking. This study examines whether GLP-1RA use reduces the risk of primary malignant brain tumor incidence in T2D patients.

Methods: Using de-identified electronic health records from 94 healthcare organizations, we identified 1,438,987 adults with T2D who initiated either a GLP-1RA or another second-line oral hypoglycemic agent (2L-OHA) between 2014 and 2024. Cohorts were balanced via 1:1 propensity score matching on 20 characteristics to yield 415,548 patients. The primary outcome was primary malignant brain tumor diagnosis; secondary outcome was any OAC diagnosis. Sensitivity analyses compared GLP-1RA users to metformin-only and insulin-only users. Additionally, we investigated the association between GLP1R mRNA expression and age-adjusted survival using 376 IDH-mutant and 182 IDH-wildtype glioma samples from The Cancer Genome Atlas (TCGA). Hazard ratios (HR) and 95% confidence intervals (95%CI) were estimated using Cox proportional hazards models.

Results: GLP-1RA use was associated with a 43% lower risk of primary malignant brain tumor diagnosis compared to 2L-OHA use (HR [95%CI] = 0.57 [0.49–0.67], P<.0001). This reduction markedly surpassed the 13% risk reduction of OAC diagnosis (HR=0.87 [0.85–0.90]; P<.0001). Sensitivity analyses demonstrated similar risk reductions when comparing GLP-1RA users to metformin-only (HR=0.54 [0.45–0.64], P=.0046) and insulin-only users (HR=0.62 [0.54–0.72], P<.001). In the TCGA analysis, above-median GLP1R expression correlated with a 49% survival improvement for patients with IDH-mutant gliomas (HR=0.51 [0.29–0.90]; P=.019), but not IDH-wildtype gliomas (P=.60).

Conclusion : GLP-1RA use significantly reduces the risk of primary malignant brain tumor diagnosis in T2D patients. The greater risk reduction compared to OACs, in addition to the survival benefit associated with GLP1R expression in gliomas, suggests a potential direct antitumorigenic effect. Despite the observational design and potential residual confounding, these findings warrant further clinical and mechanistic investigation into GLP-1RAs as potential neuro-oncological therapies.