P17.11.A SURVIVAL IMPACT AND PREDICTIVE FACTORS OF EPILEPSY IN GLIOBLASTOMA PATIENTS

Le résumé fourni par la source

Abstract BACKGROUND More than half of glioblastoma patients experience epileptic seizures before or after diagnosis. Epileptic activity may promote tumour growth, while glioma cells may increase seizure activity, a synergy linked to poor prognosis. It remains unclear which patients develop epilepsy. We aim to study the association between epilepsy and prognosis using a real-world dataset and to identify biomarkers predicting epilepsy after diagnosis. MATERIAL AND METHODS From our prospective database, all IDH wild type glioblastoma patients administered standard therapy (radiation therapy with concomitant + adjuvant temozolomide) from 2016 to 2024 at Rigshospitalet (Denmark) were included. Patients were offered genomic tumour profiling (whole-exome or whole-genome sequencing). Patients’ records were screened for a diagnosis of epilepsy and initiation of anti-seizure medication. Changes in anti-seizure medication will be used as a measure of seizure control. The cohort was divided into training (2016-21) and validation (2021-24) datasets. Candidate genetic factors comprised pathogenic and likely-pathogenic genetic variants. Uni- and multivariate Cox regression analysis with time dependent covariates, where relevant, will be performed to model time to event endpoints (overall survival, epilepsy, and failure in seizure control). RESULTS The study included 552 patients (training set, n = 368 and validation set, n = 184) of whom 217 patients (39%) had epilepsy before and 157 patients (29%) developed epilepsy after glioblastoma diagnosis. In the training set, multivariate analysis showed epilepsy being associated with poor overall survival when epilepsy occurred prior to diagnosis (HR: 1.36, 95% CI: 1.02-1.81, p = 0.04) and after diagnosis (HR: 3.04, 95% CI: 2.30-4.04, p = <0.0001). In multivariate analysis modelling time to development of epilepsy, multifocal disease was significantly associated with higher risk of epilepsy (HR: 2.50, 95% CI: 1.42-4.40, p = 0.002) and younger age was non-significantly associated (HR: 0.86, 95% CI: 0.72-1.03, p = 0.096). Other clinical factors (sex, MGMT status, corticosteroid use, extent of resection, performance status) were not significant (p > 0.10). In univariate analysis, genetic factors associated (p < 0.10) with development of epilepsy were RB1, PDGFRA, EGFR mutation, and alterations in the RTK-RAS pathway. Associated gene alterations will be tested in multivariate analysis adjusted for clinical factors. The final model will be tested in the validation set. CONCLUSION Epilepsy was associated with a poor overall survival in patients with glioblastoma. Multifocal disease was predictive of epilepsy. RB1, PDGFRA, EGFR, and RTK-RAS pathway alterations were associated with epilepsy. A multivariate model combing clinical and genetic factors and data including seizure control will be presented.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.