P11.79.b multicentre multi-component data collection to define new predictors of relapse in high grade glioma

Abstract BACKGROUND The aim of our multi-centric study is to evaluate the predictive relapse factors for high-grade glioma patients. MATERIAL AND METHODS We studied retrospectively multi-centric cohort of high-grade glioma patients, from 2009 to 2021 at three French medical centers. This data collection was carried out as part of national funding to promote the integration of AI tools into clinical oncology routine (AI.DReAM project). Clinical, pathological, genomic, and radiotherapy data, with MR images before treatment and at relapse, were collected and are being analyzed.Relapse patterns were separated into radiotherapy and radiological-related patterns. The latter were defined based on T1-gadolinium and FLAIR sequences. We distinguished local and disseminated disease. We then sub-categorized into residual growth, new local or distal lesion, multi-focal and multi-centric relapse.We underwent a uni-variate analysis with the Chi-squared test and a multivariate analysis with logistic regression to find multi-centric relapse-related factors. RESULTS Overall, 976 files were collected. In this preliminary analysis, we evaluated data from 203 patients. Out of the latter, 70% were male. 28% of the patients were aged more than 69 years. 43% of patients underwent Gross Tumor Resection, and 36% had only biopsy. Considering that of 20,7% of patients underwent Subtotal Tumor Resection, less than 10% had a resection of more than 90%.23% of patients had multi-focal disease and 16,7% had multi-centric disease. 54,6% had temporal disease making it the most frequent site, 26,6% of patients had a deep location. Ventricular contact was noted in 58% of the cases. MGMT methylation was noted in 34% of the patients.The median radiation dose received was 60Gy [33Gy-72Gy]. 98,5% of patients received Temozolomide with radiation and 80% continued it in the adjuvant setting.The median progression-free survival was 7 months, and the median overall survival was 17,6 months [3 to 92 months].The univariate analysis showed a correlation between multi-centric or bilateral initial disease and multicentric relapse pattern (p=0.015 and p=0.012 respectively). The multivariate analysis showed no correlation between our data and the multi-centric relapse. CONCLUSION The preliminary results of our study did not find an association between the analyzed data and the multi-centric relapse pattern.The challenge set is ambitious, as there are no studies yet that identify the factors associated with the multi-centric relapse pattern.Further analyses, currently in progress, are needed for the cohort of 976 patients, taking into account disease and radiotherapy volumes. The final results will be presented at the congress