Validation of a novel molecular RPA classification in glioblastoma (GBM-molRPA) treated with chemoradiation: A multi-institutional collaborative study

Background and purpose: A novel molecular recursive partitioning analysis classification has recently been reported integrating the MGMT promoter methylation (MGMTmeth) and IDH1 mutation (IDH1mut) status for glioblastoma (GBM-molRPA) patients treated with temozolomide-based chemoradiation. The current study was initiated to validate the model in a multi-institutional study. Materials and methods: Four-hundred seventy-one newly diagnosed GBM patients (validation cohort) were allocated to classes I-III of the previously reported GBM-molRPA model. Of the patients, 15.7%, 56.1%, and 28.2% patients were GBM-molRPA class I, II, and III, respectively. MGMTmeth and IDH1mut were observed in 32.3 and 8.8% of patients, respectively. In the training plus validation cohort of 692 patients, 16.2%, 60.8%, and 23.0% patients were class I, II, and III, respectively. Results: The median follow-up for survivors and the median survival (MS) of patients was 23.3 and 18.4 months, respectively. The MS for GBM-molRPA class I, II, and III was 49.7 (95% CI, 22.8-76.6), 19.2 (95% CI, 16.2-22.1), and 13.8 months (95% CI, 11.8-15.4) (P < .001 for all comparisons) in the validation cohort. In the training plus validation cohort, the MS was 58.5 (95% CI, 40.7-76.3), 21. (95% CI, 18.6-23.3), and 14.3 months (95% CI, 12.5-16.1) (P < .001 for all comparisons) for class I, II, and III, respectively. Conclusion: The GBM-molRPA is a valid model. This GBM-molRPA classification can be useful in clinics and guiding patient stratification in future clinical trials. (C) 2018 Published by Elsevier B.V.