Comparative response assessment using PET-RANO 1.0 and RANO 2.0 in vorasidenib-treated <i>IDH</i> -mutant gliomas: The VORAFET study.

Abstract

2073 Background: MRI-based response assessment may be suboptimal for capturing early treatment effects during isocitrate dehydrogenase ( IDH ) inhibition. We compared metabolic and morphological response classification using PET-RANO 1.0 and RANO 2.0 in vorasidenib-treated IDH -mutant gliomas and assessed treatment-related metabolic changes. Methods: All consecutive patients treated at Veneto Institute of Oncology and Padua University Hospital with vorasidenib (VOR) under a compassionate-use program (June 2024-March 2025) with baseline and follow-up [ 18 F]FET PET/MR were retrospectively included. Response was assessed every 3 treatment cycles according to PET-RANO 1.0 and RANO 2.0. PET-positive tumor volumes and tumor-to-background ratios (TBR max , TBR mean ) were analyzed. MRI response was evaluated using volumetric RANO 2.0 criteria. The VORAFET study was approved by the local Ethics Committee. Results: Twenty-six patients were included (10 astrocytomas, 16 oligodendrogliomas; 22 WHO grade 2, 4 non-enhancing WHO grade 3). Median follow-up from VOR initiation was 9 months (IQR 9-11). According to PET-RANO 1.0, partial metabolic responses (PET-PR) increased over time, from 20% at 3 months to 30.8% at 6 months and 63.6% at 9 months, while metabolic progression (PET-PD) was already detected in 18.2% at 9 months. In contrast, RANO 2.0 identified no partial responses at any timepoint, classifying most patients as stable disease (84.2%) or minor response (10.5%), with progression detected in only 5.3% at 9 months. Quantitative PET analysis showed an early and sustained reduction in metabolic activity. TBR mean decreased significantly at 3 months (mean difference [MD] -0.04; 95% CI -0.08 to -0.01) and continued to decline at 9 months (MD -0.07; 95% CI -0.11 to -0.02). Compared to baseline, at 9 months, TBR max decreased from 2.7 to 2.0 (MD -0.5; 95% CI -0.8 to -0.2) in all patients included. No significant changes were observed in PET volumes. Similarly, FLAIR-based tumor volume showed no significant variation (at 9 months: MD -2.2; 95% CI -5.1 to 0.7). Conclusions: PET-RANO 1.0 enabled earlier detection of treatment response and progression compared with MRI-based RANO 2.0, highlighting clinically relevant limitations of morphology-driven MRI-based assessment. These findings support the clinical integration of amino acid PET into response monitoring strategies and warrant prospective validation of PET-derived biomarkers in IDH- mutant gliomas.