IMG-99. Volumetric tumor growth rate (TGR) modeling predicts clinical outcomes in patients with Grade 2, isocitrate dehydrogenase 1/2 mutant (mIDH1/2) glioma receiving vorasidenib or placebo in the Phase 3 INDIGO trial

TGR shows emerging utility as a quantitative measure for assessing treatment efficacy in glioma. We evaluated three distinct TGR models using volumetric data from patients with mIDH1/2 diffuse glioma from the Phase 3 INDIGO trial (NCT04164901) to determine their predictive value for clinical outcomes. Data from the INDIGO trial comparing vorasidenib (n=168) with placebo (n=163) were analyzed. We assessed tumor growth patterns after treatment using: (1) a linear regression model (mL/6 months); (2) a bi-exponential model with growth parameter “g”; and (3) velocity of diametric expansion (VDE; mm/year). Individual TGRs were calculated during the first 6-month on-treatment period and correlated with progression-free survival (PFS) and time to next intervention (TTNI). Patients receiving vorasidenib demonstrated substantially lower on-treatment individual TGRs, compared with placebo, across all models (linear model: –0.037 vs 1.094 mL/6 months; bi-exponential: reduced “g” parameter, P<0.0001; VDE: –0.142 vs 2.280 mm/year). Quantile analysis revealed that patients with lower individual TGRs had notably better PFS and TTNI outcomes (P<0.0001). In Cox regression analysis, patient-specific TGR during first 6 months of treatment was strongly associated with both PFS (HR 1.046; 95% CI 1.026, 1.067; P<0.0001) and TTNI (HR 1.051; 95% CI 1.031, 1.072; P<0.0001). Holding other variables as constant, multivariate analysis identified treatment with vorasidenib versus placebo, an oligodendroglioma versus astrocytoma histological subtype, and, most notably, lower versus higher baseline tumor volume, as strong determinants for reduced individual TGRs. These findings demonstrate that individual TGRs are a robust quantitative measure for predicting outcomes in mIDH1/2 glioma. During the initial 6-month period, TGR was a strong predictor of long-term outcomes, supporting the utility of TGR as a potential surrogate endpoint in clinical trials and for treatment decision-making.