PL3.1 A phase 1, open-label, perioperative study of ivosidenib (AG-120) and vorasidenib (AG-881) in recurrent, IDH1-mutant, low-grade glioma: results from cohort 1

Abstract

BACKGROUND: Mutant isocitrate dehydrogenase (mIDH) enzymes produce the oncometabolite D-2-hydroxyglutarate (2-HG). Ivosidenib (IVO; AG-120) is a first-in-class oral inhibitor of mIDH1 being evaluated in 66 glioma patients (pts) in an ongoing phase 1 study (NCT02073994). Vorasidenib (VOR; AG-881) is an oral, potent, brain-penetrant inhibitor of mIDH1/2 being evaluated in 52 glioma pts in an ongoing phase 1 study (NCT02481154). In an orthotopic glioma model, IVO and VOR reduced 2-HG levels by 85% and 98%, respectively, despite different brain:plasma ratios (<0.04 vs 1.33). MATERIAL AND METHODS: This is a phase 1, multicenter, open-label, perioperative study (NCT03343197). Pts with recurrent, nonenhancing WHO 2016 Grade (Gr) 2 or 3 mIDH1-R132H oligodendroglioma or astrocytoma undergoing craniotomy were randomized 2:2:1 to IVO 500 mg QD, VOR 50 mg QD, or no treatment (control) for 4 wk preoperatively in Cohort 1. Postoperatively, pts continued to receive IVO or VOR, and control pts were randomized 1:1 to IVO or VOR. Tumors were assessed for mIDH1 status, cellularity, 2-HG, and drug concentration. Treated samples were compared with control pts and mIDH1 and wild type (WT) banked reference (ref) samples. Plasma and CSF 2-HG were assessed. Primary endpoint: brain tumor 2-HG concentration with IVO or VOR treatment. Pts with nonevaluable tissue were replaced. RESULTS: As of 29 Nov 2018, 26 pts (17 men, 9 women; 25 Gr 2, 1 Gr 3) were randomized preoperatively (IVO 10, VOR 11, control 5) and 25 received drug (IVO 12, VOR 13). At the data cut, 19 tumors were analyzed, with 16 evaluable. Treatment-emergent adverse events in >10% of patients (all Gr 1 or 2) were diarrhea (36%); hypocalcemia and constipation (each 20%); anemia, hyperglycemia, pruritus, headache, and nausea (each 16%); and hypokalemia and fatigue (each 12%). Mean brain:plasma ratio was 0.16 for IVO and 2.4 for VOR. Geometric mean (range) tumor 2-HG levels (μg/ mL) were: IVO (n=6), 10 (2.2-104); VOR (n=6), 6.8 (3.9-10); control mIDH1 (n=65 [4 pts, 61 ref]), 141 (4.8-909); and WT ref (n=15), 2.7 (0.46-12). Mean changes (95% CI) in 2-HG level vs control were IVO-93% (74%, 98%) and VOR-95% (82%, 99%). Updated data from Cohort 1 will be presented. CONCLUSION: In Cohort 1 of this phase 1 perioperative study, IVO and VOR were CNS penetrant and lowered tumor 2-HG levels compared with untreated samples. Cohort 2 is open and will evaluate IVO 250 mg BID and VOR 10 mg QD.