2008 Background: Somatic mutations in IDH1 and IDH2 occur in ̃80% and ̃4% of LGGs, respectively, promoting tumorigenesis via increased levels of the oncometabolite D-2-hydroxyglutarate (2-HG). Vorasidenib (VOR; AG-881) is an oral, brain-penetrant, dual inhibitor of mIDH1/2; ivosidenib (IVO; AG-120) is a first-in-class oral inhibitor of mIDH1. In this ongoing perioperative study, treatment with IVO/VOR reduced 2-HG levels in resected tumors vs untreated control tumors in patients (pts) with LGG (NCT03343197; Mellinghoff SNO 2019). We assessed the biological impact of 2-HG suppression on tumors and TIME. Methods: Pts (n = 49) with recurrent, non-enhancing, mIDH1-R132H LGG eligible for resection were randomized to IVO (500 mg QD/250 mg BID), VOR (10/50 mg QD), or no treatment, for 4 weeks preoperatively. Tumor tissue samples collected at surgery were assessed in genomic (n = 42), transcriptomic (n = 42), and immunohistochemistry (IHC; n = 43) analyses. Unpaired t-test was used for statistical comparisons. Results: Optimal 2-HG suppression (post-treatment 2-HG below the upper limit of 2-HG levels in a reference set of 15 wild-type [wt] IDH samples) was observed in 23 of 40 pts, including 9 (90%) pts receiving VOR 50 mg QD and 6 (50%) receiving IVO 500 mg QD. Of samples with valid biomarker data, those with optimal 2-HG suppression (n = 21) showed upregulation of neural differentiation-related gene expression, but downregulation of stemness-related gene expression, vs those with suboptimal 2-HG suppression (post-treatment 2-HG above upper limit of wt IDH 2-HG levels; n = 17; p < 0.01). IHC analysis of the proliferation marker Ki-67 showed a ̃2-fold decrease in Ki-67–positive cells in samples with optimal 2-HG suppression (mean 2.7%; n = 22) vs those with suboptimal suppression (5.8%; n = 16; p < 0.05). Epigenetic analysis revealed a ̃2-fold increase in mean 5-hydroxymethylcytosine (5hmC) levels in samples with optimal (0.36%; n = 17) vs suboptimal 2-HG suppression (0.2%; n = 15; p < 0.05), suggesting reversal of TET2 inhibition. IHC analysis of TIME revealed increases in mean CD3+ and CD8+ tumor-infiltrating lymphocyte levels in samples with optimal (1.05% [CD3]/0.22% [CD8]; n = 22) vs suboptimal 2-HG suppression (0.44% [CD3]/0.07% [CD8]; n = 16; p < 0.05). Optimal 2-HG suppression was associated with upregulation of gene expression related to type I interferon signaling and antigen presentation (p < 0.01). Conclusions: These data suggest that both tumor-intrinsic and -extrinsic mechanisms underlie 2-HG suppression by VOR and IVO. VOR, and IVO to a lesser extent, increased 5hmC, promoted cellular differentiation, and inhibited tumor cell proliferation; both also increased T-cell infiltration, activated interferon signaling, and increased antigen presentation capability. These data support development of VOR in combination with immunotherapy. Clinical trial information: NCT03343197.
716. Fusing Tumor Cell Vaccines Are Potent Immunogens at Least in Part through the Release of Exosome-Like Vesicles Which Load Dendritic Cells with Tumor Antigen