Ivosidenib Monotherapy Results in Deep mIDH1 Clearance in Patients with mIDH1 AML

Conference Correspondent - ASH 2017

Ivosidenib (AG-120) is a selective, oral small-molecule inhibitor of the mutant IDH1 (mIDH1) protein, which is currently in clinical testing in patients with mIDH1 acute myeloid leukemia (AML). Previous results from a phase 1 dose-escalation study have shown evidence of mIDH1 clearance in ivosidenib-treated patients with advanced mIDH1 hematologic malignancies using a next-generation sequencing (NGS) technique that had a 1% lower limit of detection; clearance of mIDH was detected in 5 of 14 patients who achieved a complete remission (CR) versus only 2 in 52 that did not. Using a digital polymerase chain reaction technique that is 50-fold to 100-fold more sensitive (lower limit of detection, 0.02%-0.04%), the current analysis assessed the impact of ivosidenib on longitudinal mIDH1 variant allele frequency (VAF), the depth of decrease in mIDH1 VAF as a molecular marker of minimal residual disease (MRD), and correlated co-occuring mutations with response in patients treated in the expansion phase of the phase 1 study. Deep mIDH1 clearance (IDH1-MC) was defined as reduction in the mIDH1 VAF to below the limit of detection for ≥1 on-treatment time points. In this analysis, results from baseline genetic profiling using a 95-gene NGS hematologic malignancies–targeted panel to determine the predictive value of co-occurring mutations were also reported.

The analysis data set included longitudinal mIDH1 VAF data from bone marrow mononuclear cells (BMMCs) for 127 patients; of these, 102 patients had relapsed or refractory (R/R) AML and received their first dose of ivosidenib ≥6 months prior to the analysis cutoff date of May 12, 2017, and 25 patients had previously untreated AML and had received ≥1 doses of study treatment on the phase 1 study. The data set for assessing longitudinal mIDH1 VAF from neutrophils included 82 R/R AML patients and 22 untreated AML patients, whereas that for the baseline co-occurring mutation included 101 R/R AML patients and 25 untreated AML patients.

Among evaluable patients who achieved CR with ivosidenib treatment (n = 30), IDH1-MC, denoting MRD-negative CR, was detected in 7 of 25 (28%) patients with R/R AML and 3 of 5 (60%) patients with untreated AML. Among untreated patients with AML who achieved CR with partial hematologic recovery (CRh), IDH1-MC was detected in 2 of 4 untreated patients, and in none of the patients with R/R AML. However, no IDH1-MC was observed in patients who did not achieve a best response of CR or CRh (R/R AML, 39; untreated AML, 14). Moreover, in the R/R AML and BMMC data set, patients with MRD-negative CR had improved duration of CR and overall survival (OS) compared with patients with CR who did not achieve MRD negativity. There was significant association of IDH1-MC between neutrophils and BMMC in R/R patients with best response of CR or CRh. The correlative mutational analysis showed no specific single-gene mutation that was significantly predictive of clinical response or resistance to ivosidenib therapy in patients with R/R AML.

The authors concluded that ivosidenib monotherapy reduces IDH1-MC allele burden in a subset of patients with R/R AML and untreated AML who achieve CR or CRh, with R/R AML patients achieving MRD-negative CRs showing improved duration of CR and OS.

Stone RM, et al. 2017 ASH. Abstract 2684.