A Comparison of Acute Myeloid Leukemia (AML) Regimens: Hypomethylating Agents Combined with Ivosidenib or Venetoclax in Newly Diagnosed Patients with IDH1 Mutations: A Real-World Evidence Study

Background: Two frontline combination regimens are available for patients with mutant IDH1 (mIDH1) AML: ivosidenib (IVO)+hypomethylating agent (HMA), and venetoclax (VEN)+HMA. Robust comparisons between these regimens have not been performed, and small post hoc analyses have been constrained by low numbers of mIDH1 patients and clustering of IDH1 and IDH2 mutations (Pollyea 2022, Perl 2022). We evaluated a large cohort of patients with newly diagnosed (ND) mIDH1 AML to gain insights into the treatment patterns, efficacy, and safety associated with these combination regimens.

Methods: We conducted a US-wide retrospective chart review of patients with ND mIDH1 AML ineligible for intensive chemotherapy (IC), treated with IVO+HMA or VEN+HMA. Treatment was initiated in community and academic settings November 2021 through November 2022, enabling a minimum follow-up of 6 months. Collected data included baseline patient and practice characteristics, treatment patterns and rationale, and efficacy and safety outcomes. Endpoints focused on events likely to occur within 6 months of follow-up, including response rates, tolerability, bridge to transplant, and acute care episodes.

Results: Among 283 patients (IVO+HMA, n=182; VEN+HMA, n=101), baseline characteristics were similar between regimens. Median age was 63 years; 65.7% were male. ELN cytogenetic risk status was 59.7% and 15.2% for intermediate and poor, respectively. Prior MDS, myeloproliferative neoplasms, or secondary or secondary-like AML mutations occurred in 37.1%. A higher proportion of VEN+HMA use occurred in the academic setting (68.3% vs 55.5% for IVO+HMA; p=0.035). Median follow-up from treatment initiation was 7.1 vs 8.1 months for IVO+HMA vs VEN+HMA.

Complete response (CR) plus CR with incomplete count or incomplete platelet recovery (CRi/p) rates were 63.2% vs 49.5% for IVO+HMA vs VEN+HMA ( p=0.025), with the difference based on the higher CR rate for IVO+HMA (42.9% vs 26.7%; p=0.007) (Figure). Median time to first bone marrow biopsy on treatment across cohorts was 56 days and median time to best response was 3.3 vs 4.1 months for IVO+HMA vs VEN+HMA ( p=0.002). A competing risks regression showed that 6-month event-free survival (EFS; CR within 24 weeks, and no relapse or death) favored IVO+HMA vs VEN+HMA (56.0% vs 39.6%, hazard ratio of 0.773; p=0.044). Bridge to transplant was achieved for 11.5% of patients with IVO+HMA vs 5.0% for VEN+HMA ( p=0.066).

Few patients in either cohort had a dose (ie, strength) change following initial ramp-up. Treatment discontinuation was 37% for both regimens. Likewise, toxicity incidence was similar except for higher febrile neutropenia (FN) rates for VEN+HMA vs IVO+HMA within 30 days of initiation (7.9% vs 1.6%; p=0.009). Patients receiving IVO+HMA had 61% lower relative risk of unscheduled acute care use in the first 12 weeks (42.9% vs 70.3% for VEN+HMA; p<0.001). VEN schedule intensity was also captured, with only 22.8% receiving the full FDA-labeled 28 days of VEN during the 28-day cycles; 44.6% did not receive >11 days of VEN per cycle, raising concerns of clinical consequences.

Median time from mIDH1 test to receipt of result was 7 days in both cohorts. Median time from diagnosis to start of treatment was 14 vs 20 days for IVO+HMA vs VEN+HMA. The biggest differential driver of regimen choice was favoring selection of VEN+HMA due to hospital protocol (30.7% vs 16.5%).

Conclusions: In a large, balanced cohort of nearly 300 patients with ND mIDH1 AML ineligible for IC, patients treated with IVO+HMA had higher rates of CR and CR+CRi/p, achieved CR faster, and had longer EFS compared to those treated with VEN+HMA. Approximately half of patients receiving VEN were unable to receive >11-day schedules, potentially impacting the regimen's efficacy. Despite the reduced intensity of VEN relative to label, patients receiving VEN+HMA had higher early incidence of FN and greater overall need for unscheduled acute care. Wait time for mutational test results was 7 days, well within the 14- to 20-day median times to treatment. Based on the efficacy seen in this real-world assessment, consistent with the efficacy demonstrated in the AGILE clinical trial of IVO+azacitadine, clinicians and hospitals should ensure early mutational testing occurs, to enable consideration of IVO+azacitadine as the frontline treatment of choice for patients with ND mIDH1 AML.