Topics Covered

1. KOMET-007 and SNDX-5613-0708: Menin Inhibitors with Intensive Chemotherapy

KOMET-007: Phase 1a/b study adding ziftomenib 600 mg once daily, starting on day 8, to 7+3 induction and continuing it through consolidation, in 99 patients with newly diagnosed NPM1-mutated (n=49; median age 60) or KMT2A-rearranged (n=50; median age 43) AML. Presented by Amer Zeidan at EHA 2026 (Abstract S130; NCT05735184).

SNDX-5613-0708: Phase 1 dose-escalation study adding revumenib to cytarabine plus daunorubicin or idarubicin for up to two induction cycles in induction-eligible adults aged 18–75 (ECOG 0–2) with newly diagnosed KMT2A-rearranged, NPM1-mutated, or NUP98-rearranged AML. Two dose levels were tested: DL1 (110/220 mg) and DL2 (160/270 mg, the approved monotherapy dose). 31 patients were treated (13 at DL1, 18 at DL2) as of October 2025. Presented by Ibrahim Aldoss at EHA 2026 (Abstract PF489; NCT06226571).

Key results (KOMET-007 median follow-up 17.6 months NPM1-m, 11.0 months KMT2A-r):

• KOMET-007 CRc 96% (47/49) in NPM1-m and 90% (45/50) in KMT2A-r, 93% (92/99) overall; MRD negativity among CRs ~85%

• KOMET-007 12-month OS 94% (NPM1-m) vs. 71% (KMT2A-r); median OS not reached in either cohort; 60-day mortality 2% and 4%

• KOMET-007 safety: grade 3 differentiation syndrome 4%, no grade 4; no ziftomenib-related QTc prolongation

• Revumenib + intensive chemotherapy, response-evaluable patients: DL1 (n=12) ORR/CRc 100%, CR 92%; DL2 (n=14) ORR 93%, CRc 86%, CR 79%; MRD-negative CR 100% (DL1) and 70% (DL2); responses similar across dose levels

• Revumenib safety: no differentiation syndrome; one grade 3 QTcF prolongation DLT; any-grade QTcF prolongation ~15% at DL1

Discussion points: Both studies start the menin inhibitor after the cytotoxic days of induction, when there is less leukemia left to differentiate. Dr. Zeidner credits that timing and the chemotherapy itself for differentiation syndrome rates of 3–4%, compared with roughly 15–30% for single-agent menin inhibitors in relapsed/refractory disease. He saw no new QTc signal and no clear delay in count recovery. He cautioned that single-arm data can only be judged against historical expectations, and that the low early mortality reflects selected patients at academic centers. Ashwin questioned how much a high CR rate adds in NPM1-mutated disease, where 7+3 already works well. Dr. Zeidner agreed the two genotypes need separate readouts. In NPM1-mutated, FLT3 wild-type disease the goal is cure with chemotherapy, so relapse and survival are the open questions, and the randomized phase 3 KOMET-017 (7+3 plus ziftomenib or placebo) will answer them. In KMT2A-rearranged disease remission is a bridge to transplant, so he asks whether azacitidine-venetoclax plus a menin inhibitor could replace intensive induction. He is leading a single-arm study of that approach (RAVEN) in younger fit KMT2A-rearranged patients. Choosing between the two drugs, he finds differentiation syndrome rates comparable in NPM1-mutated disease (about 20–25%) and more severe in KMT2A-rearranged disease. Both drugs prolong the QTc. Revumenib carries the boxed warning and somewhat more grade 3 events, and he checks ECGs weekly when starting either one.

2. HOVON156/AMLSG28-18/PASHA

Phase 3, open-label trial randomizing 768 adults with newly diagnosed FLT3-mutated AML fit for intensive chemotherapy 1:1 to gilteritinib 120 mg daily or midostaurin 50 mg twice daily, each with 7+3 induction and consolidation and followed by FLT3 inhibitor maintenance. Median age 59; FLT3-ITD allelic ratio ≥0.5 in 48%, FLT3-TKD in 21%, NPM1 co-mutation in 58%. The primary endpoint was OS. Presented by Marc Raaijmakers at EHA 2026 (Abstract LB5005; NCT04027309).

Key results (43.2-month median follow-up):

• OS (primary endpoint): median not reached in either arm; HR 1.02 (95% CI 0.81–1.28; P=0.86)

• Median EFS 51.1 vs. 19.9 months (HR 0.83; P=0.052)

• Post-CR morphologic relapse 21% (gilteritinib) vs. 36% (midostaurin) (HR 0.68; P=0.003)

• Among relapsing patients, median OS 7.2 months with gilteritinib vs. 10.2 months with midostaurin; 50% of midostaurin-arm relapses received salvage gilteritinib vs. 17% in the gilteritinib arm

• CR rates did not differ between arms; early mortality, infection rates, and myelosuppression were numerically higher with gilteritinib

Discussion points: Dr. Zeidner called it a negative trial with OS curves that overlap completely. Gilteritinib lowered relapse in patients who reached CR. Midostaurin-arm patients who relapsed could still be salvaged with gilteritinib and transplant, while gilteritinib-arm patients had no equally effective option left. He did not use gilteritinib off-label before this trial and will not now. He offered a thought experiment: if EFS had been the powered primary endpoint and crossed significance, the same data would have been called positive and a new standard, so endpoint choice drives the interpretation as much as the data do. The hosts compared this with PARADIGM, where an EFS benefit without an OS difference has already shifted practice, and with the earlier randomized phase 2 PrECOG 0905 comparison of the same two drugs. Raj noted that in myeloma an EFS gain usually carries into OS because several effective lines follow, while in FLT3-mutated AML little works after upfront gilteritinib. Dr. Zeidner added that midostaurin may now be an outdated control for FLT3-ITD disease, since many centers have moved to quizartinib.

3. Quizartinib vs. Midostaurin with Intensive Chemotherapy (Multicenter Cohort)

Retrospective cohort study across 12 US centers (2017–2025) of patients with newly diagnosed FLT3-ITD AML treated with intensive chemotherapy plus midostaurin (n=127) or quizartinib (n=88); median age 56; 89% ELN intermediate risk. Primary endpoint 1-year OS. Median follow-up was 667 days with midostaurin vs. 324 days with quizartinib. Presented by Michelle Lee at EHA 2026 (Abstract S133); Dr. Zeidner is a co-author.

Key results:

• 1-year OS 93% (quizartinib) vs. 78% (midostaurin) (HR 0.19; 95% CI 0.07–0.52)

• 1-year EFS 77% vs. 56% (HR 0.44)

• CRc 85% vs. 73%

• FLT3 clearance similar in both groups (~39–40%)

• The quizartinib group had roughly half the follow-up and fewer triple-mutated (NPM1/DNMT3A/FLT3-ITD) patients

Discussion points: Dr. Zeidner stressed the retrospective design. Most midostaurin patients were treated earlier, the quizartinib group has shorter follow-up, and the investigators chose 1-year OS to fit that follow-up. He reads the cohort as support for current practice rather than a reason to change it. For fit patients he gives 7+3 plus quizartinib for FLT3-ITD and 7+3 plus midostaurin for FLT3-TKD. He avoids quizartinib in patients with significant QTc prolongation or cardiac comorbidity and is more cautious above age 60, given higher early mortality in that subgroup on QuANTUM-First. He still uses quizartinib in older patients with the NPM1/DNMT3A/FLT3-ITD genotype, which depends heavily on FLT3 signaling and did well with quizartinib on ...