Early real-world experience with revumenib outside of a clinical trial setting: A single center retrospective review of efficacy and tolerability

Abstract

Abstract Background: Revumenib was approved by the US Food and Drug Administration on 11/15/2024 as the first menin inhibitor for treatment of relapsed/refractory (R/R) acute leukemias with a KMT2A-translocation (KMT2Ar) based on data from the AUGMENT-101 trial. Similar safety and efficacy data have now been shown for revumenib in patients (pts) with R/R NPM1 mutated AML. Additional pt subgroups may benefit from menin inhibition including those with measurable residual disease (MRD) or NUP98r, and there is potentially enhanced efficacy with combination approaches. Here, we present early efficacy and tolerability data from real-world use of revumenib. Methods We reviewed our institutional prescription data to identify pts who received >1 dose of revumenib between 12/12/2024 and 5/28/2025 outside of a clinical trial. Pts were followed from date of initial revumenib dispensing until 7/17/2025 or death. Efficacy was evaluated using European Leukemia Net 2022 criteria for complete remission (CR), CR with partial hematologic recovery (CRh), and CR with incomplete hematologic recovery (CRi). Composite CR (CRc) was defined as CR+CRh+CRi. Overall response rate (ORR) also included pts with morphologic leukemia-free state (MLFS) or partial remission (PR). Other than differentiation syndrome (DS) and QT prolongation (QTc), only grade 3 (G3) and G4 adverse events (AEs) were captured. Results As of 7/17/2025, 18 pts had received ≥1 dose of revumenib outside of a clinical trial. At the time of treatment initiation, median age was 59.5 yrs (range [r], 22.9-79). For the genomic profile, 9 (50%) pts had KMT2Ar with 6 treated in the R/R setting, 2 in the firstline setting, and 1 as maintenance post hematopoietic stem cell transplant (HSCT). 6 pts had NPM1m (33%, all R/R) and 3 had NUP98r (17%, all R/R). At least 1 co-mutation was present in 14 (78%) pts with DNMT3A, TET2, and WT1 being most common (n=4 for each). Targetable co-mutations were present in 3 pts (FLT3-ITD [n=2] and IDH1 [n=1]). Median follow up was 3.97 mo (r, 0.9-17; with n=2 on clinical trial previously). Pts were pre-treated with a median of 3 (r, 0-6) prior lines of therapy, including venetoclax in 13 (72%) and HSCT in 6 (33%). Revumenib was used as part of combination therapy in 14 (78%) pts. 13 (72%) pts received hypomethylating agents and venetoclax in combination with revumenib and 1 pt received revumenib with enasidenib. 2 pts were excluded from the efficacy analysis (no bone marrow evaluation post revumenib and use in post HSCT maintenance only setting). Of the 16 efficacy-evaluable pts, 14 were treated for morphologic marrow disease/relapse and 2 were treated for NPM1m MRD positivity. CRc and ORR (calculated among 14 pts with morphologic disease) were 50% (7/14) and 79% (11/14), respectively. The best response distribution was CR (n=2), CRh (n=1), CRi (n=4), and MLFS (n=4). For pts treated for NPM1m MRD positivity, 1/2 achieved MRD negativity by NPM1m ddPCR. Amongst responders treated for morphologic disease, for KMT2Ar-MRD was negative in 6/7 pts by flow cytometry and for NPM1m-MRD was negative in 2/3 pts by flow cytometry and 1/3 pts by ddPCR. Overall, median time to first response was 0.75 mo (r, 0.47-2.57) and median time to best response was 2.05 mo (r, 0.47-6.73). Of initial responders, 2 pts have relapsed (both KMT2Ar) including 1 with isolated CNS relapse (with marrow MRD negative CR); both continued on revumenib. 4 pts received HSCT post revumenib (2 as 2nd HSCT) and 3 pts received revumenib as post HSCT maintenance (2 as continuation, 1 only post HSCT). All 18 pts were included in the safety analysis. Grade 3/4 AEs were noted in 10 (56%) pts. Other than expected hematologic toxicities, G3 sepsis (n=2) and G3 heart failure (n=1) were noted. In pts with regular EKG monitoring, QTc prolongation was noted in 7/13 pts (3 [23%] with G3 (23%), 4 [31%] with G1/G2). DS occurred in 2/18 (11%, 1 G3 and 1 G2). AEs resulted in revumenib interruption in 8 (44%) pts and dose reduction in 2 (11%) pts. AEs did not lead to permanent revumenib discontinuation in any pt. Death occurred in 2 pts, 1 with CML-BP progression and 1 with CNS relapse, neither attributable to treatment.Conclusion: Early real-world data for revumenib confirms good tolerability with no AEs leading to discontinuations. Early efficacy data shows excellent activity across molecular and disease subsets. Longer term follow-up data, inclusive of outcomes in molecular subsets, will be presented at the meeting.