Menin inhibitors for NPM1-mutated and KMT2A-rearranged relapsed/refractory acute myeloid leukemia (AML): A systematic review and meta-analysis

Abstract Background: Approximately one-third of adult AML cases are characterized by mutations in nucleophosmin 1 (NPM1), which encodes a protein that is critical for regulating the cell cycle via modulation of histone-lysine N-methyltransferase 2A (KMT2A). Menin inhibitors (MIs) represent a novel drug class that inhibits menin-KMT2A complex formation to interrupt leukemogenesis in AML. However, the efficacy and safety profile of different MIs as mono- or combination therapies for adult patients with NPM1-mutated (NPM1m) and KMT2A-rearranged (KMT2Ar) relapsed/refractory (R/R) AML requires further characterization. Methods: We performed a systematic review and meta-analysis that searched Cochrane Library, Google Scholar, Ovid Embase, Ovid MEDLINE, Scopus, and Web of Science databases for eligible studies. We used a combination of free text and MeSH terms that included “acute myeloid leukemia” and “menin inhibitor”. We included prospective cohort studies and randomized controlled trials published in English prior to March 2025 that assessed adult patients with R/R NPM1m or KMT2Ar AML treated with MIs. Studies were excluded if they were review articles or commentaries, or if they had insufficient reporting of the primary endpoint of overall response rate (ORR). The secondary endpoints were complete response (CR), composite complete response (cCR), and adverse events (AEs). Identified studies were screened independently by two reviewers. Included studies were assessed for quality using the Downs and Black Quality Assessment Tool. The study protocol was registered on PROSPERO (CRD420251023507). Results: Our search yielded a total of 488 articles. 18 were selected for full text review, and 9 studies were ultimately selected for inclusion in our analysis. Six studies were high quality, and 3 studies were moderate quality. The included studies comprised a total of 546 patients (55.2% female) with a median age of 57.5 years who had a median of 3 prior lines of therapy. The pooled ORR was 59.7% (95% CI 47.9% – 70.6%; k=8; n=287, I2=63.9%). NPM1m AML patients had higher ORR than KMT2Ar AML patients at 57.5% (95% CI 41.7% – 72.0%; k=4; n=73; I2=31.6%) vs. 38.3% (95% CI 21.5% – 58.4%; k=3; n=66; I2=57%), although this difference was not statistically significant (p=0.074). However, there was a statistically significant difference in ORR between patients treated with MI in combination with venetoclax and azacitidine (MI+Ven/Aza) relative to patients treated with MI monotherapy at 68.0% (95% CI 46.5% – 83.8%; k=2; n=81; I2=60.3%) vs. 51.3% (95% CI 38.1% – 64.4%; k=5; n=184; I2=60.8%; p=0.011). The overall CR rate was 29.2% (95% CI 19.0% – 42.1%; k=4; n=144; I2=44.3%) with a statistically significant difference between MI+Ven/Aza and MI monotherapy at 48.1% (95% CI 31.8% – 64.7%; k=1; n=33; I2=0%) vs. 20.6% (95% CI 13.9% – 29.4%; k=3; n=111; I2=0%; p=0.003). The pooled cCR was 37.4% (95% CI 26.6% – 49.7%; k=7; n=252; I2=64.1%) and was higher for MI+Ven/Aza (42.8%; 95% CI 19.1% – 70.4%; k=1; n=48; I2=67.5%) than MI monotherapy (32.6%; 95% CI 24.8% – 41.6%; k=5; n= 182; I2= 22.4%; p<0.001). The most commonly reported AEs were nausea (38.6%; 95% CI 34.0% – 43.5%; k=6; n=417; I2=0.4%), diarrhea (36.0%; 95% CI 26.2% – 47.2%; k=4; n=338; I2=67.0%) and febrile neutropenia (32.4%; 95% CI 25.2% – 40.5%; k=7; n=415; I2=55.6%). Differentiation syndrome was seen in 18.6% (95% CI 13.7% – 24.8%; k=8; n=497; I2=49.5%), and QTc prolongation occurred in 9.1% (95% CI 3.3% – 22.8%; k=5; n=252; I2=76.6%). The pooled rate of grade 3 or 4 AEs was 60.9% (95% CI 30.8% – 84.5%; k=5; n=356; I2=93.7%) with early mortality rate of 3.3% (95% CI 0.5% – 18.2%; k=4; n=227; I2=67.9%). Conclusion: This is the first systematic review and meta-analysis evaluating the efficacy and safety of MIs in adult R/R AML patients. Pooled analysis of the included studies revealed that MIs are effective and safe for the treatment of R/R NPM1m or KMT2Ar AML. Their effect appears to be greatest in the NPM1m AML cohort and is amplified in patients treated with MI+Ven/Aza. Further investigation is required to elucidate cause for the differential response to MIs in patients with NPM1m AML relative to patients with KMT2Ar AML. Future analyses will be needed to assess whether the efficacy of MIs is due to a class effect or if particular MIs are superior.