Targeting Menin in T-Lineage Acute Lymphoblastic Leukemia

Relapsed/refractory (R/R) T-lineage acute lymphoblastic leukemia (T-ALL) has a dismal prognosis with <20% long-term survival. There is an urgent unmet need for new therapies in R/R T-ALL, as a new class of drug has not been approved for over a decade. Menin inhibitors (e.g., revumenib, ziftomenib) had remarkable single-agent activity for KMT2A-rearranged (KMT2A-r) and NPM1-mutant leukemias in early phase clinical trials. Preclinical data suggest that they may be effective in other HOXA-upregulated genetic subtypes, such as NUP98-r leukemias (Heikamp, Blood 2022). 30% of T-ALL cases are HOXA-high by their transcriptomic profile (e.g., PICALM::MLLT10, HOXA-fusions). We hypothesized that menin inhibition is an effective strategy in a subset of patients with T-ALL.

We investigated in vitro ziftomenib sensitivity of 14 primary T-ALL cells with different genetic backgrounds (4 HOXA-high) using an MTT-based assay. IC₅₀ values ranged from 0.9 to >10,000 nanomolar (nM). Concomitantly, we performed RNA-seq to assess expression of HOXA and downstream targets of menin inhibition. There was no correlation between HOXA expression and sensitivity to ziftomenib. Among 5 samples with IC₅₀ values <100 nM, only 1 had a KMT2A­­-r and 1 was HOXA-high. In contrast, 3 other HOXA-high or KMT2A-r T-ALL cases had higher IC₅₀ values. We observed similar patterns with revumenib treatment. Next, we tested in vivo efficacy of ziftomenib at different dose levels (vehicle, 75 mg/kg and 100 mg/kg; n=5 per group) in NOD scid gamma (NSG) mice engrafted with KARPAS-45 (KMT2A::AFX1) or ALL-SIL (NUP214::ABL1) T-ALL cells. After 4 weeks of therapy, we harvested hematopoietic (blood, spleen, bone marrow) and non-hematopoietic organs to assess for tumor burden and treatment toxicity, respectively. We observed significant shrinkage in spleen size and circulating leukemia (CD7+) population in ziftomenib-treated mice compared to vehicle control (p< 0.001). Histologically, bone marrow and spleens from vehicle-treated control mice demonstrated >90% involvement with T-ALL. In both models, there was a dose-dependent reduction in tumor burden, characterized by 35-50% blasts at 75 mg/kg dose and <5% blasts at 100 mg/kg dose. We did not observe significant toxicity in non-hematopoietic organs. Collectively, these results provide preclinical evidence on the efficacy of menin inhibition in a subset of T-ALL cases, which is not entirely predicted by high HOXA expression.

Next, we investigated on-target activity of ziftomenib in T-ALL primary cells with IC₅₀ <100 nM. Both at the RNA and protein level, several menin target genes (MEIS1, MEF2C, CDK6, MYC) were down-regulated upon treatment with ziftomenib. By gene set enrichment analysis, we observed down-regulation of MLL target genes in treated cells. In addition, T-cell differentiation and receptor serine/threonine-kinases were activated, while DNA replication and cell cycle checkpoint signaling were suppressed.

Based on RNA and protein data confirming on-target activity of menin inhibition and enrichment of serine/threonine-kinases in ziftomenib-treated cells, we performed whole and phosphoproteome analyses in ziftomenib-sensitive T-ALL. Several members of the CDK and MAPK families were active in these primary cells. Among the top three substrates, we identified phospho-MEF2C (S222) as a shared phosphosite for both kinase families. Since MEF2C is a down-stream target of menin, we hypothesized that sensitivity to menin inhibition in HOXA-low T-ALL cells is driven by a dependency on high p-MEF2C activity. We observed a correlation between high p-MEF2C/MEF2C ratio and lower IC₅₀ to ziftomenib in 7 different T-ALL samples. Treatment with CDK1/2 or ERK1/2 inhibitors led to decreased p-MEF2C, suggesting their role in regulating MEF2C. Ziftomenib-sensitive T-ALL cells were also sensitive to treatment with CDK1/2 and ERK1/2 inhibitors. We observed significant synergy when ziftomenib was combined with CDK1/2 (max HSA score: 19) or ERK1/2 inhibitors (max HSA score: 28). Therefore, ziftomenib-mediated reduction in MEF2C leads to reduced substrate availability for CDK1/2 and ERK1/2 kinases, which are necessary for maintenance of T-ALL.

These data provide evidence for preclinical activity of menin inhibitors in T-ALL with high p-MEF2C. Additional studies investigating combination strategies and the use of p-MEF2C as a biomarker of ziftomenib response in T-ALL are ongoing.