Complete remission of NUP98 fusion-positive acute myeloid leukemia with the covalent menin inhibitor BMF-219, icovamenib

Acute myeloid leukemia (AML) containing nucleoporin 98 (NUP98) fusions is the most common genotype among children with relapsed/refractory disease, representing a poor prognosis group.¹ NUP98 fusions occur in ~5% of cases of childhood AML, and are especially frequent in acute erythroid and megakaryocytic leukemias.^1,2 NUP98 translocations also occur in adults; in FLT3-ITD AML – a setting with frequent co-occurring NUP98 fusions – one of the more common, NUP98-NSD1, was reported in six of 97 (6.2%) cases.³ Adults with AML harboring NUP98-NSD1 fusions have a poor response to therapy, with induction failure rates following intensive anthracycline/ cytarabine-containing chemotherapy of 83% versus 36% for NUP98-NSD1 fusion-negative cases (P=0.038).³

NUP98 chimeric proteins fuse the N-terminus with various C-terminal partners, including epigenetic modifiers and transcription factors, such as HOX proteins.² Over 30 NUP98 fusions occur in malignancies including de novo and therapy-related AML, myelodysplastic syndromes, chronic myeloid leukemia, T-cell acute lymphoblastic leukemia, and mixed phenotype acute leukemia.^2,4 It is important to note that many NUP98 fusions are cryptic on conventional cytogenetic analysis; accordingly, only molecular methods such as next-generation sequencing (NGS) can reliably identify the variety of NUP98 fusions.² Although many commercially available NGS diagnostic assays capture NUP98 fusions, this is not invariably the case; likewise, ‘homebrew” NGS assays offered as laboratory-developed tests may or may not identify these fusions. Clinicians should consult with commercial vendors and/or their molecular diagnostic laboratory to determine whether a specific assay can reliably identify the various NUP98 fusions.

Preclinical models demonstrate that NUP98 fusions such as NUP98-HOXA9, NUP98-HOXD13, NUP98-JARID1A, and NUP98-NSD1, are potent drivers of leukemogenesis.^5,6 WT1 or FLT3-ITD mutations that co-occur with NUP98 fusions worsen prognosis.^1,2 NUP98 fusions elicit leukemogenesis through interaction with histone-modifying chromatin complexes.^6,7 For example, NUP98-NSD1 recruits WDR82–Set1A/COMPASS complexes, resulting in histone H3 lysine 4 tri-methylation at the HOXA genes and MEIS1 promoter.⁷ NUP98 fusions also interact with mixed lineage leukemia-1 (MLL1; aka KMT2A) chromatin complexes. For instance, NUP98-HOXA9 colocalizes with KMT2A at the HOXA/B genes, KMT2A deletion delays leukemogenesis in a NUP98-HOXA9 mouse model, and KMT2A is required for the gene expression signature observed in NUP98 fusion-positive AML.⁸

Interaction between KMT2A and the scaffold protein menin is essential for KMT2A-rearranged and NPM1-mutant leukemias.^9,10 Disruption of this interaction using small-molecule inhibitors is a promising therapy under clinical evaluation for these leukemias and early results have demonstrated complete remissions in some patients.^9,11

The association of NUP98 fusions with KMT2A complexes on promoters, together with the observation that similar leukemia-associated gene expression signatures occur in NUP98-rearranged, KMT2A-rearranged, and NPM1-mutant AML, suggests that NUP98-rearranged AML may depend on the menin-KMT2A interaction. Indeed, Heikamp et al.¹² showed that interruption of this interaction halts leukemic progression in cell lines and patient-derived xenograft models expressing NUP98-HOXA9, NUP98-HOXD13, NUP98-JARID1A, and NUP98-NSD1. Rasouli et al.¹³ examined menin-KMT2A inhibition in primary patients’ AML cells harboring NUP98 fusions, demonstrating that the non-covalent reversible inhibitor revumenib elicits differentiation and suppresses proliferation but not apoptotic death in NUP98-NSD1 AML samples. Rasouli and co-workers also showed that NUP98-rearranged leukemias containing other fusion partners, such as TOP1 and DDX10, are sensitive to menin inhibition.

These preclinical results support menin inhibition as a potential therapy for NUP98-rearranged AML but clinical proof has not been described so far, to our knowledge. Here we report a patient with relapsed NUP98-NSD1 AML who achieved a long-lasting complete remission with the covalent menin inhibitor BMF-219 (icovamenib).

The patient, a 39-year-old Caucasian male with a history that included melanoma in-situ after multiple excisions, was diagnosed with primary NUP98-NSD1-positive AML with leukocytosis (white blood cell count [WB-C]=317.6x10⁹/L), anemia (hemoglobin [Hb]=7.1 g/dL) and thrombocytopenia (platelet count=74x10⁹/L) along with significant lymphadenopathy and splenomegaly. Bone marrow biopsy showed AML with normal cytogenetics (46, XY [20]); NGS revealed NUP98-NSD1, CEBPA p.G99Vfs*63, c.295_296ins5 variant allelic frequency (VAF, 32.3%), NRAS p.Q61K, c.181C.A (VAF, 1.9%), NRAS p.G13D, c.38G>A (VAF, 38.4%), WT1 p.P132Rfs26, c395delC (VAF, 35.2%), and RUNX1 p.G365S, c.1093G>A (VAF, 49.6%), a variant of uncertain significance (VUS). Skin fibroblast germline testing was positive for CHEK2. Bone marrow studies on day 14 after induction with cytarabine and idarubicin (“7+3” regimen) showed persistent AML (10% blasts in hypocellular marrow), while a second induction course with the “7+3” regimen resulted in a first complete remission. One cycle of high-dose Ara-C consolidation then preceded hematopoietic stem cell transplantation with curative intent. Conditioning therapy with busulfan and cyclophosphamide was followed by a matched unrelated donor allogeneic hematopoietic stem cell transplant. Following the transplant, the patient was given tacrolimus, methotrexate and mycophenolate mofetil.

Five months after the transplant, the patient became cytopenic and transfusion-dependent. A bone marrow aspirate revealed 13% blasts and 46XY, del(5)(q22q35[5]//46XX[15] karyotype. Bone marrow NGS showed NUP98-NSD1, CEBPA p.G99Vfs*63, c.295_296ins5 (VAF, 7%), and RUNX1 VUS pG365S, c.1093 G>A (VAF, 9.1%).

Figure 1.Antileukemic response and normal hematopoietic cell recovery in a patient with NUP98-NSD1-positive acute myeloid leukemia treated with the covalent menin inhibitor BMF-219 (icovamenib). Horizontal dashed blue and red lines in each graph represent the lower limit of normal and upper limit of normal, respectively, for each parameter; the vertical interrupted black lines in each graph indicate the timing of administration of packed red blood cell transfusions. BM: bone marrow; RBC: red blood cells; WBC: white blood cells; PRBC: packed red blood cells.

Table 1.Select laboratory parameters during the patient’s course.

The patient was enrolled in the COVALENT-101 study, ‘’A phase I first-in-human dose-escalation and dose-expansion study of BMF-219, an oral covalent menin inhibitor, in adult patients with acute leukemia (AL), diffuse large B-cell lymphoma (DLBCL), multiple myeloma (MM), and chronic lymphocytic leukemia (CLL/SLL)” (NCT05153330), approximately 10 months following his AML diagnosis. The study had been approved by the Institutional Review Board at the site and informed consent was obtained from the patient. BMF-219 (icovamenib) is a small molecule designed to disrupt interactions of menin with various proteins, such as KMT2A and JunD, which regulate multiple transcriptional and cell-cycle regulation pathways. Preclinical data on BMF-219 showed potent cytotoxicity on various menin-dependent AML cells, as well as cell lines and patients’ samples representing other hematologic malignancies.¹⁴

BMF-219 500 mg PO once daily was initiated in continuous 28-day cycles. On day 15 of cycle 1 (C1D15) of treatment, the patient was afebrile but with a dry cough. Chest computed tomography showed peribronchial ground glass opacities and the patient was hospitalized for suspected grade 3 differentiation syndrome, and given prednisone, allopurinol, acyclovir, piperacillin-tazobactam, and azithromycin. Although the patient’s baseline cytopenias persisted at C1D15, the peripheral blood counts revealed an emerging differentiation response to BMF-219 as evident by an increase in the absolute neutrophil count from 0.42x10⁹/L on C1D1 to 0.88x10⁹/L on C1D15, and an increase in the absolute monocyte count from 0.10x10⁹/L on C1D1 to 0.24x 10⁹/L on C1D15. The patient was discharged after 2 days in hospital and received prednisone as an outpatient until C1D20, with complete resolution of his symptoms.

Select laboratory parameters at diagnosis and throughout the patient’s course are shown in Table 1. The patient’s hematologic parameters showed steady improvement with BMF-219 (Figure 1). At study entry, the patient was receiving blood products 3-4 times weekly. This frequency decreased to weekly by the end of cycle 1, and 5 weeks after treatment initiation the patient became transfusion-independent and remained so for the duration of BMF-219 therapy.

The first bone marrow assessment at C2D1 revealed a partial response (ELN2017), with a decrease in blasts from 13% at baseline to 6% (Figure 1). The marrow karyotype was normal with 46,XX [20]; the WBC count, Hb concentration and platelet count were 1.99x10⁹/L (65% neutrophils), 7.6 g/dL, and 140x10⁹/L, respectively. Bone marrow analysis at C3D1 demonstrated a complete remission with 0% blasts, no circulating blasts, and normal female donor karyotype along with recovering normal hematopoiesis (WBC count=2.68x10⁹/L [65% neutrophils], Hb=10.3 g/ dL, platelet count=165x10⁹/L). Tested by flow cytometry (sensitivity ≥10-⁵) on C3D1, the patient had measurable residual disease (~0.2% of WBC). Gene expression changes consistent with the anti-leukemic response were observed (Figure 2). C4D1 assessments demonstrated continued cytogenetic complete response with 1% marrow blasts but with minimal residual disease-negativity, bone marrow NGS without detectable mutations, a WBC count of 4.16x10⁹/L (80% neutrophils), Hb of 11.5 g/dL, and platelet count of 143x10⁹/L.

At C5D1 and C6D1, the complete remission continued with 3% marrow blasts, a WBC count of 5.21x10⁹/L (76% neutrophils), Hb of 13.9 g/dL, and platelet count of 128 x10⁹/L but positive minimal residual disease was noted (~0.7% of WBC). NGS revealed the NUP98-NSD1 fusion and cytogenetics showed 46XY, t(2;10)(p11.2;q24), del(5)(q22. q35), del(13)(q12q22)[2]//46,XX[19]. The next dose level in the ongoing COVALENT-101 dose-escalation trial had been cleared, allowing the patient’s dose to be increased to 650 mg once daily towards the end of cycle 6. The higher dose level was well tolerated but, unfortunately, the patient’s leukemia progressed, and treatment was discontinued in C7D8. Bone marrow NGS at study discontinuation showed NUP98-NSD1 and mutations in CEBPA (VAF, 30%), KRAS G13D (VAF, 34.1%), NRAS G13D (VAF, 6.2%), WT1 (VAF, 45.4%), and RUNX1 VUS (VAF, 46.7%).

Co-occurring mutations may affect the efficacy of menin inhibition in NUP98-rearranged AML. For example, Heikamp et al.¹² showed responsiveness of a NUP98-NSD1 patient-derived xenograft model that was also WT1 mutant, while another NUP98-NSD1 AML patient-derived xenograft model containing mutations in WT1, ASXL1, IDH1, BCORL1 and FLT3-ITD did not respond. Of note, Rasouli and colleagues¹³ showed that menin inhibitor monotherapy was sufficient to suppress NUP98-rearranged AML with co-occurring FLT3-ITD and WT1 mutations; however, co-administration of a menin inhibitor with an FLT3 inhibitor enhanced the antileukemic activity of either alone. Similarly, Miao et al.¹⁵ found that combinations of a menin inhibitor with kinase inhibitors targeting either CDK6 (overexpressed in NUP98-rearranged leukemia) or FLT3 strongly enhanced the antileukemic activity of menin inhibition alone in samples from patients with NUP98-rearranged AML and patient-derived xenograft models. A broader clinical assessment of menin inhibitor monotherapy, as well as combinations with other active agents, in NUP98-rearranged AML is warranted.

Figure 2.Gene expression profiling in a patient with acute myeloid leukemia containing the NUP98-NSD1 fusion under treatment with covalent menin inhibitor BMF-219 (icovamenib). RNA-sequencing analysis of bone marrow aspirates revealed differentially expressed genes before and after treatment. Gene expression levels are presented as transcripts per million. Note at cycle 3 day 1 of treatment, coincident with attainment of complete remission, the pro-leukemogenic gene expression program in the bone marrow was downregulated by more than two-fold compared to the pre-treatment screening sample. These changes included suppression of key hematopoietic transcription factors (HOXA9, HOXA10, MEIS1, MEF2C), other relevant transcription factors (WT1, TRIB1, ZFP36L2, BCL6, BCL11B, cMYC), kinases (FLT3, CDK6), and the MEN1 gene, which encodes menin. Notably, there was no noticeable upregulation of markers of differentiation as observed with non-covalent menin inhibitors. Instead, BMF-219 led to their downregulation (CD14, ANPEP, ITGAM) or maintenance of gene expression level (MNDA). Expression of the housekeeping gene HPRT1 remained essentially constant across the screening, cycle 2, and cycle 3 timepoints. C2D1: cycle 2 day 1; C3D1: cycle 3 day 1.

Footnotes

• Received August 26, 2024
• Accepted November 25, 2024

Correspondence

H.E. CARRAWAY - carrawh@ccf.org

Disclosures

AC, YM, BM, GW, CM, UA, NK, TB and SWM are current or former employees of Biomea Fusion, Inc. HEC and JN have no conflicts of interest to disclose.

Contributions