A Phase 2 Study of the LSD1 Inhibitor Bomedemstat (IMG-7289) for the Treatment of Advanced Myelofibrosis (MF): Updated Results and Genomic Analyses

Abstract

The need for therapies for myelofibrosis (MF) persists, especially in the setting of JAK inhibitor failure. Lysine-specific demethylase-1 (LSD1) is a histone demethylase critical for self-renewal potential of malignant stem cells and progenitor maturation. Bomedemstat is an oral LSD1 inhibitor clinically active in patients with MPNs. IMG-7289-CTP-102 is a global, open-label Phase 2 study evaluating bomedemstat dosed in MF patients (NCT03136185).

Eligibility criteria included IPSS int. or high-risk patients refractory, resistant, poorly controlled, or intolerant to JAK inhibition with a platelet count ≥100 x 109/L. Splenomegaly was not required. Key objectives were safety, spleen volume reduction (SVR) and total symptom score (TSS) improvement. Serial bone marrow (BM) biopsies were read centrally. 261 genes recurrently mutated in AML, MPNs and MDS, were sequenced at baseline and serially to quantify the frequencies of mutant alleles (MAF) and to identify new mutations. Dosing was individualized using platelet count as a biomarker of bomedemstat activity on megakaryocyte function targeting counts of 50-75 x 109/L.

The study is fully enrolled (N=89): 46% PMF, 33% PET-MF, 21% PPV-MF. Median age was 68 (35-88) with 52% males. Prior treatment with ruxolitinib was reported in 82%, 70% had also received ≥1 other treatments. The median baseline SV was 1354 cm3, expectedly low given the entry criteria. 37% of patients (33/89) had received ≥1 RBC transfusion prior to dosing. IPSS risks were 53% (high), 40% (int.-2), 7% (int.-1). At screening (N=89), deep sequencing (>1000bp) found mutations in JAK2 (65%), CALR (22%), or MPL (7%); 61% had ≥2 mutations of which 67% had ≥1 high-molecular risk mutation (e.g., ASXL1, IDH1/2, EZH2, SRSF2).

The median duration of treatment stands at 28 wks (2-132). In patients evaluable at 24 wks with baseline TSS ≥20, 65% recorded a reduction (mean change -19%; -81% to +52%) with 19% reporting ≥50% reduction. Of patients evaluable for SVR at 24 wks (N=50), 66% had SVR from baseline: 28% ≥20% SVR, 6% showing ≥35% SVR. Of transfusion-independent patients (N=41), 90% had stable (∆ <±1.0 g/dL) or improved hemoglobin (Hb) (≥1.0 g/dL). Of those requiring transfusions, 14% became independent. Of patients evaluable for follow-up BM fibrosis scoring (N=59), 85% improved by ≥1 grade or were stable; 53% of patients with improved BM fibrosis also saw improved Hb or required fewer transfusions.

Of 43 pts with follow-up sequencing data, the frequency of mutations in one or more alleles were reduced in 42% (with 2 patients showing complete elimination of mutant alleles); 44% were stable. ASXL1 mutations were the most commonly reduced with most consistent declines in truncating ASXL1 mutations clustered around codon 642. No new mutations were identified in any patient.

We identified in the germline at the JAK2 locus a haplotype distinct from 46/1, we call P6, composed of high-frequency polymorphic sites extending >40kb that was significantly enriched in our MF population (20%; N=88) compared to the frequency of P6 in the 1000 genome population (6.35%; n=2504) or in AML patients (5.13%; N=43). The P6 haplotype was also enriched in the germlines of MF (18.2%; N=22), PV (19.3%; N=57) and ET (15.8%; N=95) patients in the European Genome Archive MPN dataset. There was a significant increase in the likelihood of somatic JAK2 mutations on the P6 haplotype in our patient population (p = 0.023, N=196). Of the six P6 patients with 24 wk follow-up MAF, all were JAK2V617F homozygotes.

The most common non-hematologic AEs were dysgeusia (36%) and diarrhea (33%). Of the 14 related SAEs, 4 were Grade 2, 9 Grade 3 and 1 Grade 4 (thrombocytopenia). 29 pts have entered an extension study. Early terminations due to AEs occurred in 18 (20%) pts (9 related to bomedemstat) and 12 discontinued for other reasons. There were no safety signals, DLTs, or deaths related to drug. No patient transformed to AML during this study.

In conclusion, in pts with advanced disease, a high-risk mutation burden and limited treatment options, once-daily bomedemstat as oral monotherapy had an acceptable tolerability profile, relieved symptoms, reduced spleen volume and BM fibrosis while improving anemia without significant safety signals. Further, bomedemstat selects against clones with ASXL1 mutations. The impact of the P6 haplotype on disease course and response to bomedemstat is under study.