Sotatercept for anemia of myelofibrosis: a phase II investigator-initiated study

Anemia (hemoglobin <10 g/dL) is common in myelofibrosis (MF), present in about a third of patients at diagnosis and eventually developing in all patients. The Janus kinase 1/2 (JAK1/2) inhibitor ruxolitinib ameliorates splenomegaly and symptoms of MF and prolongs survival;¹ however, on-target anemia from JAK2 inhibition, especially pronounced in the first 12-24 weeks of therapy, is a significant problem. Anemia may be the most common cause of ruxolitinib discontinuation,² and frequently results in dose reduction. Spleen responses to ruxolitinib are dose-dependent and correlate with survival.^3,4 Thus, counteracting ruxolitinib-induced anemia remains an important goal.

Very recently, the JAK1/2 and activin receptor type 1 (ACVR1) inhibitor, momelotinib, was approved in the US for anemic patients with intermediate/high-risk myelofibrosis, based on the SIMPLIFY-1 and MOMENTUM trials.^5,6

Therapies currently used specifically for anemia of MF include corticosteroids, danazol, erythroid-stimulating agents (ESA) and immunomodulatory drugs, but responses are infrequent and often short-lived. Sotatercept (formerly ACE-011, Acceleron Pharma, Cambridge, MA, now Merck, Kenilworth, NJ), a novel fusion protein, is a first-in-class, activin receptor type IIA (ActRIIA) “ligand trap” that sequesters MF bone marrow-derived TGF-|3 superfamily ligands (such as Activin A and growth and differentiation factor 11) that inhibit terminal erythropoiesis via Smad signaling upon ActRIIA binding.^7,8 Sotatercept demonstrated substantial efficacy in anemic patients with |3-thalassemia and myelodysplastic syndromes (MDS).^9,10

This was a phase II, open-label, single-institution, investigator-initiated trial (clinicaltrials gov. Identifier: NCT01712308). Adults (≥18 years) with primary MF (PMF) or post polycythemia vera/essential thrombocythemia myelofibrosis (post-PV/ET) MF were eligible if they were anemic (i.e., hemoglobin [Hg]b <10 g/dL sustained over ≥84 days preceding study entry without red blood cell [RBC] transfusions, or Hgb <10 g/dL with occasional transfusions but not RBC-transfusion dependant [TD] per International Working Group-Myeloproliferative Neoplasms Research and Treatment [IWG-MRT] criteria), or RBC-TD per IWG-MRT criteria.¹¹ Sotatercept was administered subcutaneously every 3 weeks. All monotherapy patients after the first patient, who received 0.3 mg/kg, received 0.75 mg/kg or 1 mg/kg. Upon early demonstration of activity, a combination cohort was added: patients must have been on ruxolitinib for ≥6 months with a stable dose for the preceding ≥8 weeks. The sotatercept dose chosen for this cohort was 0.75 mg/kg, as most responses in the monotherapy cohort at the time had been observed at this dose. MF-directed therapies within 2 weeks of sotatercept initiation were not permitted, except ruxolitinib in the combination cohort. Patients with uncontrolled hypertension were excluded. Additional eligibility criteria are listed in the study protocol, available as a supplement. Sotatercept was held for Hgb values ≥11.5 g/ dL (resumed once the Hgb level was <11 g/dL). Concomitant use of erythroid stimulating agents or any other MF-directed therapy (except ruxolitinib in the combination cohort) was not permitted. The study was approved by the MD Anderson Cancer Center Institutional Review Board and was conducted according to the principles of the Declaration of Helsinki. All participants provided written informed consent. The study was supported by Celgene Corporation (now Bristol Myers Squibb) through drug supply and funding. BMS/Celgene had no role in the study design, data collection, analysis, interpretation, or manuscript writing.

The primary endpoint of the study was the anemia response rate, a composite of hemoglobin response in non-TD patients, and achievement of transfusion independance (TI) in RBC-TD patients. Hemoglobin response was defined as an increase from baseline Hgb level of ≥1.5 g/dL sustained for ≥84 days, without RBC transfusions (Gale criteria).¹² The baseline Hgb in anemic patients was the lowest Hgb level in the 84 days preceding study entry. In patients who were RBC-TD at enrollment, TI was defined as no RBC transfusions in any “rolling” 84-day interval during the treatment period. Secondary endpoints included duration of and time to response. All patients who received at least one dose of sotatercept were evaluable for safety.

Patients had to remain on study for ≥84 days to be efficacy-evaluable.

Online Supplementary Figure S1 shows the patient disposition. A total of 63 patients were enrolled and 56 were treated. One patient received a dose of 0.3 mg/kg for 6 cycles and is not considered further. Thirty-four patients received sotatercept monotherapy (16 at 0.75 mg/kg/dose and 18 at 1 mg/kg/dose), and 21 received sotatercept (0.75 mg/kg/dose) “added” to a stable dose of ruxolitinib. Baseline characteristics appear in Table 1. Five patients in the monotherapy cohort were treatment-naïve. Prior therapies in the remainder are available in Online Supplementary Table S1. In the monotherapy cohort, 17 patients each were “anemic” and RBC-TD at study entry. In the combination cohort, 15 patients were “anemic” and six were RBC-TD at study entry. All patients are currently off-study. The study was terminated after the commercial supporter ended investigational drug supply in December 2021. Four patients, two in each cohort, were receiving sotatercept on the study at that point.

Eight of 27 evaluable (30%) patients in the monotherapy cohort responded. Five were anemia responses (of 13 evaluable) and three, TI responses (of 14 evaluable). Six responses (4 anemia and 2 TI) occurred at the 0.75 mg/kg dose, and two (1 anemia and 1 TI) at the 1 mg/kg dose. Of the seven unevaluable patients, three had Hgb increases of ≥1.5 g/dL from baseline but came off study due to hypertension deemed related to sotatercept (N=1), or a decision to proceed to allogeneic hematopoietic cell transplantation (allo-HCT) (N=2). The median number of cycles of sotatercept was 6 (range, 1-73), and median time on-study was 4.4 (range, 0.7-75.3) months. The median time to (onset of) response was 19 (range, 1-22) days, and the median duration of response was 23.3 (range, 3.9-74.6) months. Reasons for discontinuation of sotatercept included lack or loss of response (N=14), MF progression in other aspects, e.g., splenic progression (N=6), allo-HCT (N=4), logistical/travel-related (N=3), patient decision (N=2), study termination (N=2), transformation to AML (N=1), hypertension deemed related to sotatercept (N=1) and medical complications unrelated to sotatercept (N=1).

In the combination cohort, there were six responses (of 19 evaluable patients, 32%). All were hemoglobin responses (of 14 evaluable); there were no TI responses (of 5 evaluable).

Table 1.Characteristics of the patients at baseline.

Table 2.Adverse events* at least possibly related to sotatercept.

The median number of cycles of sotatercept was 8 (range, 2-52), and the median time on-study was 5.5 (range, 1.6-57.1) months. The median time to (onset of) response was 14 (range, 6-147) days, and the median response duration was 20.9 (range, 3.7-56.8) months. Reasons for discontinuation of sotatercept included lack or loss of response (N=8), allo-HCT (N=4), MF progression in other aspects (N=2), logistical/travel-related (N=2), study termination (N=2), transition to hospice (N=1), patient decision (N=1) and loss of insurance (N=1).

Several responders in both cohorts required sotatercept doses to be held per protocol for Hgb levels ≥11.5 g/dL, with resumption of dosing when the Hgb level was <11 g/dL. Eight responders, five in the monotherapy cohort and three in the combination cohort, experienced multiple instances of this phenomenon. However, we were not able to identify a molecular biomarker or clinical factor predictive of these robust and durable responses to sotatercept.

We did not observe any consistent effects of sotatercept on other disease-related parameters, such as spleen size, symptoms, leukocyte or platelet counts, bone marrow fibrosis grade and JAK2 V617F variant allele frequency. No responder in either cohort had a detectable SF3B1 mutation at study entry; however, spliceosome genes were not sequenced as part of our institutional next-generation sequencing panel until April 2017. No responder had bone marrow ring sideroblasts (RS) at study entry.

Sotatercept was well-tolerated. Table 2 lists the adverse events (AE) felt to at least possibly be related to sotatercept. No grade 4 or 5 AE occurred. Seven patients experienced grade 3 hypertension on the study, not in the context of high Hgb levels. Hypertension (all grades) occurred in 20% of patients. Pain in the extremities (muscle, bones, joints) on the days following injection of sotatercept was common, occurring in 40%; however, most of these events were grade 1 or 2 in severity, with only two patients reporting grade 3 limb pain. There were no on-study deaths.

In conclusion, our study adds to a growing body of evidence supporting the safety and clinical activity of the activin receptor ligand traps in anemic patients with myeloid malignancies. Although sotatercept is currently being developed for the treatment of pulmonary arterial hypertension,¹³ luspatercept, an ActRIIB ligand trap, is approved for the treatment of anemia in patients with lower-risk MDS with RS, as well as those with myelodysplastic/myeloproliferative neoplasm with RS and thrombocytosis. In a phase II study in 95 patients with MF and anemia, luspatercept led to a 26.3% rate of TI during the primary treatment period (24 weeks) in the cohort of RBC-TD patients receiving a stable dose of ruxolitinib (N=38), and 50% of the patients in this cohort experienced at least halving of their transfusion burden during this time. These results have led to an ongoing, phase III, placebo-controlled trial of luspatercept (INDEPENDENCE™) in RBC-TD MF patients on a stable dose of a JAK inhibitor.¹⁴ Early data on elritercept, an investigational, modified ActRIIA ligand trap are promising, with some “trifactor” (hematopoiesis, spleen, and symptoms) responses observed.¹⁵

Footnotes

• Received August 14, 2023
• Accepted March 26, 2024

Correspondence

°JEC current address: Georgia Cancer Center, Augusta, GA, USA

P. BOSE - pbose@mdanderson.org

Disclosures

PB discloses research support to his institution from BMS, Incyte, CTI, Morphosys, Kartos, Telios, Ionis, Disc, Blueprint, Cogent, Geron, Janssen, and Sumitomo; and honoraria/consulting fees from Incyte, BMS, CTI, GSK, Abbvie, Morphosys, Karyopharm, Pharma Essentia, Blueprint, Cogent, Novartis, Jubilant, and Morphic. HK discloses research grants to his institution from AbbVie, Amgen, Ascentage, BMS, Daiichi-Sankyo, Immunogen, Jazz, and Novartis; and honoraria/ consulting fees from AbbVie, Amgen, Amphista, Ascentage, Astellas, Biologix, Curis, Ipsen Biopharmaceuticals, KAHR Medical, Labcorp, Novartis, Pfizer, Shenzhen Target Rx, Stemline, and Takeda. TK discloses research support to his institution from AbbVie, Amgen, Ascentage, Astella, Astex, AstraZeneca, BMS, Celgene, Cellenkos, Cyclacel, Delta-Fly Pharma, Genfleet, Genentech, Glycomimetrics, Iterion, Janssen, Jazz, Pfizer, Pulmotect, Regeneron, and Sellas; and honoraria/consulting fees from Abbvie, Agios, Daiichi Sankyo, Genentech, Genzyme, Jazz, Liberum, Novartis, Pfizer, PinotBio, Pulmotect, and Sanofi-Aventis. JC discloses support to his institution from Abbvie and Actuate. EJJ discloses research support to his institution, as well as honoraria/consulting fees from Abbvie, Adaptive Biotech, Amgen, Bristol Meyers Squib, Ascentage, Genentech, Novartis, Pfizer, and Takeda. NP discloses research support to his institution from the US Department of Defense, other financial or non-financial interests in Dan’s House of Hope, and honoraria/consulting fees from Abbvie, Aplastic Anemia and MDS International Foundation, Aptitude Health, Astellas Pharma US, Blueprint Medicines, BMS, CancerNet, CareDx, Celgene, Cimeio Therapeutics AG, Clearview Healthcare Partners, CTI BioPharma, Curio Science, DAVA Oncology, EUSA Pharma, Harborside Press, Imedex, Immunogen, Intellisphere, Magdalen Medical Publishing, Medscape, Menarini Group, NeoPharm, Novartis Pharmaceuticals, OncLive, Pacylex, Patient Power, PeerView Institute for Medical Education, Pharma Essentia and Physicians’ Education Resource. MA discloses research support to his institution from Daiichi Sankyo, Breast Cancer Research Foundation, Astra Zeneca, Oxford Biomedical UK, Eterna, Senti Bio, Pinot Bio and Syndax; payments for advisory board or data safety monitoring board participation from Cancer UK, Leukemia and Lymphoma Society, Aptose, German Research Council, National Cancer Institute, CLL Foundation and Eterna; and stocks or stock options from Reata, Aptose, Eutropics, Senti Bio and Chimerix. NGD discloses research support to his institution from Daiichi-Sankyo, Bristol- Meyers Squibb, Pfizer, Gilead, Servier, Genentech, Astellas, AbbVie, ImmunoGen, Amgen, Trillium, Hanmi, Trovagene, FATE Therapeutics, Novimmune, Glycomimetics, and KITE; and consulting fees from Daichii-Sankyo, Bristol-Meyers Squibb, Pfizer, Gilead, Servier, Genentech, Astellas, AbbVie, ImmunoGen, Amgen, Trillium, Arog, Novartis, Jazz, Celgene, Syndax, Shattuck Labs, Agios, KITE and Stemline/Menarini. All other authors have no conflicts of interest to disclose.