Eltrombopag for Moderate Aplastic Anemia and Unilineage Cytopenias: Dosing, Long-Term Follow-up, Clonal Evolution and Somatic Mutation Profiling

Eltrombopag (EPAG) received FDA approval for treatment of refractory severe aplastic anemia (rSAA) in 2014, based on our phase I/II dose escalation trial of single agent EPAG for patients failing one or more treatment cycles with ATG/cyclosporine (Olnes NEJM 2012; Desmond Blood 2014). There is no standard treatment for patients with moderate aplastic anemia (MAA) or hypo-productive uni-lineage cytopenias (MAA/UC), conditions that can also impact on morbidity, mortality and quality of life. To explore the safety and effectiveness of EPAG in MAA/UC, we conducted a phase II study of EPAG given at escalating doses from 50-300mg/day (25-150mg/day for East Asians) through a primary hematologic response endpoint at 16-20 weeks (NCT 01328587). 34 patients enrolled between February 2012 and March 2017. 27 had never been treated with ATG/CSA IST. Responding patients could continue EPAG treatment on an extension arm. The drug was well-tolerated in 33/34 patients, with 1 patient coming off study at 10 weeks for nausea and vomiting. 25 patients reached the maximal dose. The median duration of follow-up in all patients was 16 months, and 27 months in responding patients. 17 of 34 (50%) of patients met criteria for response at the primary endpoint in at least one initially protocol-qualifying lineage (Hb 1.5 gr/dL increase in Hb or > 50% reduction in transfusions), including a patient with RSP19-mutated Diamond-Blackfan anemia. 7/24 with severe thrombocytopenia had a platelet response (>20,000/ul or transfusion-independence). 2/13 with both severe anemia and thrombocytopenia had bi-lineage responses at the primary endpoint, and an additional 5 went on to bi-lineage responses during the extension period. 3 patients without a response to EPAG were later treated with ATG/CSA and responded. EPAG was discontinued in 11/17 (65%) responding patients upon achievement of robust blood counts (Hb > 10mg/dL, platelets > 50,000/uL, and ANC > 1000) or stable blood counts for 6 months in 1 patient (6%) after a median duration of drug administration of 8 months (2-14 months) (Fig 1). In contrast to our prior experience in rSAA, the majority of patients still being followed on study after drug discontinuation (8/10) needed to have EPAG re-initiated for declining counts at a median of 6 months (2-38) later. All 8 responded again, with 4/8 able to discontinue EPAG after a second robust or stable response. 2/34 patients (6%) developed marrow cytogenetic abnormalities while on drug in contrast to 16/83 (16%) rSAA patients treated with EPAG in our prior studies. Both MAA patients were responders and neither had dysplastic changes or increased blasts. Patient # 5 had +8 in 7/20 metaphases at 22m, went off drug and relapsed, and EPAG was restarted off protocol. Repeat cytogenetics on EPAG were normal. Patient #20 had del13q in 5/20 metaphases at 9.5m, EPAG was stopped, and repeat cytogenetics off drug were normal. Of note, patient #12 had a robust response and had been off drug for 25m when counts declined and BM revealed dysplastic changes with no increase in blasts and normal cytogenetics. The acquisition and selection of somatic mutations, particularly in myeloid candidate genes (MCG) recurrently mutated in MDS/AMLhas been proposed to be an initiating step in clonal evolution. We performed targeted next generation exome sequencing (NGS) of 66 MCG and additional genes previously found to be somatically-mutated in SAA (Yoshizato, NEJM, 2015) pre-EPAG treatment and at the primary response endpoint of 16-20 weeks. Only 5 patients showed somatic mutations in these genes at baseline (SETBP1, CBL, SF3B1, PPMID, EP300). There were no significant changes in VAF. 2 mutations became detectable on EPAG (BCOR, and DNMT3A transiently), and 1 disappeared (SF3B1). In summary, administration of EPAG at escalating doses up to of 300mg/day was well-tolerated and 50% of patients with MAA/UC had clinically-meaningful responses, including those not previously treated with IST. The responses were durable, often robust, although frequently required ongoing EPAG treatment, in contrast to the experience in rSAA. Clonal cytogenetic evolution was rare, with no instances of chromosome 7 abnormalities, and there was no consistent expansion of MCG somatically-mutated clone size in patients during EPAG treatment. Figure. Figure. Disclosures Winkler: National Institute of Health: Research Funding. Young:CRADA with Novartis: Research Funding; GlaxoSmithKline: Research Funding; National Institute of Health: Research Funding. Dunbar:National Institute of Health: Research Funding.