Eltrombopag for bone marrow failure and Diamond-Blackfan Anemia

National Heart, Lung, And Blood Institute

Investigators

Linked publications & trials

Abstract

We developed a translational and clinical research program investigating the efficacy and safety of in vivo stimulation of hematopoiesis via the drug eltrombopag, a small molecular that binds to the c-mpl receptor. Murine, our own rhesus macaque studies, and human data all suggest that the endogenous hormone thrombopoietin can stimulate primitive hematopoietic stem and progenitor cells to proliferate without terminal differentiation. The small molecule oral thrombopoietin mimetic eltrombopag was initially developed to overstimulate platelet production from marrow megakaryocytes to compensate for platelet destruction in immune thrombocytopenia purpura (ITP). We originally developed a protocol utilizing eltrombopag in a phase 1/2 clinical trial for patients with refractory severe aplastic anemia. We reported that eltrombopag had efficacy in this setting with 44% (11/25) of patients having clinically significant hematologic responses (Olnes et al, NEJM 2012). We then reported additional safety and efficacy data on an expanded cohort of 43 patients, confirming an overall response rate or 40% with 3 to 4 months of treatment, including tri- and bilinear responses (Desmond et al, Blood, 2014). The majority of patients who remained on eltrombopag in an extension study continued to show improvement, and most eventually had significant increases in neutrophil, red cell, and platelet lineages. Those with robust near-normalization of blood counts had drug discontinued at a median of 28.5 months after entry and the vast majority maintained stable counts off eltrombopag. A subsequent trial in refractory SAA (13-H-0133) asked whether more prolonged administration of eltrombopag to patients with refractory SAA, for 6 months instead of 3 months, would improve response rate and rescue a larger fraction of refractory patients. The response rate at the primary 6 month end point was 50%, not significantly improved over the initial trial. However, 25% of patients (5/20) did not reach response criteria at 3 months but were responders at 6 months, and thus were salvaged by the more prolonged treatment. We combined both cohorts for molecular analyses asking whether EPAG impacts on clonal progression, and whether HPSC carrying mutations in specific genes linked to myeloid malignancies or other clones defined by new somatic mutations are specifically stimulated by EPAG treatment. We found no evidence for specific somatically-mutated clonal expansions related to EPAG treatment. The cytogenetic progression rate in both trials combined is now 19%. Of note, all patients that developed monosomy 7 or other deleterious chromosome 7 abnormalities did so within the first 3-6 months, suggesting potential preferential stimulation of a pre-existing clone. Other cytogenetic abnormalities appeared later, were often transient whether on or off drug, and were not accompanied by dysplastic changes or leukemic progression(Winkler et al, Blood, 2019). This program resulted in FDA approval of eltrombopag in August 2014 and European commission approval in 2015 for eltrombopag treatment of refractory aplastic anemia. This was the first new drug approved for aplastic anemia in decades, and the first drug approved specifically for the refractory aplastic anemia patient population. In addition, a collaborative trial carried out with Dr Neal S Young and his Branch documented that EPAG added to standard immunosuppression for untreated severe AA resulted in improved outcomes (Townsley et al, NEJM, 2017), leading to a 2nd FDA/EHA approval for EPAG in AA based on our studies. We have now completed analyses and published results from 34 patients enrolled in a clinical trial (11-H-0134) examining safety and efficacy of eltrombopag in patients with moderate aplastic anemia or unilineage cytopenias. The response rate was 50% at the primary endpoint of 4 months, including in a patient with the inherited Diamond-Blackfan Anemia (DBA). Patients tolerated doses up to 300mg/day without significant toxicity. We noted a lower rate of cytogenetic progression (6%) than in the prior trials in severe refractory aplastic anemia, and no patients with chromosome 7 cytogenetic progression. 65% of patients were able to have EPAG discontinued for a robust multi lineage response, however the majority needed to have the drug restarted to maintain counts, in contrast to our prior experience in refractory severe aplastic anemia (Fan et al, Blood Advances, 2020). Based on data generated in all three trials, we made the observation that eltrombopag acts as a clinically-relevant iron chelator, raising blood levels of iron and resulting net iron loss in a cohort of patients on the drug long term (Zhao et al, Blood 2018). A much larger cohort has now been analyzed and compared to marrow failure patients not treated with EPAG, documenting rapid and clinically-significant iron unloading in patients on long-term EPAG. Response rates and relapse rates were not impacted by initial iron status, suggesting that in aplastic anemia the activity of EPAG is linked to HSC stimulation not iron unloading. However, several patients have required oral iron supplementation while on long-term EPAG to avoid clinically-relevant iron deficiency, a finding of clinical relevance that is being submitted for publication. We hypothesize that the surprising activity of EPAG in reversing anemia in the inherited ribosomopathy DBA may be due to the potent intracellular chelating activity of EPAG, because recent laboratory studies suggest that erythroid development is inhibited in DBA due to slowed protein synthesis in erythroid progenitors, with a resulting imbalance in global chain production versus heme biosynthesis, leading to free heme/increased intracellular iron and toxic accumulation of reactive oxygen species. We have designed and now initiated a clinical trial to investigate the safety and activity of EPAG in DBA. Patient accrual finally began in mid FY21, after delays due to restrictions on patient travel and clinical care due to the COVID-19 pandemic. We have enrolled 15 patients and have already seen clinical responses. We are carrying out correlative laboratory and imaging studies to assess iron status and mechanism of EPAG action. We have also noted that the initial DBA patient responding to EPAG in our prior trial was a mosaic, with likely somatic reversion in an early HSC, resulting in a fraction of wild type HSC hematopoietic output. Despite this mosaicism, the patient remained severely anemic and transfusion-dependent prior to EPAG, suggesting that mutant developing erythroid cells could inhibit wild-type cells within erythroblastic islands. We have explored this hypothesis in a murine competitive repopulation model collaboratively with Janis Abkowitz at the University of Washington, documenting a marked inhibitory effect of mutant DBA cells on WT cells in experimental murine transplant chimeras. We are now initiating single cell RNASeq and genotyping of developing erythroid cells from both DBA patients (including the mosaic patient) and mosaic mice to investigate the likely mechanism and pathways that are involved. These studies have great relevance for the development of gene therapies for DBA, which would not be effective if residual mutant cells can inhibit wild type erythropoiesis.