Background
Investigator-initiated, prospective, randomized, placebo-controlled, double-blind, multicenter phase 3 trial (NCT02773225; EudraCT 2014-000174-19). N=85 with moderate aplastic anemia (MAA) and clinically significant cytopenia, treated first line. Eltrombopag added to horse ATG plus cyclosporine A (CSA) is standard of care in transplant-ineligible severe aplastic anemia, but no consensus standard exists for moderate disease, where ATG is generally not used and practice ranges from supportive care to single-agent CSA. EMAA is the first phase 3 to test whether adding the thrombopoietin receptor agonist eltrombopag to CSA improves trilineage response in MAA.
Results
Interventions and follow up: Arm A: Eltrombopag + cyclosporine A, first line (n=41); dose and schedule per protocol, NR in the primary report
Arm B: Placebo + cyclosporine A (n=44)
Crossover: after the week-24 response assessment and unblinding, placebo-arm patients without CR switched to eltrombopag plus continued CSA
Primary endpoint: Overall response rate (CR + PR) at week 24, in patients evaluable for response (n=75)
mFollow up: NR
Results: ORR at week 24 (primary): 71.4% (95% CI 54.8–83.8) with eltrombopag+CSA vs 42.5% (95% CI 28.5–57.8) with placebo+CSA, P=.011 (Fisher exact)
ORR after crossover: 73.5% by week 48 in placebo-arm patients who added eltrombopag, P=.009 (McNemar, week 24 vs week 48)
CR vs PR breakdown: NR
Relapse rate: no significant difference between arms
Failure-free survival: no significant difference between arms
Arm B: Placebo + cyclosporine A (n=44)
Crossover: after the week-24 response assessment and unblinding, placebo-arm patients without CR switched to eltrombopag plus continued CSA
Primary endpoint: Overall response rate (CR + PR) at week 24, in patients evaluable for response (n=75)
mFollow up: NR
Results: ORR at week 24 (primary): 71.4% (95% CI 54.8–83.8) with eltrombopag+CSA vs 42.5% (95% CI 28.5–57.8) with placebo+CSA, P=.011 (Fisher exact)
ORR after crossover: 73.5% by week 48 in placebo-arm patients who added eltrombopag, P=.009 (McNemar, week 24 vs week 48)
CR vs PR breakdown: NR
Relapse rate: no significant difference between arms
Failure-free survival: no significant difference between arms
Adverse events
Grade ≥3 (any): NR
Hematologic: NR
Hepatic (eltrombopag class effect): NR
Clonal evolution / cytogenetic abnormality: NR
Discontinuation: NR
Overall: eltrombopag was well tolerated with no new safety concerns
Hematologic: NR
Hepatic (eltrombopag class effect): NR
Clonal evolution / cytogenetic abnormality: NR
Discontinuation: NR
Overall: eltrombopag was well tolerated with no new safety concerns
Conclusions
Adding eltrombopag to cyclosporine A as first-line therapy for moderate aplastic anemia with clinically significant cytopenia significantly improved trilineage hematologic response at week 24, raising ORR from 42.5% to 71.4%. Patients who did not achieve CR on 24 weeks of CSA alone and then had eltrombopag added reached an ORR of 73.5% by week 48, supporting eltrombopag as either upfront combination or delayed add-on. Relapse and failure-free survival did not differ between arms, so the demonstrated benefit is on depth and rate of early response rather than on durability.
Key Limitations
N=85 is small for a phase 3, and the primary analysis was restricted to the 75 patients evaluable at week 24, so the effect estimate rests on a modest denominator with wide confidence intervals. Week-24 ORR is a surrogate: relapse rate and failure-free survival, the outcomes that matter over years in aplastic anemia, did not differ between arms, which raises the question of whether faster response translates into durable benefit or transfusion independence. Mandatory crossover after week 24 permanently forecloses a clean long-term randomized comparison. The control arm was single-agent CSA, which is reasonable for moderate disease but is not universal practice. Clonal evolution to MDS/AML and PNH clone expansion, the central long-term safety concerns with thrombopoietin receptor agonists in bone marrow failure, are not addressed at this follow-up duration and require extended surveillance. Neither median follow-up nor granular toxicity rates were reported.
Clinical Context
Eltrombopag is FDA approved in combination with standard immunosuppressive therapy for first-line severe aplastic anemia, on the basis of the NIH phase 1/2 experience, and the randomized RACE trial established hATG + CSA + eltrombopag as the transplant-ineligible severe AA standard. Moderate aplastic anemia has had no evidence-based standard of care: guidance from NCCN, the British Society for Haematology, and EBMT/severe aplastic anemia working party has generally advised observation for asymptomatic patients and CSA-based immunosuppression for those with transfusion-dependent or otherwise clinically significant cytopenias, with ATG reserved for severe disease. EMAA is the first randomized phase 3 evidence in this population and supports adding eltrombopag to CSA up front, or adding it at 24 weeks in non-responders. Eltrombopag is not separately labeled for moderate aplastic anemia, so use in this setting is off-label pending regulatory or guideline uptake.