Adding eltrombopag to immunosuppressive treatment with cyclosporine as front-line therapy for moderate aplastic anemia improves trilineage hematologic response: One-year analysis of the randomised, placebo-controlled, double-blind Phase III trial eltrombopag for moderate aplastic anemia (EMAA Trial)

Abstract

Abstract Introduction: Eltrombopag (EPAG) combined with horse antithymocyte globulin and cyclosporine A (CSA) when given as first-line treatment of patients (pts) with severe aplastic anemia (SAA) improves rate, rapidity, and strength of hematologic response. No consensus exists on the optimal treatment for moderate aplastic anemia (MAA), and to date, there has been no randomized trial on the efficacy and safety of EPAG in MAA pts. The EMAA trial (Eltrombopag for Moderate Aplastic Anemia) compared placebo with EPAG as a first-line therapy for pts with MAA who also received CSA as standard immunosuppression (NCT02773225; EudraCT2014-000174-19). Methods: The EMAA trial is an investigator-initiated, prospective, randomized, placebo-controlled, double blind, multicenter, phase III trial. It included 85 treatment-naïve pts with MAA (full analysis set [FAS]) requiring treatment due to severe cytopenia in at least one cell lineage or transfusion dependency. Pts were randomized to receive placebo (n = 44) or EPAG (150 mg/day) (n = 41) until week (wk) 24. CSA was given at 5 mg/kg/day in both arms, adjusted to trough levels of 150–250 ng/mL. The primary endpoint was the proportion of pts with trilinear hematologic response at wk 24 and was assessed at wk 24 in 75 pts (40 placebo, 35 EPAG; per protocol set [PPS]) followed by unblinding of therapy. In pts in the placebo arm without a complete response (CR) by wk 24, EPAG (150 mg/day) was added from wk 25 to 48. Results: The median age was 53 (interquartile range [IQR] 35-65) years; 47.1% were female. The median (IQR) baseline platelet, neutrophil and hemoglobin concentrations were 22x10⁹/L (14-29x10⁹/L), 0.98x10⁹/L (0.64-1.29x10⁹/L), and 9.0 g/dL (8.2-9.8 g/dL), resp. Patient characteristics in both arms were well balanced. Ten pts discontinued before the 24-wk evaluation (n=4 placebo, n=6 EPAG). In the PPS, the overall response rate (ORR), including PR and CR, after 24 wks was significantly higher in the EPAG arm (71.4%; 25/35; 95% confidence interval [CI], 54.8%-83.8%) than in the placebo arm (42.5%; 17/40; 95% CI, 28.5%-57.8%)(p = 0.011, Fisher's exact test; odds ratio [OR] of ORR 3.325; 95% lower confidence limit (CL) for OR 1.354). At 24 wks, the proportion of pts with CR was 11.4% (4/35) in the EPAG arm and 5.0% (2/40) in the placebo arm. Pts with transfusion-dependent MAA had a significantly lower chance of reaching the primary endpoint (OR 0.18; 95% CI 0.05-0.63, p=0.007). In the FAS, the ORR after 12 wks was 48.8% (20/41; 95% CI 34.3%-63.5%) in the EPAG arm and 25% (11/44; 95% CI 14.4%-39.6%) in the placebo arm (p = 0.02; OR for ORR 2.821; 95% lower CL 1.202). After response assessment at wk 24 and unblinding, pts in the placebo arm without CR (n=38; 23 NR, 15 PR) received EPAG in addition to continued CSA from wk 25. This group achieved an ORR of 73.5% at wk 48 (25/34 pts with an available 48 wk assessment: 9 NR, 14 PR, 11 CR, 4 not assessed; p=0.009, McNemar test, comparing ORR at 24 vs. 48 wks, i.e. before vs. after adding EPAG). Thus, adding EPAG to placebo arm pts without CR at wk 24 increased both the ORR and the strength of response. The cumulative incidence of trilineage response was superior in pts who had been initially randomized to EPAG (HR 1.71; p=0.039). The safety analysis included 77 pts with ≥ 10 wks of therapy: 70 pts experienced a treatment-emergent AE (TEAE) until wk 10, and 58/77 from wk 10 to 24 (until wk 10: 95.1% and 86.1% in the placebo and EPAG arms, resp.; after wk 10 to wk 24: 68.3% and 83.3% in the placebo and EPAG arms, resp.). Serious AEs (SAE) were reported by 8/77 pts (10.4%) until wk 10 and 9/77 pts (11.7%) between wk 10 and 24 (10.4% and 11.1% in the placebo and EPAG arms, resp.). Three SAE (1 thrombosis, 1 pulmonary embolism, 1 pyrexia) were assessed as related to the study medication (2 EPAG, 1 placebo arm). Conclusions:Adding EPAG to CSA treatment for first-line treatment of MAA significantly improves trilineage hematologic response at wk 24. Addition of EPAG in pts without CR after 24 wks of single-agent CSA treatment also improves the trilineage hematologic response rate. The cumulative incidence of trilineage response was significantly better in pts randomized into the EPAG arm. EPAG was well tolerated, and no new safety concerns were identified. EPAG combined with standard CSA therapy should be the preferred initial treatment for pts with MAA and could become the new standard of care for MAA, administered orally on an outpatient basis.