Eltrombopag Added to Standard Immunosuppressive Treatment Improves Long-Term Outcomes As Front-Line Therapy for Severe Aplastic Anemia: Final 2-Year Analysis of EBMT-Saawp Race Study

Introduction

The RACE study (NCT02009747) compared standard Immuno-Suppressive Treatment (IST) (as horse antithymocyte globulin plus ciclosporin A) ± eltrombopag [EPAG] as front-line treatment for Severe Aplastic Anemia (SAA). Primary analysis has shown that triple therapy improved rate and quality of hematological response, with higher rates of complete response (CR) at 3 months (21.9% vs. 9.9%, primary endpoint), and of overall response (OR) at 3 months and CR and OR at 6 months. Here we report the final analysis of this 2-year follow-up, prospective, randomized, phase III study, focusing on the long-term outcomes.

Methods

The trial population consisted of 197 treatment-naive patients randomized to receive either standard IST (arm A; n=101) or standard IST + EPAG (arm B; n=96) at the dose of 150 mg/d from day +14 until 6 months (m) (or 3m, in case of early complete response). Here we report final data on overall survival (OS), disease-free survival (DFS; events: death, no response at 6m, myeloid malignancy, relapse, transplant), event-free survival (EFS; events: as for DFS, plus new AA treatment), relapse (competing events: death, transplant, myeloid malignancy), and evolution to either myeloid malignancy or clinical paroxysmal nocturnal hemoglobinuria (PNH) (competing events for both: death and transplant). We report multi-variable analysis (MVA) hazard ratios (HR) for arm B compared to arm A, adjusted for age and disease severity.

Results

All the 197 patients were tracked and analyzed; 5 patients discontinued the study prematurely (withdrawn consent to data collection). A total of 22 deaths were observed, 14 in arm A and 8 in arm B; all deaths were due to SAA and its expected complications (mostly infectious or bleeding). For the remaining 170 patients reaching the 2-year end of study visit, the median follow up was 24 months (range 21.4-28.1). The 2-year OS was 91.4% (95% CI, 85.8-97.1%) in arm B vs 86.0% (95% CI, 79.2-92.8%) in arm A. In MVA, the HR for arm B was 0.54 (95% CI, 0.28-1.04, p=0.064); only disease severity was found to affect OS. The 2-year DFS was 54.7% (95% CI, 44.7-64.7%) in arm B vs 36.6% (95% CI, 27.2-46.0%) in arm A. In MVA, the effect of treatment arm varied in time: HR=0.50 (95%CI, 0.38-0.66, p<0.001) on average across the 2-year period, being significant only in the periods 0-6mo (HR=0.41) and 6-12mo (HR=0.61); both age and disease severity affected DFS. The 2-year EFS was 48.4% (95% CI, 38.4-58.5%) in arm B vs 32.7% (95% CI, 23.5-41.8%) in arm A (p<0.001). As for DFS in MVA the effect of treatment arm (on average across the 2-year period, HR=0.54 [95% CI, 0.43-0.69], p<0.001) was not significant after 12 months; only age reached significance for EFS. The cumulative incidence of relapse was comparable in the two arms, with a 2-year estimate of 22.7% (95% CI, 12.9-32.6%) in arm B vs 18.0% (95% CI, 6.7-29.3%) in arm A (adjusted HR=1.05 [95% CI, 0.87-1.27], p=0.60). The risk of clonal evolution remained negligible, with 1 patient in arm A and 2 in arm B developed karyotypic abnormalities. The 2-year cumulative incidence of clinical PNH was 1.1% (95% CI, 0-3.2%) in arm B vs 8.1% (95% CI, 2.7-13.5%) in arm A (adjusted HR=0.12 [95% CI, 0.04-0.33], p<0.001).

Conclusion

We demonstrate that improved rate and quality of hematological response resulting from the addition of EPAG to standard IST leads to improved 2-year OS, DFS and EFS, in absence of increased risk of secondary myeloid malignancies. While even longer follow up is appropriate to confirm the role of triple therapy in the scenario of disease-eradicating treatments for SAA, these data prove that EPAG on top of standard IST should be the preferred initial therapy for all SAA adult patients not eligible for first-line hematopoietic stem cell transplantation.