Clinical Characteristics and Response Rates to Eltrombopag for Primary and Secondary Thrombocytopenia after Allogeneic Hematopoietic Stem Cell Transplant (HCT)

IntroductionThrombocytopenia after HCT is multifactorial in etiology and frequently requires prolonged transfusions to minimize bleeding risks. Eltrombopag (TPO) is a thrombopoietin agonist that is FDA-approved for use in the setting of ITP. Given encouraging studies in aplastic anemia, TPO is increasingly used after HCT in the setting of poor engraftment and thrombocytopenia; however, published data is limited to small retrospective cases/series. We present our single institution experience.MethodsWe retrospectively reviewed HCT patients and identified the clinical characteristics and outcomes of those who received TPO post HCT from 1/1/13 to 10/1/18. We excluded patients with documented relapses at the time of TPO initiation.Results17 patients were identified; 9 had primary failure of platelet recovery (<20K without transfusion) and 9 had secondary failure (decline to <20K or recurrence of transfusions). Median ANC was 1.31 (2 patients in the primary failure group had ANC of 0). Only 2/17 had matched sibling transplants; the remainder underwent MUD (7/17) and mismatch (3/17) using anti-thymocyte or haplo (5/17) with post-transplant cyclophosphamide. 12/17 received IVIG prior to TPO, and 11/17 were on systemic immunosuppression. 7/17 (41%) patients had platelet response on TPO (>30K without transfusion preceding 7 days). Of 7 responders, 5 were started as outpatients. Median time to response was 39 days (range 10 – 118). In 5 patients with unidentified relapse at TPO start, TPO preceded diagnosis of relapse by median 62 days (range 22-281). Of the 9 patients who started TPO as inpatient, 7 subsequently died with median time from TPO to death 195 days (range 18-398). No significant toxicities to TPO were identified. The median time from TPO outpatient prescription to start was 22 days due to insurance delays (range 8-40).ConclusionWhile causality is difficult to establish, TPO was associated with improvement of platelet counts in a subset of patients, with responses most likely to occur in outpatients with viral reactivation, GvHD, or primary poor engraftment. In 5 patients with unidentified relapse, start of TPO preceded a diagnosis of relapse by median 62 days. This should alert clinicians to assess for relapse when considering TPO. The start of TPO in inpatients generally reflects grave clinical outlook given low response rates and high number of subsequent deaths.