Poor graft function (PGF) is a life-threatening complication that occurs in 5–27% of the patients following allogeneic hematopoietic stem cell transplantation (allo-HSCT) [1, 2]. Management strategies, including the use of growth factors [3], CD34+-selected stem cell boost [4], mesenchymal stem cell (MSC) transfusion [5], and second allo-HSCT [6], are not effective for all patients.

Eltrombopag, a c-mpl receptor agonist, is an effective treatment for immune thrombocytopenic purpura (ITP) and thrombocytopenia after transplantation [7, 8]. In a recent phase I/II study of 92 patients with severe aplastic anemia (SAA) [9], eltrombopag plus standard immunosuppression resulted in 94% hematological response rate. Considering the similarity between PGF and SAA, we speculated that eltrombopag is also effective against PGF. This retrospective analysis included 12 consecutive patients receiving eltrombopag for secondary PGF after allo-HSCT during a period from February 2016 to October 2017. Secondary PGF (sPGF) was defined as: cytopenia in at least two lineages (platelet < 20 × 109/L, neutrophil < 0.5 × 109/L, hemoglobin < 70 g/L), and/or with transfusion requirements beyond day + 28, with full donor chimerism, without relapse or severe graft versus host disease, and lasting at least for 14 consecutive days [5, 10].

Clinical characteristics of the subjects were summarized in Table 1. All 12 patients responded poorly to previous treatments, including growth factors (n = 12), MSCs (n = 2), and decitabine (n = 2). All but one patient were transfusion-dependent. The median duration was 116 (35–1000) days from transplantation to sPGF diagnosis and 59 (30–180) days from sPGF diagnosis to eltrombopag treatment. Eltrombopag was started at a dose of 25 mg/d for 3 days and then increased to 50 or 75 mg/d. Median duration of eltrombopag treatment was 8 (2–23) weeks. Total dosage was 2487.5 (700–10,500) mg.

Table 1 Clinical Characteristics of the 12 sPGF patients
Full size table

The overall response rate (ORR) was 83.3% (10/12). Eight patients achieved complete response (CR), as defined by platelet ≥ 50 × 109/L, neutrophil ≥ 1.0 × 109/L, and hemoglobin ≥ 90 g/L, without blood cell transfusion or granulocyte colony stimulating factor for ≥ 7 consecutive days [5]; the time from eltrombopag initiation to achieving CR was 29 (10–49) days. Two patients achieved partial response, as defined by hematopoietic engraftment of at least two lineages (platelet ≥20 × 109/L, neutrophil ≥0.5 × 109/L and hemoglobin ≥70 g/L) but not fulfilling CR criteria.

The follow-up was 18.5 (3–37) months post transplantation. Among the 10 responding patients, median platelet count was 18 (5–27) × 109/L vs 74 (30–117) × 109/L prior to and after treatment (P = 0.00008; Fig. 1a). Median neutrophil count was 0.51 (0.28–0.69) × 109/L vs 1.84 (0.78–4.90) × 109/L (P = 0.0015; Fig. 1b). Median hemoglobin was 88 (63–123) vs 101 (78–134) g/L (P = 0.0001; Fig. 1c). The response lasted to the last follow-up (10–18 months) in all 8 subjects who achieved CR.

Fig. 1
figure 1

Blood cell counts prior to and after eltrombopag treatment. The analysis included only 10 responding patients. a Median platelet count was 18 (5–27) × 109/L vs 74 (30–117) × 109/L before and after the treatment (P = 0.00008). b Median neutrophil count was 0.51 (0.28–0.69) × 109/L vs 1.84 (0.78–4.90) × 109/L (P = 0.0015). c Median hemoglobin level was 88 (63–123) vs 101 (78–134) g/L (P = 0.0001)

Full size image

Eltrombopag was well tolerated by all 12 patients. There were no treatment-related mortality and no evidence of cataract, thrombosis, or any other grade 3/4 toxicities. Upon the last follow-up, 9 subjects were PGF-free; 9 had normal blood cell counts. The 12-month overall survival rate after transplantation was 83.3% (95% CI: 62–100%).

With increasing application of alternative donor in transplantation, especially haploidentical HSCT and cord blood transplantation, PGF has become a major obstacle contributing to higher non-relapse mortality. Eltrombopag, as a stimulator of platelet production, promotes the proliferation of megakaryocytes by binding with thrombopoietin receptor (c-mpl) [7], also can promote hematopoiesis along all three lineages. Indeed, clinical trials have establishedefficacy of eltrombopag against ITP, thrombocytopenia after transplantation, as well as SAA [7,8,9]. Considering the fact that all patients in the current study failed previous treatments for sPGF, the ORR (83.3%) and CR (66.7%) are encouraging. Another important finding is the relatively long duration of the response after eltrombopag withdrawal. The current study represents the first case series of using eltrombopag for secondary PGF after allo-HSCT. Due to the retrospective nature of the study and the small sample size, the results must be considered preliminary and should be verified by randomized controlled trials in the future.

In summary, we showed that eltrombopag could produce a rapid and sustaining response in patients with sPGF after allo-HSCT who failed treatment with conventional treatments. This finding is particularly interesting considering the increasing use of alternative donor HSCT and high rate of non-relapse mortality due to PGF.