For refractory or chronic children with ITP whose risk of severe bleeding would grow even with the second-line treatments, a strong desire has persisted for therapeutic agents with distinct action mechanisms. The discovery and development of a human recombinant TPO (rh-TPO, the first generation of TPO) proved that the activation of TPO receptor can increase thrombopoiesis of megakaryocytes, facilitating TPO-based therapies as treatment for ITP [22–24].

However, the initial clinical trials demonstrated that healthy volunteers who received rh-TPO became severely thrombocytopenic because of cross-reactivity between autoantibodies to rh-TPO and endogenous TPO. This result led to the development of new agents that stimulate TPO receptor but led to little immunogenic adverse effect. The newly developed TPO-R agonists (the second generation of TPO) belong to either TPO non-peptide mimetics (eltrombopag) or TPO peptide mimetics (romiplostim) that increase platelet production by promoting the maturation of BM megakaryocytes through activation of TPO receptor signaling.

The properties of eltrombopag (daily p.o. medicine) and romiplostim (weekly SC injection) are presented in Table 2. The approved administration of eltrombopag is 50 mg/day initial dose and 75 mg/day maximum in Europe and the United States, but for the East Asian patients, the applied dosage was reduced to 12.5 mg/day initial and 50 mg/day maximum because of ethnic difference of drug responsiveness [25].

In the prospective and randomized studies for adult with chronic ITP, 109 of 135 patients (79%) showed significant increases of platelet counts, decreases of concurrent drugs, and lower demand of rescue therapies [26]. Therapeutic effects were not significantly influenced by prior treatments, previous splenectomy, or pretreatment platelet counts (Fig. 2).

For romiplostim, the initial dose is 5 μg/kg/week SC and adjusted up to 10 μg/kg/week at maximum with no marked racial difference of drug responsiveness. Similarly to eltrombopag, the efficacy of romiplostim was approximately 80% for adult with chronic ITP. In a randomized and double-blinded clinical trial for 125 patients [27], the percentage of patients who attained reduction or discontinuation of concurrent drugs was 87% for the romiplostim cohort and 38% for the placebo cohort. Moreover, a randomized clinical trial for patients who were refractory ITP without splenectomy, the subsequent execution rate of splenectomy was 9% for the romiplostim cohort and 36% for the standard therapy cohort, suggesting the possible avoidance of splenectomy by incorporating TPO-R agonists to treatment for refractory ITP [28].

The most common adverse effects of TPO-R agonists were headache, although nasopharyngitis and liver dysfunction were also observed for eltrombopag [26, 27, 29–31]. Thrombosis was reported to have a low rate of incidence, but its causal relation remains unclear [28, 31]. As long-term adverse effects, major concerns are the development of myelofibrosis, depletion of hematopoietic stem cells, and induction of other bone marrow abnormalities including malignant diseases. Further investigation of those areas of concern is expected to lead to better understanding.

Recently, many reports have described the efficacy and safety of TPO-R agonists for children with chronic ITP who had been treated unsuccessfully with the first-line treatments (Table 3) [32–38]. Patients included children (1–17 years old) with chronic/refractory ITP who received romiplostim or eltrombopag at doses adjusted to less than the maximum dose to maintain platelet counts at least >50 × 10⁹/L. In 2011, the first randomized clinical trials with romiplostim were reported: short-term observation (12 weeks) for 18 and 22 patients for whom the respective pretreatment duration was 2.4 and 2.5 years [32, 33]. These studies showed that the efficacy to maintain >50 × 10⁹/L of platelet counts without rescue medication was 83.5–88% for patients with romiplostim, as compared to 0% for those with placebo. Nevertheless, the small number of patients and the short-term duration of treatment of these studies might limit the generalization of their conclusions. More recently, a larger study of 62 children (romiplostim, 42; placebo, 20) during 24 weeks of treatment duration revealed that the efficacy of romiplostim was 52% for the romiplostim group as compared to 10% for the placebo group [38]. Another randomized double-blinded and subsequent open-labeled study with eltrombopag has demonstrated its efficacy as 40% for the eltrombopag group, compared to 3% for those of placebo [37]. The efficacy of eltrombopag increased further to 80% (70/87 patients) of efficacy during the following 24 weeks of the open-labeled period. These two recent studies confirmed that both romiplostim and eltrombopag TPO-R agonists are effective for chronic/refractory ITP in children. For these periods, no drug resistance or autoantibody against intrinsic TPO was detected.