Immune thrombocytopenia (ITP) is an acquired bleeding disorder characterized by a reduced platelet count of less than 100 × 10⁹/L, with an estimated incidence of 2–5/100 000 children per year [1]. Since 2014, the China National Medical Products Administration (NMPA) has approved five TPO-RAs for treating ITP in China. Among these, eltrombopag (ELT) and avatrombopag (AVA) are the two most evidence-based and available oral medications for pediatric ITP; both demonstrated promising efficacy and tolerability in clinical trials [2, 3]. Because no head-to-head studies have been conducted, making direct comparisons between the two drugs is difficult. To address this gap, we conducted a multicenter, retrospective, observational cohort study to evaluate the efficacy and safety of ELT versus AVA in pediatric patients with ITP in China, to offer personalized treatment guidance based on factors such as the severity of bleeding, medication adherence, and family economic conditions.

Patients younger than 18 with a confirmed diagnosis of ITP who received treatment with ELT or AVA in 3 Children's Specialty Hospitals from April 2017 to December 2023, with medication duration for at least 2 months and with a total observation period of more than 6 months. For this retrospective study, ethics approval was obtained from the institutional review boards of Beijing Children's Hospital, Capital Medical University (Ethics approval NO. [2024]-Y-098-D).

The initial dosage of ELT was determined by age and body weight. For children aged 1–5 years, the starting dose was 1.5 mg/kg/day. In those aged 6–17 years, the dose was 37.5 mg/day for individuals weighing less than 27 kg and 50 mg/day for those weighing 27 kg or more. Similarly, the starting dose of AVA was 10 mg/day for children aged 1–6 years and 20 mg/day for those aged 6–18 years.

The current study defines several key definitions and outcomes [4]:(1) Complete Response (CR): a platelet count ≥ 100 × 10⁹/L and no bleeding within 2 months. (2) Response (R): a platelet count of 30–100 × 10⁹/L, with at least a twofold increase from baseline and no bleeding within 2 months. (3) No Response (NR): a platelet count < 30 × 10⁹/L, less than a twofold increase from baseline, or bleeding after dose titration of ELT or AVA within 2 months. (4) Overall Response (OR): the combined number of patients achieving CR and R. (5) Durable Response (DR): a platelet count > 30 × 10⁹/L with at least a twofold increase from baseline at 6 months. (6) Remission: a platelet count ≥ 30 × 10⁹/L at 12 months. (7) SRoT: a platelet count ≥ 30 × 10⁹/L for at least 6 months off ELT or AVA treatment. (8) Time to typer: defined as the period from the initiation of ELT/AVA therapy to the commencement of dosage tapering within the first 12 months. (9) Rescue therapy, including IVIG, platelet transfusion, and high-dose corticosteroid pulse therapy was initiated when the platelet count dropped below 20 × 10⁹/L or the bleeding score reached Grades 3–4. (10) Concomitant treatment including administering rituximab (RTX), prednisone, and immunosuppressants (Sirolimus, Rapamycin, Mycophenolate mofetil) when the start of the TPO-RAs, furthermore, RTX is excluded from combination therapy after 12 weeks of use, and IVIG after 3 weeks of use. The comparison of typer rates between the two groups was conducted by the Log-rank test, with statistical significance defined as p value < 0.05.

Totally 253 patients treated with TPO-RAs were initially evaluated. After excluding cases of secondary ITP and those with incomplete clinical data, 233 patients were ultimately enrolled, comprising 199 in the ELT group (EG) and 34 in the AVA group (AG). Among the 233 patients, 34 were newly diagnosed ITP (NITP, ITP duration of < 3 months), 78 were persistent ITP (PITP, ITP duration of 3–12 months), and 119 were chronic ITP (CITP, ITP duration of > 12 months) (Figure S1). In the EG, 52.7% (105/199) were male, compared with 47% (16/34) in the AG. The median age was 5.0 years (range 0.2–16 years) in the EG and 5.2 years (range 0.2–13.2 years) in the AG. The median duration of TPO-RAs treatment was 24 months (range 2–62 months) for the EG and 9 months (range 2–22 months) for the AG. The median platelet count at the initiation of the EG was 17 × 10⁹/L (IQR:9 × 10⁹/L, 23 × 10⁹/L) and 14 × 10⁹/L (IQR: 6 × 10⁹/L, 25 × 10⁹/L) in the AG, without significant differences. The median bleeding score prior to initiating TPO-RAs was 1 (range 0–4) in both groups. The types of previous treatments before TPO-RAs were similar between the two groups, with a median of 3 (range 1–6) in EG and 3 (range 1–7) in AG. Prior to receiving ELT treatment, 91 patients had been treated with RTX and 163 with high-dose dexamethasone (HDD). In the AVA group, 19 patients had received RTX and 28 had received HDD. None of the patients in either group had undergone splenectomy before initiating TPO-RAs therapy. There were no statistically significant differences between the two groups, except for the TPO-RAs treatment duration (Table S1).

Throughout the trial, OR was attained by 137 patients (69%) in the EG and 25 patients (74%) in the AG. Among them, CR was achieved by 90 patients (45%) in the EG and 20 patients (61%) in the AG. Furthermore, NR was observed in 62 (31%) patients in the EG and 9 (26%) patients in the AG, without significant differences between the two groups (p = 0.304). The median time to response (TTR) of EG was 15 days (range: 2–64 days) and 8 days (range: 1–32 days) of AG (p < 0.001). As the treatment progressed, 124 patients (66%) in the EG and 19 patients (68%) in the AG achieved a DR at the 6th month, without significant differences (p = 0.881). At the 12th month, remission was attained by 43 out of 174 patients (25%) in the EG and 4 out of 13 patients (31%) in the AG, without significant differences (p = 0.316). The median time between initiating TPO-RAs and tapering was 5.8 months (range: 0.5–29 months) in the EG and 5.1 months (range: 0.5–9 months) in the AG, without significant differences (p = 0.782). Throughout the observation period, 41 patients (21%) in the EG showed SRoT after discontinuing ELT without additional ITP therapy. In the EG, 38 patients switched to AVA, and five switched to hetrombopag due to NR, while no cases of switching were reported in the AG (Figure 1A). Following 2-month treatment, the median platelet counts increased to 58 × 10⁹/L (IQR: 23 × 10⁹/L, 119 × 10⁹/L) in the EG and 103 × 10⁹/L (IQR: 18 × 10⁹/L, 234 × 10⁹/L) in the AG, revealing a significant difference between the two groups (p < 0.001) (Figure 1B). There were 64 patients (32.2%) in the EG, and 11 patients (32.4%) in the AG tapered within 6 months; 108 patients (54.3%) in the EG and 14 patients (41.2%) in the AG were tapered within 12 months, without significant differences (p = 0.133) (Figure 1C).

Initially, there was no significant difference in the percentage of patients reporting Grades 2–4 bleeding symptoms: 27.6% in the EG and 32.3% in the AG. The incidence of Grades 2–4 bleeding events decreased to 6.0% in the EG and 5.8% in the AG after 2 months of treatment, and this proportion further declined to nearly 0% at the 6th month in both groups. By the 12th month, no bleeding symptoms were reported in either group (Figure 1D). At baseline, 56.8% of patients in the EG and 67.7% in the AG were concurrently administered ITP drugs, including RTX, corticosteroids, sirolimus, rapamycin, and metformin. However, the use of concomitant medication gradually decreased to 36% and 18% at 2-month, respectively. This reduction was more pronounced in the AG. Subsequently, the concomitant medication decreased to nearly 0 at the 6th and 12th months, with no significant difference between the two groups (Figure 1E). According to a study including two groups, rescue therapy, was initially required by 31.7% of patients in the EG and 41.2% of patients in the AG. As the treatment progressed, the platelet counts increased, bleeding improved, and the need for rescue therapy decreased in both groups. After 2 months of treatment, the percentage decreased to 4% and 6% in the EG and AG, respectively. Following 6 months of treatment, none of the patients required rescue therapy (Figure 1F).

Previous studies have shown that the two groups of patients exhibit consistent baseline characteristics and similar efficacy, including improvements in dose reduction, bleeding management, concomitant medication, and rescue treatment. However, comparing the safety profile of these medications is as essential as assessing their efficacy. The overall tolerance was favorable, with all adverse events classified as Grade 1 or 2 according to CTCAE 5.0, and no Grade 3 or higher adverse reactions were observed. In the EG, 6.5% (13/199) of patients reported elevated liver enzymes, and 3% (6/199) reported iron deficiency anemia. In the AG, 2.9% (1/34) of patients experienced headaches. Both groups demonstrated an increase in platelet counts, with incidence rates of 17% (34/199) in the EG and 47% (16/34) in the AG (Table S2).

ELT and AVA are currently the two most commonly used oral TPO-RAs drugs for treating pediatric ITP. Comparing the efficacy and safety of these two drugs is crucial for informed decision-making in the management of pediatric ITP. Our real-world study shows that both drugs had similar efficacy in CR, OR, NR, taper rate, reduction of bleeding symptoms, concomitant medication, and rescue treatment, which are comparable and similar to the efficacy and tolerability described in the literature [5, 6]. Notably, this study included 36 NITP patients who received ELT or AVA treatment and were either refractory to or dependent on first-line therapies, which suggests that thrombopoietin receptor agonists may be a safe and effective option for patients with NITP who do not respond to first-line treatments or who are dependent on corticosteroids. Another noteworthy point is that the response time for ELT was longer than AVA, indicating that AVA acts significantly faster. However, AVA also carries a risk of excessive platelet elevation, requiring careful dose adjustment in patients. Further extensive randomized controlled trials are needed to optimize the use of these therapies.