Efficacy Findings in a Phase 3, Randomized Trial of Eltrombopag Vs. Standard First-Line Treatment for Newly Diagnosed Immune Thrombocytopenia in Children

Background: First-line treatments for patients (pts) with newly diagnosed immune thrombocytopenia (ITP) include corticosteroids, intravenous immunoglobulin (IVIg), and anti-D globulin. Eltrombopag (epag), a thrombopoietin receptor agonist, was FDA approved for children with chronic ITP in 2015. The efficacy of epag in the newly diagnosed phase of pediatric ITP is unknown.

Methods: The Pediatric ITP Newly diagnosed pts Epag vs Standard therapy (PINES) trial, NCT03939637, is an investigator-initiated, prospective, open label, randomized, multi-center trial sponsored by the ITP Consortium of North America (ICON) and funded by Novartis. Pts ages 1-<18 with primary ITP, ≤3 months from diagnosis, with platelet count <30 x10⁹/L who required pharmacologic treatment per the treating clinician were randomized 2:1, stratified by age and prior treatment status, to receive the experimental treatment, epag, or investigator's choice of one of 3 standard first-line therapies (SOC): prednisone, IVIg, or anti-D globulin at specified doses. The primary endpoint of ‘response’ was ≥3 of 4 platelet counts >50 x10⁹/L during weeks 6-12 without rescue treatment. This intent-to-treat analysis includes data from all randomized pts for the first 12 weeks of the study. Secondary endpoint analyses include data from all evaluable pts who received at least 1 dose of assigned treatment. The WHO Bleeding Scale and Modified Buchanan Scale (MBS) were used to assess bleeding severity. The parent-proxy reported KIDS ITP Tool (KIT) scores were used to measure health-related quality-of-life (QoL). A one-sided z-test, at alpha=0.025, tested the superiority of epag vs SOC for the primary endpoint. Two planned interim analyses utilized O’Brien-Fleming efficacy and futility monitoring boundaries. The Cochran-Mantel-Haenszel test compared the proportion of pts with high bleeding scores (WHO ≥2 or MBS ≥3) between the two arms. A t-test compared the number of rescue therapies and the absolute change in KIT score between the two arms.

Results: The epag arm had a statistically significant higher proportion of responders as compared to the SOC arm (p=0.0023; z-score=3.04) in the second planned interim analysis, crossing the monitoring boundary for efficacy for the primary endpoint. The primary endpoint of platelet response was achieved by 51/78 (65%) pts in the epag arm, compared with 13/40 (33%) pts in the SOC arm (p=0.0007). Trial accrual was closed early for efficacy per DSMB recommendation.

Between May 2019 – January 2024, 122 pts were randomized across 23 institutions; 4 pts were subsequently deemed ineligible, and 118 were included in the intent-to-treat analysis (n=78 epag, n=40 SOC). Of the 118, 46 were aged 1-<6 years, 42 aged 6-<12 years, and 30 aged 12-<18 years. Forty-six pts received upfront treatment on the study, and 72 pts had treatment failure prior to enrollment. Median platelet count at enrollment was 4 x10⁹/L (range 1 – 23) in the epag arm and 8 x10⁹/L (1 – 28) in the SOC arm. Median WHO bleeding score at enrollment was 2 (range 0 – 3) and MBS was 3 (range 0 – 3) in the epag arm and 1 (0 – 3) and 2 (0 – 3), respectively, in the SOC arm. In the SOC arm, treatments prescribed were prednisone (n=29) and IVIg (n=11).

The proportion of pts with a high bleeding score in the epag vs SOC arm was similar at weeks 1-4 and week 12. The proportion of pts who received rescue therapy was lower in the epag arm [15/78 (19%)] than the SOC arm [18/39 (46%)] (p=0.002). The mean absolute change from baseline in parent proxy-reported KIT overall scores in the epag arm vs SOC arm was 8.7 vs 10.1 at 1 week (p=0.45), 13.4 vs 10.7 at 4 weeks (p=0.24), and 15.6 vs 11.2 at 12 weeks (p=0.14), consistent with a clinically meaningful improvement in QoL at all time points in both arms.

There were 20 AEs grade 3 or higher (including 6 SAEs) in the epag arm, and 6 (3 SAEs) in the SOC arm during the first 12 weeks. The most common AEs were headache (3 epag, 3 SOC) and epistaxis (1 epag, 2 SOC). Drug-related SAEs occurred in 6 pts (epag – 2 elevated LFTs, 2 headache (HA); SOC – 1 allergic reaction, 1 HA). There were no thromboembolic events. There was 1 intracranial hemorrhage in the epag arm.

Enrolled pts are completing 12 months of follow up on the study per protocol.

Conclusion: In pediatric pts with newly diagnosed ITP requiring pharmacologic treatment, epag leads to a significantly higher rate of a durable platelet response in the absence of rescue treatments as compared with standard first-line therapies.