Updated Results from the Single-Arm, Open-Label, Long-Term Efficacy and Safety Study of Subcutaneous (SC) Romiplostim in Children with Immune Thrombocytopenia (ITP)

Background: Romiplostim, a thrombopoietin (TPO) receptor agonist approved for children and adults with chronic ITP, was evaluated in children with ITP in a ≤3-year open-label trial. Interim results were previously reported (Grainger et al., Blood 2017 130:2334). Here we present updated results as of 27 Mar 2019.

Methods: Eligible children from 17 countries with ITP for ≥6 months and screening platelet count ≤30×10⁹/L (or uncontrolled bleeding) received SC romiplostim (1 μg/kg titrated to 10 μg/kg to maintain platelet counts of 50-200×10⁹/L). In Europe, bone marrow was evaluated at baseline and after 1 (day 365 ± 4 weeks) or 2 (day 730 ± 4 weeks) years. The primary endpoint was % time with a platelet response (platelet count ≥50×10⁹/L, no rescue therapy in preceding 4 weeks) in months 0-6.

Results: A total of 203 patients (pts) received ≥1 dose; the median (interquartile range [IQR]) age was 10 (6-13) and median (IQR) platelet count 14 (7-23.5×10⁹/L). The median (IQR) duration of treatment was 145 (39-156) weeks, median (IQR) % of time with a platelet response in months 0-6 was 50% (17-83%), with 88% (179/203) of pts having a platelet response at least once (Fig. 1A). In all pts, the median (IQR) % of time with an increase in platelet counts ≥20×10⁹/L above baseline from week 2 until the end of treatment was 79% (39-92%). Median and lower quartile platelet counts were both consistently >50×10⁹/L from week 12 and 48, respectively, and did not vary by age. Eleven pts maintained platelet counts ≥50×10⁹/L without ITP medications (including romiplostim) for ≥24 weeks; median (IQR) time to onset was 50 (24-80) weeks after starting romiplostim. During the study, 60 (30%) pts received rescue therapy, typically within weeks 1-36, and 3 underwent splenectomy.

With a total exposure of 428.7 patient-years, median (IQR) average weekly romiplostim dose over the entire study was 6.9 (4.6-8.9) µg/kg; 8.5 (5.0-10.0) μg/kg at 1 year (n=144) and 6.0 (3.0-10.0) μg/kg at 2 years (n=129; Fig. 1B). Self-administration was initiated in 68% of pts. Ninety-five pts (46.8%) discontinued treatment: reasons included lack of efficacy (n=43 [21.2%]), patient request (n=15 [7.4%]), adverse event (AE; n=9 [4.4%]), and neutralizing antibodies (NAb; n=7 [3.4%]). AEs occurred in 192 pts (94.6%); the most frequent were epistaxis (38.4%), headache (37.9%), and nasopharyngitis (36.9%). Serious AEs (SAEs) occurred in 59 (29.1%) pts, including epistaxis (5.9%), decreased platelet count (4.4%), and thrombocytopenia and NAb (2% each); 8 pts had treatment-related SAEs (NAbs [n=4], headache and abdominal pain [each n=2], and presyncope [n=1]). Bleeding occurred in 69% of pts over the study, decreasing over time, with bleeding in 18% of pts from week 144 to the end of treatment. Bleeding-related AEs occurring in >10% of pts were epistaxis (38.4%), petechiae (23.6%), hematoma (20.7%), contusion (19.2%) and gingival bleeding (10.3%). CTCAE grade ≥3 bleeding events occurred in 20 pts (9.9%) and included epistaxis (n=9 [4.4%]), and persistent ITP (n=3 [1.5%]). In pts with no evidence of NAbs at baseline, there were 7 cases of NAbs to romiplostim (2 were transient) and 1 transient NAbs to TPO; only 1 pt, with NAbs to romiplostim, had a reduced therapeutic effect.

Of 75 European pts with evaluable baseline bone marrow biopsies [modified Bauermeister scores: grade 0 (no reticulin, n=16), 1 (fine fibers, n=54), or 2 (fine fiber network, n=5)], 27 had evaluable on-study biopsies after 1 year and 36 after 2 years (Table). Of these, 5 pts developed increased reticulin at year 1 and 17 at year 2. One pt had an increase in modified Bauermeister score from baseline of ≥2 grades (increase from grade 0 to 2), 4 pts had an increase in 1 grade, 1 a decrease in 2 grades, and 3 a decrease in 1 grade in year 1. In year 2, 15 pts had an increase in 1 grade and 3 pts a decrease in 1 grade. No pts developed collagen and no bone marrow abnormalities were detected.

Conclusion: Over the course of the study with >30 months of treatment, on a median dose of 6.9 μg/kg, 88% of children had a platelet response, median platelet counts were ≥20×10⁹/L above baseline 79% of the time and >50×10⁹/L from week 12. An important new safety signal over 429 patient-years was the 3.4% of NAbs to romiplostim; children develop NAbs with romiplostim more frequently than adults. Bone marrow findings showed that children, like adults, did not develop clinically important fibrosis.

View largeDownload slide

View largeDownload slide

Close modal