A Phase 3, Randomized, Double-Blind, Placebo-Controlled Study to Determine the Safety and Efficacy of Romiplostim in Children with Immune Thrombocytopenia (ITP)

Introduction:  Approximately 20% of children diagnosed with ITP have chronic thrombocytopenia, some requiring immunosuppressive treatments.  The TPO receptor agonist romiplostim could be a treatment option for symptomatic children with ITP. 

Methods:  In this phase 3, double-blind study, children with ITP for ≥6 months were randomized to weekly romiplostim or placebo (2:1) for 24 weeks, with the dose adjusted weekly from 1‑10 μg/kg to target platelet counts of 50‑200×10⁹/L.  Platelet response was defined as platelet counts ≥50×10⁹/L any week of weeks 2-25 (excluding 4 weeks after rescue medication use).  The 1° endpoint, durable platelet response, was defined as ≥6 weekly platelet responses during the final 8 weeks, weeks 18-25.  Overall platelet response was defined as ≥4 weekly platelet responses during weeks 2-25.  Antibodies to romiplostim or TPO were assayed at baseline and weeks 12 and 25.  The Kid's ITP Tool (KIT), in which higher scores reflect improved quality of life, was administered to children and their parents. 

Results:  Patients (n = 62) had median (range) age 9.5 (3-17) years, ITP duration 2.1 (0.5‑10.7) years, and baseline platelet counts 18 (1-94) × 10⁹/L; 44% were male.  Sixty-six percent of patients were Caucasian, 13% African American, 8% Asian, and 15% other.  For the 1° endpoint, measured in the final 8 weeks after 16 weeks of dose titration, rates of durable platelet response were romiplostim: 52% and placebo: 10% (p < 0.002) (Table).  Rates of overall platelet response were romiplostim: 71% and placebo: 20% (p = 0.0002), and rates of any platelet response were romiplostim: 81% and placebo: 55% (ad hoc p = 0.03).  For romiplostim, the median (range) time to first platelet response was 4.5 (1-20) weeks and the median average weekly dose was 3.9 (0.9-8.1) μg/kg.  Median platelet counts with romiplostim were ≥50×10⁹/L from week 8 on (Figure).  In the final 8 weeks, non-cutaneous bleeding rates were romiplostim: 33% and placebo: 47%, and rescue medication use rates were romiplostim: 24% and placebo: 30%.  Over the course of the study, the rates of grade ≥3 bleeding were romiplostim: 5% and placebo: 11%; rates of any bleeding and non-cutaneous bleeding were similar for romiplostim and placebo.  We also examined a composite bleeding episode endpoint, defined as clinically significant bleeding, ie, CTCAE grade ≥2, OR the use of rescue medication to prevent such bleeding.  The rate for these episodes for weeks 2‑25 was less with romiplostim than placebo (8.1 vs 18.4 per 100 patient-weeks, p<0.0001).  No patients withdrew due to adverse events (AE).  Rates of serious AEs (SAEs) were romiplostim: 23.8% (10 patients) and placebo: 5.3% (1 patient).  SAEs with romiplostim consisted of 2 cases each of contusion, epistaxis, and headache, and 1 case each of petechiae, thrombocytosis, nephrotic syndrome, nausea, vomiting, bronchiolitis, fever, urinary tract infection, and seizure.  These SAEs often occurred with related comorbidities.  Only the headache and thrombocytosis SAEs, which occurred in the same patient, were considered treatment‑related by the investigator.   There were no thrombotic events, no malignancies observed, and no findings indicating bone marrow dysplasia or fibrosis in the only bone marrow biopsy performed.  None of the 42 romiplostim-treated patients developed neutralizing antibodies to either romiplostim or TPO.  The parent impact score of the KIT showed significant quality-of-life improvement with romiplostim (p = 0.04) in the mixed-effects repeated measures analysis using a general linear model.

Conclusions:  In children with symptomatic ITP of ≥6 months duration, romiplostim was effective in inducing high rates of durable and overall platelet responses.  There were no new safety signals.  By the final 8 weeks of the study, noncutaneous bleeding was decreased with romiplostim.  Romiplostim may be a viable treatment option for children with symptomatic chronic ITP.

 

Table.       

	Placebo (N = 20)	Romiplostim (N = 42)
Durable platelet response, x/n (%), all patients	2/20 (10%)	22/42 (52%)	P < 0.002
By baseline age
·         1 to <6 years	1/4 (25%)	3/8 (38%)
·         6 to <12 years	1/9 (11%)	10/18 (56%)
·         12 to <18 years	0/7 (0%)	9/16 (56%)
Overall platelet response, x/n (%), all patients	4/20 (20%)	30/42 (71%)	P = 0.0002
By baseline age
·         1 to <6 years	2/4 (50%)	5/8 (63%)
·         6 to <12 years	1/9 (11%)	15/18 (83%)
·         12 to <18 years	1/7 (14%)	10/16 (63%)

 

 

Figure