Efficacy and safety of arimoclomol in Niemann-Pick disease type C: Results from a double-blind, randomised, placebo-controlled, multinational phase 2/3 trial of a novel treatment

Abstract

Niemann-Pick disease type C (NPC) is a rare, genetic, progressive neurodegenerative disorder with high unmet medical need. We investigated the safety and efficacy of arimoclomol, which amplifies the heat shock response to target NPC protein misfolding and improve lysosomal function, in patients with NPC. In a 12-month, prospective, randomised, double-blind, placebo-controlled, phase 2/3 trial (ClinicalTrials.gov identifier: NCT02612129), patients (2-18 years) were randomised 2:1 to arimoclomol:placebo, stratified by miglustat use. Routine clinical care was maintained. Arimoclomol was administered orally three times daily. The primary endpoint was change in 5-domain NPC Clinical Severity Scale (NPCCSS) score from baseline to 12 months. Fifty patients enrolled; 42 completed. At month 12, the mean progression from baseline in the 5-domain NPCCSS was 0.76 with arimoclomol vs 2.15 with placebo. A statistically significant treatment difference in favour of arimoclomol of -1.40 (95% confidence interval: -2.76, -0.03; P = .046) was observed, corresponding to a 65% reduction in annual disease progression. In the prespecified subgroup of patients receiving miglustat as routine care, arimoclomol resulted in stabilisation of disease severity over 12 months with a treatment difference of -2.06 in favour of arimoclomol (P = .006). Adverse events occurred in 30/34 patients (88.2%) receiving arimoclomol and 12/16 (75.0%) receiving placebo. Fewer patients had serious adverse events with arimoclomol (5/34, 14.7%) vs placebo (5/16, 31.3%). Treatment-related serious adverse events (n = 2) included urticaria and angioedema. Arimoclomol provided a significant and clinically meaningful treatment effect in NPC and was well tolerated.

Keywords: NPC clinical severity scale; Niemann-Pick disease type C; arimoclomol; biomarker; double-blindplacebo-controlled; heat shock protein.

FIGURE 1

A, CT‐ORZY‐NPC‐002 trial design. B, Patient flow. Completers analysis set excluded two patients in the arimoclomol group who did not have assessments at 12 months. PK, pharmacokinetic

FIGURE 2

5‐domain NPCCSS score observed change from baseline at month 12: A, Overall (N = 50; full analysis set); B, in patients aged ≥4 years (n = 44); C, in patients receiving miglustat (n = 39). The solid line represents least‐squares mean estimates ± SE based on data obtained while patients were exposed to study treatment. The mixed model for repeated measures included the main effect of baseline and stratum, respectively, and interaction between treatment and visit. Change from baseline and absolute estimates correspond to the at‐baseline overall average patient. Numbers of patient are presented for each time point. CI, confidence interval; NPC, Niemann‐Pick disease type C; NPCCSS, Niemann‐Pick disease type C Clinical Severity Scale

FIGURE 3

Biomarker analyses. A, Change in HSP70 in PBMCs from months 0 to 12 in arimoclomol‐treated patients (n = 11; P = .001); B, Change in unesterified cholesterol level at month 12 (between‐group difference: P = .096); C, Change in serum cholestane‐triol level at month 12 (between‐group difference: P = .225); D, ratio of plasma Lyso‐SM‐509 to baseline (between‐group difference: P = .043). Error bars show SE. Wilcoxon signed‐rank test was used to assess significance within treatment group (HSP70). For unesterified cholesterol and serum cholestane‐triols, between‐group analysis of covariance was conducted with baseline, stratum and treatment as covariates. Estimates were adjusted to reflect baseline distribution. HSP70, heat shock protein 70; NPC, Niemann‐Pick disease type C; PBMC, peripheral blood mononuclear cell