Introduction
Niemann-Pick disease type C (NP-C) is a rare lysosomal lipid storage disease characterized by neurological deterioration [
1,
2] with constant progression over time [
3‐
5]. NP-C is caused by autosomal recessive mutations in either one of the two genes,
NPC1 or
NPC2, which encode proteins involved in the regulation of normal intracellular lipid trafficking [
1,
6]. It is estimated to affect 1 case in every 100,000–120,000 live births [
1,
7].
In very rare cases of the severe perinatal (systemic) form of NP-C, patients typically die from liver failure within the first months of life [
8,
9]. However, NP-C most frequently presents during middle-to-late childhood, and an increasing number of cases are being detected among adolescents and adults [
10]. The symptomatology and rate of disease progression of NP-C are strongly influenced by age at onset of neurological manifestations, and different clinical forms have been described on this basis [
3,
11]. The early-infantile form arises at <2 years of age, the late-infantile form at 2 to 5 years, the juvenile form at 6 to <15 years, and the adolescent/adult form at ≥15 years [
2,
6,
11‐
13].
Paediatric forms of NP-C tend to feature initial hepatosplenomegaly; an episode of neonatal cholestatic icterus may have occurred [
1,
8,
14]. Later on, neurological manifestations begin to overshadow systemic symptoms. Early delay in motor milestones is often seen in the early-infantile form. Signs of vertical supranuclear gaze palsy (VSGP) are frequent early neurological manifestations, but frequently go undetected until later. Patients may present with clumsiness and progressive cerebellar ataxia. Over time, progressive dysmetria, dystonia, pyramidal signs, dysphagia, dysarthria, cataplexy and/or epileptic seizures, and cognitive impairment often develop [
1,
2,
13,
15]. Typically, patients with early-onset neurological manifestations experience more rapid decline and a lower life expectancy than those with later-onset manifestations [
1,
2].
Miglustat was approved in Europe for the treatment of progressive neurological manifestations in adult patients and paediatric patients with NP-C in January 2009, and has subsequently been approved in a number of other countries, based on data from preclinical studies [
16] and clinical trials showing that it can stabilize neurological disease and/or slow its progression [
17‐
21]. Data from a retrospective observational study of miglustat efficacy in a large cohort of NP-C patients aged between 0 and 32 years demonstrated greater beneficial effects of miglustat on neurological disease in adolescents and adults than those seen in children with the earliest forms of NP-C [
4].
Data on the therapeutic effects of miglustat in paediatric patients in clinical practice settings are relatively limited [
22‐
27], and evidence from patients with the early-infantile form are particularly scarce. There is therefore an ongoing need for further clinical experience data on the use of miglustat in children, particularly with regard to disease-specific disability assessments. In addition, there are few data on the response of specific neurological manifestations such as epileptic seizures and cataplexy to miglustat therapy.
We report data from a prospective open-label cohort study evaluating disease progression and response to miglustat therapy among all treated paediatric patients with NP-C diagnosed in French hospitals. Findings based on NP-C disability scale assessments, brain imaging and other follow-up assessments conducted according to international disease management recommendations are presented [
2].
Discussion
Assessments of key parameters of neurological disease progression based on the published NP-C disability scale indicated either stabilization or improvement of neurological manifestations in 1/8 early-infantile, 6/8 late-infantile, and 1/3 juvenile-onset NP-C patients who received miglustat in this multicentre, open-label cohort study. Beneficial therapeutic effects were seen more frequently in patients with late-infantile/juvenile neurological disease onset than in those with early-infantile onset.
In agreement with previous reports, visceral disease (prolonged neonatal cholestasis and/or hepatosplenomegaly) was more prevalent among early-infantile onset patients than in later-onset patients in this paediatric cohort [
1,
26]. Neonatal cholestasis healed in all NPC1 patients before miglustat therapy, but portal hypertension persisted in one patient during miglustat treatment. While visceral disease was not a focus of disease monitoring in this study, miglustat did not appear to have any effect on hepatosplenomegaly (data not shown) or pulmonary disease (patient 5 died from alveolo-interstitial complications after 13 months of miglustat treatment).
While the NPC2 patient had cholestasis and hepatosplenomegaly from 1 month of age (before initiation of miglustat), she developed hepatic failure during miglustat therapy. It is not possible to ascertain whether miglustat might have worsened the liver disease in this patient as a similar course of disease is quite common in NPC2 [
29].
In general, our findings appear in line with those from previous clinical trial data [
4,
18,
19] and case reports [
23,
25,
27] on the effects of miglustat on neurological disease manifestations in paediatric NP-C patients. Based on clinical assessments in a 24-month study of miglustat in children aged 4–12 years, Patterson et al. reported stabilization of SEM, an accepted marker of early neurological deterioration in NP-C, in 67% of patients throughout therapy [
18]. Ambulation (measured using the standard ambulation index) was stabilized in 80% of patients, and swallowing (patients’ ability to swallow various substances) remained stable in 90% [
18]. Based on the same NP-C disability scale as that employed in the current study, a retrospective analysis of miglustat efficacy in 66 NP-C patients aged between 0 and 32 years (mean ± SD, 9.7 ± 7.6 years) showed that ambulation, manipulation, language and swallowing were stabilized or improved in 75% of patients during an average of 18 months of therapy [
4]. Stratification of patients according to age indicated that beneficial effects were greater in juvenile, adolescent and adult patients than in those with disease onset before 6 years of age [
4].
Pineda et al. reported clinical experience with the use of miglustat in a paediatric cohort of 16 Spanish NP-C patients, comprising five with the early-infantile form, four with the late-infantile form, and seven with the juvenile form [
26]. As in the current study, efficacy assessments were based on an NP-C specific disability scale, albeit a modified version that included scores for the presence of epilepsy and ocular movements. Similar to our findings, the Spanish cohort study indicated that patients with the late-infantile and juvenile-onset forms were more likely to show improvements or stabilization of neurological disease during miglustat therapy compared with patients with severe, early-infantile onset [
26]. However, Spanish early-infantile onset patients who showed deterioration during miglustat therapy were at an advanced stage of the disease before starting therapy, while those with a better evolution had started therapy at the youngest ages. Overall, the Spanish data suggested that better treatment effects might be expected when treatment was initiated early, before “irreversible neurological damage” [
2,
26].
In our cohort, there did not appear to be as strong a correlation between patients’ disability scale scores before treatment and subsequent changes during therapy. However, a short interval between neurological disease onset and the start of miglustat therapy and/or young age at treatment start was associated with a better initial therapeutic outcome in one early-infantile onset patient: in patient 2, who initially improved, the delay between neurological disease onset and initiation of miglustat was only 4 months, and miglustat was commenced at 9 months of age. Other early-infantile patients, four of whom worsened, had a mean (range) delay between neurological disease onset and start of treatment of 1.8 (0.9–2.8) years, and a mean (range) age at treatment start of 2.3 (range 1.7–3.6) years. In the late-infantile and juvenile-onset patients the mean (range) delay to therapy was 3.8 (1.0–7.8) years in patients who were stable or improved after treatment, and 4.8 (1.5–9.6) years for those who worsened. These observations appear to support the argument for starting miglustat treatment earlier, as soon as possible after the onset of neurological symptoms and especially in the early-infantile onset patients.
Data from the current cohort are in line with previous data on the high prevalence of epilepsy and cataplexy in late-infantile and juvenile (but not early-infantile) onset patients [
1,
2]. Published data on the possible therapeutic effect of miglustat on cataplexy and epilepsy are very scarce. Zarowski et al. have previously reported a complete cessation of cataplectic activity in a young male patient with juvenile-onset NP-C [
32]. In our study, miglustat did not appear to prevent the occurrence of, or to systematically improve, cataplexy or epilepsy among the small number of patients in the late-infantile and juvenile-onset subgroups. Limited data from the Spanish paediatric cohort study indicated that the onset of epilepsy and its resistance to symptomatic pharmacotherapy may result in worsening of patients’ scores on the NP-C disability scale [
26]. In our series, the presence of pre-existing epilepsy, or its onset during miglustat therapy, did not appear to affect neurological outcome when seizures were stabilized using anti-epileptic therapies. Anti-epileptic drugs employed included sodium valproate, lamotrigin and levetiracetam. Carbamazepine, oxcarbazepine and vigabatrin were avoided as they could promote myoclonias. Phenytoin was also not used in order to avoid possible cerebellar adverse effects.
Auditory acuity remained stable in this series, and no patient experienced worsening of electrical peripheral neuropathy during miglustat therapy.
Magnetic resonance imaging showed white matter abnormalities in NPC1 patients with each of the age-at-onset forms. In general, discrete posterior periventricular white matter abnormalities were followed by more diffuse changes resembling delayed myelination or demyelination among these patients. Cortical or subcortical atrophy tended to appear first in the infantile-onset forms (although it was also present in later-onset forms). Cerebellar atrophy was present in relatively few cases (two early-infantile and three late-infantile patients).
Magnetic resonance spectroscopy identified some abnormalities, including low NAA and/or high Cho with high Cho/NAA ratio. At short echo time, a high myo-inositol peak was observed before therapy and at Month 18 of follow-up in patient 4 (worsened) and patient 20 (stabilized). It is known that progressive neurodegenerative diseases are associated with a decrease of the NAA peak, which is considered to be a marker of neuronal viability, and by an increase of the Cho peak, which is considered to be a marker of membrane destruction or gliosis. Nevertheless there was no consistent pattern of change over time or in response to miglustat in our series. A low NAA peak was associated with cerebral atrophy in 8/18 cases but not with clinical worsening in all of these. This contrasts with a previously published case series based on three adult NP-C patients treated with miglustat for 24 months, where mild clinical improvement or stabilization concurrent with sustained decreases in cerebral Cho/NAA ratio were observed [
24].
While our imaging findings are of value in that they add to the relatively limited amount of published data from longitudinal imaging studies in paediatric NP-C patients, it is notable that there were no apparent correlations between MRI or MRS findings and clinical disease course during miglustat therapy. It is possible that methodological and data limitations in our cohort preclude a definitive conclusion on the utility of this imaging technique. MRS analyses for French paediatric patients were conducted at several different sites by several technicians, and according to varied analysis protocols. MRS data follow up beyond 24 months were only available for 4/20 patients, which makes it difficult to assess long-term changes. However our findings do not favour the use of high Cho peak or Cho/NAA ratios as objective markers of therapeutic effect in paediatric patients, as has been proposed for adult patients [
24].
A possible correlation between the evolution of neurological manifestations (based on changes in NP-C disability scores) and cerebral hypometabolism (measured using positron emission tomography [PET]) was previously reported based on data from Spanish juvenile- and infantile-onset patients treated with miglustat [
26]. Cerebral hypometabolism was stabilized when miglustat appeared to slow the progression of neurological symptoms, and progressive hypometabolism correlated with increasing disability scores [
26]. Nevertheless, PET is unlikely to be of practical use for routine clinical monitoring due to the limited availability of equipment.
Mild or moderate gastrointestinal disturbances were frequent during miglustat therapy, but usually resolved within the first 3 months of treatment. In addition, gastrointestinal adverse events were easily managed in most cases by the adoption of dietary alterations, by progressive initiation of miglustat treatment, or by the use of symptomatic therapy (e.g. loperamide). In particular, dietary modifications such as reduced consumption of dietary sucrose, maltose and lactose have been shown to improve the gastrointestinal tolerability of miglustat, and to reduce the magnitude of any changes in body weight, particularly if initiated at or before the start of therapy [
33,
34]. Finally, observed factors that appear to contribute to reduced treatment compliance among the youngest patients include the bitter taste of oral miglustat therapy, and the lack of a paediatric galenic form.
A decision to stop miglustat was taken for two late-infantile patients and one juvenile patient because of persistent adverse events (e.g. asthenia or anorexia) and clinical judgment of insufficient beneficial effects on disease progression. Such choices are made on a case-by-case basis with collaborative discussions between medical staff and parents, as well as detailed consideration of patients’ clinical evolution and quality of life, well in line with the updated recommendations from an expert panel [
35].
In spite of the small number of patients and the relatively short period of follow up in the French paediatric NP-C cohort, and in recognition of the invariably progressive course of neurological deterioration in untreated patients [
3,
5], we conclude that miglustat can improve or stabilize neurological disease progression in paediatric patients with NP-C, particularly those with the late-infantile and juvenile-onset forms. Our data from early-infantile onset patients, who generally exhibit greater symptom severity and more rapid progression of neurological manifestations, indicate that commencement of miglustat at approximately 2 years of age has no sustained global effect on the natural course of the disease. A shorter delay between the onset of neurological manifestations and the start of miglustat therapy was associated with a better initial therapeutic outcome in one early infantile-onset patient in this cohort. However, this patient later exhibited worsening of neurological disease after 2 years of age. More clinical experience in early-infantile onset patients treated at the very beginning of their neurological disease over a longer period is required to more fully assess the therapeutic effects of miglustat in this group.
Current guidelines for the clinical management of NP-C propose that miglustat treatment should be initiated at the onset of neurological signs [
2,
35]. However, among patients with early-infantile NP-C, the invariable occurrence and high frequency of systemic symptoms particularly neonatal cholestasis and hepatosplenomegaly, often leads to a diagnosis before the appearance of neurologic signs or developmental delay. Regular and thorough clinical examination of these patients, if possible combined with cerebral MRI, could detect neurological problems at their very beginning, leading to earlier initiation of treatment with miglustat.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
MTV, FS, TBV, BH, BC and HO conceived the study and participated in its design. BH coordinated the study. BH, MTV and TBV drafted the manuscript. MTV and PL carried out the filipin and molecular genetic studies. All authors participated in collection of data, and have read and approved the final manuscript.