Metreleptin for the treatment of progressive encephalopathy with/without lipodystrophy (PELD) in a child with progressive myoclonic epilepsy: a case report

Lipodystrophy syndromes are a group of rare, highly heterogeneous disorders characterized by generalized or partial lack of adipose tissue and deficiency of the adipocyte-secreted hormone leptin [1‐4]. Lipodystrophies are inherited or acquired, and are classified into four major subtypes: congenital or acquired generalized lipodystrophy, and familial or acquired partial lipodystrophy [2, 3, 5]. Multiple congenital lipodystrophy phenotypes have been identified, each with distinct clinical features [5‐7]. Conditions associated with variants in the BSCL2/seipin gene include progressive encephalopathy with/without lipodystrophy (PELD), also called Celia’s encephalopathy (MIM#615924) and congenital general lipodystrophy type 2 (CGL2) [2, 5]. PELD is associated with BSCL2 variants that cause skipping of exon 7 and is characterized by the development of progressive myoclonic epilepsy (PME) at a young age and severe progressive neurological impairment, which is rapidly fatal in many, but not all, cases [8‐14].

Without effective treatment, the prognosis for patients with PELD is poor because of the intractable epilepsy and rapid neurodegeneration [10, 11, 13]. Preliminary data suggest that leptin replacement therapy, combined with dietary intervention, may provide some benefit in patients with PELD [9]. Metreleptin, a recombinant analogue of leptin, has been shown to improve metabolic abnormalities in patients with lipodystrophy, including pediatric patients with CGL2, and is currently the only drug indicated specifically for the treatment of generalized lipodystrophy [2, 6, 15]. However, due to the extremely rare nature and particularly poor prognosis of many congenital lipodystrophies, little is known about the utility of metreleptin in affected children with severe neurological involvement [9]. Here, we report the effects of metreleptin therapy on the metabolic and neurological complications of PELD in a child with PME.

The male patient was born at term by natural delivery to non-consanguineous parents in Macedonia in 2013, following a normal pregnancy. In the first month of life, he presented with generalized hypotonia. At 12 months, he developed monthly tonic seizures, which were well-controlled with valproic acid.

At 4 years, the patient had hypercholesterolemia and hypertriglyceridemia, with partial improvement of cholesterol and triglyceride levels after 3 months of hypolipidemic nutrition (Table 1). Oral glucose tolerance testing (OGTT) at 1.75 mg/kg revealed normal glucose tolerance (baseline: 75 mg/dL, 2 h: 82 mg/dL), although the homeostasis model assessment of insulin resistance (HOMA-IR) value was high (2.6). The glycosylated hemoglobin (HbA1c) level was 29 mmol/mol. Abdominal ultrasound revealed mild hepatosplenomegaly with increased liver echogenicity related to mild steatosis; echocardiography revealed mild left ventricular hypertrophy with normal function. His neurological condition continued to deteriorate: he was able to walk but was frequently falling due to his seizures, he had severe intellectual disability, and the EEG showed a progressive slowing of the background activity with generalized spike and wave associated with myoclonic, myoclonic-atonic seizures. He required hospitalization on several occasions due to prolonged seizures. Therefore, a vagal nerve stimulator was implanted with a rapid (1 month) up-titration of the stimulation parameters. Within that month, the patient experienced a consistent reduction in seizure frequency; while he was still experiencing up to 20 seizures/day he had no more episodes of status epilepticus. The detailed neurological manifestations of this case have previously been reported [16]. Four months after implantation, the seizures reappeared, with a transient deterioration of EEG background activity, and perampanel was initiated. Daily seizures (up to 60 seizures/day) persisted, but the patient had no episodes of status epilepticus.

When he was 5 years old (10 months after perampanel was initiated), he was hospitalized with new episodes of prolonged seizures with mixed tonic and myoclonic components. BMI was at the 10th percentile. Subcutaneous metreleptin 0.06 mg/kg/day was initiated and was well tolerated. After 2 months’ therapy, his overall lipid profile improved (Table 1). OGTT was normal (baseline: 59 mg/dL, 2 h: 75 mg/dL) and HOMA-IR had decreased to 0.33. The HbA1c was static at 29 mmol/mol; neurological status was stable and without prolonged seizures. Serum leptin level (0.4 ng/mL) was normal-to-low for his BMI (0.25–3.20 ng/mL) prior to metreleptin; after 3 month’s therapy, leptin was 0.4 ng/mL before and 10.8 ng/mL after the daily dose.

After 1 year of treatment with metreleptin, HOMA-IR had increased to 6.02, but due to a lack of supply the patient had not been receiving therapy for the preceding month. Abdominal ultrasound showed mild hepatomegaly and an improvement in hepatic steatosis. The patient had grown 8.13 cm/year and his BMI was below the third percentile.

Adherence to therapy during treatment was ensured by a direct hospital supply of medication. No adverse effects have been reported or found during follow-up, including from results of blood examinations, ultrasound, and clinical evaluation. In the subsequent 2 years, he did not experience any prolonged epileptic seizure leading to hospitalization. He still experienced 2–3 seizures/month, which resulted in sudden falling down, and daily (up to 30 seizures/day) myoclonic seizures involving mainly the perioral muscles. An EEG did not show any evidence of neurological modification from the pre-treatment period.

The metabolic results observed in our patient were similar to those demonstrated by Brown et al. in pediatric patients with different types of lipodystrophy [15]. In their report, metreleptin treatment outcomes both in the short (1 year) and long term (mean 5 years) were better in adolescents (aged 12–18 years) than in children; this result is probably influenced by the small number of the children patients included in the study [15]. The clinical evidence to date confirms studies in animal models – leptin-treated ob/ob mice and n-SREBP-1c lipodystrophic mice – that showed leptin’s hypoglycemic food intake-independent effect, attributable to improvements in peripheral insulin sensitivity [17‐20]. Even though the exact mechanism of action of metreleptin has not yet been fully elucidated, metreleptin seems to contribute to body weight loss by improving insulin sensitivity rather than increasing energy expenditure.

Metreleptin-associated improvements in the lipid profile of patients may be mediated by increased expression of enzymes and transcription factors involved in fatty acid oxidation and decreased expression of those regulating fatty acid synthesis, as demonstrated in rodent models [21‐23].

We previously described the neurological condition of our patient [16] before metreleptin treatment. After treatment, we saw a clear benefit from an epileptogenic point of view, with a reduction in seizure burden and a drop in hospitalizations. The neurological impairment remains stable. These results are in line with a previous case reported by Araújo-Vilar and colleagues, who demonstrated a clear slowdown in neurological regression with metreleptin treatment in combination with a diet rich in polyunsaturated fatty acids in a child with PELD [9].

It should be noted that, as in-depth histological evidence of neuronal inclusion and molecular studies were not carried out, we cannot rule out the possibility that this child had a variant of CGL2, and not PELD. However, in a previous report of similar cases, Opri and colleagues reported two cases of CGL2 that were due to variant c.974dupG in BSCL2 [11]. Further research by Sanchez-Iglesias and colleagues showed that this is the same mechanism as the c.985C > T variant in BSCL2 that causes PELD [13]. Moreover, this PELD variant leads to the same adipose and neurological phenotypes as described in the original report by Opri and colleagues, leading Sanchez-Iglesias et al. to conclude that the patients reported by Opri et al. actually had PELD and not CGL2 [13].