CNNM2 Heterozygous Variant Presenting as Hypomagnesemia and West Syndrome: Expanding the Spectrum of CNNM2 Gene-Related Epileptic Disorders.

Dear Editor,

West syndrome (WS) is one of the few most common early infantile epileptic encephalopathy. Although structural etiology contributes to a significant proportion of WS in developing countries, at the same time monogenic variants also play a critical role in the pathogenesis.[1] More than 60 EIEE genes have been identified till date, out of which the majority can cause west syndrome and the number is increasing every year.[234] Mutations in the CNNM2 gene originally described to be associated with renal hypomagnesemia and recently also have been reported to be associated with seizures and developmental delay. However, mutations in the CNNM2 gene have never been associated with WS. Here, we are reporting for the first time a 4-month-old girl with WS caused by de novo novel heterozygous pathogenic mutation in the CNNM2 gene.

A 4-month-old female infant firstborn to a non-consanguineous couple with an uneventful perinatal period, presented with flexor epileptic spasms occurring in the clusters, especially at sleep-wake transition for 1 week. There was no family history of any neurological disorder. Her developmental milestones, anthropometry, and physical examination were normal. Blood sugar, arterial blood gas (pH-7.35, pCO2-32, pO2-84, HCO3-23), serum calcium (9.4 mg/dl), serum sodium (137 mEQ/l), serum phosphate (4.6 mg/dl), serum potassium (4.2 mEQ/l), fractional excretion of potassium (4.2%), serum parathyroid hormone (46 pg/ml), and vitamin D levels were within normal limits except serum hypomagnesemia (0.46 mmol/L, range: 0.7–1.15 mmol/L). Serum biotinidase levels, Tandem mass spectrometry, and urinary organic acid profile were also normal. Her electroencephalography (EEG) showed a modified hypsarrhythmia pattern [Figure 1] and magnetic resonance imaging (MRI) of the brain was normal. Ultrasonography of the abdomen did not show any nephrocalcinosis.

Figure 1

A sedated electroencephalogram at 4-months of age (international 10-20 system, average montage, sensitivity: 7μV/mm, sweep speed: 30 mm/sec, low filter: 1 Hz and high filter 70 Hz) showing disorganized, asynchronous, chaotic high amplitude slow waves activity with interspersed multifocal spikes suggestive of Hypsarrhythmia

She was started on intravenous magnesium supplementation, but the hypomagnesemia and epileptic spasms persisted. The seizures did not respond to an adequate trial of pyridoxine, pyridoxal phosphate, folinic acid, and biotin. Subsequently, she was started on intramuscular adrenocorticotropin hormone (ACTH) injection and her epileptic spasms reduced by around 80% in frequency. Later, she also developed the arrest of developmental milestones and seizures of other semiology like tonic and focal clonic seizures and required multiple antiepileptic drugs (valproate[50 mg/kg/day], levetiracetam[50 mg/kg/day], clobazam). The hypomagnesemia also persisted, along with increased fractional excretion of magnesium (8.4%, normal 0.5-4%). Next-generation sequencing detected a novel heterozygous missense variant in exon 1 of the CNNM2 gene (c.82G>A) that results in the amino acid substitution of Threonine for Alanine at codon 28 (p.Ala28Thr; ENST00000369878.4). The in silico predictions of the variant was found to be possibly damaging by PolyPhen-2 (HumDiv) and deleterious by SIFT, LRT, MutationTaster2, and Mutation Assessor. The final PROVEAN score of the protein variant was also found to be deleterious. The evolutionary conservation of the putative pathogenic mutation was also assessed by the HomoloGene tool. The variant could not be detected in any of the parents on Sanger sequencing and most probably is of de novo origin. The parents were counseled regarding the nature and prognosis of the disease. After 1 year of follow-up, she was found to have microcephaly, global developmental delay (the developmental quotient of 45 on the developmental assessment of Indian infants-2), central hypotonia, and brief infrequent tonic seizures (1--2 episodes/week). The follow-up EEG at 16 months of age showed frequent spikes and polyspikes-wave discharges [Figure 2]. However, no repeat neuroimaging could be performed on the follow-up to look for the evolution of any new abnormality.

Figure 2

A sedated electroencephalogram at 16-months of age (international 10-20 system, bipolar montage, sensitivity: 7.5μV/mm, sweep speed: 30 mm/sec, low filter: 1 Hz and high filter 70 Hz) showing asymmetric and asynchronous generalized (frontal dominant), high amplitude spike-slow wave complexes with frequent polyspikes

Hypomagnesemia is a known etiology for seizures in neonates and infants, but in most cases, it is associated with hypocalcemia, as hypomagnesemia causes resistance to parathormone secretion. However, in such cases, correction of electrolyte abnormality leads to control of seizures and normal developmental outcome.[5] But in rare monogenetic disorders caused by mutations in TRPM6 and TRPM7 genes may cause seizures, migraine, intellectual disability, and depression, along with hypomagnesemia. Hypomagnesemia, seizures, and mental retardation 1 (HOMGSMR1) (OMIM*607803) is another recently described disorder caused by both monoallelic and biallelic mutations in the cyclin M2 (CNNM2) gene. Till date, seven kindreds from various parts of the world have been described by three different study groups.[678]

The Cystathionine-β-synthase (CBS)-pair domain divalent metal cation transport mediators (CNNMs) are involved in maintaining Mg2+ homeostasis. Among them CNNM2 is ubiquitously expressed in the brain, lung, spleen, testes, liver, and heart, thus mutations in this gene are likely to cause developmental delay and seizures. CNNM2 encoded protein either itself is an Mg2+ transporter or regulates proteins involved in its transport. The mutation in exon 1 in our case most probably distorted the large 64 amino acid signal peptide located at the N-terminal extracellular domain or CBS-par domain in the cytosolic region and led to conformational changes in the protein. These domains have been described in the literature to be essential for Mg2+-ATP binding or phosphatases of regenerating liver (PRL) binding and thus mutations affecting these crucial domains are likely to cause pathogenic loss of function mutations.

Arjona et al.[6] were first to describe six children with this rare illness from five different German kindreds in 2014. All of them had de novo heterozygous missense mutations, except one kindred who had homozygous recessively inherited mutations from consanguineous parents and had severe presentation beginning from the neonatal period. Severe phenotype with neonatal-onset epileptic encephalopathy refractory to multiple antiepileptic drugs and brain malformations has also been reported by Accogil et al.[7] in another recessively inherited case.

Four out of five patients with heterozygous mutations described by Arjona et al.[6] had seizure onset between 4 and 12 months like our case. All these cases subsequently developed limited motor skills and social communication and even one patient developed characteristic features of autism.

None of the reported cases till date has been described to manifest epileptic spasms like our case and it widens the clinical spectrum of epilepsy associated with the CNNM2 mutations. Valproate, lamotrigine, topiramate, and clobazam were various antiepileptic drugs tried in previous cases.[67] Three reported cases with homozygous mutations till now had drug-resistant epilepsy, like our case. Relatively more severe epilepsy in our case despite heterozygous mutation, that is, presence of West syndrome might be because of the characteristics of the novel mutation. In all reported children with CNNM2 mutations, seizures persisted despite multiple efforts attempting at correcting hypomagnesemia and required multiple antiepileptic drugs. These cases also had normal serum calcium levels, no evidence of nephrocalcinosis on kidney ultrasound, and renal sodium wasting as measured by inappropriately high fractional excretion of magnesium, despite low serum magnesium level, resembling our case. Even on follow-up, all cases continued to have low serum magnesium levels.

One female case with a heterozygous variant reported by Arjona et al.[6] had a mild intellectual disability, apart from myoclonus, muscle spasm, and dysesthesia from 16 years of age, without any definite clinical seizures in early childhood. A similar mild clinical presentation has also been reported by Khalaff-Nazall et al.,[8] in multiple family members spanning over three generations in a family harboring heterozygous CNNM2 mutation with autosomal dominant inheritance. Apart from mild impairment in fine motor skills like dyslalia and dyslexia, they had no other neurological abnormalities. However, all of them had hypomagnesemia and hypermagnesuria, refractory to oral magnesium supplementation. The same authors described another family cohort with heterozygous CNNM2 variants presenting with a magnesium independent neurodevelopmental phenotype with intellectual disability, autism spectrum disorder, and cone-rod dystrophy causing visual impairment, but no clinical seizures.[8] Thus, clinicians need to remember although CNNM2 mutations associated disorder has been initially described as an autosomal recessive disorder, it can present with neurological symptoms, even with epileptic spasms and low serum magnesium levels, in the heterozygous state and accordingly should provide appropriate genetic counseling to parents.

To conclude, the current case highlights the importance of considering the possibility of underlying CNNM2 mutations when the child presents with epileptic spasms and persistent hypomagnesemia. Clinicians also need to rule out hypomagnesemia in children having an epileptic spasm, without any apparent definite cause clinically and should order a renal ultrasound, if they obtain low serum magnesium levels.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.