The Expanding SCN8A-Related Epilepsy Phenotype

Commentary

The Phenotypic Spectrum of SCN8A Encephalopathy.

Larsen J, Carvill GL, Gardella E, Kluger G, Schmiedel G, Barisic N, Depienne C, Brilstra E, Mang Y, Nielsen JEK, Kirkpatrick M, Goudie D, Goldman R, Jähn JA, Jepsen B, Gill D, Döcker M, Biskup S, McMahon JM, Koeleman B, Harris M, Braun K, de Kovel CGF, Marini C, Specchio N, Djémié Y, Weckhuysen S, Tommerup N, Troncoso M, Troncoso L, Bevot A, Wolff M, Hjalgrim H, Guerrini R, Scheffer IE, Mefford HC, Møller RS; for EuroEPINOMICS RES Consortium CRP. Neurology 2015;84:480–489.

OBJECTIVE: SCN8A encodes the sodium channel voltage-gated a8-subunit (Na_v 1.6). SCN8A mutations have recently been associated with epilepsy and neurodevelopmental disorders. We aimed to delineate the phenotype associated with SCN8A mutations. METHODS: We used high-throughput sequence analysis of the SCN8A gene in 683 patients with a range of epileptic encephalopathies. In addition, we ascertained cases with SCN8A mutations from other centers. A detailed clinical history was obtained together with a review of EEG and imaging data. RESULTS: Seventeen patients with de novo heterozygous mutations of SCN8A were studied. Seizure onset occurred at a mean age of 5 months (range: 1 day to 18 months); in general, seizures were not triggered by fever. Fifteen of 17 patients had multiple seizure types including focal, tonic, clonic, myoclonic and absence seizures, and epileptic spasms; seizures were refractory to antiepileptic therapy. Development was normal in 12 patients and slowed after seizure onset, often with regression; 5 patients had delayed development from birth. All patients developed intellectual disability, ranging from mild to severe. Motor manifestations were prominent including hypotonia, dystonia, hyperreflexia, and ataxia. EEG findings comprised moderate to severe background slowing with focal or multifocal epileptiform discharges. CONCLUSION: SCN8A encephalopathy presents in infancy with multiple seizure types including focal seizures and spasms in some cases. Outcome is often poor and includes hypotonia and movement disorders. The majority of mutations arise de novo, although we observed a single case of somatic mosaicism in an unaffected parent.

Larsen, Carvill, and colleagues report a spectrum of epileptic encephalopathy phenotypes associated with the gene SCN8A, relatively recently identified in the epilepsy genetics literature as a gene responsible for epilepsy and intellectual disability. As with many newly emerging genes associated with epilepsy and epileptic encephalopathy, the SCN8A spectrum is quite broad and still expanding. Since initial reports often reflect the most severe cases of epilepsy associated with a new gene, it is difficult to predict the full range of this spectrum. Thus, there is great value in the detailed characterization of cases from well-curated cohorts of patients with epilepsy.

Even though it is not a population study, the fact that the seven cases they report derive from an impressive cohort of 683 individuals with epileptic encephalopathy reflects the role of SCN8A as one of many important genes for early-onset epilepsy. To those outside the field, this 1% may not seem substantial, but epilepsy—even epileptic encephalopathy—is genetically heterogeneous, with a multitude of genetic causes collectively accounting for the growing list of epileptic encephalopathies with identifiable genetic etiology (1–4). To add to the phenotypic characterization of SCN8A encephalopathy, the authors also included ten additional cases referred for this study. The work thus represents the collaborative strength of some of the major groups who have been devoting their careers to epilepsy genetics.

One feature of the genetics of SCN8A encephalopathy worth noting is that, as with other epilepsy genes and their mutations, the types and locations of SCN8A mutations are varied. The authors outline very reasonable criteria for their assumptions of pathogenicity for each case. Clinical neurologists should take heed that the mere mention of a variant in a gene associated with epilepsy does not necessarily mean that that gene is the cause of a patient's epilepsy. The majority of the mutations are missense variants predicted to be pathogenic, including an apparent mutation hotspot at amino-acid position 1872 (more detail is available in a subsequent review by Wagnon and Meisler [5]). While there is already some functional evidence for gain of channel function and resultant neuronal hyperexcitability for some previously described mutations, additional functional work on some of the mutations is needed and can be correlated with the clinical data presented by these authors.

The phenotypes of the 17 patients considered to have mutations in SCN8A are quite variable. While many began with focal seizures in the first few months of life, the majority of seizure types eventually seen included many generalized seizure types, including generalized tonic-clonic seizures, epileptic spasms, myoclonic seizures, and absence seizures. A main discussion point concerns differences between patients with mutations in SCN8A and Dravet syndrome, a more distinct sodium channel–related epilepsy. Indeed, patients with SCN8A encephalopathy seem to have an earlier and more varied onset, perhaps a broader range of seizure types, and less association between fever and seizures than patients with Dravet syndrome. While medications such as carbamazepine and its derivatives would be a common first choice of treatment for focal seizures, as was the case for some of the patients, the presence of generalized seizure types that could worsen with carbamazepine seemingly complicates the treatment choices for patients with SCN8A encephalopathy.

The SCN8A phenotypes presented raise some important questions that can be answered as more patients are diagnosed with SCN8A encephalopathy, which they undoubtedly will be:

• 1) The MRI atrophy seen in three cases is interesting, and over time we will see if this is a feature specific to a subset of patients with mutations in SCN8A and is disease-related as opposed to being referable to medications or other causes.

• 2) The original proband in the 2012 Veeramah et al. study died with a diagnosis of SUDEP (6), and a question raised by that report is whether SCN8A-related epilepsy really confers a higher rate of SUDEP than other refractory epilepsies, a matter of clinical concern when families face a new diagnosis of SCN8A-related epilepsy. Since two patients in this series are deceased, and because one patient died with seizures and the other was considered SUDEP, the authors raise the point that this association bears further study as more SCN8A cases emerge (3). Some patients had normal development prior to seizure onset, while others were already delayed; by the time of reporting, all were moderately or severely intellectually disabled—one might conclude that this reflects the natural history of SCN8A-related epilepsy, but the fact that some appeared normal at onset raises the possibility that earlier intervention in future cases, identified more quickly in the modern genomic medicine era, could exact some benefit not only on seizures but on developmental outcome.

At this point, though we do not yet have evidence supporting specific treatment on the basis of SCN8A diagnosis, in this era of precision medicine, we strive to find such treatments. Having our patients identified with their mutations, and ideally having functional data or strong predictors of the functional consequences of mutations, will be the first step to designing clinical trials for genetically defined conditions. This is a major argument in favor of the authors' “genotype first” approach. These authors show that “genotype first, then carefully phenotype” is an effective strategy in the research realm that should quickly translate into the clinical realm. The practice of genetic testing in epilepsy will continue to increase, not only in children with severe, early-onset epilepsy and epileptic encephalopathy but also in older children with perhaps less severe epilepsy but clinically significant comorbidities and even in adults with epilepsy and intellectual disability. In this setting, one would predict more cases of SCN8A-related epilepsy to emerge over time, perhaps with a milder phenotype than described by Larsen, Carvill, and colleagues. Treatments targeting one set of SCN8A mutations and epileptic encephalopathy will likely have applicability to a much larger group of patients. These authors have provided an excellent example of a phenotyping strategy that can be applied in the future to these additional SCN8A mutation-positive patients.