Page 45 - ILAE_Lectures_2015

microdeletion, seen in 12 patients in the classic 2009 report by Helbig13. The same CNV has
now been studied in a 246-case series, 28% of whom had seizures14; neuropsychiatric
manifestations were common, major congenital malformations were not. Further genotype-
phenotype correlation with long-established CNVs associated with epilepsies will become
possible as increasing numbers of cases are reported. Such phenotypic delineation will
facilitate genetic screening and interpretation for future practice. CNVs have now also been
identified in genetic generalised epilepsies with intellectual disability15.

At the leading edge of current research in epilepsy genetics, and the most productive tool in
terms of gene discovery, is whole exome sequencing (WES). With falling costs and
increasing availability, an ever wider range of epilepsies, more or less homogeneously
grouped, have been subjected to WES. Discoveries are being reported at a pace too great to
meaningfully list each individually. Progress has been most dramatic for the epileptic
encephalopathies. Though individually rare, the encephalopathies account for an important
part of the burden of the epilepsies. Their genetic tractability is probably because they are
often caused by variants of large effect, which is perhaps not surprising considering the
severity of the phenotype. A set of genes and related pathways responsible for a number of
epileptic encephalopathies was reported using trio exome sequencing16 – several known
‘epilepsy genes’ were identified, and a number of novel candidates were proposed. One
candidate, DNM1, was then confirmed by merging data from consortia17, illustrating the
frequent need for large numbers of patients to formally declare involvement of a given gene.
Mutations in many genes have been identified in the epileptic encephalopathies, including
for example AARS18, KCNA219, STX1B20, PURA21, WWOX22, SLC13A523, DOCK724 and
SZT225 among many others. Some of these conditions have distinctive features, but many do
not.

One consequence of this observation is that candidate gene selection in the epileptic
encephalopathies is a challenge, making gene panels for clinical genetic diagnosis of limited
value, compounded by the rapid pace of gene discovery: a gene may not be considered a
candidate for the panel, or not be included because it was not linked with epilepsy at the time
of panel design. A further implication is that genotype-phenotype correlation is needed, but
will also be challenging, and may need newer phenotyping tools accessing data not typically
used in clinical phenotyping26. Moreover, given the richness of the emerging data, there is
considerable scope for data mining and novel analytic methods, some to predict new genes
for epileptic encephalopathy27, with methods also to prioritise genes28,29. The greatest promise
lies perhaps in the identification of pathways implicated across sets of epileptic
encephalopathies, such as the mTOR pathway30, that may already have possible treatments
or repurposable drugs, or that might point the way to new generic treatments relevant across
epileptic encephalopathies linked by shared mechanisms16,31.

WES, and other methods, have also been successful in identifying the cause(s) of some rare
conditions which may feature epilepsy as part of a phenotype. Examples include alternating
hemiplegia of childhood (due in 80% of cases to de novo mutation in the ATP1A3 gene32),
in which rare condition perhaps 50% of affected individuals have seizures; DOORS
syndrome, which is very rare, due in about 50% of cases to mutation in TBC1D2433, and for
which genotype-phenotype correlation may yet show it can be considered in some cases an
epileptic encephalopathy; and epilepsies with other comorbid features such as migraine or
movement disorders, for which implicated genes include SCN1A, CACNA1A, ATP1A2,
SLC2A1, PRRT2, STXBP1 and FOXG134-38. Of considerable interest are the epilepsies with
associated language or speech disorder – these are broad summary terms for aspects of the
phenotype that have often been characterised in great detail, within the epilepsy-aphasia
spectrum. Mutations have been identified in the NMDA receptor NR2A subunit-encoding
gene GRIN2A in Landau-Kleffner syndrome, electrical status epilepticus in sleep