09 EEG of the pediatric epilepsy syndromes

The 2022 ILAE position papers define each epilepsy syndrome by a structured set of features, and the EEG is written into that structure. For each syndrome the definitions list mandatory features, which must be present, alerts, which are findings that should make you reconsider the diagnosis, and exclusionary features, which rule it out. Three papers cover the pediatric range: syndromes with onset in neonates and infants (Zuberi 2022), syndromes with onset in childhood (Specchio 2022), and the idiopathic generalized epilepsies (Hirsch E 2022). The ILAE's epilepsydiagnosis.org site presents the same material syndrome by syndrome, with an EEG tab on each page.

What this changes for the reader

Because the definitions are explicit, the EEG can do more than support a clinical impression: it can confirm a syndrome when the mandatory features are present, and it can argue against one when an alert or an exclusion is met. That means the report's clinical correlation can be specific, and it also means the reader has to know the criteria rather than a list of classic patterns. A pattern that is characteristic but not mandatory, such as hypsarrhythmia in infantile epileptic spasms syndrome, cannot be used to withhold a diagnosis, and a finding listed as an alert, such as the absence of spike-wave after adequate hyperventilation in an untreated child with suspected absence epilepsy, has to be reported as such.

Matching the study to the question

The definitions also tell you which recording answers which question. The discharges of SeLECTS and the sleep-activated spike-wave of DEE-SWAS and EE-SWAS live in sleep, so ask for a study that captures it. Childhood absence epilepsy is provoked by hyperventilation in most untreated patients, so a study without adequate hyperventilation has not tested for it. Spasms are diagnosed by their ictal EEG, so the question needs a recorded event, which usually means prolonged video-EEG. Ordering a routine awake record for each of these three questions is the commonest way to get an uninformative negative, a point Module 8 makes about sleep and Module 2 makes about activation procedures.

The two infantile-onset syndromes in this module are developmental and epileptic encephalopathies, and their EEG features are set out in Zuberi 2022.

EIDEE

Early infantile developmental and epileptic encephalopathy is the 2022 term that brings together the earlier eponymous syndromes of the neonatal and early infantile period. Its EEG is a burst suppression pattern or multifocal discharges, on a background of diffuse slowing. The neonatal background categories of Module 7 are the vocabulary for describing it, and burst suppression there carries the same definition: invariant, abnormally composed bursts with interburst voltage below 5 µV and no variability or reactivity.

Infantile epileptic spasms syndrome

Two findings are described. The interictal record classically shows hypsarrhythmia, a chaotic, high-voltage, disorganized background with multifocal spikes, but the definition also admits multifocal or focal discharges. The ictal record of a spasm shows a high-amplitude generalized slow wave followed by low-amplitude fast activity, the electrodecrement. The two pitfalls in the table are the ones that change management. First, hypsarrhythmia is not required and may be absent early in the course, so an infant with clinical spasms and a non-hypsarrhythmic record should not have treatment delayed while a repeat study is awaited. Second, a normal EEG during recorded spasms argues against the syndrome; if the event that the family calls a spasm is captured and the EEG does not change, the diagnosis needs to be reconsidered and the events re-examined. Both pitfalls are reasons why the right study for suspected spasms is one that records an event.

Three childhood-onset syndromes account for much of the pediatric EEG workload, and each has a defining pattern with a defining pitfall.

Childhood absence epilepsy

The discharge is generalized 3 Hz spike-wave, with a range of 2.5 to 4 Hz, and it is provoked by hyperventilation in most untreated patients; photic stimulation triggers it in 21% (Hirsch E 2022). The pitfall runs the other way: no spike-wave after 3 minutes of well-performed hyperventilation in an untreated child makes the syndrome very unlikely. ILAE 2022 lists this as an alert; epilepsydiagnosis.org treats it as exclusionary. Either way, the report must state whether hyperventilation was adequate, because a poor effort leaves the question open, which is the reporting rule Module 2 sets.

SeLECTS

The discharges are high-amplitude centrotemporal sharp-and-slow waves, activated in drowsiness and sleep, with a horizontal dipole: frontal positive and temporo-parietal negative (Specchio 2022). The dipole is the reason the discharge can phase-reverse with opposite polarities in different chains, a montage effect Module 1 prepares you for. The pitfall is that the same discharges occur in healthy children: in the prospective study cited in Module 5, 4.2% of healthy children had rolandic discharges (Borusiak 2010). Centrotemporal spikes support the diagnosis in a child with the right seizures; they do not make it in a child without.

SeLEAS

Self-limited epilepsy with autonomic seizures shows multifocal, high-voltage spike or sharp-and-slow waves, often posterior, that shift between recordings. The pitfall is to read a focus that has moved as evidence of a structural lesion. The shifting is the pattern.

Two childhood syndromes are defined by patterns that a routine awake record can miss, and both come with a rule about what is required.

Lennox-Gastaut syndrome

The definition requires two patterns: generalized slow spike-wave below 2.5 Hz and generalized paroxysmal fast activity of 10 Hz or more, the latter often seen in sleep (Specchio 2022). Both are required; one alone is not enough. Slow spike-wave on its own is a nonspecific finding of several encephalopathies, and paroxysmal fast activity on its own can be seen in other tonic-seizure disorders, so a record that shows only one should be described as showing that pattern, with the syndrome held open rather than assigned. Because the fast activity is often confined to sleep, a study that captures sleep is again the study that can answer the question.

DEE-SWAS and EE-SWAS

Developmental and/or epileptic encephalopathy with spike-wave activation in sleep is the 2022 name for the conditions formerly called ESES and CSWS. The pattern is slow spike-wave, at about 1.5 to 2 Hz, markedly activated in non-REM sleep. The pitfall is a number that generations of readers learned: the ILAE definition sets no spike-wave index threshold, and the old 85% cutoff is not required. What the definition asks for is marked activation in sleep in a child with the clinical picture, including regression, so a record with sparse discharges awake and near-continuous discharges in non-REM sleep meets the EEG criterion without any percentage being calculated. The mechanism, a cortex that discharges into the synchrony of non-REM sleep, is the same one Module 8 describes for sleep activation in general, taken to its extreme.

The common thread

Both syndromes are defined in part by patterns that live in sleep, and both have a rule that stops the reader short of the diagnosis on a partial finding: two patterns for Lennox-Gastaut, no index for the SWAS encephalopathies. Know the rules and the report will say exactly what was and was not seen.

Juvenile myoclonic epilepsy is the idiopathic generalized epilepsy of adolescence, and its EEG is defined in Hirsch E 2022.

The pattern

The interictal discharge is generalized polyspike-wave and spike-wave at 3 to 5.5 Hz, faster on average than the 3 Hz spike-wave of childhood absence epilepsy and often with more than one spike preceding the slow wave. A photoparoxysmal response occurs in more than a third of patients, so photic stimulation is not a formality in an adolescent with morning jerks; it is one of the activation procedures that can bring out the defining discharge, and the report should say it was done and how the patient responded.

The pitfall

The discharges can look focal or asymmetric. On a single record a polyspike-wave burst may be larger over one hemisphere, or a fragment may appear over one region only, and an unwary reader will describe a focal discharge and set off a focal-epilepsy workup, including imaging, in a patient with a generalized syndrome. Module 8 draws the line: a single asymmetric burst does not make a focal epilepsy, whereas a stable lead from the same side across repeated discharges is what raises secondary bilateral synchrony. Read the whole record, count the bursts, and describe the asymmetry as an asymmetry of generalized discharges when that is what it is.

Why it matters at the bedside

The distinction between a generalized and a focal epilepsy changes the treatment conversation, and an EEG report that mislabels a generalized discharge as focal can push a patient toward the wrong path. As with the other syndromes in this module, the safeguard is to describe exactly what was seen, in the standard terms, and then to place it against the definition rather than against a remembered picture.