08 Interictal epileptiform and non-epileptiform abnormalities

A discharge that has passed the six IFCN criteria from Module 5 still has to be described before it can be interpreted, and the description is what the next reader, and the clinician, will act on. Five elements make it complete: type, location, field, prevalence and state.

Type. By the IFCN glossary, a spike lasts 20 to under 70 ms and a sharp wave 70 to 200 ms (Kane 2017). Measure the duration at the base of the transient, not by impression. The distinction is worth recording because the vocabulary is standard, but it carries little clinical weight: a sharp wave over the left temporal region means much the same as a spike there. Location and field matter more than the width of the wave. Note also whether the transient stands alone or is followed by a slow wave (spike-and-slow-wave) or is repetitive (polyspikes).

Location and field. Name the electrode of maximum negativity, found by phase reversal or referential amplitude as in Module 1, and then the electrodes over which the field spreads. A discharge with a maximum at C4 and a field over F4 and P4 is a different finding from one confined to C4 and T4, and a horizontal dipole with a positive end deserves its own sentence, because it is one of the clues to the centrotemporal discharges of Module 9.

Prevalence and state. Say how often the discharge occurred during the record and in which state it appeared. A single sharp wave in drowsiness and a discharge that recurred every few seconds throughout sleep are different findings with different weight. Stating the state also protects the reader from the commonest over-read, since drowsiness and sleep are where the normal variants of Module 5 live.

A description built from these five elements lets the impression follow without argument, and lets a later reader compare records across time. A description that omits any of them leaves the clinician to guess.

Epileptiform discharges are broadly focal or generalized, and the split matters because it points toward different syndromes and different studies. A focal discharge has a maximum at one electrode or region, with a field that falls away from it. A generalized discharge appears over both hemispheres at once, usually with a frontal or fronto-central maximum, and is described by its frequency and by the activation that brings it out: sleep, hyperventilation or photic stimulation. The main generalized patterns are spike-wave, polyspike-wave and slow spike-wave, and their frequencies are what separate the syndromes of Module 9, where the 3 Hz spike-wave of childhood absence epilepsy, the 3 to 5.5 Hz polyspike-wave of juvenile myoclonic epilepsy and the slow spike-wave below 2.5 Hz of Lennox-Gastaut syndrome are set out in the ILAE 2022 definitions.

Secondary bilateral synchrony

Not every bilateral discharge is generalized in origin. A focal generator, most often frontal, can drive both hemispheres so quickly that the discharge looks generalized on a routine read. The clue is consistency: the same side leads, by a fraction of a second, discharge after discharge, or the discharge is repeatedly asymmetric with the same hemisphere larger. That is secondary bilateral synchrony, and it should be raised as a possibility whenever generalized-looking discharges have a stable lead or a stable asymmetry, because it changes the syndrome under consideration and the imaging question. The reverse error also happens: the outline notes that in juvenile myoclonic epilepsy the generalized discharges can look focal or asymmetric on a given record, so a single asymmetric burst does not make a focal epilepsy.

What the split changes

A focal discharge raises the question of a focal epilepsy and, depending on its location, morphology and the child's history, of a structural cause. A generalized discharge raises the idiopathic generalized and developmental and epileptic encephalopathy syndromes, whose EEG signatures Module 9 sets out. The description therefore needs to say which it is, and, when the record cannot decide, to say that too.

Many focal and generalized discharges appear only in drowsiness and sleep. The centrotemporal discharges of SeLECTS are the classic example: they are activated in drowsiness and sleep, and a fully awake record can be normal in a child who will show abundant discharges within minutes of falling asleep. Generalized discharges behave the same way in many children, and the spike-wave of DEE-SWAS and EE-SWAS is, by definition, markedly activated in non-REM sleep (Specchio 2022). The mechanism is the shift in thalamocortical rhythms that comes with sleep onset: the same circuits that generate spindles become more synchronized, and a cortex with a lowered threshold discharges into that synchrony.

What this means for the read

A routine record that did not capture sleep is a weaker negative. It can still be reported as normal, but the report must say that sleep was not obtained, because the clinician will otherwise read "normal" as "no discharges" when the record never sampled the state in which discharges occur. This is one of the elements of the technical description that Module 2 insists on.

What this means for the recommendation

When the question is a syndrome whose discharges live in sleep, ask for the study that samples sleep. Module 9 makes the pairing explicit: sleep for SeLECTS and for spike-wave activation in sleep, hyperventilation for absence, and a recorded event for spasms. A second routine study without sleep answers the same non-question twice.

The other direction

Sleep also brings out the normal variants of Module 5, so a discharge found in drowsiness has to pass the same six criteria as one found awake. State is part of the description precisely so that the reader can weigh it.

Slowing is the other great family of abnormality, and it points to a different set of causes from epileptiform discharges. The reading question is always twofold: is the slowing focal or diffuse, and is it continuous or intermittent?

Focal polymorphic delta

Irregular, mixed-frequency delta over one region that is continuous and does not react to state or stimulation suggests a structural lesion on that side. The physiology is a cortex that is deafferented or damaged: the normal faster rhythms are lost and what remains is slow, disorganized activity. The report should describe the slowing by location, frequency, continuity and reactivity, and the clinical correlation should say what it suggests. If the child has not had neuroimaging, recommending it is part of the read. Focal slowing that is intermittent and reactive is a weaker finding, and post-ictal slowing after a focal seizure is expected and transient, so the history and the timing of the study matter.

Diffuse slowing

Slowing of the whole background for age suggests encephalopathy, and the causes are metabolic, toxic, post-ictal or diffusely structural. "For age" is the operative phrase: a 4 Hz posterior rhythm is normal at 18 months and slow at 10 years, and the normal values by age in Module 4 are the reference. Diffuse slowing is a state of the whole brain, so the impression should not name a location, and the clinical correlation should list the causes the history makes likely rather than pick one. Sedating medication is a common cause in the hospital, which is why Module 2 requires the medications to be stated.

Why the distinction matters

Focal continuous slowing sends the child toward imaging. Diffuse slowing sends the clinician back to the metabolic panel, the medication list and the events of the preceding days. Placing each finding on the right side of the epileptiform versus non-epileptiform line, and on the right side of the focal versus diffuse line, is the skill this module's sign-off tests.

Three further non-epileptiform findings recur in pediatric records, and each is easy to misplace.

Intermittent rhythmic delta

Rhythmic, monomorphic delta that comes in runs is a different finding from polymorphic slowing, and its location changes its meaning. Occipital intermittent rhythmic delta activity (OIRDA) is a childhood pattern, often seen with absence epilepsy; when you find it, look through the rest of the record, and especially through hyperventilation, for generalized 3 Hz spike-wave. Frontal intermittent rhythmic delta activity (FIRDA) is usually a nonspecific sign of encephalopathy and should be read alongside the background rather than as a localizing finding. Both are rhythmic and both are intermittent, so the distinction from an evolving ictal pattern rests on the absence of evolution, a point Module 10 develops.

Asymmetry

A hemisphere with lower voltage than the other has two kinds of explanation, and the report should name both. The first is cortical dysfunction on the low side. The second is something between the cortex and the electrodes that attenuates the signal: a subdural collection, scalp swelling, or any fluid layer. After a head injury the second explanation is at least as likely as the first, and the recommendation follows from that. An asymmetry can also be a montage or electrode problem, so check impedances and the referential montage before you commit.

Breach rhythm

Over a skull defect, the bone that normally attenuates the signal is missing, so the activity beneath is higher in voltage and sharper than on the other side. That is a breach rhythm, and it is not epileptiform. Its sharp contour is the trap: faster rhythms with sharpened peaks over a craniotomy site fail the field and morphology criteria of Module 5 once the skull history is known. Ask about surgery before calling anything over a scar.

Placing each finding

OIRDA and FIRDA are rhythmic slowing; asymmetries are amplitude findings; breach is a technical consequence of anatomy. None is epileptiform, and describing them as such is one of the routes to misdiagnosing epilepsy that Module 5 warns about.