Interictal Epileptiform Patterns

Description

Focal interictal epileptiform discharges (IEDs) appear as a variety of waveforms but commonly have four characteristic features: A sharply contoured component, electronegativity on the cerebral surface, disruption of the surrounding background activity, and a field that extends beyond one electrode. IEDs do not always have all four features and not all activity with these features is epileptiform, but these features are a useful framework for considering the various IED waveforms.

The sharply contoured component of an IED can be either a spike or a sharp wave and may occur alone, as a collection of several sharply contoured waves in succession, or a sharply contoured wave incorporated in a complex with an after-going slow wave. Regardless of whether a sharply contoured wave occurs alone or with other waves, it is usually slightly asymmetric with a steeper rise to the peak than fall to the baseline, but this asymmetry may be evident only with close inspection (Fisch, 1999). A collection of several sharply contoured waves in succession is termed a polyspike and is less common than spikes and sharp waves. Polyspikes appear as a brief burst of paroxysmal fast activity (PFA) because the component spikes have durations consistent with fast frequency activity. Essentially, polyspikes differ from PFA in their duration and a burst with five or more spikes could be considered a brief burst of PFA, but there is no standard delineation. As discussed in Chapter 18, focal PFA may be as brief as 250 milliseconds. When a slow wave follows the polyspike, the overall complex is polyphasic; and when the slow wave follows a single sharply contoured wave, the overall complex may be triphasic or polyphasic. When the complex is triphasic, the first and third phases are respectively the sharply contoured wave and the slow wave, which are both negative, and the second phase is the deflection below the baseline between these two waves (Niedermeyer, 1999a). This triphasic waveform is the essential classic for epileptiform waves, but, as described, it is not requisite for IEDs. Some sharply contoured epileptiform waves have a low-amplitude, initial positive deflection and an overshoot of the baseline of the downslope, so the wave has three phases without an after-going slow wave. With an after-going slow wave, a fourth phase is present with much longer duration, and the overall sharp and slow wave complex is polyphasic. IEDs may be isolated spikes and sharp waves that are monophasic or diphasic.

Regardless of which elements are included, focal IEDs disrupt the background activity by having a waveform that is dissimilar to the components of the surrounding activity. This difference usually is based on the IED’s frequency components, which includes the faster activity that produces the sharp contour, but it is also due to a reduction in the number of superimposed frequencies, which gives IEDs a simpler, cleaner waveform than the more disorganized admixture of frequencies surrounding it. Although IEDs often also disrupt the background in their higher amplitude, they are not sufficiently likely to have high enough amplitude to use this feature as a reliable identifying characteristic. Focal IEDs typically have fields that include three or four electrodes, but may have fields that include much broader distributions. However, fields limited to one electrode are highly unlikely after infancy.

Rolandic (also called centrotemporal) IEDs are a type of IED that is distinctive because of certain features and an association with a specific epilepsy syndrome, benign epilepsy with centrotemporal spikes. This syndrome has also been called benign rolandic epilepsy (Lombroso, 1967). The most characteristic IED feature is
an exception to the polarity feature of the four listed IED features. Rolandic IEDs may include a surface electropositive component. This positive wave is synchronous to a negative wave and produces a discharge orientation that is tangential to the cerebral surface. The negative potential of other types of IEDs is the negative end of a dipole that is oriented perpendicularly to the cerebral surface within the crown of a gyrus. The positive end of such IEDs is not visible at the scalp because it is subcortical. In contrast, rolandic discharge’s presentation in a tangential orientation results in both ends of the dipole appearing across the scalp. The negative end is an IED over the central to midtemporal region and typically includes the C3/C4 to T3/T4 electrodes, and the positive end is an IED over the midfrontal region and typically includes the F3/F4 to Fz electrodes (Graf and Lischka, 1998; Legarda et al., 1994).

Although the common conceptualization of rolandic discharges is as single tangential dipole, and magnetoencephalography and EEG source analysis has supported this understanding with one dipole localized to within the precentral gyrus, EEG source analysis using different models indicate two dipoles that each are oriented with only one end at the scalp, and functional magnetic resonance imaging (fMRI) also identifies more than one active region (Archer et al., 2003; Pataraia et al., 2008; van der Meij et al., 2001; van der Meij et al., 1993a). Regardless of the number of sources, the rolandic dipole that is evident with scalp EEG is oriented along the rolandic fissure. Rolandic discharges have several other common features including more often occurring with a triphasic waveform than other IEDs, occurring in short runs of 1.5 to 3 Hz, a greater increase in the number of occurrences with drowsiness than occurs with other IEDs, and a decrease in frequency with hyperventilation (Degen and Degen, 1992; Fisch, 1999; Nicholl et al., 1998). They are unilateral in 70% of occurrences and bilaterally symmetric in the remainder (Zhao et al., 2007). When unilateral, they typically alternate between hemispheres with an overall shifting symmetry (Niedermeyer, 1999b). The negative end of the dipole usually has a higher amplitude and typically is between 130 and 200 μV (Kellaway, 2000). Although often called rolandic spikes, rolandic IEDs have an average duration of 74 to 88 milliseconds, so they technically are sharp waves (Kellaway, 2000; van der Meij et al., 1993b).

Unlike rolandic IEDs, occipital IEDs do not have particularly distinguishing waveform features but sometimes have an occurrence pattern related to visual stimulation. Especially in some focal occipital epilepsy syndromes, occipital IEDs occur most predominantly when visual fixation is not present. This is termed fixation-off sensitivity (FOS) and may be demonstrated by any interruption of fixation, as will occur with closure of the eyes, absence of ambient light, Frenzel lenses, and ganzfeld stimulation. FOS may occur with occipital epilepsies both with and without associated neurologic deficits and also sometimes occur with occipital epilepsies that produce generalized IEDs, including generalized FOS IEDs (Kurth et al., 2001). Like rolandic IEDs, occipital IEDs may be either unilateral or bilateral and often occur in repetitions. The repetitions are rhythmic and usually have a frequency between 2.5 and 4 per second (Talwar et al., 1992). Sleep has no consistent effect on occipital IEDs.

Positive IEDs occur with two conditions other than benign epilepsy with centrotemporal spikes, and both also have localization near the vertex. These conditions are the syndrome cherry red spot myoclonus and the development of periventricular leukomalacia, which commonly occurs after a germinal matrix hemorrhage in neonates. Positive discharges that occur as repetitions in the range of 10 to 20 Hz are typical of cherry red spot myoclonus (Engel et al., 1977).

Secondary bilateral synchrony (SBS) is an uncommon pattern of bilaterally synchronous spike and slow wave complexes that arise from a focal source (Tukel and Jasper, 1952). Thus, SBS is a focal pattern because it is not initially generalized. It becomes generalized with the spread of a focal discharge over a large, bilateral field while assuming a repeating spike and slow wave waveform. The initiation and termination usually has an asymmetry consistent with the hemisphere of onset, but truly symmetric onsets also occur. SBS’s spike and slow wave complexes characteristically have a frontal predominance, a frequency less than 3 Hz, and an irregular waveform. Focal IEDs also may occur independently at the SBS source without propagation into SBS, and the source most commonly is a frontal lobe and almost always is neocortical (Blume and Pillay, 1985). However, these focal IEDs must be distinguished from fragments of truly generalized IEDs, which may have phase reversals at F3 and F4 with a shifting asymmetry. Sometimes EEGs include both SBS and MISD. Overall, SBS is an uncommon pattern. A review of an EEG database with 10,410 EEGs recorded over 8 years identified 57 (0.5%) instances of SBS (Blume and Pillay, 1985).

Distinguishing Features

• Compared to Alpha Activity’s Wicket Spike or Mu Rhythm Fragment

Fragments of arciform activity, which includes both the wicket rhythm and mu rhythm, can resemble IEDs because of the sharply contoured component preceding the rounded component. The appearance is similar to a diphasic spike and slow wave complex. Furthermore, wicket rhythm fragments occur over the temporal region, which is a common site for IEDs, and mu rhythm fragments occur centrally and may be mistaken for rolandic IEDs. Distinguishing these fragments of normal activity from IEDs relies on identifying the more extended wicket or mu rhythms within the same portion of the EEG. Finding a normal rhythm of repeated waves that have a waveform similar to the wave in question is strong evidence against the isolated wave being an IED. Without the presence of a longer-lasting and
neighboring wicket rhythm or mu rhythm, a suspicious wave cannot be ascribed to be a fragment. Fragments rarely occur without the occurrence of the rhythms.

• Compared to Artifact

Several types of artifact appear similar to IEDs, and the ones that are most similar are electrocardiographic (ECG), electromyographic (EMG), electrode pop, and lateral rectus spike with or without eye flutter.

ECG artifact and an IED may similarly disrupt the EEG background activity. Moreover, ECG artifact usually is diphasic or triphasic with a phase that has a duration within the spike range. When ECG artifact occurs with a highly regular interval or can be compared to a channel specifically dedicated to an electrocardiogram, differentiating it from IEDs is straightforward. When a highly regular interval is present, careful scrutiny of the waveform and location is useful. ECG artifact almost always occurs in channels that include electrodes that are low on the head, especially ear electrodes. When a wave occurs in such a channel and has a perfectly unchanging waveform and period, it is likely to be ECG artifact. IEDs show greater variation between occurrences than ECG artifact even when they recur as the same wave type; that is, IEDs vary more in amplitude, duration, and contour than ECG artifact.

Only when EMG artifact occurs as individual muscle potentials is it similar to IEDs. This may occur as lateral rectus spikes or photomyogenic responses. Such potentials differ from IEDs by being more sharply contoured and having shorter durations. Unlike EMG artifact, spikes, which are the sharpest and briefest of IEDs, commonly are long enough in duration to separate the rising and falling phases. The rising and falling phases of EMG artifact usually are immediately adjacent to each other when viewed with usual EEG review time scales, so the wave appears as a vertical line. Like ECG artifact, EMG artifact also typically is unchanging across occurrences and tends to manifest within certain regions. Lateral rectus spikes are lateral frontal and photomyogenic responses are broad frontal.

Electrode pop is due to the sudden flow of current due to a stored voltage across the gap between one electrode and the subjacent skin. Therefore, only one electrode is involved, and a field that is so small is very rare for IEDs. The waveform of electrode pop also is different from spikes by having a much steeper rise and much slower fall. Electrode pops more closely resemble right triangles and IEDs, despite the slightly steeper rise than fall, more closely resemble isosceles triangles.

When the slow wave artifact of ocular flutter occurs in combination with the faster frequency artifact from eyelid movement, a compound wave results that can appear similar to a bifrontal spike and slow wave complex. The field for this artifact is at the frontal poles, which is an unusual location for a spike and slow wave complex because generalized spike and slow wave complexes are the most common bifrontal IEDs and their phase reversals usually are centered at the F3 and F4 electrodes. Nevertheless, this location is possible for focal IEDs. However, focal spike and slow waves typically occur at one frontal pole and do not have a bilateral symmetry. Spike appearance also may distinguish these waveforms. Muscle artifact is less consistent in its waveform across occurrences than epileptiform abnormality, which may vary slightly but typically is highly similar. Another differentiating feature depends upon behavioral state. IEDs usually occur in states beyond light drowsiness, which is the state for ocular flutter. Even when the IEDs occur only with drowsiness, they continue to occur into stage 2 non–rapid-eye movement (NREM) sleep. Eyelid flutter ceases with deepening sleep.

Brief myogenic potentials from the lateral rectus muscle and the slow wave produced by lateral gaze ocular artifact produces another compound wave that appears similar to an IED. The similarity is enhanced by the lateral rectus spike, which results from a single motor unit potential and is, therefore, relatively stereotyped across occurrences like the spike of an IED. The artifact also occurs in the anterior temporal region, which is a region that often produces focal IEDs. Distinguishing lateral rectus spikes from IEDs depends on the lateral rectus spike’s consistent low-amplitude, maximal field only at the F7 and F8 electrodes, and association with lateral gaze ocular artifact, which typically includes out-of-phase deflections at the two anterior temporal regions. IED spikes typically vary more in their amplitude and location, even if the variation is only one electrode distance. A shifting asymmetry between F7 and F8 is not helpful because some individuals with temporal lobe epilepsy have bilateral independent temporal IEDs.

• Compared to Benign Epileptiform Transients of Sleep

Benign epileptiform transients of sleep (BETS) are more likely to be mistaken for focal IEDs than other transients because of their epileptiform features, field centered over the temporal lobes, and occurrence during sleep, a state with greater IED frequency. Distinguishing them from IEDs is much easier when they recur. Compared to BETS, IEDs tend to vary more in their waveform with inconsistent amplitudes and durations. IEDs also are more likely to have a prominent after-going slow waves, with an amplitude equal to or greater than the spike or sharp wave. Furthermore, focal IEDs tend to have smaller fields, which usually are limited to an electrode and its nearest neighbors, and IEDs also may have co-localized but independently occurring focal slowing. BETS typically have broader fields that may extend across an entire lateral region. BETS are normal and not associated with focal slowing, so the presence of slowing indicates the region is abnormal and, therefore, is supportive of the discharges in question being IEDs. However, abnormal slowing does not exclude the possibility of BETS. The occurrence of the same transient during wakefulness also is a distinguishing feature and eliminates the
possibility of BETS. When a discharge consistent with a BETS occurs only once during a recording and no other suspicious finding is present, determining whether the transient is a BETS with absolute certainty may not be possible. In such instances, the interpreter may rely on a basic tenet of EEG interpretation that undercalling abnormality is preferable to misidentifying normal as abnormal. Based on this tenet, the interpreter may describe the transient within the body of the EEG clinical report, state that the EEG was normal in the conclusion, and include that the suspicious transient may have been a BETS as a comment after the conclusion.

• Compared to Beta Frequency Activity and Breach Effect

The amplitude and frequencies of beta frequency range activity may spontaneously change during an EEG, and the coinciding of a sudden rise in amplitude and increase in frequency may appear as a spike. This occurs more commonly when a breach effect is present because of the breach region’s greater amplitude and faster frequency components. Distinguishing IEDs from variations in the surrounding beta activity depends on identifying elements within the beta frequency background that have waveform similarities to the possible IED. If elements are similar and differ only in amplitude, then the wave in question is likely a variation in the background. The matter is complicated when a breach effect is present because IEDs may occur within a region of breach effect as a consequence of the trauma that led to the skull defect and because the trauma may have produced focal slowing within the breach region. This combination of abnormal slowing and increased fast activity favors the occurrence of activity that appears similar to spike and slow wave complexes. Identifying such activity as breach related depends on finding similar sharp and slow activity occurring independently within the breach region.

• Compared to Lambda Waves

The characteristic wide triangular waveform and occurrence only with visual exploration distinguishes lambda waves from IEDs. IEDs more often are sharper than lambda waves and usually are not affected by eye opening or closure. Fixation-off sensitivity is the exception to this. Another differentiating feature is the effect of sleep. IEDs usually become more frequent during sleep, which is a state during which lambda waves do not occur.

• Compared to Needle Spikes

The occurrence of occipital spikes in the context of a history of blindness or a large scotoma from early life leads to a question of whether the spikes are needle spikes related to the visual dysfunction or IEDs. Occipital location and increased frequency during sleep are not distinguishing features because both IEDs and needle spikes may have these features. Waveform provides the best means to differentiate these patterns; however, significant overlap exists between these two patterns during mid-childhood. The potential waveform differences are most prominent in early childhood and adolescence when needle spikes are sharper than typical IEDs and they lack an after-going slow wave (Niedermeyer, 1999a). When spikes occur without a slow wave, they may be identified as IEDs if they resemble independently occurring sharp waves or spikes that are followed by slow waves. Thus, similarity with a more clearly identifiable IED is helpful. When occipital spikes occur in the context of later-life onset blindness, they are much more likely to be associated with seizures and, therefore, are not true needle spikes (Fisch, 1999).

• Compared to Paroxysmal Fast Activity

Both focal and generalized IEDs may occur as polyspikes, which are a train of spikes that may or may not be followed by a slow wave. When an after-going slow wave is not present, polyspikes are highly similar to PFA, and the only difference is duration. Classic polyspikes, as sometimes occur with after-going slow waves, usually consist of only several spikes, so they rarely last longer than 250 milliseconds. Focal PFA often has a duration longer than 250 milliseconds. This duration distinction between polyspikes and PFA is arbitrary and may be artificial because the two patterns have similar clinical significance.

• Compared to Positive Occipital Sharp Transients of Sleep

The unvarying wide triangular waveform and consistent absence of an after-going slow wave distinguishes positive occipital sharp transients of sleep (POSTS) from IEDs. Even when they are sharp waves, IEDs usually are asymmetric and more sharply contoured and shorter duration than POSTS. Although IEDs occur more frequently in sleep, they typically occur in wakefulness as well. Therefore, identification of a similar wave during wakefulness may help identify the wave in question as an IED and not a POSTS.

• Compared to Vertex Sharp Transients

Although IEDs at the vertex are rare, they may occur, and parasagittal IEDs that occur within the region of off-center vertex sharp transients (VSTs) are not rare. Regardless of the exact localization, IEDs typically have a sharper contour and lower amplitude than VSTs. VSTs typically stand above the background activity as a characteristic feature and IEDs stand out because of their sharpness and
sometimes polyphasic waveform. When the IED is a rolandic spike, its negative field may be within the possible distribution of VSTs but it has the additional differentiating features of its characteristic waveform and tangential dipole with a frontal positive potential. Like the VST, classic rolandic spikes are triphasic; however, the first and third phases of VSTs are more symmetric in amplitude and duration. The occurrence of multiple discharges in a short time also can be differentiating. VSTs occur in runs of immediately adjacent discharges, which differs from successive rolandic spikes because of the brief periods of background activity that separate rolandic spikes.

Co-occurring Patterns

When IEDs reflect a focal epilepsy with structural abnormality, co-localized focal slowing or attenuation may be present because of the dysfunction produced by the lesion or malformation and this may be independent to the epileptogenic process. However focal epilepsies are not always associated with underlying structural abnormality and do not always have co-localized focal findings.

IEDs occur in wakefulness and both stages of sleep but with a considerable difference in number and some difference in waveform. The stages of NREM sleep have the greatest density (discharges/time) of IEDs and the density increases with the later stages of NREM. REM sleep has the lowest density of IEDs, and wakefulness is between the two other states (Malow et al., 1999). In contrast, IEDs are most focal in REM sleep and have the broadest fields in NREM sleep.

Clinical Significance

IEDs are a sign of epilepsy, but like all diagnostic signs they are not infallible because they have less than full specificity. However, the most central clinical question rarely is whether the patient has epilepsy in the sense of having experienced epileptic seizures, but whether the patient continues to have epilepsy and is going to have seizures in the future. Many large series have addressed this question, and the results are disparate. A meta-analysis of 19 studies that included 4,288 patients found the sensitivity of IEDs varied from 0.20 to 0.91 and the specificity varied from 0.13 to 0.99 (Gilbert et al., 2003). Differences in thresholds for abnormality among the readers accounted for 37% of the variance, and no other factor was found to explain the remainder. Part of the divergence of results may be related to different IED locations having different associations with epilepsy. In children, for example, focal IEDs over the temporal regions have a higher association with seizures than focal IEDs over the central regions. Overall, the meta-analysis found that the predictive value of an IED as a sign of epilepsy was greater when the EEG interpretation criteria were more restrictive. This supports the conventional wisdom of erring toward under-identifying epileptiform abnormality when the EEG is less than fully convincing.

Within healthy control populations, IEDs are present in 0% to 2.6% of individuals. Including 12 studies of healthy adults that have a combined population of 59,496 participants produces an overall positive rate of 0.2% (Jabbari et al., 2000; Trojaborg, 1992). However, this positive rate includes both true and false positives because many of the studies did not follow the participants over time to determine if seizures later developed. A retrospective study of 13,658 military recruits who had EEGs to screen eligibility for aircrew training reported spontaneous IEDs in the EEGs of 25 recruits (0.2%) and a photoparoxysmal response without spontaneous IEDs in another 44 (0.3%) (Gregory et al., 1993). Of the 38 recruits who were surveyed between 5 and 29 years later, only 1 definitely developed epilepsy, but another recruit died in a seizure-suspicious motor vehicle accident. In a separate series that included 2,947 asymptomatic air force cadets, 14 (0.5%) had EEGs with IEDs and none had a seizure over 10 to 15 years of follow-up (Everett and Akhavi, 1982). In contrast, a series that included 2,015 air force applicants reported IEDs in 38 (1.9%) and 1 applicant with a normal EEG had a seizure during the 15 years of follow-up (Ribeiro, 1994). Overall, IEDs are rare in asymptomatic adults and are likely to be false positives and not indicative of epilepsy later developing. These results may be different for children.

Hospital and EEG laboratory populations have higher rates of IEDs among patients without epilepsy than healthy control populations, but this is likely due to obtaining the series from a group that has sufficient neurologic abnormalities to warrant obtaining an EEG even if seizures have not occurred (Fisch, 1999). One EEG laboratory study of 521 patients over 10 years identified IEDs in the EEGs of 64 patients (12.3%) with 47 of these patients having acute or progressive cerebral disorders (Sam and So, 2001). A U.S. National Institutes of Health review of 6,497 patients without epilepsy identified 142 (2.2%) had IEDs (Zivin and Marsan, 1968). This group was followed clinically over time, and 20 (14.1%) subsequently developed epilepsy. Those with progressive neurologic diseases were most at risk.

Within the epilepsy population, approximately 90% have IEDs. However, the IEDs may be inconsistently present across one individual’s multiple EEGs. Overall, any one EEG of someone with epilepsy has about a 50% chance of demonstrating at least one IED (Goodin and Aminoff, 1984

Only gold members can continue reading. Log In or Register to continue