Spontaneous intermittent generalized attenuations (SIGAs) are defined as a transient decrease in amplitude of electroencephalography (EEG) activity in response to a physiologic process, external stimuli, or as a result of a pathologic condition^{Reference Noachtar, Binnie, Ebersole, Mauguiere, Sakamoto and Westmoreland1}. However, these are infrequently described and are not included in standard EEG terminology^{Reference Hirsch, LaRoche and Gaspard2}. These EEG attenuations have been previously described in children, mainly as an epileptic phenomenon in syndromic epilepsies (e.g. Lennox–Gastaut syndrome, West syndrome) also known as electrodecremental events (EDE)^{Reference Wong-Kisiel and Nickels3}. However, the significance of SIGAs has not been identified in adults^{Reference Stern4}. Diffuse EEG attenuations in adults have been noted by electroencephalographers without definite clinical associations or formal study. Anecdotal evidence suggests there may be an association with the triphasic waves, generally seen in the setting of metabolic encephalopathy^{Reference Kaplan5}.

We have observed that severe metabolic encephalopathies appear to be associated with SIGAs in hospitalized patients. We, therefore, hypothesize that EEG attenuations in adults may be of prognostic value in the assessment of metabolic encephalopathies. We seek to further investigate their relationship to triphasic waves and other markers of metabolic encephalopathy, and to characterize underlying etiologies.

This is a prospective case series. We reviewed EEG and clinical features from 22 consecutive adult patients who underwent EEG as part of their diagnostic evaluation and demonstrated SIGAs. SIGAs were defined as transient, infrequent, and intermittently generalized decreases in EEG voltages lasting at least 500 msec, with at least 50% decrease from baseline amplitude, and were felt to clearly be discrepant from background activity. Patients taking anesthetic agents and those with preexisting epilepsy were excluded. All pertinent medications are listed in Table 1. Demographic information, baseline and discharge modified Rankin Scale (mRS), diagnosis at discharge, EEG altering medications, intensive care unit (ICU) admissions, relevant imaging, mental status, discharge destination, pertinent lab values, and metabolic abnormalities were noted. Each patient’s EEG was then re-evaluated by an EEG-certified neurologist. The neurologist began by identifying if attenuations were indeed present, duration of the attenuation (msec), background amplitude and frequency, presence of spikes and/or triphasic waves, and temporal relationship to these discharges. Ethics approval was obtained from our local biomedical research ethics board.

Table 1: Relevant clinical information

The assessment includes 22 patients, 7 males, and 15 females. Examples of EEG attenuations from Patient 16 and Patient 18 (with triphasic waves) are shown in Figures 1 and 2, respectively. Mean patient age was 73.7 (SD = 11.0) at admission. Mean mRS score at baseline was 1.4 (SD = 1.4), and was 4.8 (SD = 1.5) at discharge (Table 1). Twelve of the 21 patients (54.5%) died during their course in hospital as a result of their illness. Four patients (18.2%) did not have a change in mRS score from baseline to discharge. The mean discharge mRS was 4.8 (SD = 1.5). Eighteen patients were drowsy at the time of their EEG, 2 were in a coma, and the other 2 were described as nonspecifically encephalopathic. Twelve patients (54.5%) were admitted to the ICU or coronary care unit during their time in hospital. Only three patients were discharged home. Other clinical information is further described in Table 1. Mean SIGA duration was 1300 msec (SD = 403 msec). Fourteen of the 22 (63.6%) patients showed IGAs immediately following a triphasic wave. All patients except one showed a slowing of background activity. Epileptiform activity was seen in four patients. Other EEG findings are detailed in Table 2. Underlying diagnoses were variable and included both primary brain pathologies (stroke, hemorrhage, seizures; n = 14) and systemic illness (metabolic encephalopathies n = 8). The primary brain pathology group had a mean discharge mRS of 5.1 (SD = 1.5) and the systemic illness group had a mean discharge mRS of 4.4 (SD = 1.8).

Figure 1: EEG with attenuations and tiphasic waves from Patient 16 post-cardiac arrest.

Figure 2: EEG with attenuations and triphasic waves from Patient 18, who had an intracerebral hemorrhage.

Table 2: EEG data

As EEG attenuations have previously not been well described in adults, we sought to understand their clinical associations and prognostic significance. The most striking finding from this review is the strong association with mortality and persistent disability at discharge, with a majority of survivors discharged to some form of supportive care rather than home. While these outcomes are difficult to distinguish from the underlying disease prognosis, the association appears to support this EEG finding as a marker of severe brain dysfunction.

The underlying pathophysiology of SIGAs remains unclear. Generalized attenuation is likely associated with transitory dysfunction in the setting of reduced cortical synchronicity^{Reference Andraus and Alves-Leon6}. The duration of the attenuations may support potassium channel dysfunction, secondary to a burst release of calcium^{Reference Amzica7}. The length of the attenuations was mostly around the 1300 msec mark, with variance as short as 600 msec and as long as 2200 msec. Length of attenuation did not seem to correlate to the prognostic outcome. In children, it is hypothesized that diffuse attenuations are triggered by the activation of the arousal circuit in the setting of diffuse brain dysfunction^{Reference Losey and Ng8}. SIGAs often followed a triphasic waveform, and most patients showed triphasic waves at some point during the recording. The apparent association with triphasic waveforms supports the idea of an underlying metabolic encephalopathy, and triphasics with attenuations may represent a more advanced stage of encephalopathy.

Limitations of this study include a lack of control group and small sample size. The aim of the current study was to describe this EEG finding and the associated clinical and electrographic features, as well as outcomes. Causation and independent prognostic value cannot yet be determined from this study.

From this data, it seems probable that SIGAs is a meaningful EEG finding that should be noted and reported by electroencephalographers, and that they are likely associated with encephalopathies and poor outcomes.