Conference Coverage

Noninvasive EEG may speed diagnosis of West syndrome



Several EEG abnormalities that are readily measured noninvasively using scalp electrodes show considerable promise as epileptogenicity biomarkers of particular interest for speedier diagnosis of West syndrome, Hiroki Nariai, MD, reported at the International Epilepsy Congress.

Dr. Hiroki Nariai, pediatric neurologist at UCLA Bruce Jancin/MDedge News

Dr. Hiroki Nariai

Among the most promising of these potential EEG biomarkers for development of epilepsy are ictal or interictal high-frequency oscillations (HFOs) at 80 Hz or more, along with cross-frequency coupling of HFOs and delta wave activity, according to Dr. Nariai, a pediatric neurologist at the University of California, Los Angeles.

West syndrome, the most common epileptic encephalopathy during the first 2 years of life, has diverse etiologies. For example, in 250 infants with West syndrome enrolled in the United Kingdom National Infantile Spasms Consortium, a cause was identified in 64%. The etiology was genetic in 14% of subjects, a structural-congenital anomaly in 11%, tuberous sclerosis – a genetic-structural abnormality – in 10%, stroke in 22%, a metabolic defect in 5%, and infection in 2% (Epilepsia. 2015 Apr;56[4]:617-25).

West syndrome is rare, with an estimated prevalence of roughly 1 per 6,000 live births, but the associated mortality is high: 31% on average. And West syndrome often brings severe neurodevelopmental morbidity, with normal or near-normal intelligence present in only 25% of survivors.

An intensive search is on for an objective, reliable diagnostic biomarker – be it electroencephalographic, biochemical, or perhaps a neuroimaging finding – because the clinical diagnosis of West syndrome is highly subjective. It relies upon the triad of epileptic spasms, developmental regression or psychomotor delay, and hypsarrhythmia, which is a chaotic, disorganized, patternless form of brain electrical activity. And while that description makes hypsarrhythmia sound as if it should be easily recognizable, in fact that’s often not the case: Interrater reliability was poor in a study of six pediatric EEG experts at four centers who viewed 5-minute-long EEG samples obtained from 22 patients with infantile spasms (Epilepsia. 2015 Jan;56[1]:77-81), Dr. Nariai noted at the congress, sponsored by the International League Against Epilepsy.

“The clinical trial is maybe not so useful,” he observed.

The hunt for a reliable biomarker is further fueled by evidence that early diagnosis and treatment of West syndrome and other etiologies of infantile spasms makes a real difference. Indeed, investigators found in the United Kingdom Infantile Spasms Study that increasing lag time from onset of spasms to initiation of treatment was associated in stepwise fashion with significantly lower IQ at 4 years of age. While infants who started treatment within 7 days of onset of the seizure disorder had a mean IQ of 76.2 at age 4 years, those with an 8- to 14-day lag time between symptom onset and treatment averaged an additional 3.9-point decrement in IQ. A 15- to 30-day delay was associated with a 7.8-point reduction in IQ, compared with the reference group, while the decrease in IQ averaged 11.7 points in infants with a 1- to 2-month lag time and 15.6 points in those with a lag time of more than 2 months (Epilepsia. 2011 Jul;52[7]:1359-64).

At the start of the decade, Dr. Nariai and other investigators demonstrated that pathologic HFOs recorded during invasive EEG monitoring in conjunction with epilepsy surgery served as a reliable biomarker of epilepsy. While this was an important observation, a biomarker obtained through invasive monitoring during brain surgery clearly has very limited clinical applicability. But more recently, Dr. Nariai and his coinvestigators in the Tuberous Sclerosis Complex Autism Center of Excellence Network reported that noninvasive detection of interictal HFO fast ripples in the 250-500 Hz range via scalp EEG showed promise as a biomarker of epilepsy. Sensitivity of this far more practical approach to the detection of fast ripples was excellent, whether analyzed visually or by automatic detector (Clin Neurophysiol. 2018 Jul;129[7]:1458-66).

Moreover, in a recent, not-yet published study that Dr. Nariai and coworkers conducted in 24 infants with active epileptic spasms and 6 controls, noninvasive objective measurement of HFO rate using scalp EEG had an 83% sensitivity and 100% specificity for active epileptic spasms, while the modulation index of HFO and delta coupling in the 3-4 Hz range showed 74% sensitivity and 86% specificity.

If future studies validate the utility of detection of HFOs above a defined threshold or another noninvasively obtained EEG biomarker for diagnosis of epilepsy, the same strategy would presumably also be applicable for monitoring response to antiepileptic therapies, thereby eliminating the traditional trial-and-error approach to treatment. This would be a particularly important application in patients with West syndrome, where it’s believed that the electrical activity itself is contributing to the progressive – and often rapid – loss of cognitive function and behavioral disturbances. Thus, unlike in most other forms of epilepsy, the treatment goal isn’t merely to suppress the seizures, but also to achieve disease modification by eliminating the underlying subclinical EEG abnormalities, he explained.

A reliable biomarker would also be a boon in selecting the best participants for clinical trials of new antiseizure therapies.

Dr. Nariai reported having no financial conflicts regarding his presentation. His work is funded by research foundations and the National Institutes of Health.

Next Article: