Subscribe to eTOC

Polygenic Risk Scores May Help Identify Patients at Higher Risk of Common Types of Epilepsy

Article In Brief

Using polygenic risk scores from a European-ancestry genome-wide association study, researchers quantified the genetic burden in patients with generalized and focal epilepsy.


(1) Polygenic scores for the two main classes of epilepsy (generalized and focal) were derived from the largest GWAS in epilepsy to date (International League Against Epilepsy Consortium on Complex Epilepsies, 2018). (2) PRS were calculated in patients with generalized or focal epilepsy and in population controls and (3) tested in their ability to identify significant differences of common variant burden among groups. (4) The UK and Vanderbilt biobanks were available to test the behavior of the PRS in individuals ascertained by ICD-10 codes for epilepsy, while the Biobank Japan was available to test the performance in a non-European population.

Genetic risk variants associated with common types of epilepsy were significantly higher in many groups of patients compared to controls, especially for generalized epilepsy, according to a new paper published in the October 14 online edition of Brain.

“This is the first proof that genetic factors play a measurable risk in common forms of epilepsy and the first study to show that a genetic test could become clinically useful for generalized and focal epilepsy,” lead investigator Dennis Lal, PhD, assistant professor and assistant staff at Cleveland Clinic Genomic Medicine Institute and Neurological Institute, told Neurology Today.

“This could open avenues for clinical decision-making in the future,” Dr. Lal said. “For an individual with new-onset epilepsy, it is critical to differentiate between a focal versus generalized epilepsy to inform the selection of the first-line anti-seizure drug.”

Dr. Lal referred to a study last year that found that individuals with epilepsy had about 30,000 genetic risk variants, which is higher than the general population.

“In our study, we wanted to assess how predictive those risk variants were in independent cohorts by developing polygenic risk scores (PRS), which combine the effect sizes of these variants into a single score that can stratify affected and healthy individuals.”


“This study changes the perception of the clinical utility of GWAS analyses in epilepsy—from being simply of biological interest to providing potential clinical utility.”—DR. SAM BERKOVIC

Study Design, Findings

To conduct their analysis, the researchers quantified the common genetic burden in patients with generalized epilepsy or focal epilepsy, using PRS from a European-ancestry genome-wide association study (GWAS) in generalized and focal epilepsy.

Two European cohorts with generalized or focal epilepsy were recruited through Epi25, an international multicenter epilepsy genetics research consortium that includes the Cleveland Clinic Epilepsy Center. An isolated Finnish population was also recruited from Epi25.

Patients were matched by ancestry with controls recruited from several sources including the Partners HealthCare Biobank and the FINRISK study. The researchers replicated the PRS findings in individuals assessed with ICD-10 codes for epilepsy in the UK Biobank and Vanderbilt Biobank. They used Biobank Japan to test the PRS performance in a non-European population.

The total cohort was 630,598 individuals and the study took two years to complete.

In the two European cohorts, the PRS scores for generalized epilepsy were significantly higher in 2,256 patients with generalized epilepsy cases than in 20,435 population controls (p=2.25x10-70).

Similarly, the PRS for focal epilepsy was significantly higher in 3,449 patients with focal epilepsy than in population controls (p = 5.74x 10-19).

When the researchers tested whether the PRS scores from the European population would work in an isolated Finnish population, only the PRS for generalized epilepsy was significantly higher in 112 Finnish patients with generalized epilepsy compared with 1,559 population controls (p= 3.11x10-4).

Another important finding was that PRS could distinguish the genetic variant loads between generalized epilepsy and focal epilepsy on a cohort level. The generalized epilepsy PRS was significantly higher in the European patients with generalized epilepsy than those with focal epilepsy (p=1.64 10-15).

The researchers observed the most striking results in patients with the highest PRS ranking in the generalized epilepsy cohorts. There was a significant 4.63-fold increase of patients with generalized epilepsy at the top 0.5 percent.

Dr. Lal commented that the four-fold enrichment is similar to the risk found in PRS studies of cardiovascular disorders, mental disorders, or diabetes.

“Interestingly, most of those PRSs are based on GWAS studies that are three to five times larger than ours,” he said. “Given the already strong predictive power of the generalized epilepsy PRS, we hope that with larger epilepsy GWAS studies, the epilepsy PRS will become even more predictive.”


“This is the first proof that genetic factors play a measurable risk in common forms of epilepsy and the first study to show that a genetic test could become clinically useful for generalized and focal epilepsy.”—DR. DENNIS LAL

Dr. Lal is optimistic that in the next ten years, researchers will learn a lot from polygenetic tests and that these types of assessments could become a clinically routine biomarker to test for the risk of epilepsies.

Expert Commentary

“PRS is not anticipated to be used as a widely applied screening test but it could potentially be useful when confronted with a patient reporting spells or paroxysmal events of uncertain type,” said Michael A. Rogawski, MD, PhD, professor of neurology at the University of California, Davis School of Medicine, who was not involved in the study.

“A higher PRS would raise the index of suspicion that the events may be epileptic and that anti-seizure drugs ought to be considered. In addition, because it can sometimes be difficult to distinguish between focal and generalized epilepsy, and the distinction has practical consequences for the drugs we use for treatment, the discovery that PRS can aid in making the distinction is an exciting development,” he said.

However, Dr. Rogawski added, “We are not there yet. I see this paper as a steppingstone to a future where we might use PRS as an aid in the work-up of common epilepsies that are not associated with specific causative mutations. We will still use gene panels, chromosomal microarray analysis, and whole exome sequencing for the occasional rare patient, such as those with an early onset childhood epilepsy or epileptic encephalopathy, likely due to a causative epilepsy mutation.”

Dr. Rogawski wasn't surprised that there was a higher association with PRS scores with generalized epilepsy rather than focal. “We generally think of generalized epilepsies as genetically based whereas focal epilepsies are often due to acquired neurological insults such as brain trauma, stroke, or tumors possibly in conjunction with genetic factors.”

The study had several strengths including replication of the findings in separate cohorts of patients and the use of a control group of patients with type 2 diabetes in which the PRS scores were no different than in the general population.

A limitation of the study was that it was conducted retrospectively at the population level and “requires prospective replication before it can be considered for clinical use,” said Dr. Rogawski.

“This study suggests that combinations of common variants increase the risk of epilepsy. The findings don't direct us to a novel biological pathway in epilepsy but rather provides a possible biomarker of how likely a person is to develop certain types of epilepsy,” said Erin L. Heinzen, PhD, Herbert Irving assistant professor of pathology and cell biology in the Institute for Genomic Medicine at Columbia University.

Dr. Heinzen noted that a number of previously published heritability studies clearly show that epilepsy is highly genetic. “This study suggests that some of that genetic risk can be predicted by looking at the genotypes at multiple common variants.”

Dr. Heinzen was surprised by the extent to which combinations of common variants predict the risk of epilepsy. “I would not have expected the pattern to be so clear in this number of patients.”

“We have come a long way in epilepsy genetics but there is still a lot we do not know,” she said. “Many variants in many genes cause epilepsy. This means we have to study very large numbers of people to be able to tell the difference between variants that are benign and those that may cause the disease.”

But at least one epilepsy researcher thought the findings could be clinically useful. “This study changes the perception of the clinical utility of GWAS analyses in epilepsy—from being simply of biological interest to providing potential clinical utility,” Sam Berkovic, MD, laureate professor and director of the Epilepsy Research Centre at the University of Melbourne (Austin Health) in Australia said by email.

Dr. Berkovic was surprised by the poor performance of PRS on Biobank samples, “with crude (ICD codes) classification, which emphasizes the importance of accurate classification and epileptology.”

Dr. Berkovic explained that the original PRS scores were developed from the International League Against Epilepsy Consortium on Complex Epilepsies, which contains subjects of largely Western and Central European origin.

“The replication of the findings was good in cohorts of similar origin (Epi25, Cleveland) but not in those of different origin (Finland, Japan). The issue of ethnic matching will become increasingly important for the equitable deployment of these measures.”

Dr. Lal noted, however, that in the past few years, the Epi25 consortium has focused more on recruiting samples from non-European cohorts.


Drs. Lal, Rogawski, Heinzen, and Berkovic had no competing interests.

Link Up for More Information

• Leu C, Stevelink R, Smith AW, et al. Polygenic burden in focal and generalized epilepsies Brain 2019; Epub 2019 Oct 14.
    • Abou-Khalil B, Auce P, Avbersek A, et al; for the International League Against Epilepsy Consortium on Complex Epilepsies. Genome-wide mega-analysis identifies 16 loci and highlights diverse biological mechanisms in the common epilepsies Nat Commun 2018;9(1):5269.
      • Feng Y-CA, Howrigan DP, Abbott LE, et al; for the Epi25 Collaborative. Ultra-rare genetic variation in the epilepsies: A whole-exome sequencing study of 17,606 Individuals Am J Hum Genet 2019;105(2):267–282.