LOS ANGELES -- Low racial diversity has limited genome-wide association studies (GWAS) in epilepsy, researchers found, with implications for prognostication and management decisions.
Across all published GWAS on epilepsy, 71.5% of tested individuals were of European ancestry, 14.3% East Asian, 0.5% African, and 13.7% without reported ancestry.
None of the participants were of South Asian, Native American, or Hispanic ancestry, Chethan Rao, DO, of the University of Maryland in Baltimore, reported at the American Epilepsy Society annual meeting.
Diversity levels were even lower among people with epilepsy in these studies, as only about 8% were non-European (6.4% East Asian and 1.5% African ancestry).
That appeared to be meaningful in terms of findings.
The more diverse studies found more single nucleotide polymorphisms (SNPs) that linked to epilepsy conditions. Data that included only participants of European ancestry comprised 66% of all studies but contributed only 21% of the significant SNPs. The 16% of studies that included two or more ancestries, however, contributed to 58% of the significant SNPs identified.
"So because so many of the studies so far that have been done were European ancestral patients, the more non-European patients we involve in these studies, the more we're realizing that there are other statistically significant SNPs and clinically significant data that we can use for management of patients and diagnosis," Rao told ľֱ.
GWAS results are incorporated into polygenic risk scores and used for prediction of adverse events, he noted.
For example, "there are some pharmacogenomic studies that rely on genetic data to be able to say is this person likely to develop a severe side effect from a sodium channel blocker," Rao said. "Some of the Asian ancestry patients are more likely to have severe rashes like Stevens-Johnson syndrome from sodium channel blocker medications. So using genome wide association studies and large amounts of genetic data can help with knowing who are those patients who should be."
The findings don't mean the data gathered thus far haven't been useful, but they do have to be taken in context, commented David Labiner, MD, of the University of Arizona in Tuscon. "It's important to have diverse populations to make the results generalizable."
And it's certainly not a challenge isolated to GWAS studies in epilepsy. NIH and FDA have been pushing for diverse populations in clinical trials as well.
"The problem is recognized," Labiner said. "It's just a matter of how do you make progress in improving it."
Historical wrongs have made some groups understandably sensitive to their genetic material being collected, he noted.
"There's a sensitivity, particularly among Native Americans, about genetic studies because things have been done in that population, at least in my state, without their consent and then the data were released without the consent of the tribe that was studied," he acknowledged. "We can't ignore the history when we have difficulty recruiting those populations to participate."
The study examined all published GWAS on epilepsy included in the National Human Genome Research Institute-European Bioinformatics Institute as of August 2024. These included a total of 293,725 epilepsy cases and 3.2 million controls. Notably, 79% of all epilepsy GWAS participants were recruited from the U.K.
The studies spanned genetic/idiopathic generalized epilepsy (childhood absence epilepsy, juvenile myoclonic epilepsy, and others) and focal epilepsy (self-limited epilepsy with centrotemporal spikes, focal epilepsy with hippocampal sclerosis, and others). The researchers screened the cohorts for sample overlap and removed duplicates.
Ancestries were categorized per the into African, admixed American (Native American or Hispanic), East Asian, South Asian, and European categories.
Disclosures
The researchers disclosed no relevant relationships with industry.
Labiner reported no relevant relationships.
Primary Source
American Epilepsy Society
Rao CK "Lack of diversity in epilepsy genome-wide association studies hampers discovery of single-nucleotide polymorphisms" AES 2024; Abstract 2.524.