NEW YORK — Researchers from the Institute for Molecular Medicine in Finland, or FIMM, presented data on new studies related to the genetics of epilepsy and drug treatment adherence during separate talks at the European Society of Human Genetics meeting, held online this week.
In the first study, the researchers proposed that polygenic risk scores could be used to help diagnose epilepsy, and to distinguish epileptic individuals experiencing seizures from those whose seizures have another cause. In the second study, they described a genome-wide association study to search for genetic factors underlying adherence to prescribed medications.
Both studies relied on data collected via the FinnGen project, an international effort to collect genotype and health registry data on 500,000 people in Finland by 2023. The University of Helsinki's FIMM is coordinating the project, which commenced in 2017 and according to its website currently has 356,000 individuals with genotype and health registry data available. The project has relied on Thermo Fisher Scientific's Axiom array platform to genotype participants.
Researcher Henrike Heyne provided an overview of the work on epilepsy during a presentation on Sunday. While common, epilepsy remains difficult to diagnose, and about a quarter of cases are misdiagnosed at first, she said. Heyne and colleagues aimed to determine if polygenic risk scores could be used to better diagnose patients at first presentation. At the time they embarked on the study, FinnGen had genotype and digital healthcare data available on about 269,000 Finns. The health registry data also contains information on antiepileptic drug purchases.
The researchers relied on genetic loci discovered through recent International League Against Epilepsy genome-wide association studies to calculate their risk scores. When they screened roughly 10,000 people who had been identified as having epilepsy versus the rest of the repository, they found their scores were elevated in epileptic individuals versus healthy controls. The scores were significantly higher in subjects with adolescent myoclonic epilepsy, but Heyne also noted this group represented the largest proportion of cases in the ILAE consortium's cohort.
Heyne also said that individuals presenting with an unclear seizure but high scores were more likely to progress to epilepsy during their lifetime, particularly in those below the age of 40.
"This is true throughout their lifetime, but the risk is much higher after a seizure event," said Heyne. "This really shows polygenic risk scores as a potential biomarker in yet another disease."
In a separate presentation on Monday, Matta Cordioli, a Ph.D. student at FIMM, described efforts to use FinnGen data to obtain a genetic understanding of patients' lack of adherence to prescribed medication. He noted that only about half of patients with chronic diseases adhere to their medication, and that poor adherence is known to increase mortality. He also said that it accounts for a significant amount of drug acquisition costs in healthcare systems. The aim of his study was to investigate genetic risk factors for adherence, as well as correlated health and behavioral traits.
To do so, Cordioli and colleagues tapped into the FinnGen data, particularly FinnGen's drug purchase register, which has data on medication purchases dating back to 1995. They then calculated adherence by dividing quantities of purchases by the number of days between purchases, factoring in the prescribed daily dose. They found adherence tended to be lower in general medications, such as statins and blood pressure medication, but was higher for more specific medications for a more severe condition, such as breast cancer.
Next, they ran an association for each of the medications, but did not turn up specific loci related to adherence. However, they did replicate their approach and findings in UK Biobank data for statins. The UK Biobank data also had more comprehensive data available on drug adherence, he noted. Cordioli also uncovered a positive correlation between adherence, such as educational achievement, meaning those with a genetic predisposition for educational achievement were more adherent. Meantime, those with a genetic predisposition for risk-taking were less adherent.
"Genetics data represents a powerful and valuable tool, because it allows us to identify which traits are correlated and possibly drive adherence," Cordioli said during his talk.
Cordioli and colleagues at FIMM plan to continue the research by replicating the association studies' results for other medications in UK Biobank data, as well as to undertake Mendelian randomization to determine what other genetic traits might be risk factors for nonadherence.