Kaiser Permanente study finds clues in the electronic medical record
A Kaiser Permanente analysis found that neurological side effects from the anti-seizure medication phenytoin are more common in patients with certain genetic variants. These same genetics were also associated with lower adherence to treatment, suggesting they may play a role in patients finding the right medication.
The study, published Aug. 30 in the journal Pharmacogenetics and Genomics, affirmed the previously identified association between specific genetic variants and phenytoin blood concentrations. Variants of the CYP2C9 gene can cause patients receiving the same dose of phenytoin to have blood levels of the drug that are too high or too low.
The researchers also examined electronic patient records and found that genetic variation can affect medical care in surprising ways; patients may not take their medications as prescribed and may have more difficulty finding the ideal drug and dose, leaving them vulnerable to uncontrolled seizures.
This suggests patients and doctors may be responding to patients’ pharmacogenetic variants even without knowing their specific genetics, said lead author Alison Fohner, PhD, adjunct researcher at the Kaiser Permanente Division of Research who was an informatics fellow sponsored by The Permanente Medical Group at the time of the study.
“This is the first big real-world view of how phenytoin pharmacogenetics is affecting patients in mainstream care,” said Fohner, who is also an assistant professor of epidemiology at the University of Washington. “This kind of research in health care systems like Kaiser Permanente is vital to moving the field forward.”
The study used genetic data collected from more than 100,000 Kaiser Permanente Northern California (KPNC) patients who agreed to fill out a survey and share a blood or saliva sample from which their genotype was mapped. The researchers found that patients with the lowest predicted activity of the CYP2C9 gene were more than twice as likely to experience neurologic side effects, such as slurred speech and loss of balance. These same patients were almost twice as likely to not take their medications as prescribed, potentially because of side effects.
The research does not offer specific clinical advice to physicians, as it would be unusual for them to have broad genetic information about their patients, said senior author Catherine Schaefer, PhD, a research scientist at the Kaiser Permanente Division of Research and executive director of the Research Program on Genes, Environment and Health. This study shows that having information about genetic variants in drug-metabolizing genes may be important to prescribing drugs that are effective, safe, and tolerated, she added.
“There are many factors physicians weigh to decide which medication to prescribe and at what dose,” Schaefer said. “Knowing whether the patient has a genetic variant that leads to higher or lower concentrations of the drug in their blood could be useful information for balancing the pros and cons of prescribing a particular medication at the standard dose.”
This study paints a valuable picture of what genetics can mean for patients undergoing treatment for scary and often-chronic conditions, Fohner said. The longer it takes to get these patients a therapy that works, the longer they are at risk for serious side effects or ongoing seizures, and the more disheartening their treatment odyssey. “The more information we have to get these patients the best medication and the best dose, the better,” she said. “Genetics can help us do that.”
The Clinical Pharmacogenetics Implementation Consortium (CPIC) currently lists 23 guidelines on gene-drug combinations with enough evidence to influence clinical practice. The CYP2C9/phenytoin combination is on that list, which is why the researchers chose it to analyze, Fohner said. CPIC is expected to update its recommendation for phenytoin dosing soon. As the largest study to-date on phenytoin pharmacogenetics, this study will likely play an important role in updating those guidelines, Fohner said.
This study used de-identified patient genetic data from the Resource for Genetic Epidemiology Research on Adult Health and Aging (GERA) cohort, which is now part of the Kaiser Permanente Research Bank (KPRB). The research bank has grown to contain data from more than 340,000 Kaiser Permanente members who have volunteered their health information, which is maintained securely and without their names. The KPRB continues to actively recruit volunteers among Kaiser Permanente members to provide samples and health information.
The original development of the GERA cohort in KPNC was funded by the Robert Wood Johnson Foundation, the Wayne and Gladys Valley Foundation, the Ellison Medical Foundation, Kaiser Permanente Community Benefit Programs, and a grant from the National Institutes of Health. This study was supported by the Kaiser Permanente Delivery Science Fellowship Program and a Kaiser Permanente Division of Research Healthcare Delivery and Policy grant.
Other authors included Vincent Liu, MD, MSc, Brian L. Lawson, PhD, Dilrini K. Ranatunga, and Khanh K. Thai, MS, from the Kaiser Permanente Division of Research; Aline T. Jelalian, MD, of the Kaiser Permanente Northern California Department of Neurology; Allan E. Rettie, PhD, of the University of Washington Department of Medicinal Chemistry; and Neil Risch, PhD, and Akinyemi Oni-Orisan, PharmD, PhD, of the University of California Institute of Human Genetics and University of California, San Francisco, Department of Clinical Pharmacy.