Skip to main content

A pharmacogenomic exploration of adverse drug reactions in epilepsy

Periodic Reporting for period 1 - PGXOME (A pharmacogenomic exploration of adverse drug reactions in epilepsy)

Reporting period: 2017-09-01 to 2018-08-31

The aim of this project is to enhance our understanding of the genetic factors that underlie the risk for adverse drug reactions (ADRs) to anti-epileptic drugs (AEDs). Epilepsy is a common disease characterized by recurrent seizures affecting 50 million people worldwide. For most people with epilepsy, long term treatment with AEDs are necessary and 40% do not respond to the first drug, leading to a lifelong odyssey towards effective treatment. Further, for each AED a person is treated with there is a considerable risk that they will have a serious and intolerable side-effect. ADRs can be minimized by careful titration towards therapeutic dose however some rare ADRs, particularly the various skin rash and hypersensitivity syndromes, have been shown to have a genetic component. It is hoped that the results of this project will provide clinically useful genetic markers of ADRs that can be used by clinicians to screen people with epilepsy and help prescribe optimal, safe treatment targeted to the individual thereby improving medication safety in epilepsy.
In this project we explored the effects of common and rare genetic variation on ADRs attributed to epilepsy treatment in a European cohort of people with epilepsy.

We investigated the contribution of common variation to ADRs using a combination of GWAS and polyenic risk score (PRS) analysis. We set out to determine whether an individual subjects polygenic risk for a given trait, calculated from aggregating the effect sizes of many genetic variants, could predict the likelihood of developing an ADR. Specifically we tested a) whether a PRS for schizophrenia (SCZ-PRS) was associated with mild to psychotic behaviroual disorders in people who are treated with levetiracetam, and b) whether a PRS for body-mass-index (BMI-PRS) was associated with weight gain observed in people who are treated with sodium valproate. Our results, while preliminary, show a gender-specific effect for SCZ-PRS on levetiracetam-induced behavioural disorders and an age-related effect for BMI-PRS on valproate-induced weight gain.

We also investigated the contribution of rare variation to ADRs using exome sequence data. By using available exome sequences of cases and controls in the EpiPGX Consortium, we tested whether there was a burden or rare, damaging mutations in specific drug target genes that may explain a) severe skin rashes (such as Stevens Johnson syndrome), b) mild to psychotic behavioural disorders and c) hyponatraemia. Our results indicate that there is no significant genetic burden on specific drug target genes for the ADRs in question. However the results may be confounded by sample size.

Meanwhile, we also contributed towards additional analyses of epilepsy genomics in collaboration with the ILAE-CCE, EPI25 and EuroEPINOMICS, that have furthered the understanding of the underlying mechanisms of focal and generalised epilepsies. These accumulated findings are crucial to understand the genetic contribution to ADRs in epilepsy and to aid clinicians in making safer, optimal treatment decisions on behalf of their patients.

We have showcased the work of this study, through poster and oral presentations, at international conferences and meetings in the Netherlands, Austria, Belgium and Columbia. Manuscripts have been written for analyses that have been fully completed, while we continue to pursue collaboration with others in order to enhance and replicate preliminary findings from this project.
However, despite the relative success of GWAS to identify genetic predictors of hypersensitivity there are few other predictors of drug-induced ADRs. This project proposed to move beyond traditional GWAS into new territory for epilepsy pharmacogenomics by exploring the contribution of polygenic risk and rare variants in a two-pronged study design.
Genome-wide association studies have made strides in identifying common variation associated with disease. Polygenic risk scores (PRS) combine effect sizes of thousands of genetic variants into a single aggregated value. Assessing the distribution of PRS values for a given trait in a population has been shown to have immense power in distinguishing cases versus controls for disorders such as schizophrenia, hypertension, and breast cancer as well as anthropological traits such as height and BMI. Despite the known success-stories of the PRS, their predictive power and utility has not been fully investigated in the context of ADRs common to epilepsy. This project has attempted to address this question and our preliminary findings are important because they suggest a useful role for PRS in identifying ‘at-risk’ subjects for unnecessary adverse events.
Looking ahead, we expect the analyses described in this action to continue beyond the lifetime of the reporting period. The researchers are actively extending invitations for collaboration to others in the epilepsy genomics community and much of the work performed to date is currently undergoing peer-review for publication.
In summary, the results of this project support further research into understanding the genetic contribution, be it polygenic or driven by single variants of large effect, to ADRs in epilepsy. If our preliminary findings can successfully be replicated in larger cohorts, then it will herald a new and exciting avenue in epilepsy pharmacogenomics. We endeavor to advance genetic testing in the clinical setting, expand the array of genetic tests available to aid clinicians, thereby reducing rates of drug discontinuation due to adverse events and ultimately improving patient safety.
PGXOME poster at Vienna ECEE