New framework improves genetic prediction of drug response and side effects

New framework improves genetic prediction of drug response and side effects

A UCLA study has described a new framework that researchers say would improve the prediction of genetics in determining how well a patient would respond to commonly prescribed medications as well as the severity of any side effects.

Published in JournalCell genomicsThe study found that data from large libraries of sequenced human genomes and other biological data, called biobanks, can provide new insights into the genetic architecture of responses to widely used drugs.

Study lead author and UCLA bioinformatics Ph.D. Candidate Michal Sadowski said the most common method for analyzing the genetics of drug response is pharmacogenomic studies on genotyped participants in randomized controlled trials. However, these studies have small numbers of participants, are expensive and, depending on the drug, sometimes aren't even feasible, Sadowski said.

Genetic data in biobanks offers several advantages. These libraries can also be analyzed at lower cost along with sequenced genetic data from large populations, including people on and off certain medications. While biobank data cannot replace randomized controlled trials, they can unlock new information that will improve future studies and advance the evolving field of using genetics to predict treatment outcomes, Sadowski said.

We hope that in the future this will allow clinicians and patients to weigh the benefits and risks of treatment in a more personalized way and make more informed and timely decisions to engage in treatment. We hypothesize that analysis of biobank data will be most useful for widely prescribed medications. “

Michal Sadowski, UCLA Bioinformatics Ph.D. candidate

The study, overseen by UCLA neurology, computational medicine and human genetics professor Noah Zaitlen and Uchicago genetic medicine professor Andy Dahl, used genetic data from more than 342,000 people in the UK biobank. Researchers analyzed how their genetic makeup affected their response to four of the world's most commonly prescribed drugs: statins for high cholesterol, metformin for type 2 diabetes, warfarin for blood clots, and methotrexate for autoimmune diseases and cancer.

Sadowski and his colleagues sought to determine how much genetic variation played in the variability in response to these drugs and which specific genes were involved.

“If a lot can be explained by genetics, then genetics can be used as a good predictor of how you will respond to the drug,” Sadowski said. "Say you want to take statins because of your cholesterol levels. Your doctor can look at your genetics and give you an opinion, including potential side effects. If you have predictors that say you'll respond well and there's a small chance that you'll want to have side effects, it's probably a good choice to start treatment."

For example, the study identified 156 genes that may drive variation in the effects of statins on LDL cholesterol levels. Overall, about 9% of the variation in drug response was attributed to genetic differences from person to person.

Additionally, the study found that gene-drug interactions may also influence the predictive power of a genetic risk tool known as a polygenic score. Polygenic scores are used to summarize the combined effect of a large number of genetic variants to estimate a person's risk of developing a particular trait or disease. The models used to generate these assessments must be trained on genetic data from large populations of people and have important limitations, including relying primarily on data from people of European ancestry.

Sadowski's study found that the accuracy of polygenic polygenic assessments in clinical contexts was likely subpar because they included data from both statin and non-statin users.

"We were surprised to see that polygenic predictors produced such significant differences in performance between people who are on drugs," Sadowski said. “We were also surprised by the magnitude of drug-specific heritability for some outcomes.

The study has several limitations, with future work needed to improve the reliability of inference from biobank observational data and to understand the limitations of genetic risk prediction.

Sources: