Polygenic risk scores as predictors of topiramate's effect in treating alcohol use disorder
In brief
Drinking patterns led topiramate response estimates; a genetic score added information
In trial data from 278 people with alcohol use disorder, a model using average drinks per day, heavy-drinking days and a genetic score for time until relapse captured 30% of variation in estimated topiramate response. Drinking measures were the strongest predictors, and the model classified 80% as likely responders; whether it predicts benefit in new patients remains unknown.
- Journal
- Alcohol and alcoholism (Oxford, Oxfordshire) (Q2)
- Published
- 20 September 2026
- Study design
- Randomized controlled trial
- Evidence level
- Level 1, High (CEBM 1b)
- Authors
- Albert J Arias, Carole Siegel, Eugene Laska, Joseph Wanderling, Brendan Ho, Richard Feinn, et al.
- PMID
- 42822858
- DOI
- 10.1093/alcalc/agag064
Why clinicians should know about it
- Picked for Health Informatics (top studies of the week, 4 October 2026): Machine‑learning PRS model predicts topiramate response
Abstract
INTRODUCTION: Topiramate is efficacious for treating alcohol use disorder (AUD), but treatment response varies considerably. To advance precision treatment, we used machine-learning methods integrating clinical trial and genetic data to estimate expected response to topiramate. METHODS: Random forest (RF) regression models were used to estimate expected topiramate response using randomized clinical trial data from 278 participants and 23 pre-treatment features: four alcohol polygenic risk scores (PRS), nine alcohol drinking measures, and 10 clinical and sociodemographic variables. RF importance scores quantified each feature's contribution to predictive performance. Likely Responders (LRs) were identified using quintiles of predicted treatment response and counterfactual placebo response, with LRs defined as quintiles in which randomized trial data demonstrated topiramate superiority over placebo. Variable Selection Using Random Forests (VSURF) was used to eliminate less informative and redundant predictors. RESULTS: Two pre-treatment drinking measures consistently had the highest importance scores across all VSURF levels. The final parsimonious model retained three predictors: average drinks per day, percent heavy drinking days, and the PRS for "Time Until Relapse," with bias-corrected R2 = 0.30. LRs comprised the top four predicted response quintiles (80% of the sample) across models. Compared with unlikely responders, LRs had lower baseline drinking severity and higher "Time Until Relapse" PRS. CONCLUSIONS: The "Time Until Relapse" PRS contributed meaningfully to estimating expected response, LR stratification, and observed topiramate treatment effects, although pre-treatment drinking measures remained the strongest predictors. These findings support incorporation of pharmacogenomic markers into machine-learning models to advance precision prescribing for AUD.
Abstract as published, via PubMed.
For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.