Computational Drug Repurposing Approaches Can Help Identify New Medications for Opioid Use Disorder

Opioid use disorder (OUD) is a problematic pattern of opioid use that causes significant impairment or distress.¹ In 2022, an estimated 6.1 million Americans had a diagnosis of OUD, with opioid-involved overdose deaths nearly doubling from 49,860 in 2019 to 81,806 in 2022.^1,2 Three medications are safe and effective for OUD: naltrexone, buprenorphine, and methadone. However, attrition and relapse rates are high among patients who use these medications. A need for additional safe and effective treatments for OUD is apparent, but the extensive drug discovery process is a barrier in producing timely solutions. A team of scientists from Case Western Reserve University in Cleveland Ohio have examined data-driven computational drug repurposing approaches to identify potential treatments for OUD.³

Computational approaches for drug repurposing involve systematic analysis of an array of different data. Strategies include obtaining data from genomic signature matching, molecular docking and drug-target interaction prediction, network/graph-based analysis, machine/deep learning, and Mendelian randomization. Once researchers identify potential drug candidates for OUD by way of 1 or a combination of these computational methods, they evaluate selected drugs using real-world data from patient electronic health records (EHR). The authors note that although observational studies of this nature may garner strong results, using patient EHR data can have limitations such as over-, mis-, or under-diagnosis; unmeasured or uncontrolled confounders; self-selection; reverse causality; and other potential biases.³

The authors discussed a retrospective case-control study published in Molecular Psychiatry which evaluated the top 20 drug candidates for OUD using de-identified population-level EHR data of more than 72.9 million patients across all 50 states.⁴ Based on their analysis of patient EHR data, 5 medications (tramadol, olanzapine, mirtazapine, bupropion, and atomoxetine) were associated with significantly better OUD remission among patients with OUD. Causality cannot be inferred based on this data alone, but researchers may consider these drug candidates for evaluation in future randomized-controlled trials.³

Computational drug repurposing approaches can be used to identify promising candidate drugs, several of which have shown potential clinical effectiveness for improvements in OUD remission in real-world populations. However, challenges exist in employing these strategies, including the limited availability of comprehensive, integrated, and standardized databases for OUD. Interpretability is also challenging because OUD is a multifactorial disease, and individuals with OUD may require more personalized treatment solutions. Regardless, drug repurposing may be promising in discovering potentially effective treatment for OUD by using large publicly available datasets and computational methodologies.

References

1. Preventing opioid use disorder. Centers for Disease Control and Prevention. May 8, 2024. Accessed May 20, 2024. https://www.cdc.gov/overdose-prevention/prevention/preventing-opioid-use-disorder.html.
2. Drug overdose death rates. National Institutes of Health. May 14, 2024. Accessed May 20, 2024. https://nida.nih.gov/research-topics/trends-statistics/overdose-death-rates.
3. Gao Z, Ding P, Xu R. IUPHAR review – Data-driven computational drug repurposing approaches for opioid use disorder. Pharmacol Res. 2024;199:106960. doi:10.1016/j.phrs.2023.106960
4. Zhou M, Wang Q, Zheng C, John Rush A, Volkow ND, Xu R. Drug repurposing for opioid use disorders: integration of computational prediction, clinical corroboration, and mechanism of action analyses. Mol Psychiatry. 2021;26(9):5286-5296. doi:10.1038/s41380-020-01011-y