Stimulant use in the United States is on the rise, and amphetamine-type stimulants are the second most-used illicit drugs in the world. Along with this rise in stimulant use has been a corresponding increase in overdose deaths related to stimulants, as well as increases in serious and lasting health effects for stimulant users.
All of this has highlighted the need to find effective treatments for stimulant use disorder. To date, there have been no medications approved for treating stimulant use disorder, but finding a medication that would work could make a significant difference not just in treatment outcomes but also in encouraging people to engage in treatment to begin with.
“Drug repurposing,” which identifies new uses for previously approved drugs with established safety profiles, offers a faster and more efficient way to find new treatments for diseases compared to traditional drug development or discovery. But this process can be complex and time-consuming . . . for humans. Artificial intelligence (AI), on the other hand, can process and analyze vast amounts of biomedical data quickly, speeding this process up dramatically.
This study, part of the NIDA Clinical Trials Network study CTN-0114, aimed to put an AI-driven drug discovery framework to work on finding a medication that is already FDA-approved and might work as a treatment for amphetamine-type stimulant use disorder (ATSUD). In a previous study, this same AI-driven framework had discovered that ketamine appeared to improve outcomes for patients with cocaine use disorder – would the AI model also identify ketamine as an effective treatment for ATSUD?
The first step was exploring potential drug candidates for ATSUD. The AI model constructed a knowledge graph, a visualization of relationships between different entities, by integrating multiple biomedical databases and identifying FDA-approved drugs with potential for ATSUD treatment through a systematic analysis of interactions within the knowledge graph.
From there, researchers selected the top 10 ranked drugs as potential candidates. They then reviewed results from clinical trials to see how well these drugs had worked at treating ATSUD. Based on these results, they selected ketamine as their target drug for the rest of the study.
Researchers then analyzed 100 million patient electronic health records (EHR) to look at the association between ketamine and ATSUD remission in clinical cases. Finally, they analyzed the potential mechanisms of action of ketamine in the context of ATSUD, looking at both genetic and molecular factors.
Patients included in the analyses all had diagnosed ATSUD and had either received anesthesia (n=3663) or been diagnosed with depression (n=4328) (two common reasons patients might be given ketamine). Researchers looked at how many patients who had received ketamine had achieved ATSUD remission within a year.
Ketamine for anesthesia in ATSUD patients was associated with greater ATSUD remission compared with other anesthetics. Similar results were found for ATSUD patients with depression when comparing ketamine with antidepressants, including bupropion/mirtazapine (two medications that have previously shown limited efficacy in treating ATSUD).
Analysis of how ketamine might work to treat ATSUD found that it targets several ATSUD-associated pathways. Ketamine’s interaction with certain genes also highlights its potential to modulate critical neurotransmitter systems, like dopamine and serotonin, which are involved in the reward pathways that contribute to addiction.
In conclusion, the researchers’ AI-driven drug discovery framework identified clinician-prescribed ketamine as a promising treatment for ATSUD.
Future work, including randomized controlled trials, is needed to confirm this finding and to better understand the underlying mechanisms and potential adverse effects.
Stimulant use disorder remains a significant challenge for clinicians, as there are very few evidence-based behavioral interventions and no approved medications. In the U.S., fewer than 20% of people using publicly funded programs for substance use disorders are receiving treatment specifically for stimulant use disorders, even though we know the prevalence of stimulant use disorders is increasing.
There are many barriers to receiving behavioral-type interventions for ATSUD, however, including stigma, lack of awareness that these options exist, insufficient treatment resources in a community, and personal barriers like fear of legal consequences or loss of employment.
Effective medication treatments, however, could remove a lot of these barriers and potentially attract more people to seek care for their ATSUD.