At the center of the review is the endocannabinoid system, the body’s own lipid signaling network that cannabis hijacks so effectively. The system’s principal receptor, CB1, is one of the most densely expressed G-protein-coupled receptors in the mammalian brain, sitting presynaptically on axon terminals throughout the cortex and striatum. When endogenous cannabinoids such as anandamide and 2-arachidonoylglycerol are released, they travel backward across the synapse and dampen neurotransmitter release, acting as a short-range feedback brake on neural communication. Tetrahydrocannabinol, the main intoxicating component of cannabis, mimics these messengers but with far greater potency and duration, flooding the same circuits and disrupting the finely tuned timing they normally regulate. The review emphasizes that chronic exposure produces dysregulation of CB1 signaling and of the enzymes that synthesize and degrade endocannabinoids, and that this dysregulation maps directly onto the cortico-striatal circuits that govern reward learning, habit formation, and impulse control.

Equally important is the system’s role in stress biology. The endocannabinoid system modulates the hypothalamic-pituitary-adrenal axis and acts within the amygdala and prefrontal cortex to buffer the physiological response to stress. Withdrawal from cannabis, by contrast, is characterized by irritability, anxiety, sleep disturbance, and heightened stress reactivity, a picture consistent with a temporarily depleted endocannabinoid buffer. The review frames craving and relapse as failures of this buffer under load, which explains why stress is such a powerful trigger for relapse in people trying to cut down. This circuit-level account also explains why the disorder shares neurobiological ground with other addictions, while retaining features, such as the uniquely dense CB1 expression in cortical areas, that make it pharmacologically distinct.

Genetics adds a second, complementary layer of evidence. Large-scale genome-wide studies have identified polygenic vulnerability to cannabis use disorder involving loci such as FOXP2, NPTX1, CHRNA2, and PDE4B. Each of these genes points to a different biological process. FOXP2 is a transcription factor best known for its role in speech and motor circuit development, suggesting that neurodevelopmental pathways shape susceptibility. NPTX1 encodes a neuronal pentraxin involved in excitatory synapse maturation, implicating glutamatergic plasticity in the remodeling of reward circuits. CHRNA2, a nicotinic acetylcholine receptor subunit, hints at crosstalk between cholinergic signaling and cannabinoid systems, while PDE4B, a phosphodiesterase that degrades cyclic AMP, ties vulnerability to intracellular signaling and neuroimmune function. The authors stress that no single variant is decisive; rather, risk is distributed across many loci of small effect, which is precisely the kind of architecture that demands large datasets and careful stratification before it can be translated into the clinic.

The therapeutic section of the review is where the framework becomes concrete. Several mechanism-based candidates have now reached early-phase clinical trials, and although none has yet secured regulatory approval, the pattern of results offers proof of concept that targeting the endocannabinoid system can move clinical endpoints. Fatty acid amide hydrolase inhibitors, which block the enzyme that degrades anandamide, aim to amplify the brain’s own cannabinoid signaling in a controlled, spatially restricted way, in contrast to the global receptor activation produced by smoked cannabis. The logic is elegant: rather than flooding every synapse, the drugs strengthen endocannabinoid tone only where and when neurons are already releasing the messengers, potentially easing stress and withdrawal without producing intoxication.

Perhaps the most striking new class is the signaling-specific CB1 modulators, exemplified by the compound AEF0117. Traditional CB1 antagonists failed in the clinic years ago, largely because complete blockade of the receptor produced psychiatric side effects, including anxiety and depression risk. AEF0117 belongs to a generation of molecules designed to be more selective: it blocks the receptor’s signaling toward certain intracellular pathways while leaving others intact, aiming to suppress the reinforcing effects of THC without dismantling the receptor’s role in mood regulation. Early trials suggest this approach can attenuate the subjective reward of cannabis and reduce use, a result that, if confirmed in larger studies, would represent the first mechanism-based pharmacotherapy for the disorder. Alongside these targeted agents, cannabidiol has shown promise in reducing craving and reinforcement, and the review notes that selected repurposed drugs, developed for other indications, have also demonstrated proof-of-concept effects on withdrawal and use.

What elevates the review beyond a catalog of drug candidates is its insistence on heterogeneity. Cannabis use disorder is not one illness but a family of presentations that differ by biological sex, age of onset, psychiatric comorbidity, and patterns of consumption. Sex differences are particularly well documented: women appear to progress from casual use to disorder more rapidly, report different withdrawal profiles, and may respond differently to interventions, a pattern echoed in preclinical work showing sex-biased changes in gene expression networks after adolescent cannabis exposure in animal models. Comorbid anxiety, depression, and psychosis-spectrum symptoms further complicate the picture, since the same endocannabinoid circuits implicated in addiction also modulate mood and cognition. A treatment that works for a young man with heavy daily use and no comorbidities may fail for a woman in her thirties with anxiety-driven relapse, and the review argues that recognizing this is the precondition for precision medicine rather than an obstacle to it.

The precision psychiatry vision the authors sketch is therefore integrative by design. A patient’s endocannabinoid neurobiology, genetic risk profile, clinical presentation, and pattern of use would together inform which mechanism-based treatment is most likely to help. Someone with high stress-driven craving might be matched to an agent that boosts anandamide signaling; someone whose reinforcement circuitry is particularly THC-sensitive might receive a signaling-specific CB1 modulator; someone with strong glutamatergic genetic loading might benefit from interventions aimed at synaptic plasticity. This is not yet possible with current evidence, and the review is careful to frame it as a research agenda rather than a clinical protocol. But the building blocks, validated genetic loci, circuit-level neurobiology, and a growing pipeline of mechanism-specific compounds, now exist in the same place at the same time, which has not been true before.

The unmet clinical need gives this agenda urgency. Rising potency of cannabis products, expanding legalization, and growing global prevalence mean that the number of people with the disorder is climbing even as treatment options remain limited to behavioral therapies, which help many patients but leave a substantial fraction relapsing. The review, published open access with a permanent DOI and supported by funding from Taiwan’s National Science and Technology Council, makes the case that the field should stop treating cannabis use disorder as a monolithic problem awaiting a single blockbuster drug. Instead, it should be approached the way other chronic, heterogeneous conditions have been: by mapping the molecular subtypes, matching treatments to mechanisms, and letting biology, rather than trial-and-error, guide therapy. If that program succeeds, the era in which no approved medication exists for one of the world’s most common substance use disorders may finally be drawing to a close.

Subject of Research: Endocannabinoid neurobiology, genetic vulnerability, and emerging pharmacotherapies for cannabis use disorder

Article Title: Towards precision psychiatry for cannabis use disorder: integrating endocannabinoid neurobiology, genetics, and therapeutic innovation

Article References: Chan, L.-W., Chen, I., Yu, S., & Huang, Y.-C. (2026). Towards precision psychiatry for cannabis use disorder: integrating endocannabinoid neurobiology, genetics, and therapeutic innovation. Translational Psychiatry. https://doi.org/10.1038/s41398-026-04512-6

Image Credits: AI Generated

DOI: 10.1038/s41398-026-04512-6

Keywords: cannabis use disorder, endocannabinoid system, CB1 receptor, precision psychiatry, FAAH inhibitors, AEF0117, cannabidiol, genetics, addiction, withdrawal, Translational Psychiatry, pharmacotherapy