OUD remains a major public health crisis in the U.S. Even with evidence-based, lifesaving treatments for patients with OUD, gaps persist in treatment retention, accessibility, and tolerability. The immediate priority is not replacing MOUD but determining whether these agents can safely and effectively augment care. Efforts for novel treatments attempt to engage beyond the µ-opioid-receptor, exploring the broader circuitry including reward and incentive salience, stress, hyperkatifeia, neuroplasticity, learning, and comorbid mood/anxiety disorders. Despite differences in each class’s mechanism of action, one potential convergence is the mesolimbic reward circuitry, particularly on cue-reactivity (70). Early preclinical and clinical evidence suggests ketamine could mediate cue-responsive mesolimbic activity (71), CBD may attenuate cue-induced craving and attentional bias (35, 36), psychedelics potentially affect neuroplasticity in which cue-driven associations can be relearned (72), and GLP-1 RAs have been shown to suppress incentive motivational properties of cues in other substances (60). Consequently, cue-induced craving could serve as a shared mechanistic target and a potential common endpoint for future trials. Though promising, the evidence base clearly varies significantly across the four therapeutic classes and further research is needed to elucidate mechanism, safety, and efficacy.

The distribution of evidence across these classes is shaped substantially by drug scheduling, reflecting perceived risk and historical trends (Table 1). Ketamine, GLP-1 RAs, and CBD are FDA-approved for existing indications, clinically accessible, and have a larger body of evidence with randomized trials, off-label clinical experience, and large observational datasets. Psychedelics and ibogaine remain Schedule I - a classification that imposes restricted drug supply, constrained funding, and institutional barriers and has limited the evidence base largely to small, frequently unblinded or observational studies. Ibogaine has the least rigorous controlled evidence of the compounds reviewed, reflecting regulatory policy rather than lack of biological promise. Nevertheless, the safety concerns associated with ibogaine are substantial and warrant careful consideration. Notably, however, much of the existing safety signal derives from “underground” use, reflecting substantial patient demand and the limited availability of regulated alternatives. Regulated research access would permit both higher-quality characterization of efficacy and risk and the conduct of studies under medical supervision. A recent executive order in April 2026 allocates $50 million to states for psychedelic research and accelerated pathways through FDA approval (73). From this perspective, the regulatory pathways, funding, and infrastructure available to investigators are themselves determinants of the evidence base, and clarifying the role of these agents in OUD will depend on the capacity to study them under adequately monitored conditions rather than on current perceptions of risk alone. Scheduling reform alone, however, may not substitute for methodological rigor, as the ketamine and GLP-1 literatures demonstrate, as the American Psychiatric Association has called for an evidence-driven approach (74).

For these therapies, limitations are important to consider. First, there exists an expectancy bias despite blinding, where perceptible drug effects such as dissociation, psychoactivity associated with psychedelics, ketamine, and cannabinoids, and gastrointestinal side effects associated with GLP1s unblind a priori allocation. Furthermore, there is a shifting phenotype of opioid use disorder, as reflected in evolving chemical potency and users, such that foundation trials precede the fentanyl-dominant era, so their current applicability is unclear (20, 21). In terms of generalizability, there may be sample skew and underrepresentation of minorities in novel therapies, especially given previous breaches of trust (e.g., Tuskegee). Furthermore, this narrative review demonstrates that existing research varies significantly in terms of endpoint, dosing, type of drug, patient population, and length of follow-up and requires standardization.

Addressing these limitations requires a two-pronged approach: coordinated research methodology and policy reform. Trials should incorporate active placebo or comparators approximating the agent’s effects, formally assess and report blinding integrity rather than assume it, and prioritize blinded, objective endpoints over self-report. Given small samples and high placebo response, the sequential parallel comparison design, which re-randomizes placebo non-responders and pools across phases, offers improved efficiency (75). Concurrently, streamlining approval pathways, funding, and trial infrastructure for Schedule I agents, but also all therapeutics under study, is a precondition for this rigor in OUD: regulatory reform enables better studies, and better studies inform sound regulation.