A Canadian pilot trial aimed to test oral cannabinoids in people living with HIV but managed to enroll only 10 of 205 individuals approached, with cannabis-related stigma as the most commonly cited reason for refusal. The study provides no efficacy data but offers critical lessons about why cannabinoid research in HIV populations remains extraordinarily difficult, even where cannabis is fully legal.

People living with HIV on antiretroviral therapy often experience persistent low-grade inflammation that contributes to cardiovascular disease, liver disease, and accelerated aging. Cannabinoids have shown anti-inflammatory promise in preclinical models, but translating that promise into human evidence requires clinical trials that can actually enroll participants. This feasibility report from Montreal exposes a sobering reality: cannabis stigma remains a more formidable obstacle to cannabinoid research than the legal and regulatory frameworks that once defined it, raising urgent questions about how we design and communicate cannabis research to the very populations it aims to serve.

People living with HIV on suppressive antiretroviral therapy often experience residual chronic inflammation associated with elevated risks of cardiovascular, hepatic, and neurocognitive complications. Preclinical and observational data have suggested that cannabinoids, particularly CBD and THC, may exert anti-inflammatory and anti-fibrotic effects relevant to this persistent inflammatory burden. CTNPT 028 was designed as a pilot randomized trial to test whether a larger definitive trial of oral cannabinoids could be successfully conducted in this population, using a single Canadian HIV clinic as the study site. The mechanistic rationale drew on experimental model data showing cannabinoid-mediated modulation of immune cell activation and inflammatory cytokine pathways.

The trial’s key finding was the severity of its recruitment challenge: only 10 of 205 approached individuals (4.9%) consented, far below the 26-participant target. Cannabis-related stigma was the most frequently cited reason for refusal, followed by excessive study visit requirements and reluctance to undergo a cannabis washout period. Among the 10 enrolled, medication compliance was 100% and retention was 80%, but two participants (20%) required safety-related withdrawal for transaminitis and aggravation of pre-existing anemia. Regulatory delays, including an 11-month wait for a Cannabis Research License from Health Canada, contributed to a roughly four-year gap between initial funding and first enrollment. The authors conclude that future trials must address stigma proactively, reduce visit burden, and streamline regulatory processes.

Cannabis Is Legal. So Why Won’t HIV Patients Join Cannabis Research Trials?

In 2018, Canada made cannabis legal for everyone. By 2021, researchers at one of Canada’s largest HIV clinics had to approach 205 patients just to find 10 willing to enroll in a study of cannabis-based medicines, and the most common reason for refusal was stigma. That finding alone makes CTNPT 028 one of the more important feasibility reports published in cannabinoid research this year, not for what it discovered about pharmacology, but for what it revealed about the gap between legal access and social acceptance. The authors deserve credit for reporting their recruitment funnel with unusual transparency: they do not dress up a 4.9% consent rate as a stepping stone, and they document every layer of delay, from the 11-month Cannabis Research License application to the protocol deviations they had to accept just to reach double digits. This kind of honest negative-leaning reporting is exactly what the field needs and rarely gets. But honesty about what went wrong does not automatically yield a blueprint for what to do next. The paper catalogues barriers through informal verbal feedback from patients who declined, which is a bit like diagnosing a patient’s chest pain by asking them in the hallway how they feel. You will learn something real, but you will miss the ECG. Without a validated, systematic instrument applied across all 205 approached individuals, we cannot rank barriers with the confidence needed to redesign the protocol effectively.

The deeper clinical tension here is that this study tried to answer a biological question while the actual problem was social. It is like designing a trial of a new medication but discovering that most patients refuse to even pick up the prescription because they are embarrassed to be seen walking out of that part of the pharmacy. The pharmacology never gets a fair test. Meanwhile, the absence of pre-specified feasibility thresholds leaves us in an oddly indeterminate position. The study documents an 80% retention rate and 100% compliance among those who stayed, which sounds encouraging in isolation. But without a predefined finish line, we cannot officially declare whether the race was won or lost. This is a common gap in pilot feasibility studies, but it matters more here because the stakes of scaling up are high: two of ten participants (20%) were withdrawn for safety reasons, including transaminitis and aggravation of anemia. In a ten-person, open-label study without a control arm, we cannot attribute those events to the study drugs with any certainty. But a 20% withdrawal rate for safety in a cohort this small is a signal that no responsible investigator should ignore when planning a larger trial.

If a patient asked me what this study means for them, I would say it means the science is not ready yet, and the reason it is not ready has less to do with molecules and more to do with how we design research for communities that carry compounded stigma. If a colleague asked, I would say the 20% safety withdrawal rate deserves serious attention in the next protocol, and that community co-design is no longer optional. If a policymaker asked, I would point to the eleven months Health Canada took to issue a research license and ask whether that timeline is compatible with the stated goal of building an evidence base for cannabinoid therapeutics. CTNPT 028 teaches us something durable: legalization changes what is legal, but it does not automatically change what people are willing to discuss, be seen doing, or participate in. Research infrastructure for stigmatized interventions must be built from the social context up, not imposed from the regulatory framework down.

For clinicians following the cannabinoid-HIV evidence arc, this study sits squarely at the infrastructural stage: it tells us whether the research itself can be conducted, not whether the intervention works. The companion papers from CTNPT 028, published separately, report safety, tolerability, and exploratory biomarker data from the eight participants who completed the study, but those results must be interpreted in the context of a sample that fell far short of statistical adequacy. The current paper is most useful as a diagnostic of the research landscape itself, highlighting that post-legalization recruitment strategies require fundamentally different approaches than those that preceded cannabis law reform.

From a pharmacological standpoint, the 200 mg daily CBD dose and the low-dose THC:CBD combination (2.5 mg/2.5 mg) used in this trial occupy the lower end of doses explored in non-HIV cannabinoid research. The two safety-related withdrawals, involving transaminitis and worsened anemia, warrant attention in future protocols, particularly given that many antiretroviral regimens carry their own hepatotoxic potential and may interact with CBD’s inhibition of cytochrome P450 enzymes. Clinicians considering cannabinoid adjuncts for HIV patients should ensure regular liver function monitoring and be attentive to drug-drug interactions with protease inhibitors and non-nucleoside reverse transcriptase inhibitors that share metabolic pathways with CBD.

This is a feasibility and process outcomes report for a prospective, open-label, randomized pilot clinical trial, published as a communication in the Journal of Personalized Medicine. It sits below efficacy trials in the evidence hierarchy and is designed to inform trial design rather than clinical decision-making. The most important inference constraint is that no efficacy or definitive safety conclusions can be drawn from this paper, as it reports only on enrollment, compliance, retention, and regulatory timelines.

Prior observational cohort data have suggested that cannabis-using people with HIV may exhibit lower markers of systemic inflammation, and a substantial body of preclinical work has demonstrated anti-inflammatory and anti-fibrotic properties of THC and CBD in experimental models. However, the translational gap between these findings and rigorous human evidence has been wide, with very few interventional trials completed in this population. CTNPT 028 confirms the difficulty of closing that gap and extends the literature by documenting, for the first time in a post-legalization Canadian context, the specific magnitude and nature of recruitment barriers. Unlike earlier feasibility studies in other therapeutic areas, this report identifies cannabis stigma as an ongoing impediment despite full legal access, a finding that challenges assumptions about how legalization affects willingness to participate in clinical research.

The most consequential analytic choice was the decision not to pre-specify feasibility thresholds for enrollment rate, compliance, or retention. Had the investigators defined, for example, a minimum 50% enrollment rate as a success criterion, the 4.9% consent rate would have constituted a clear, unambiguous failure by the trial’s own standards. Similarly, had refusal reasons been collected using a validated quantitative instrument administered systematically to all 205 approached individuals, rather than through informal verbal exchanges with a subset of decliners, the barrier analysis could have yielded ranked, generalizable data suitable for directly informing protocol modifications. These design choices did not alter the raw numbers but substantially affect how confidently we can interpret and act on the findings.

The most likely overinterpretation of this study is to cite the 100% compliance rate and 80% retention as evidence that the trial design was feasible. These numbers apply only to the 10 individuals who enrolled, a group that represents fewer than 5% of those approached. Extrapolating post-enrollment adherence to the broader target population ignores the most critical finding: the vast majority of candidates refused to participate at all. Equally problematic is reading the two safety-related withdrawals as proof that cannabinoids are unsafe in people with HIV. In a 10-person, open-label study without a placebo arm, these events cannot be causally attributed to the study drugs. Finally, this paper does not report any efficacy data; readers seeking information about whether cannabinoids reduce inflammation or affect the HIV reservoir must consult the separately published companion papers from the same trial.

CTNPT 028 contributes rare, granular documentation of the barriers to conducting cannabinoid clinical trials in people with HIV, demonstrating that cannabis stigma, visit burden, and multi-year regulatory delays are the primary obstacles to enrollment even in a fully legal jurisdiction. It does not establish whether cannabinoids are effective or safe for HIV-associated inflammation. For clinical practice, this study changes nothing at the bedside but changes a great deal about how the next generation of trials should be designed.

Does this study prove that CBD or THC helps with HIV-related inflammation?

No. This paper reports only on whether the trial was feasible to conduct, not whether the cannabinoids worked. Efficacy and biomarker data were published separately in companion papers and involved only eight participants, far too few to draw definitive conclusions about effectiveness.

Why was recruitment so difficult if cannabis is legal in Canada?

The most commonly cited reason for declining was cannabis-related stigma or discomfort. Other significant barriers included the high number of required study visits and unwillingness to stop using cannabis during a washout period before the study began. Legal access, it turns out, does not automatically translate into willingness to participate in clinical research involving cannabis.

Were there safety concerns with the cannabinoid capsules?

Two of the ten participants were withdrawn for safety reasons: one developed elevated liver enzymes (transaminitis) and another experienced worsening of pre-existing anemia. However, without a placebo control group in a study this small, it is not possible to determine whether these events were caused by the cannabinoid capsules or occurred independently.

Should people with HIV start taking CBD or THC based on this research?

This study does not provide a basis for starting cannabinoid therapy for HIV-related inflammation. People with HIV who are interested in cannabinoids should discuss them with their physician, particularly regarding potential interactions with antiretroviral medications that share liver metabolic pathways with CBD.

Have thoughts on this? Share it: