Psilocybin for Cocaine Use Disorder: Peter Hendricks on a Trial Ten Years in the Making

An Overview of the Study and Findings

We found that psilocybin plus psychotherapy elicits more cocaine abstinence compared to placebo plus psychotherapy.

Josh Hardman, Psychedelic Alpha: Can you briefly describe the study and what you found?

Peter Hendricks: We recruited 40 people with cocaine use disorder from the Deep South of the US who were motivated to quit. These participants were representative of the population of people with cocaine use disorder, meaning they were predominantly socioeconomically disadvantaged and Black.

We provided 4-5 sessions of psychotherapy, which included client-centred and cognitive-behavioural approaches, before administering one large dose of psilocybin (25 mg/70 kg) or active placebo (100 mg diphenhydramine). We then provided another 5 sessions of psychotherapy and then conducted assessments 90 and 180 days after the end of treatment.

We found that psilocybin plus psychotherapy elicits more cocaine abstinence compared to placebo plus psychotherapy. Specifically, those who received psilocybin reported more days of abstinence, a greater likelihood of complete abstinence, and longer time to first lapse after drug administration compared to those who received placebo.

Ten Years in the Making

I thought there was a good chance that this would be the last study of my career. So, we proceeded very deliberately, with full awareness that we were working with a sensationalised drug and a vulnerable population.

Hardman: I understand participants were recruited between May 2015 and August 2023. Does the timeline reflect challenges in recruitment?

Hendricks: In part, yes. People who use major stimulants can be quite challenging to recruit and retain in clinical trials. This is well known in the field. More importantly though, this trial was very modestly funded and when conceived around 2014, we were aware that the study could be controversial. The political climate was not at all as supportive as it is now, and the truth is that many of my colleagues thought I had lost my mind or that I was the second coming of Timothy Leary. I thought there was a good chance that this would be the last study of my career. So, we proceeded very deliberately, with full awareness that we were working with a sensationalised drug and a vulnerable population. We wanted to do things right, prioritising ethics, safety, and methodological rigour over speed. Now, once we developed some comfort and confidence and began to hit our stride, COVID came into the world and complicated recruitment.

I do think it’s worth mentioning that those of us who have done some of this early work with psychedelics and addiction have each taken some time to collect our data. This includes Michael Bogenschutz’s groundbreaking work with alcohol use disorder (AUD) and Matt Johnson’s pioneering work with smoking cessation. I think we’re all hoping that recent developments, including the executive order, will allow this work to move at a faster speed going forward.

Avoiding WEIRD Samples

This is fundamentally a matter of collaboration rather than extraction. Science has for too long been extractive. We must be collaborative.

Hardman: Many psychedelics studies tend to recruit higher-income, predominantly white populations. As you mentioned, the majority of your participants were Black and socioeconomically disadvantaged. Do you have any advice for other researchers looking to recruit less ‘WEIRD’ and more representative populations?

Hendricks: I’ve worked with vulnerable populations since I graduated from college, first working at a long-term homeless shelter before starting graduate school in 1999. It is something I’ve done for a long time, so I don’t really know any other way. These individuals are disproportionately affected by mental health conditions, so my first reaction to the psychedelic studies you mention was, ‘What in the world are they doing to recruit so many really WEIRD participants?’ If I specifically wanted to target WEIRD individuals, maybe a study of psychedelics for C-suite executives or something, someone would have to show me how that’s done. I would probably have to move to New York or Los Angeles or the Bay Area, for starters. Things are much different in the Deep South of the US.

But look, there’s a legacy here that’s important to acknowledge. Psychedelics have long been popular among the college-educated elite, the intelligentsia, and, as you know full well, especially in recent times, the ultra-wealthy. When a psychedelic study is recruiting, these are the people on the lookout. These are the same people who’ve read Michael Pollan’s book, who’ve followed the media coverage, and who in all likelihood enter a study with a very strong set of expectations about what psychedelics might do.

I don’t see any reason to exclude these individuals, but there’s also no reason our samples should be comprised of them either. The field needs to do better to recruit representative samples, and a few ideas come to mind. Establish relationships with community partners. Earn the community’s trust and respect by conducting your work in a collaborative and transparent manner. Doing good for your local community should be at the centre of your mission—if you’re just dead set on collecting your data, you won’t be successful. Perhaps whatever PR efforts go into courting private donations from high-net-worth individuals can be redirected to establishing lasting, meaningful collaborations with community partners working with populations with the greatest need. This is fundamentally a matter of collaboration rather than extraction. Science has for too long been extractive. We must be collaborative.

Blinding and Expectancy

[O]f the 40 participants in this trial, only 3 had ever used psychedelics before, and of those 3, 2 were people in their 50s who had a vague memory of one-time experimental use in their teenage years.

Hardman: How did 100 mg diphenhydramine perform as an active placebo? Around half of the participants in the active placebo group correctly guessed their allocation. Do you view that as successful? Relatedly, what do you make of 2 participants who received psilocybin but thought they got placebo?

Hendricks: I think a placebo that cannot be distinguished from the active drug any better than chance is a good placebo. This shows effective blinding. One explanation for our finding is that of the 40 participants in this trial, only 3 had ever used psychedelics before, and of those 3, 2 were people in their 50s who had a vague memory of one-time experimental use in their teenage years. This speaks again to the importance of recruiting representative participants versus a sample of psychedelic enthusiasts with significant experience with these drugs. Placebos are easily distinguished by those who’ve been there before.

Now, there’s been much brouhaha about unblinding in psychedelic research. One thing we need to acknowledge at the start here is that the integrity of the blind is very rarely assessed in studies of psychiatric medications, and when it is, it’s almost always poor. This is the case with ADHD medications, antipsychotic medications and cannabinoids, among others. One question might be why we’re holding psychedelics to a higher standard. Why is the assumption that unblinding is an issue unique to psychedelics?

What I can tell you for now is, when the placebo is easily distinguished, at least in trials of depression, the difference between the active psychedelic and placebo is likely inflated. Given that our placebo was not easily distinguished, I’m not especially concerned about inflated effect sizes here. A bigger issue might be that the therapists correctly guessed treatment allocation 39 of 40 times, but the literature isn’t sufficient to inform how this might affect outcomes. In any event, I think the field is wrestling with the unblinding issue, whether it’s warranted or not, and we’re likely to see some creative studies designed to mitigate functional unblinding in the future.

As for those 2 participants, I think this speaks to the fact that dose and subjective effects are imperfectly correlated. Sometimes people receive large doses and have little to report, and sometimes people receive small doses—or placebo—and report “complete mystical experiences.” Subjective effects and clinical outcomes are also imperfectly correlated. Case in point, of those 2 participants, one reported complete abstinence after drug administration. And I, the primary therapist, believed that this person received the placebo as well. A strict word limit prevented me from discussing this in the manuscript, but this one participant, I believe, presents a counterfactual to the notion that findings are primarily explained by unblinding and expectancy.

Of course, expectancy could have nevertheless played an important role here. Just as it does with all psychiatric medications, and just as it does for any treatment in real-world contexts.

Hardman: What was learned, if anything, from measuring expectancy (and credibility) at multiple time points? Do you have any thoughts about how to integrate that data into efficacy analyses in the future?

Hendricks: Participants in both groups rated their interventions as highly credible before drug administration. Shortly after drug administration, ratings of credibility increased slightly in both groups, though a little more so in the psilocybin group. By the end of treatment, both groups rated their interventions as highly credible, though there was a trend for the psilocybin group to rate their intervention as slightly more credible.

Participants in both groups also held high expectations for the effectiveness of the intervention before drug administration. From shortly after drug administration through the end of treatment, participants in the psilocybin group reported slight increases in expectations of effectiveness. Over this same period, participants in the placebo group reported slight decreases in expectations of effectiveness.

In short, there was just no evidence that participants who received the placebo were disappointed and demoralised, an outcome that has been suggested in psychedelic trials. Instead, there were only modest shifts in expectancies between the two groups, and expectancies remained high throughout the trial. In a secondary analysis, I would be interested in better understanding how and why these ratings shift over the course of treatment, whether they are predictive of certain outcomes, and whether they are predicted by certain variables. I hope these analyses would contribute to our understanding of how expectancies might affect outcomes, and I think this is something all scientists should consider. Of course, expectancies are a shifting target, and I would expect that they operate differently in different populations and different treatment settings, so I wouldn’t be surprised if the field yields inconsistent or conflicting findings.

Psychotherapy and Integration

I would still prefer that participants be eager to terminate treatment than to feel neglected, or to terminate treatment without a sense of closure. 

Hardman: Your protocol includes lots of psychotherapy, with 4 or 5 two-hour-ish prep sessions and once-weekly one-hour-ish integration sessions for five weeks after the dose. Do you think this intervention would be safe and effective without this support?

Hendricks: I think so, yes, though this is obviously an empirical question. For this first study of psilocybin for cocaine use disorder, we felt it was better to offer more rather than less psychotherapy, especially considering the vulnerable nature of the population. As an ethical matter, we wanted to provide as much psychotherapy as we felt was feasible to maximise safety, at a minimum—we certainly didn’t want to insufficiently prepare participants for the experience or just walk away after drug administration. That said, qualitative feedback from participants suggests we could streamline the process, and I would guess that most participants were ready to move on by the end of treatment. I would still prefer that participants be eager to terminate treatment than to feel neglected, or to terminate treatment without a sense of closure.

Anecdotally, I think many participants would have been content to terminate therapy after the first integration session—they got the message and were ready to hang up the phone.

Hardman: How important do you believe the integration sessions were? We noticed that psilocybin participants spent more days in integration than placebo, and only 15/20 placebo participants completed integration, versus all of those in the psilocybin group. Can you tell us about that?

Hendricks: This too is an empirical question. Again, for this first study we erred on the side of caution and provided as much psychotherapy as we felt feasible for safety and ethical reasons. If there is indeed a window of plasticity, I would assume that integration sessions play a rather crucial role. But if this treatment resolves ambivalence and leaves people completely committed to abstinence, integration may not matter much. Again, it’s an empirical question worth pursuing.

Psilocybin participants spent more days in integration on average, and this is because everyone in this group completed the full course of psychotherapy as compared 75% of the placebo group. I am cautious in making a bigger deal of this than I should, but in some ways, this is remarkable because treatment dropout is a major issue in the treatment of cocaine use disorder. Perhaps an advantage of psilocybin treatment is its ability to increase treatment engagement.

Also being cautious not to overstate this, but because psilocybin participants spent more days in integration, they were observed for a longer period of time overall than placebo participants and thus had more opportunities to relapse. If anything, this allowed for a more stringent test of our hypothesis.

Mechanism of Action

Hardman: You served as lead therapist. Do you have any thoughts on the psychological mechanism of action, or anything else you observed or learned from that position in the study?

Hendricks: I think psychological flexibility is a solid hypothesis. I saw what appeared to be the resolution of ambivalence secondary to insights around the incompatibility of cocaine use with deeply held personal values and beliefs, and enhanced “thought-action repertoire,” or engagement with a wider range of more adaptive behaviour to cope with negative affect, life stressors, and the urge to use.

Adherence

I would emphasise that this was an efficacy trial, and for this reason, it was not designed to approximate treatment in real-world contexts. Pragmatic trials are important next steps.

Hardman: What was adherence like? You mention in the study that 4 psilocybin participants and 3 placebo participants completed a fifth prep session “to address non-adherence to the study protocol”; is there any further colour you can provide there?

Hendricks: Adherence was generally quite good, but if you examine the flow chart, you’ll see that only a minority of people who completed the telephone screen were consented, and only a minority of those consented were randomised. No shows were by far our biggest challenge. I think this is the nature of conducting a clinical trial targeting cocaine use disorder, though surely, we can learn from this experience and improve. The larger point, however, is that our participants were highly motivated and able to endure the demands of a clinical trial and were probably not representative of routine clinical populations. This is a common limitation of clinical trials.

We required that participants report 7 days of abstinence and test negative on a urine drug screen to participate in the drug administration session. This was a standard washout period that we used, among other reasons, to ensure safety. Of the 7 people you mentioned, some had trouble abstaining for 7 days and required a bit more support. Others had difficulty communicating to coordinate logistics and required another session to explore their continued engagement with the study and address any barriers to further participation.

I would emphasise that this was an efficacy trial, and for this reason, it was not designed to approximate treatment in real-world contexts. Pragmatic trials are important next steps.

Durability

Anecdotally, I think for many people the reaction to the psilocybin experience was, “Enough is enough. I’m never doing this again. Once and for all, I’m done with cocaine.”

Hardman: Were you surprised by how durable the effect appears to be?

Hendricks: I guess I was, only because addiction is, by nature, a chronic, relapsing condition. At the same time, cocaine use, like all drug use, ultimately does represent a choice. It is a concrete behaviour, and if sufficiently motivated, people have shown time and again that they can achieve prolonged abstinence. There are countless success stories from outside the world of psychedelics to support this idea.

Anecdotally, I think for many people the reaction to the psilocybin experience was, “Enough is enough. I’m never doing this again. Once and for all, I’m done with cocaine.” This is consistent with the notion of quantum change, or the Ebenezer Scrooge-like experience of sudden, dramatic, and permanent change that others like William Miller of Motivational Interviewing fame have observed in their non-psychedelic work.

More research is needed, of course, but if the effect for cocaine use disorder continues to prove durable, then I am excited at the prospect of a treatment option that these patient populations very much need and deserve.

Adverse Events

Hardman: The adverse event table reports a very low incidence of nausea, headache, visual perceptual changes, and other common AEs we expect to see in psilocybin studies. Do you have any thoughts on why these are lower than seen in the broader literature? 

Hendricks: I cannot say I know for sure why the incidence was lower, as you say, but as this was the first trial of cocaine use disorder, I’m not sure comparison with the broader literature is warranted, especially considering the differing participant characteristics we discussed earlier.

I found the participants in this trial to be extraordinarily resilient. Many had endured hardships beyond compare. Adverse childhood experiences, trauma, incarceration, homelessness, and so on. A 6-hour psilocybin session in a comfortable and supportive setting may have been a walk in the park by comparison. That’s only my conjecture for now, but I would offer that people with cocaine use disorder have a grittiness that is often overlooked.

Hardman: Is there any more colour on the reported suicidal ideation in one psilocybin participant?

Hendricks: This participant was very disappointed about a cocaine lapse after a long period of abstinence and expressed passive suicidal ideation in a moment of frustration. No intent, no plan, and not uncommon in this population. As you might imagine, cocaine use disorder can elicit feelings of hopelessness. The suicidal ideation was resolved with no sequelae, and we did not believe it was related to the intervention.

Hardman: And, any comment on spontaneous orgasm?

Hendricks: This was reported by one of the women participants to one of the secondary therapists, also a woman. The participant was, apparently, quite amused at this outcome. We would obviously have no way of confirming the veracity of this report, but we were very serious about accounting for all potential adverse events, including those indicative of abuse potential. Worth noting, spontaneous orgasm is not an uncommon outcome for a range of medications with serotonin activity, so this shouldn’t come as a great surprise. Still, it’s the first instance of a spontaneous orgasm to occur during the drug administration session, to my knowledge

Additional Doses?

If a second (or third, or fourth, etc.) dose increases the likelihood of this outcome, then I think we should explore this.

Hardman: It seems you achieved excellent efficacy from one dose. Do you think there is a case for a second dose, and did any participants in the active arm request it?

Hendricks: I would maintain that the ideal outcome is complete abstinence, and only 6 of the 20 participants in the psilocybin condition reported this outcome. If a second (or third, or fourth, etc.) dose increases the likelihood of this outcome, then I think we should explore this.

We did include a question previously used by Roland Griffiths in his research, asking if participants would be willing to receive psilocybin again, and if so, under what circumstances. We’ll be reporting these data in a separate publication, but the sneak peek is that participants were most likely to say that they would rather not receive psilocybin again but would be willing to consider doing so again in a therapeutic context if it was clinically indicated.

Next Steps

Hardman: What’s next?

Hendricks: A large multisite trial or trials that can inform a new drug application—a priority in light of the executive order—pragmatic trials, and those that allow us to optimise dosing regimens and psychotherapy protocols. Much work to be done, I would assume, and well after I’ve had my say.