Josh Hardman, the founder of Psychedelic Alpha, reviews the published findings of COMPASS Pathways’ Phase 2b clinical trial and its upcoming Phase 3 program. Psychedelic Alpha is an independent newsletter and community that strives to empower a diverse constellation of individuals and organizations with the knowledge, network and nuance to make an impact within the field of psychedelic medicine and beyond.
It has been a year since COMPASS Pathways released the initial findings of its Phase 2b clinical trial of a single-dose of its synthetic psilocybin, COMP360, alongside psychological support for treatment-resistant depression (TRD). The market responded to the study’s mixed results by punishing COMPASS’ stock price, with most other psychedelic drug developers suffering collateral damage from which the sector has yet to recover. Since that initial readout, many have hoped that a peer-reviewed publication and a well-designed Phase 3 program might provide clarity and instill confidence in the protocol’s efficacy and safety.
Earlier this month, a peer-reviewed write-up of COMPASS’ Phase 2b study was published in the New England Journal of Medicine. With 233 participants it becomes the largest published psilocybin trial to date.
Three weeks prior to the publication, the company announced its Phase 3 program at a virtual Capital Markets Day. Four years after it received a Breakthrough Therapy Designation, COMPASS is taking its psilocybin into the final pre-approval phase of clinical trials.
Exploring these two milestones in tandem surfaces some of the key questions arising out of the company’s Phase 2b trial and how it might address them via its imminent Phase 3 program.
COMPASS’ Phase 2b: Efficacy and Safety
We already had a pretty good idea of COMPASS’ Phase 2b results: they were announced a year ago. But, their peer review and publication in NEJM gives us a little more detail, and offers an opportunity to revisit the headline findings and resulting discussions.
The trial, which involved 233 participants, administered a single dose of COMP360 psilocybin alongside “psychological support” (a term that the company quite conspicuously prefers over “therapy”) at one of three doses: 1mg, 10mg or 25mg. The 1mg dose serves as a placebo, for all intents and purposes. Participants in the trial were suffering from treatment-resistant depression, which generally describes patients who have not demonstrated adequate response or remission following two or more treatments.
The results, in short: at week 3, 37% of participants in the 25mg arm were responders and 29% were in remission. One in five were sustained responders at week 12. These are positive results, given the treatment-refractory nature of TRD. But, they’re certainly not stunning.
(Note: “response” was defined in the trial’s protocol as a decrease of at least 50% from baseline in the Montgomery-Asberg Depression Rating Scale (MADRS) total score, while a “sustained response” was defined as a week 3 response sustained through week 12. When announcing the topline results back in November 2021, management chose to highlight a higher figure of 24.1% sustained responders, which was arrived at by “[u]sing a definition of sustained response that is consistent with other TRD studies.”)
Compare this efficacy profile to that of another therapy being developed by psychiatrist Aaron Beck back in the 1960s, while Leary et al. were proselytizing psychedelics: Cognitive Behavioural Therapy (CBT). NYU Langone’s Rick Zeifman pointed out that CBT appears to yield more impressive results in TRD patients than the present study. According to a 2018 meta-analysis Zeifman cited, response and remission rates to CBT were higher at a one-year follow-up than those demonstrated in this Phase 2b study at week 12, with a greater portion of responders from the get-go.
Despite the arguably underwhelming results—which are likely a byproduct of unreasonably inflated expectations—having a broader treatment armamentarium is certainly not a bad thing. Nevertheless, identifying ways to boost the effectiveness of its COMP360 psilocybin therapy will likely be a key consideration in the company’s Phase 3 program.
Aside from discussions around efficacy, much of the reaction to this trial has focused on the incidence of serious adverse events (SAEs).
Given the nature of the intervention, it’s not surprising to see adverse events like headaches, nausea and dizziness among the majority of participants. Given the treatment population, we also might expect to see some level of SAEs like suicidal ideation, which are unfortunately more prevalent in TRD sufferers. The task is to determine what portion of these SAEs emerge in response to the treatment. Despite the small sample size limiting the power of analysis, a correlation between the size of the dose and the incidence of SAEs (which included suicidal ideation as well as self-injury) can be observed.
As psychedelics researcher Dr. Sandeep Nayak noted on Twitter, inspection of the supplemental data shows that the handful of patients displaying suicidal behaviour in the 25mg group were all non-responders. This fact has also repeatedly been emphasized by the company. This lends credibility to a theory that excess incidence of suicidal behavior, intentional self-injury and suicidal ideation might be rooted in feelings of hopelessness among non-responders, especially when viewed in the context of the immense hype surrounding psychedelic-assisted therapies.
Indeed, one academic involved in the Phase 2b trial, Dr. Sidney Zisook of UC San Diego, explained that a “significant minority” of participants demonstrated “profound disappointment with the trial”, more so than he had seen in other studies. He explained that some of these treatment-refractory patients “looked to this study with great excitement and enthusiasm”. However some participants, presumably non-responders, “slept” or “napped through [the treatment session]”, and were “profoundly disappointed”.
Nayak’s Johns Hopkins colleague, Dr. Natalie Gukasyan, also forwarded this idea, positing that desperation heading into trials, and demoralisation resulting from non-response, could explain these higher rates of suicidal ideation among this sub-group.
This certainly isn’t a new phenomenon: researchers like Tehseen Noorani have been publicly commenting on it for over two years, drawing on the experiences seen in earlier studies like Johns Hopkins’ fifteen-person psilocybin for smoking cessation pilot study. It raises an interesting question: how can trial sponsors, like COMPASS Pathways, mitigate or control for the impact of expectancy, especially given the increasing salience of psychedelic-assisted therapies in popular media and culture?
COMPASS Reveals Phase 3 Program
The company’s Phase 2b is the largest published trial of psilocybin to date, but the company’s planned Phase 3 program will significantly outsize it. The program expects to enroll nearly 950 participants across two trials, which will take place at around 150 sites in 14 countries. The two trials will run concurrently, with both placebo and dose-ranging trial designs employed.
The first, COMP 005, will have two arms: one dose of 25mg COMP360 psilocybin versus placebo (unlike the Phase 2b trial, an actual placebo will be used). Here, COMPASS will look to see whether it can replicate the response seen in the 25mg arm of its Phase 2b trial, which was the most effective of the three doses. It will also look at how the safety profile of its 25mg dose of COMP360 psilocybin shapes up to a placebo.
The COMP 006 trial, meanwhile, will administer a second dose of COMP360 psilocybin at week 3, with the same dose-ranging design as the Phase 2b study (25mg, 10mg, or 1mg). Patients will be randomized to doses at a 2:1:1 ratio, reminding us that the 25mg dose is where COMPASS is placing the majority of its chips. This is unsurprising given the existing evidence base, which includes the fact that the company’s Phase 2b trial showed that response to the 10mg dose was comparable to 1mg at week 6. But COMPASS is likely hoping to rule-out whether two 10mg doses might be a worthwhile endeavor.
The primary question this trial will seek to answer, though, is whether a second dose of COMP360 psilocybin increases the portion of participants that respond to the treatment, or whether it improves the quality of response.
Across both trials, it’s unclear whether the amount and nature of “psychological support” patients will receive, will differ from the Phase 2b.
We might expect to see topline data from these studies at the end of 2024 and mid-2025, respectively.
Commercial Considerations: The Role of Technology in COMPASS’ Business Case
So far, we have focused on the key clinical questions that the company’s Phase 3 program will seek to address. But, given the costly nature of psychedelic-assisted therapies, commercial questions abound for companies like COMPASS.
Some companies might attempt to reduce the cost of their protocols by reducing the amount of therapy, or even attempting to replace preparatory and/or integration sessions with a digital therapeutic, for example. They might also aim to reduce the number of therapists involved in the delivery of the protocol, or trial group therapy. While COMPASS often uses the word “therapist”, singular, in its public-facing communications, its published trial shows that two therapists are present at dosing and integration sessions. The company might also develop shorter-acting or non-hallucinogenic psychedelics (or, to use an increasingly common descriptor, “psychoplastogens”).
Not all of these avenues are open to COMPASS’ COMP360 program—which is positively psychedelic—but there are still a bevy of cost-optimising measures available. Not least of these is the embrace of technology.
Today, COMPASS Pathways is keen to position “digital tools” as one of the core pillars of its COMP360 psilocybin therapy, touting a suite of painfully punny tools like its “COMPanion” online therapist platform and “myPathfinder” patient support app. Like some of its peers, COMPASS is presumably hoping that—aside from augmenting the safety and efficacy of its therapy—these tools might also cut costs at varying stages of the patient journey.
To illustrate this point, let’s look at two examples of how this might play out at very different ends of the patient journey.
Theoretically, one of the most propitious ways to boost efficacy and slim down costs is to develop a method of weeding out responders from non-responders as early as possible. In this vein, COMPASS was keen to showcase “Chanterelle”, its program that seeks to apply natural language processing (NLP) in pursuit of this hefty endeavor. The program slurps up qualitative and quantitative data and spits out a prediction of how likely a participant is to be a responder.
According to data shown during the Capital Markets Day (and in a preprint publication), Chanterelle is able to identify responders with a great degree of accuracy, but—and this is crucial—only after the first dose of COMP360 has been administered. Given that the vast majority of costs associated with the single-dose version of this therapy have been spent by that point, it’s akin to Booking.com predicting that you will enjoy your vacation only after you’ve picked up your complimentary coconut water at checkin-in, tested out the firmness of the mattress, and dipped your toes in the pool.
Of course, COMPASS will look to bring forward the predictive window of Chanterelle, which might be expected to only improve as the sample size increases. Might the AI-based tool also be able to predict treatment-emergent serious adverse events, too?
At the other end of the patient journey, companies like COMPASS have teased their plans to use technology on a longer-term basis to identify people who may be slipping into relapse through the analysis of digital biomarkers, for example.
As alluded to earlier, COMPASS might hope that multiple administrations of its COMP360 psilocybin will boost the efficacy of its therapy, or its durability. Indeed, on the final page of the company’s latest presentation, COMP360 is envisioned as an “episodic” treatment, with 1-3 administrations provided over six months.
This might make more commercial sense for COMPASS (as Goldman Sachs analysts crudely asked in a much-memed report four years ago, “Is curing patients a sustainable business model?”), though it will be a careful balance between demonstrating the cost-effectiveness of the therapy and achieving a sustainable business model.
Not everyone is excited about these technologies, though. In a way reminiscent of the chanterelle mushroom’s fruity aroma and peppery taste: some are sweet to the prospect of augmenting psychedelic therapy with technology, while others are concerned by the quantity and sensitivity of data collection that these technologies rely on.
This ambitious Phase 3 program is not only a milestone for COMPASS Pathways, but for psychedelic research and development more broadly. While MAPS’ late-stage trials of MDMA-assisted therapy for PTSD have de-risked psychedelic clinical research, this forthcoming program represents the latest-stage investigation of a classic psychedelic which will be trialed in almost ten times as many patients as the former.
It’s worth considering that the company’s titular technology—the compass itself—was among the “Four Great Inventions” developed in ancient China over two thousand years ago. It took about a thousand years for this navigational technology to find its way to Europe. Perhaps there’s an analogy to be made to the apparent therapeutic qualities of psilocybin, known to many for millennia, but only recently to the “West”? Anyhow, it’s clear that COMPASS is hoping to lean heavily on a suite of contemporary technologies as an adjuvant to its potential first-in-class therapy.
But, as evidenced in the publication of the company’s Phase 2b trial, many fundamental questions remain. It will be fascinating to observe how COMPASS navigates these questions in the coming months and years as it embarks on this enormous endeavor.