Perspective: FDA Advisory Committee vote on MDMA-AT
I have been supporting the strategy of MAPS over the last 6 years. I am no longer employed by MAPS (since December 2023). After Tuesday’s results, I feel the need to share my perspective.
The views expressed below are my own.
The non-binding rejection of MDMA Assisted Therapy for use to treat PTSD by the independent advisory committee to the FDA brings us a deeply humbling moment for Lykos, MAPS, and our entire field. It demonstrates how fragile our collective work is. Victory is not a given. Every step has been hard won over nearly 40 years. Let us not take the progress that has been made for granted. Now is the time that we must be our best, most responsible, honest, and transparent “psychedelic” selves. We must also remember that we are in this together. Lykos and MAPS are imperfect. Yet they have done painstaking work over decades to move the needle. This is incrementalism inside of a megalithic bureaucratic institution. People who have jumped in over the last few years might not realize how difficult, slow, and at times grim it has been.
This is a major challenge, yet it is not impossible to overcome it. And harder things have been overcome.
Here are some of my thoughts:
Lykos was incredibly prepared to answer difficult questions at this meeting. They did not get to. Thankfully, they have the answers to many of the questions posed yesterday, many of which have been discussed - some for years - over the course of rigorous FDA negotiations. It is possible that issues raised were not new to the FDA and were already being considered, as well as vetted for relevance.
The panel had 10 days to review and develop an opinion about data that the FDA has been reviewing for decades. While Lykos has done an excellent job bringing the FDA along - demonstrated by collaboration in the Phase III design itself - they have not accomplished the same shared understanding of this new modality throughout the mainstream mental health or pharmaceutical establishment. The FDA presentation yesterday demonstrated a clear depth of understanding that the advisory panel did not have. Lykos could do better in transparently and richly communicating about their journey.
The panel harped on several factual inaccuracies and seemed to swirl into group think. They brought in old stigma and assumptions that don’t hold water, like that people will become addicted to MDMA or somehow develop a desire to do other drugs (no evidence from the trials), that diversion will be a huge problem (when the drug is likely to be 10-100x the cost of what it is on the street, and under tightly regulated distribution), that Lykos intentionally screened out difficult to treat PTSD patients (when 26.9% (28/104) of participants had moderate PTSD, and 73.1% (76/104) of participants had severe PTSD in Phase 3), that this treatment seems to “only work for white people” (when 53% of participants in MAPP2 identified as people of color), and that there was insufficient data about MDMA’s physical health risks (outside of trials exploring the therapy, nearly all existing research about MDMA is about that because of how NIDA has operated for decades). It was painful to watch the sheer lack of context or competency from the experts on the panel.
Specifically, the problem with functional blinding is well understood and methods to mitigate the potential bias are well established. The issue exists with other drugs and interventions. It was the subject of Doblin’s 2001 dissertation. It took 16 years to land on the study design that advanced to this new drug application, done with extensive input from FDA. MAPS PBC went through a special protocol assessment to proactively deal with this issue to ensure the study design was the best it could be. The question remains: what design would be more effective in functionally blinding psychedelic trials? Or does this issue simply mean that no research of this kind is valid, ever? This is a real challenge for anyone doing this type of research. There was no intelligent conversation on the panel about grappling with this reality, and it seemed like most of the panelists did not know this history. Instead they held a stale perspective that, without better blinding, data is insufficient. This perspective simply won’t work for our field. Study design experts: please weigh in here.
Expectancy bias: first, the idea that this study could be so biased it would sway that large of an effect size is hard to believe, yet the panel seem convinced of this without providing evidence or explaining the math to back their logic. Lykos did an analysis on differences in outcomes in the group that had taken MDMA previously vs the group that hadn’t - and found no statistically significant variation in outcomes. This is key - while one may be able to argue bias in selection criteria, the data shows that it doesn't actually bias the outcomes. About 10% of the US population has tried MDMA. Balázas Szigeti said “MDMA’s effect size are large enough to accommodate both some expectancy and a drug effect. For example, in one phase 3 trial, the between-treatment difference was a 0.9 standardized mean difference (SMD). That is a LARGE effect. Even if half of it is expectancy, that is still a 0.45 SMD. For context, the difference between placebo and SSRIs is 0.3 SMD.”
Ultimately, I believe this is a new class of drugs and that the FDA is not fully tooled to correctly assess this treatment. This is because they do not regulate therapy. Points were made by the panel about how this treatment is done as a combination: The drug cannot be separately evaluated from the therapeutic component. This confounded the panel. A lot of public comments were about the therapy. Nobody on the panel could opine on therapy as doing so is completely outside of their expertise. All the convo on therapy added to the air of confusion.
Missing data: Thankfully, Lykos has a lot of what people were drilling down on (like results based on gender and race). For example: BIPOC inclusion in clinical trials: While Lykos had 53% BIPOC participants in MAPP2 (second phase 3 trial), that was not the case in earlier trials. There is a general lack of representation in clinical trials in the USA. Lykos developed a health equity plan and improved their numbers greatly, demonstrating their willingness to correct oversights and invest in achieving higher than average participation in trials.
Potential for harm/misconduct: in every medical intervention, there is potential for harm. This is true for participants in all psychedelic assisted therapy trials. The risks that exist in vulnerable populations need to be seriously considered and mitigated. Some people will be worse off after receiving psychedelic assisted therapy. It is the unfortunate truth, and we should speak honestly about it. The fact that the potential for harm exists is not a reason to shut down this treatment option for the millions it could help, and should be considered against the existing risks to the patient population, including rate of suicide. These risks require appropriate guardrails to be in place including channels to report harm and accountability measures (as is done in general therapy practice, which has on average 7-12% of participants reporting violations annually). Reducing potential harms are all more possible in a regulated environment than a criminalized one.
The comments on physiological safety (liver and heart) were frustrating as adverse effects are well characterized. Shout out to Matt Baggot for his comments on how widely understood MDMA safety is. Lykos will have to do a better job at establishing and reporting on this data.
And finally: I found the fact that the folks associated with Psymposia did not disclose their shared affiliation deeply unethical. At least three of the negative public comments were a well cultivated, planned, and unified perspective that is bent on halting progress with psychedelic research that they deem unethical - from individuals that do not have clinical experience or healthcare training. The fact that they did not disclose this flies in the face of their moral agenda.
A lot of work will be done over the next months to clear up the misunderstandings and respond in detail to every concern raised at the advisory committee meeting. Lykos will continue on the path of requesting FDA approval and are expecting the final result from FDA by August 11th. May the teams involved feel our support.
[Edit: I have learned that two people that are closely associated with Psymposia are not actively affiliated, so I shifted my final point from saying “five of the public comments..” to “three of the public comments”.]
