Last Thursday, in a surprise move, the FDA made 89 previously unpublished complete response letters (CRLs), which it had issued to sponsors since 2024, available via its openFDA database. Going forward, the agency says that all CRLs will be published “promptly after they are issued to sponsors.”
Our Medical Advisor, Michael Haichin, was quick to spot the news, and trawled through the database in search of the CRL sent to Lykos Therapeutics last August when the agency rejected its MDMA for PTSD new drug application (NDA). Indeed, he found it, and we broke the news on Thursday morning.
Here, we take a much closer (~5,000-word) look at the CRL and speak with several insiders and observers…
***
- Three Strikes Precluded Approval
- Failure to Capture ‘Positive’ Adverse Events
- Lack of Durability Data
- Selection Bias
- Beyond the Dealbreakers: Additional Comments and Recommendations
- With the CRL Public, Can Lykos Avoid a New Phase 3?
***
Three Strikes Precluded Approval
The ten-page letter begins by setting out the three broad issues that precluded the application’s approval. Here, we take each in turn.
Failure to Capture ‘Positive’ Adverse Events
The first issue outlined by the agency was what it describes as a failure to adequately collect and report adverse events that occurred during dosing sessions that study staff deemed to be “positive” or “favorable”.
In the wording of its CRL, as well as in other documents surrounding the NDA, FDA is insistent that it made it clear to Lykos that it must collect and report all events, regardless of whether study staff or the sponsor deemed them to be neutral or positive.
That information, the agency says, “is necessary…to assess signals of abuse potential and patient impairment in the clinical trials in order to adequately describe the drug effects in labeling and inform appropriate monitoring for the safe use of midomafetamine.”
The MAPP2 protocol made it clear, according to FDA, that study sites should understand an adverse event to mean (deep breath!): “any medical occurrence in a participant, including any abnormal sign (e.g., abnormal and clinically significant physical exam finding, laboratory finding, ECG result, or vital sign), symptom, or disease, temporally associated with the participant’s involvement in the research, whether or not considered related to participation in the research.”However, FDA says that, during its sponsor inspection, it identified that the definition of an adverse event in the MAPP2 training information and safety manual was not consistent with this.
Slide presentations used in site training meetings, and the safety manual that Lykos distributed to study sites, described an adverse event as “any undesirable, unfavorable, inappropriate, or untoward medical occurrence in a patient while participating in a clinical trial… NOT positive or favorable effects”, according to the CRL.
The agency describes this as a “systematic training of not reporting ‘positive’ or ‘favorable’ effects as AEs”, and says that this “raises concerns over the reliability of the safety data.”
It’s hard to disagree with the agency here, as the materials explicitly directed sites to “NOT” view “positive or favorable effects” as AEs, which runs counter to the agency’s requests and guidances. (The agency, as if scolding a schoolchild, begins one sentence in this section of the CRL: “As you were specifically advised”!)
Without this data in hand, FDA says that Lykos has not adequately characterised the safety of MDMA, its acute effects, the duration of impairment, or signals of abuse potential. “This information is necessary to appropriately label and provide recommendations for the safe use of midomafetamine”, it says.
This is all quite well-trodden ground, as it became clear during the FDA Advisory Committee that it was something of a major concern to the agency. (See our June 2024 reporting for more.)
But the CRL offers something new in this realm, too. It reads: “FDA inspections also identified several unreported adverse events for at least two sites, which increase our concerns about the reliability of safety data.” More on that later.
In short, a failure to collect adverse events that it deemed positive or favourable, despite clear guidance and reminders from the agency, in combination with the FDA uncovering “several unreported adverse events for at least two sites” during inspections, led it to have substantial concerns regarding the reliability of the safety data.
Responses to this section of the CRL, and the general topic of ‘positive adverse events’, are varied.
Some have suggested that the agency’s concept of ‘adverse events’ should not include things that are conventionally deemed to be positive, such as feelings of euphoria or elated mood. This group appeals to semantics, arguing that a ‘positive adverse event’ is an oxymoron.
It’s hard to take this argument very seriously. Sure, what is or is not an ‘adverse event’ is an imperfectly worded regulatory construct, but the agency has made it quite clear what it expects to see in this realm. That includes the collection and reporting of what might be conventionally deemed to be positive effects, which inform abuse potential (drug-seeking behaviour outside of the confines of medical practice) assessments.
Regardless of where you stand, regulators like FDA and EMA have made this very clear.
Despite this, Lykos acknowledged in its Advisory Committee (AdComm) briefing document (PDF) that it conducted “no direct assessment of the prosocial or other subjective effects related to potential for nonmedical use of MDMA in the Phase 3 program.”
It is unclear why Lykos did not collect such data. We have spilt plenty of ink speculating on this in the past, wondering why the company would apparently ignore FDA’s Abuse Potential Guidance (published in January 2017, which specifically mentions euphoria-like response as a “key observation” in abuse potential assessments) and the reminders it issued to the company.
Last August, the company told Psychedelic Alpha that it had interpreted adverse events to mean “untoward or unfavorable”, and as such “did not capture effects deemed positive, favorable, or neutral”. There was still no explanation as to how it could have arrived at that interpretation and, given that the company seems to have a policy of silence these days, we are no closer to understanding.
The one person who is very keen to speak to press, much to Lykos’ continued chagrin, is MAPS founder Rick Doblin. For that same article, last August he told us that he had “no idea” why Lykos didn’t gather data on positive adverse events. “I should have paid attention at that level of detail”, he told us at the time, “but [I] just assumed that all the data FDA wanted to see was being gathered.”
Jennifer Mitchell, a UCSF neuroscientist and lead author on both Phase 3 publications, said that her group typically collects and quantifies positive adverse events to evaluate addiction potential, when studying CNS drugs. “In the MDMA phase 3 long term follow up study, addiction potential was collected via a dedicated questionnaire that asked questions about both craving and use of MDMA post-study, and these data were reported to the FDA”, she added.
No one was surprised to see this featured prominently in the CRL. But many were surprised to learn of those two sites that had “several unreported adverse events”, according to FDA inspections.
Corine de Boer, who served as Chief Medical Officer of Lykos between summer 2020 and winter 2022, told Psychedelic Alpha that she was interested and ‘somewhat alarmed’ to see that mentioned in the CRL. She says that they “clearly made the agency less comfortable with the overall reliability of the safety data.”
She thinks that the company’s failure to report “a whole category” of adverse events “most likely made the inspectors dig a little deeper.”
A psychedelic researcher and trialist, who preferred not to be named, was also surprised by the new material. “Unreported AEs at at least two sites was material and news to me”, they told us. “That’s a big deal.”
But those who claim to be familiar with the unreported adverse events at those sites tell Psychedelic Alpha that they were ‘trivial’, and that ‘making a fuss about them is ridiculous’. One individual said that the FDA conducted far more inspections of Lykos’ program than is usual, and said that what they found ‘wasn’t much’.
If this is true, many might expect Lykos to publicly explain themselves, so as to dispel speculation. But, given the company’s aversion to speaking publicly, this seems less likely than ever.
All in all, many people outside of the psychedelics bubble that we spoke with see the agency’s conclusion as quite reasonable, here.
Amanda Wiggans, a strategy and insights manager at consulting firm ZS, for example, said that “ultimately the question of their data reliability and reporting was the dealbreaker, and a pretty reasonable one in my opinion.” “The clear, probably obvious takeaway for other developers is that everything needs to be reported, erring on the side of overreporting, which shouldn’t have been a surprise for Lykos”, she added.
In terms of its stipulations for a future Phase 3 study, FDA says that Lykos must “capture all abuse-related adverse events, regardless of perceived emotional valence”, and that the company should create standardised discharge-readiness criteria for treatment sessions, which should include both psychological and physiological assessments…
Sign-in or join Pα+ to continue reading this article…
Join Pα+ Today
Independent data-driven reporting, analysis and commentary on the psychedelics space: from business and drug development through to policy reform and culture.
Already a member? Log In
✓ Regular Bulletins covering key topics and trends in the psychedelics space
✓ Regular articles and deep dives across psychedelic research, policy and business
✓ Interviews with insiders
✓ Monthly interactive database and commentary on psychedelic patents
✓ Quick-take analysis of major developments
✓ A Library of primers and explainers
✓ Access to our full back catalogue