Why the FDA Declined MDMA Assisted Therapy for PTSD

In August 2024 the U.S. Food and Drug Administration declined to approve midomafetamine (MDMA) capsules paired with psychological support for post-traumatic stress disorder, issuing a complete response letter to the sponsor, Lykos Therapeutics. The agency did not rule that MDMA fails to treat PTSD. It concluded that this specific application, built on the phase 3 trials known as MAPP1 and MAPP2, could not establish substantial evidence of effectiveness or adequately characterize safety, and it asked for at least one additional well-designed study. Reading the letter closely shows why those are separate findings, and why a refusal is not the same as a scientific verdict.

What a complete response letter actually is

A complete response letter, or CRL, is the FDA's formal notice that it has finished reviewing a marketing application and will not approve it in its current form. It is not a permanent ban and not a judgment on a molecule. It is a list of deficiencies the sponsor must resolve before a drug can be marketed, together with the agency's recommendations for how to resolve them. For most of the FDA's history these letters stayed private, visible only to the company. The MDMA letter is unusual because the FDA later released the full text, giving anyone a chance to read the reasoning rather than the press summaries. Structure matters when you read one: approvability issues, the reasons for the refusal, come first, followed by additional comments that are not by themselves grounds for rejection.

The June 2024 advisory committee vote foreshadowed the outcome. The Psychopharmacologic Drugs Advisory Committee voted 2 to 9 that the available data did not show the treatment is effective, and 1 to 10 that its benefits, even with a proposed risk management program, outweighed its risks. Advisory votes are non-binding, but here they tracked the concerns the letter would spell out.

The data-integrity problem

The most serious findings concerned whether the safety data could be trusted. During an inspection, the FDA found that site training materials and the safety manual told staff that an adverse event did not include "positive or favorable effects," while the study protocol defined an adverse event far more broadly. In a trial of a euphoriant with known abuse potential, systematically not recording pleasant or mood-altering experiences as adverse events is a serious gap, because those very experiences are the signals regulators use to gauge abuse potential and impairment. The agency also identified adverse events at study sites that had not been reported at all. Its conclusion was direct: the studies failed to adequately characterize the drug's acute effects, the duration of impairment, or signals of abuse potential.

This is why the letter reads less like a debate about effect size and more like a chain-of-custody problem. If the record of what happened to participants is incomplete or inconsistently defined, the numbers built on top of it become hard to interpret, however large the treatment effect appears.

Unblinding and expectation bias

The second cluster of problems went to whether the efficacy result meant what it seemed to. Roughly 40 percent of enrolled participants reported prior MDMA use, well above what national survey data would predict for the target population. People who already know what MDMA feels like are far more likely to recognize when they have received it rather than placebo. That is functional unblinding: the blind is technically in place, but participants, and even the therapists rating their symptoms, can infer the assignment. Combined with strong expectations of benefit in a community that had advocated for the therapy, the agency concluded it could not separate a true drug effect from the effect of knowing you received the drug.

A related concern was durability. The pivotal trials measured outcomes at 18 weeks, eight weeks after the last dose. PTSD is chronic, so the agency wanted evidence the benefit lasts. The long-term follow-up study used a single visit scheduled anywhere from six months to two years out, drew from a self-selected subset of participants, and allowed other treatments in the interim, so it could not establish durability.

What the FDA asked for

The letter is specific about remedies, which is the practical value of reading it. The FDA recommended a new randomized, double-blind trial with blinded long-term follow-up and prespecified criteria for retreatment; steps to minimize bias, including excluding or limiting participants with prior psychedelic use, measuring expectancy at baseline, and formally assessing whether participants and raters guessed their assignment; and capturing every abuse-related adverse event regardless of whether it felt good. It flagged gaps that were not reasons for refusal on their own, such as incomplete cardiac safety testing and open questions about how much of the benefit came from the psychotherapy itself rather than the drug. The agency even suggested an independent third-party audit of study records, including session recordings, to surface unreported events.

Why a rejection is not the end

A CRL closes a review cycle; it does not close the file. The sponsor can run the requested trial, address the deficiencies, and resubmit, and the earlier data can still count toward a future application once its reliability is shored up. The distinction worth holding onto is between the molecule and the dossier. Nothing in the letter says MDMA cannot help people with PTSD. It says this particular evidence package, with these blinding and data-integrity weaknesses, did not meet the standard the law requires before a drug is sold to the public. Those are different questions, and conflating them is the most common misreading of a regulatory rejection.

This article is educational and not medical advice.