MDMA therapies hit a roadblock – what's next?

Roughly one year ago, thousands of people gathered in Denver, Colorado, for the largest psychedelic conference in history. The mood was electric, with most attendees confident that the US Food and Drug Administration (FDA) was on the verge of approving its first psychedelic drug.

But last week, the FDA dealt a devastating blow to supporters of psychedelic therapies. It rejected the hallucinogen MDMA as a treatment for post-traumatic stress disorder (PTSD), citing concerns about safety and the validity of clinical trial results. The decision is a pivotal moment for psychedelic science and raises questions about what – if any – future these drugs have in medicine.

The California-based company Lykos Therapeutics has published two phase III clinical trials showing that MDMA, along with talk therapy, significantly improved symptoms of PTSD. The trials, which involved almost 200 adults with moderate-to-severe PTSD, found that between 33 and 46 per cent of those treated with three doses of MDMA were in remission from the condition two months later. The same was true for less than a quarter of the trial participants who had received only talk therapy.

At face value, these results are remarkable. Only two medications have been approved for treating PTSD in the US, both of which lead to remission in just 20 to 30 per cent of cases. But an advisory committee to the FDA expressed doubts about the studies at a meeting in June, ultimately voting 9 to 2 that the research had not proved MDMA’s efficacy.

One of the committee’s chief concerns was the lack of blinding, a technique that prevents participants from knowing which treatment group they are in. Approximately 90 per cent of participants receiving MDMA and 75 per cent of those given a placebo were able to accurately guess which treatment they received. This means the placebo effect could have led those in the MDMA group to expect, and thus report, an improvement in symptoms.

Another concern was the lack of details about the potential risks of MDMA. Lykos Therapeutics submitted incomplete information about the drug’s effects on the heart and also failed to collect data on its risk of abuse.

An independent research organisation called the Institute for Clinical and Economic Review also reported that it had ethical concerns about the trials. The institute found that the researchers had selected therapists, and in some cases participants, from communities that were already interested or involved in the use of psychedelics as medicine. A Lykos representative says this is untrue and that “many of the therapists had no experience with psychedelic-assisted therapy”. The report also said that the people involved in the research who had familiarity with MDMA viewed the drug “more like a religious movement than like pharmaceutical products”. According to the institute, this led some participants to feel pressured to report good outcomes on MDMA and suppress bad ones. The Lykos representative says, “Patients were encouraged to report any and all adverse events accurately over many timepoints throughout the study.”

While all of these issues probably contributed to the FDA’s rejection of the drug, they don’t spell the end of psychedelic research. For one, the FDA has asked Lykos Therapeutics to conduct another phase III trial of MDMA-assisted therapy, suggesting the agency is still open to such treatments. Researchers can also learn from this decision.

“Creating rigorous, replicable science is extremely important, and psychedelics are no exception,” says David Yaden at Johns Hopkins University in Maryland. “If the advisory committee had better reasons to trust the data, we may have looked at a very different result.”

In the meantime, Lykos Therapeutics is asking the FDA to reconsider its decision. “The FDA’s request for another study is deeply disappointing,” Amy Emerson at Lykos Therapeutic said in a statement, noting the new research will take several years to complete. She also said that the company maintains that many of the FDA’s concerns could be addressed with existing data.