Participants in Lykos’s trials were supposed to be blind to whether they received MDMA or a placebo, but a survey — conducted at the FDA’s urging — showed that most correctly guessed their group. Knowing they got the drug may have biased participants to expect it would help them, and “it is reasonable to assume” that this had an effect on the results, FDA staff wrote in the review.

The agency also flagged potential safety concerns surrounding MDMA, including cardiovascular, liver and abuse risks of a substance that the Drug Enforcement Administration lists in the most restrictive category of drugs. The FDA noted that it asked the company to collect reports associated with abuse, but researchers failed to capture effects such as “euphoria” or “elated mood,” information that could gauge the risk of abuse.

Further complicating the evaluation, the FDA wrote, was that it couldn’t measure how much benefit might have come from the therapy, which all participants received.

The review comes ahead of an FDA advisory committee meeting, scheduled for June 4, which will consider whether Lykos’s data show MDMA-assisted therapy is effective in treating PTSD.

Lykos CEO Amy Emerson said the company — formerly known as MAPS Public Benefit Corp. — stands by the quality and integrity of its research and development. “Given there has not been a new PTSD medication approved in the United States in over two decades for the 13 million people living with PTSD, we are deeply committed to making a difference by bringing a potential new treatment to patients,” she said in a statement.

The successful clinical trials for MDMA have kindled excitement about the prospect that psychedelics could emerge as game-changers for a variety of mental-health and substance-use disorders. Many doctors and veterans groups have praised the benefits of MDMA for treating PSTD, a debilitating condition that afflicts about 5 percent of U.S. adults.

“Our patients report that this treatment has been life-changing, leading to disease remission and drastically improved quality of life,” doctors and researchers at New York’s Icahn School of Medicine at Mount Sinai wrote in a comment to the FDA last week. “Many have tried all the gold-standard treatments, and tell us that this is the best, most effective treatment they have ever received.”

The FDA’s concerns about the potential for bias in the MDMA results echo those raised by the nonprofit Institute for Clinical and Economic Review, which concluded in a draft report that the evidence was too murky to determine a net benefit.

The difficulty of blinding a trial of a psychedelic drug was well known to the FDA. Early on, the agency and Lykos considered designing a trial to give some participants a low dose of MDMA — instead of a placebo — as a control group to compare with those getting a full dose. But Lykos “noted occurrences of increased anxiety and difficulty tolerating the medication” in mid-stage trials, and the company and regulator ultimately settled on a design they knew would be susceptible to participants knowing whether they got the drug or placebo.

Lykos also conducted a follow-up study to determine how long the benefits of its treatment lasted, and the data showed a durable effect: Participants who got the MDMA fared better than those who got the placebo more than six months later.

That could mean there’s a “true treatment effect” for the MDMA-assisted therapy, the FDA wrote, but it also flagged “major limitations”: many participants dropped out of the study, others were evaluated at widely ranging intervals and some used other therapies over that time. These factors, combined with the potential bias from participants knowing they received the MDMA, make it hard to interpret the longer-term results, FDA wrote.

Lykos, in a separate briefing document, said there is “substantial evidence of effectiveness” of its treatment and described MDMA as having “low potential for nonmedical use” as it proposes, citing “extensive risk mitigation measures.”

The FDA said that if the drug is approved, restrictions would be required to minimize risk of serious harm to patients impaired by the medication.

Among the proposed restrictions: MDMA could be dispensed only in certain health care settings with a doctor available while the drug is administered; at least two health care providers to monitor a patient’s physical and psychological state for at least eight hours and until the patient is discharged; emergency care plans; and discharge to an accompanying adult after each session.

Such restrictions would also include a patient registry to track negative physical and mental effects. The agency said it is considering the requirement of a post-approval study to gauge the safety risk of MDMA.

Lykos, in its submission ahead of the meeting, acknowledged it was working with the FDA on restrictions, which could also include packaging of single doses only and oversight of distribution by the DEA.

The proposed restrictions are complicated and strict, illustrating the unprecedented decision facing the FDA in approving a potentially dangerous drug, said Aaron S. Kesselheim, a professor of medicine at Harvard Medical School.

He cautioned that even such restrictions don’t always ensure the risk of drugs will be flagged. He pointed to a study that showed the FDA could not determine whether restrictions on long-acting opioids, which have contributed to the nation’s addiction crisis, reduced improper prescribing or better patient outcomes.

The agency’s risk protocols can be a helpful first step, Kesselheim said. “Then it needs to be implemented and integrated well into the regulatory system.”