A Round Peg in a Square Hole

The challenges psychedelic drugs face navigating the FDA regulatory system

A Round Peg in a Square Hole: The Challenges Facing Pharmaceutical Companies in Their Quest to Medicalize Psychedelics, with Joyce Tsai, PhD

Many folks I’ve encountered in the psychedelic community believe the FDA have suppressed psychedelic research at the bidding of Big Pharma. That just doesn’t make sense to me. Why would FDA want to suppress a drug they’ve fast tracked? And Big Pharma? I’m betting that companies like Merck and Bayer are biding their time until smaller companies like Compass have done all the work to get their drug approved by FDA and then they will pounce. The behemoths will acquire that drug one way or another, probably by making Compass shareholders very rich.

On the other hand, confidence in the FDA has eroded in part because of its role in the opioid crisis. According to the AMA Journal of Ethics in 2020, the FDA failed to “regulate false marketing claims by opioid manufacturers.” FDA failed to “require adequate and well-controlled clinical trials for opioids and FDA conducted “poor management of conflicts of interest between FDA staff and industry.” Here’s a timeline of FDA responses to the opioid crisis. Corruption can happen within good systems. Our government is home to some corrupt politicians. But I don’t think abandoning democracy is the way to get rid of them.

I think people’s concerns about FDA are based more on what they don’t know about bringing a drug to market in the USA than what they do know. And it seems to me that complaining about FDA suppressing psychedelic drug development is kind of like being upset the DMV won’t license your rocket ship.

This had been on my mind for a long time and though I’d tried to understand the system based on public information, there were still many gaps in my understanding. And then, to my delight, I met Joyce Tsai.

In late 2024 a friend of ours offered us two of four tickets to the opera that he had to forfeit. We settled in next to two women and being New Yorkers, we asked each other what we do. The gal sitting next to me, Joyce Tsai, rather haltingly explained she led clinical sciences—the group that designs clinical trials--for Compass Pathways, a biotech company developing psychedelic medicines. I could tell she wasn’t expecting me to know much about the subject, much less be annoyingly curious about her work. Indeed, I could barely contain myself as the lights went down and the curtain went up, but there is no talking in the opera. No whispering either. You can’t even clear your throat or the whole row in front of you turns around and gives you a dirty look.

But during intermission, and then after the second act as I walked her to the subway, I couldn’t stop asking questions. I’d read about RFK Jr’s accusations that FDA “aggressively suppressed” the marketing of psychedelics. That smelled like conspiracy thinking to me. But I didn’t know. I wanted to understand better how the FDA works, and what psychedelic companies face in their quest to bring their drugs to market. Eventually I cajoled Joyce into a lunch, and proposed we work on an analysis describing the challenges. Turns out Joyce is a leader in drug development for nervous system disorders, with over 20 years’ experience in biotech and pharmaceutical clinical development, medical affairs, and scientific assessments. She is the coauthor of numerous papers drawn from her clinical work. (To see her Psilocybin-related papers search here.) I didn’t realize until I met her that I’d read some of her papers during the research phase of HAGT and I was curious about how she got into working on psychedelics.

“At first I wondered if I would be niched into a fringe ghetto by working with psychedelics,” she told me. “And I wondered if it would affect my career. But I was getting a little bored with the opportunities I’d had to work on drugs like SSRIs. They are interesting, and the disease states are interesting, but you get repeats. How many programs can you work on with the same kind of compound? So, when a recruiter for Compass Pathways called me, I was intrigued. With conventional drugs you know how to design the studies, you know the typical safety issues for the class of drugs. You have a roadmap. But [the work] can sometimes feel very cookie cutter. With psychedelics, this round peg/square hole problem is interesting. I must ask the same questions, but I can’t use the same methods to answer them. I feel like I am participating in something that is breaking new ground. It’s not cookie cutter, that’s for sure.”

I learned from our conversations that Joyce has a formidable understanding of drug development and an uncanny grasp of acronyms which I have mostly avoided using because, well, acronyms. We didn’t talk about the specifics regarding the drug she worked on at Compass, because she has a confidentiality clause in her contract. (Since then, Joyce as taken a job with Mycomedica Life Sciences.) But we did cover the challenges psychedelics face in broad strokes. This article was built from publicly available information like FDA documents, books, and newspaper articles, interviews with Joyce, and then a lengthy back and forth for accuracy.

For starters, Joyce explained to me that the FDA wants new drugs and places a higher priority on a drug that addresses a life-threatening disease, like Covid, or where a drug has the potential to meet an “unmet need,” like treatment-resistance depression—especially, Joyce explained to me, if the drug worked via a new and different molecular mechanism than any current approved drugs, which psychedelics most certainly do. That’s why, over the last five years, FDA has fast tracked a handful of psychedelic drugs because they have the potential to address an unmet need. (BTW, FDA doesn’t do blanket fast tracks, like “if it’s a psychedelic it’s automatically fast tracked.” They respond to a particular drug that a company wants to eventually sell.)

Globally around 5% of adults suffer from depression. It’s a potentially fatal disease yet the drugs we treat it with right now don’t work for everyone. What’s attractive about psychedelics is they clearly do something. “We know the compounds are active in the central nervous system,” said Joyce. “The question is, are they therapeutic? And are they safe?” That’s what psychopharmaceutical companies hope their studies will show. But to get their drug approved in the USA (and there are a host of other regulatory agencies worldwide should a pharmaceutical company want to enter those markets), they must prove efficacy and safety to the FDA.

The Food and Drug Administration is ultimately a consumer protection agency. Its roots go back to 1906’s Pure Food and Drug Act, which prohibited interstate commerce in adulterated and misbranded food and drugs, stuff like Benjamin Bye’s Combination Oil Cure, which falsely promised to cure head and neck cancer. (For a terrific history of the legal and scientific tussles that went into passing this act, I highly recommend Deborah Blum’s The Poison Squad, 2018, which I reviewed for the WSJ here). The act was vaguely worded and full of loopholes; the law required that drugs meet standards of strength and purity, but it was on the government to prove whether they worked or not. It was overhauled in 1938 according to Ms. Blum, “following the deaths of dozens of children who were poisoned by a cough syrup legally sweetened with the antifreeze ingredient diethylene glycol.” In 1963 congress passed another amendment that demanded a company must prove both safety and efficacy of their drug based on controlled studies and receive approval from the FDA before marketing. Stricter labeling requirements followed, as well as a law that allowed the agency to regulate medical devices after 200,000 women reported injuries from a birth-control device, another that called for anti-tamper labels (in response to the Chicago Tylenol murders where caps were laced with cyanide), and the passage of the orphan drug act which allowed the FDA to encourage R&D of drugs needed to treat rare diseases. The takeaway here is, the agency evolves in response to national need and emergency.

In terms of psychedelics, that’s begun at FDA, with the issuance of their guidance for the psychedelic industry. But it’s a long road ahead. Moving a psychedelic drug along the FDA regulatory pathway is like putting a round peg in a square hole. The specific assumptions regarding long-term or chronic disorders like depression, and study methodologies employed by the FDA are based on conventional drugs that are administered chronically (as in long-term use for persistent effect) and that can be blinded at least somewhat successfully in clinical trial. (We’ve qualified blinding here because some drugs have symptoms that give it away, like getting the shakes on antipsychotics.) Psychedelic drugs are a very different animal. For example, they may only need to be administered once, and blinding? It’s super hard to do. I mean, most people know when they are tripping. So, we are talking about a very different type of medicine than what the FDA is built to deal with.

As it stands now, a company with a psychedelic drug must go through the same series of data-gathering steps to get their product to market as a company with a new antacid. Let’s look at those steps and see what challenges a psychedelic company faces.

Step 1: Discovery and Development

This is development of a new chemical entity or a formulation that can be granted intellectual property rights. For the purposes of argument, let’s look at a hypothetical example for a psilocybin-like drug. Psilocybin is an existing compound—it’s already discovered, so a new chemical entity might be a synthetic form of psilocybin. A formulation might be a unique delivery system, like an intranasal spray that contains psilocybin. In which case, a new drug delivery system needs to have a reason to exist. So, the benefit of a psychedelic intranasal spray may be bypassing the stomach, and possibly lessening the chance of nausea. Once a company has a product with a patent, it is ready to start preclinical studies.

Step 2: Preclinical Studies

Before a drug can be tested on people, researchers must determine whether the drug has the potential to cause serious harm. There are two types of preclinical research: In Vitro, which is research conducted outside a living organism, like on isolated cells, and In Vivo, which is defined as research conducted on a living organism. In drug development, that means animal studies.

In vitro studies may involve testing the drug with other substances in solution, or with specific living cell types grown in the lab. These studies are useful for testing interactions like what receptors a drug binds to, and what, if any liver enzymes are involved in breaking down the drug. Liver metabolism studies are important because they can identify risks for drug-drug interactions and potential efficacy and safety issues. For example, I take a calcium channel blocker for blood pressure, but the label on my meds says not to eat grapefruit because when the drug makers were doing their preclinical studies, they learned that grapefruit interferes with an enzyme that processes the drug in my body.

In vivo studies look at whether the drug has the desired activity (in animals), and ask whether the drug damages organs, cause tumors, or poses reproductive risks. “All of this,” Joyce says, “is to understand what the potential risks to people might be.”

Psychedelics are particularly challenging when it comes to conducting in vivo behavioral pharmacology studies, according to Joyce. These are studies that look at whether a psychedelic drug is having an expected effect in animals. For example, researchers need to determine whether a rodent is at a tripping dose. One way they do that is by observing something called the head twitch response. That doesn’t say what the rodent is experiencing, only that it is experiencing the drug and twitching its head in a characteristic way.

Since depression is one of the focuses of psychedelic drug development, let’s look at how researchers determine efficacy. Mouse models include the forced swim: how quickly does a mouse give up swimming? Or tail suspension: how long before they give up hope of escape? Or the mild chronic stress model, where a mouse is kept in brighter light than it likes or has a slightly tilted cage. Or the mouse bullying models, where a mouse is bullied by a more dominant mouse. In response to these stressors, the mouse may explore his surroundings less, become less social with other mice, and consume less sugar water, all considered signs of depression. Researchers see if they can rescue the mouse with the psychedelic drug.

Researchers also identify dosing limits on animal models to determine toxicity levels, including long term chronic dosing (although the FDA may have a conversation with a psychedelic company regarding less frequent administration, said Joyce, since chronic dosing may not be a psychedelic drug’s model—though it might be for a microdosing). After the designated period of dosing, researchers will euthanize the animal and dissect it in search of possible drug-induced abnormalities. They also must do reproductive toxicology studies in animals to determine whether the drug effects the normalcy of sperm production, pregnancy, and resultant litters. Any concerning findings, like say a cardiac effect in dogs, would need to be followed up with additional studies.

So, let’s say you have the studies to prove your psychedelic drug formulation has a reason to exist, that it seems to work with mouse models, and your preclinical toxicology studies show your drug does not cause significant damage to organs or reproduction even after administered in repeated high doses. The next steps are clinical trials, which are studies done on people, but those studies cannot begin in the USA until the company completes an IND, an Investigational New Drug application.

Step 3: Getting the IND

The point of the IND is to get approval to move forward with testing in humans. The process starts with scheduling a pre-IND meeting between you—the drug company--and FDA where you discuss whether the drug is ready to go into people, and whether your first study on people is appropriately designed for the potential risks. FDA will provide a meeting date sometime in the next two months. A month prior to the meeting, you will need to submit a briefing book that includes your final questions for FDA, like does FDA agree with the proposed doses to be studied? And your reasons why FDA should agree, along with a summary of data that backs up your argument as to why FDA should agree, and the outline or synopsis of the first clinical trial in humans that you are proposing. Any psychedelic company that has done or is doing a study on human subjects, and that would include Lycos, Compass, and Usona, to name a few, will have gone through this process. In the pre-IND meeting, FDA will answer your questions (or provide a written response if they think no discussion is needed after reviewing your briefing book) and provide feedback on whether you are ready to submit a full IND application. If FDA determines you are not ready to start experimenting on people, at least you will find out what info they perceive to be missing.

The next step is a formal IND submission. The IND includes the full manufacturing process and related studies demonstrating the drug’s reproducibility and its shelf life. Also included are all preclinical studies you’ve done, and the full protocol for the proposed first study in humans. So, for example, Joyce told me a typical first-in-human study might have, say 4 dose groups of 8 people, where 6 get drug and 2 get placebo. Dosing would start at the lowest dose, and if it is deemed safe and well tolerated, the study would go to the next dose, and so on.

“Such a study would look at pharmacokinetics [what the body does to the drug],” said Joyce, “and in the case of psychedelics, it would also look at pharmacodynamics” (what the drug does to the body--in this case, the intensity of the psychedelic experience, if any, at each of the tested doses). The IND also includes an investigator brochure that summarizes all the known data about the drug and key summaries of what was done and any meaningful findings to date. This includes published information from other studies that is relevant to your planned human trials. The point of this is to identify risks that may have popped up in diverse studies but may not have occurred during your preclinical work. In the case of psychedelics, that would typically include things like heart palpitations, panic attacks, and nausea. The investigator brochure is regularly updated and briefs the doctors who will conduct the studies of your drug. It’s scientific crowd sourcing. Finally, you must include an informed consent document for study participants (written at the 7^th grade reading level) that describes the study and potential risks of participating.

FDA typically has 30 days to review protocols for new studies in humans and let you know if you may proceed. If they find potential safety concerns in the IND, or they think the study poses too much risk to human participants, or they disagree with the study design they will either not let you proceed and require that you make changes to the protocol and then resubmit it, or they will recommend changes before proceeding.

Throughout the drug development process, companies conduct meetings with FDA for scientific guidance. For example, a company doing a study on a psychedelic drug for treatment resistant depression might ask FDA if they agree on the selection criteria, like specific symptoms or diagnostics that are used to select patients. The meetings are about asking FDA to judge what is permissible when novel situations arise in the drug development process. They are not about holding a company’s hand and guiding them to a market landing.

If the FDA is satisfied that participants will not be put at significant safety risk by your drug, and that your study design is reasonable, then a company can move on to clinical research.

Step 4: Clinical Research

This is where the rubber hits the road, and fitting psychedelics into standard clinical trial designs is problematic, to say the least. Before the trial/s begin, researchers must have determined who qualifies to participate in the study, how many people will be in it, how long the study will last, whether there will be a control group, how patients will get the drug and in what dosage, and how and what data is collected and assessed, reviewed, and analyzed.

Research on women has been a standard for a long time, but especially in Phase 1 healthy volunteer studies (which we will discuss) researchers were historically worried about the “variability” that different phases of the menstrual cycle might introduce. So, to keep the science simpler, they assumed women are the same as men, even though they excluded women as being too different. A succinct historical overview is here.

Throughout the clinical trial process, FDA will continue to monitor the safety of your drug. And not only your drug, but others like it, because all companies doing drug research in people need to file safety reports regularly and even if nothing bad happens in your study, if another company studying the same or similar drug has a very concerning adverse effect happen, like suicidal ideation, FDA might pause or stop all studies with the same or similar drugs and try to figure out if the adverse effect is by chance, or drug-related, or arising from the study design. Depending on the outcome of the analysis, FDA might lift the hold, or require protocol modifications, or extend the hold.

Cartoon by Paul Noth

One of the most significant problems for psychedelic passage through the FDA regulatory process is meeting clinical trial standards. The gold standard for clinical research is the randomized double-blind placebo-controlled study. In this study type, some participants get active drug, and some get a placebo, but no one, not the researchers, not the participants, is supposed to know who got the active drug until after the study is completed. However, most people just know when they are tripping, and they break the blind. To deal with this problem, some researchers use not a placebo but another drug, like niacin which causes sensations, but not ones like psychedelics. Others have designed very low doses that are not expected to have efficacy and intermediate doses that cause a trip, but not as intense a trip as believed to be needed. These alternative study models may work well—they are, however, novel, so it is still unclear how they will measure up to the double-blind standard. Some trials are open label, where the placebo is eliminated, and everyone gets the psychedelic. However, folks may bring an expectation of benefits to the experience if they know they are getting the drug, which makes it harder to evaluate the treatment.

Another huge challenge is dosing. Let’s look at a psychedelic food example. If you look on the back of a psychedelic chocolate bar, it might say 1-3 squares is a microdose, 3-5 squares are a therapeutic dose, and 5-8 squares are a spiritual dose. That won’t fly with the FDA. They need to know exactly how much drug you are getting in each square, and that every square in every psychedelic chocolate bar contains the same amount. They need to know how long that chocolate bar is good under what storage conditions, and when it will expire. They also need to know what else is in that chocolate bar, how many people got sick after eating that chocolate bar, and why. And of course, they need to know that what you want to market as a therapeutic dose meaningfully reduces symptoms. That’s not easy. A psychedelic dose is subjective: people may respond differently to the same dose size for a variety of reasons (including SSRI use. See my substack essay on this here). With conventional drugs, studies typically test a few different doses based on the highest doses that were well-tolerated in healthy volunteer studies, or dose by body weight.

If a drug can be reliably made, passes all the tests in animals, and FDA agrees to the clinical research plan, which includes figuring out how thorny problems like blinding and dosing are going to be addressed, then the drug can begin studies on people.

Phase 1: Clinical Research in Healthy Volunteers

There are three stages of clinical research, called phases. Phase I development is about characterizing how the human body absorbs, distributes, breaks down, and eliminates the drug (aka clinical pharmacokinetics). An important distinction here is Phase 1 studies focus on healthy volunteers while later study phases focus on patients—people with the disease the drug manufacturer hopes to treat.

Phase 1 studies explore the effects of dosing, what’s called “first in man” or single-ascending studies. Basically, a group of 6 to 12 participants receive a small dose and if that dose is tolerated after a medical review, the dose is increased with the next group, and so on until the highest planned dose is reached.

These studies may also test whether there are differences in how quickly the drug is absorbed, broken down, and eliminated, based on things like whether it is taken after a high fat meal. (Those of you in the underground know that if you take magic mushrooms after a fatty meal like Osso Bucco it may take longer for the drug to come on.) Finally, if the planned treatment regimen for a medical condition is expected to require multiple doses over time—and a two or three dose schedule is common in academic psychedelic research--then a multiple dose study in small groups of healthy volunteers must be done as well.

A key purpose of a Phase 1 study is safety, and that is tricky when it comes to psychedelics. Drug safety is determined using a therapeutic index, which is a ratio that compares at what concentration in the bloodstream a drug becomes intolerable and at what concentration the drug is effective. For example, the label on your 200mg Ibuprofen bottle may say take 1 tab every 4 to 6 hours, but do not take more than 6 pills over a 24-hour period. That defines the therapeutic window the ibuprofen maker had shown in large scale trials to be basically safe before their drug went to market. Psychedelics, however, have a narrow therapeutic window that can vary from person to person. For example, if you call a bad trip a bad side effect, that might be considered by FDA to be intolerable and dose-limiting. But there’s no telling if someone is going to have a bad trip, which may have very little to do with the dose anyway. “Psychedelics may turn the whole notion of therapeutic indexing on its head,” said Joyce. “It’s one of the many challenges, but it’s also what makes it interesting.”

Phase 1 studies are the minimum healthy volunteer studies that are usually required before beginning patient studies. They let you know what doses to test in patients, and whether to administer the drug with food. Of course, with psychedelics there is a wealth of anecdotal experience, from the indigenous people who have a long history of using psychedelics in ceremony and medicine to a robust underground in the USA, but a lot of that experience is not officially written up in the medical literature or was not rigorously recorded. So how to take advantage of anecdotal dosing and safety knowledge to develop psychedelic drugs more efficiently is another round peg in a square hole problem.

Phase 1 research continues even while drug development goes into patient clinical trials. There will be additional healthy volunteer studies using a radioactive version of the drug to track exactly how it is absorbed, distributed, metabolized, and eliminated in the human body, and possibly additional human studies looking at other risks identified in test tube, animal, or earlier human studies such as potential specific drug-drug interactions, cardiac function, safety in people with existing liver or kidney damage, and risk of addiction or abuse. That last type of study is yet another problem psychedelics face. Psychedelics are typically not addictive or abused in the way that, say, heroin is. But like heroin, psychedelics like LSD, psilocybin, ayahuasca, MDMA, and now DMT are “Schedule 1” – denoting no acceptable medical use. So, the question becomes whether psychedelic drugs need to be tested for risk of addiction even though addiction historically has not been a problem. The answer is yes. Because according to Joyce, the FDA has made clear that new psychedelic molecules will need to have Human Abuse Liability studies, because “FDA has said ‘We already know how LSD, psilocybin, etc. are abused.’”

Phase 2: Clinical Trials with Patients

Phase 2 clinical trials are usually the first studies in patients—that is, people with the disease the drug is supposed to help--and the trials look, again and as always, at efficacy and safety. These are usually modest-sized studies, probably 40 to 250ish people. Sometimes Phase 2 drug development is split into “a” and “b” studies. Phase 2a studies are typically considered “proof-of-concept” and may be open-label or double-blind controlled studies, while Phase 2b studies are almost always controlled double-blind studies conducted to narrow down the dose(s) to be tested in the larger Phase 3 studies. Phase 2b studies are expected to demonstrate statistically significant efficacy. That means demonstrating that any improvements vs. placebo are more likely than just chance. As clinical trials in patients are super expensive to conduct, this approach may be taken to reduce the risk of running large expensive Phase 3 trials before you think the drug probably works.

Many of what Joyce calls “academic” studies—the ones made so famous in Michael Pollen’s book How to Change Your Mind—are open-label studies in patients, without placebo or negative control group, and done by academic researchers with synthetic psilocybin. Results of these relatively small open-label studies may provide preliminary data on efficacy and safety, but usually much larger controlled studies are needed to demonstrate that they work and are generally safe.

Government has historically funded large drug studies and they are among the most important and influential studies in medicine because they seek to answer questions that aren’t particularly in the interest of drug companies to ask. Recently Health Secretary RFK Jr said in a hearing that “We are launching clinical trials…on [psychedelics] now at FDA, I think there are 11 clinical trials at the VA going on at this point…This line of therapeutics has tremendous advantage, given the clinical setting, and we are working very hard to make sure that happens within 12 months.” (Thanks, Psychedelic Alpha for the transcription.) It seems he is talking about the Veteran’s Administration-funded research on MDMA, which he intends to legalize in 12 months. You can find those studies by going to clinicaltrials.gov and filling out the search. For Condition/disease type “PTSD.” For Other terms type “veterans.” For Intervention/treatment type “MDMA.”

But back to drug companies…

Drug companies that want to go into Phase 3 development after completing Phase 2 studies need to present their data to FDA and propose a Phase 3 development program, in an “end of Phase 2 meeting.” The regulators want to make sure that the existing data supports the risks of testing the drug in a lot more people.

Phase 3: Clinical Trials with Patients

These studies are large, with 300 to 3000 volunteers who have the disease being treated. They are multi-centered, which means the study is conducted at multiple medical centers or clinics, all of which follow the same protocols. The duration of study participation may last several months and up to a year—that’s the requirement (it’s longer for child participants). Phase 3 studies typically have a more generalizable patient population with the disease, including those with common comorbidities (if they don’t pose additional risks or confound data interpretation), and they must use the final form of the drug to be marketed (the exact tablet or capsule or device).

The basic purposes of Phase 3 studies are (1) efficacy: does the drug work? (2) tolerability: what are the adverse reactions and are they tolerable? And (3) safety: are there short-term and long-term consequences or increases in risks for serious consequences?

To answer those questions, sponsors must show in two separate studies that the drug almost certainly works. They must characterize the side effects or adverse events and show that the side effects are tolerable for most people. Phase 3 programs also include a long-term study to show that there are no worrying long-term consequences after 6 months and 1 year of treatment. Basically, FDA wants to know everything that could go wrong, as that ends up on the drug label. It’s to inform not just what bad things might happen, but what might happen, period. Psychedelics are very challenging in this regard. For example, if you are taking an antacid and have a hallucination, the FDA would consider that an adverse reaction. And if several people hallucinate on that antacid, FDA will almost certainly say no way is an antacid that makes you trip going to be approved. But in the case of psychedelics, the hallucinations may contribute to the drug’s efficacy. “You are essentially using chemistry to evoke a psychological state that may be healing,” said Joyce. “But having a psychedelic experience confuses everything.” (And it contributes to the growing interest in a hallucination-free “psychedelic-like” drug and the prospect of microdosing.)

Phase 2 and Phase 3 trials ask how the drug affects different groups of people (like gender and race). Limiting factors, like whether the trial pays or there is a history of abuse by researchers among some demographic groups can affect the diversity of participants. Maybe that’s why so many psychedelic trials are populated by well-off (aka white) people who are informed about the “psychedelic renaissance,” (aka have read How to Change Your Mind). The responsibility falls on the petitioners to “show that in all groups the data goes in the same direction,” said Joyce. “Your sample size has to be big enough to prove with 90% confidence this drug is working and that most subgroups of people respond similarly.”

Phase 3 trials are not that different from Phase 2, only much larger and with more rigorous statistical criteria. But the larger the trial size, the more adverse reactions pop up. Rare side effects may occur that didn’t in the smaller, Phase 2 studies. Additionally, because Phase 3 are longer studies, long-term side effects may be detected. It’s important to get side effect and safety data in large studies because once a drug is approved, even more thousands to millions of real patients may take it. (Some drugs have been taken off the market after very rare but life-threatening consequences were identified. “Remember fen-phen for weight loss?” asked Joyce; others have had their label modified to emphasize the risks. That’s the case with OxyContin.)

A particular problem psychedelics face when it comes to side effects is measuring aftereffects. In a clinical trial setting where multiple doses of some classes of drugs that affect the brain are administered, companies are obliged to follow up with patients for three weeks after their last dose to see if they are withdrawing or experiencing cravings. However, the rating scales for measuring those feelings are designed for specific drug classes. The abuse liability rating scales for SSRIs, Benzos, and alcohol aren’t applicable to psychedelics. So, a company must create or modify a rating scale and ask FDA if it is okay to use it.

Demonstrating that patients are truly safe to go home after tripping in the clinic is another example where we don’t yet have purpose-built assessment tools, Joyce said. In addition to showing that patients don’t experience suicidality because of what they may have experienced during their trip, researchers also need to show that the psychedelic experience is fully over and that there are no remaining effects before allowing a patient to go home. “We don’t have specific tools to do this. Do we use a schizophrenia interview that asks about hallucinations, delusions, paranoia? Do we use a general cognitive impairment assessment that is typically used for Alzheimer’s patients? How can we ensure that patients are unimpaired before we send them home?”

There are concerns about potential “flashbacks” long after the psychedelic has left the patient’s system, explained Joyce. It will be necessary to follow patients for at least a year and to show little-to-no-risk for flashbacks. “Just relying on patients to report adverse effects will not be sufficient; there needs to be more systematic assessment.”

Usually, a drug needs to be tested in 1500-2000 people by the end of Phase 3, with at least 300 people tested for 6 months and 100 tested for a year, for a company to apply for marketing approval. “But what does it even mean to test a psychedelic for several months or a year?” asked Joyce. “Unlike conventional drugs whose dosing regimens are meant to keep a level of drug constantly in the body, psychedelics might be taken only once or a few times during those months to years.” FDA and other regulatory agencies have made it clear that whatever the dosing regimen, patients need to be followed for at least 3 months to see how well the psychedelic is working, and out to a year for safety. It’s a sticky problem and only a few psychedelic products are in Phase 3 trials. (I track updates on the pharmaceutical industry at Psychedelic Alpha.)

Companies have addressed this problem of dosing regimens by giving the minimum number of doses they believe should work in the short-term, and then follow the patients for safety and efficacy as part of a long-term study, typically allowing for limited retreatment if needed. For example, the nonprofit USONA Institute has a study on psilocybin and major depressive disorder. They have employed a combination of short-term studies during which the dosing happens, followed by longer studies where patients are tracked as if they were taking the drug every day, and includes the possibility of a potential retreatment.

A unique aspect of psychedelic drug therapy is the therapy. Psychedelics are commonly tested alongside psychological support for illnesses like depression and it can come in multiple forms. “Supportive psychotherapy” is aimed at helping the patient get through—and get the most out of—the experience. “Directive psychotherapy” is when a patient and her therapist together decide to go for psychedelic-assisted psychotherapy where the drug is used to help the psychotherapy be more effective. But the therapeutic piece of a psychedelic drug petition is problematic. Not only does the FDA not regulate psychotherapy, but if it did, what therapeutic modalities would be okay? Because if you look at the bios of psychedelic therapists and guides, their certifications are all over the place, from social work to breathwork to trendy concepts like Internal Family Systems. So, while psychotherapy in conjunction with psychedelics may be the path to optimal care—indeed, to not provide therapeutic support may cause harm—it’s simply not what the FDA does. FDA is not empowered to evaluate therapy. FDA’s mandate is to regulate drugs, not the practice of medicine, under which psychotherapy falls. The addition of therapy, says Joyce, “confounds their ability to evaluate the effect of the drug, which is their charge. That’s the problem.”

Yet FDA could approve a label that states the psychedelic is intended to work alongside psychotherapy, but they want to make sure that whatever you are doing in your trial, you are doing it the same exact way for every patient, and you are doing it in such a way that you and they can tease out the effect of the drug vs the psychotherapy. And it’s on the sponsor to demonstrate that the drug has a beneficial effect either on its own or the effect of the combination of drug and therapy clearly improves what is seen with psychotherapy alone.

Step 5: FDA Review

Let’s say your psychedelic drug makes it this far. You get great results in two separate and independent large studies that each show with 90% probability that your drug works, and it seems to be well-tolerated and generally safe after exposing 1,500 to 2,000 people to it. Then a company must propose to submit a New Drug Application (NDA) to FDA in a “pre-NDA meeting” where a summary of all the relevant materials is reviewed and the FDA agrees (or doesn’t agree) with moving forward with the actual NDA submission.

The New Drug Application must include all the relevant materials, the human and animal data, information about manufacturing, pharmokinetics, and toxicology. The FDA has 90 days to accept the application or not—it may not if the application is determined to be incomplete (or deemed to be not as agreed during the pre-NDA process). But if the NDA is accepted for review, then an FDA review team goes to work evaluating the research, typically even re-crunching the numbers from the clinical trials to verify the company’s results, or analyzing the data in other ways that FDA statisticians think are needed. They may physically inspect the company that designed and led the studies, some of the clinical sites that conducted the studies, and sometimes even the facility where the drug was manufactured.

FDA reviewers include a team of experts from various fields: medical doctors, statisticians, chemists, microbiologists, pharmacologists, and others; and each expert reviews the NDA through the lens of their discipline. They write up evaluations which are shared with team leaders.

Occasionally advisory committees are asked to weigh in if outside expertise is needed. Advisory committees are almost always needed the first time a new drug of a new drug class is up for approval or the first time a new treatment is being considered for a new disease that previously had no treatment. Psychedelics, which represent a new class of drugs, can be expected to require FDA Advisory Committee review. Advisory committees are a potential weak spot in the FDA system because, as Joyce explained, finding experts in their various fields who have never interacted with a drug company before is difficult, and hiring advisors who are not experts in their fields may be unwise. “FDA has been criticized for including people with industry interactions. [Even though] while on a committee you can’t collaborate with anyone working on the drug. But the standard of perfectly clean people? That’s what is difficult.”

Over the last nine months that I’ve been doing presentations of my book, Have a Good Trip, I’ve met a few people who are under the impression that there is a revolving door between FDA and industry. Joyce told me that while she has known of some junior level scientists going from industry to FDA, for physicians, it’s far more common to see the flow the other way, from FDA to industry. “As junior staff, they went to FDA first to get the first-hand understanding of regulatory thinking. Then after a few years, they were able to get a much higher-level job in industry. But I’ve not known of senior scientific/medical leadership go from industry to FDA.”

The FDA is obliged to decide within 12 months. That timeline is secured by something called a PDUFA fee (Prescription Drug User Fee Amendment) to the tune of about $4 million. Per drug. This is a kind of expediting fee that “everyone builds into their budget,” says Joyce. Historically, companies claimed the FDA was taking too long to review drug applicants, which was a problem because companies have deadlines on their patents—they expire 20 years from the initial filing—and the subsequent exclusivity those patents ensure. That doesn’t mean you can’t get exclusivity for your drug after the patent expires, but it’s more limited. In 1982, drug companies made a deal with the FDA to pay this fee to boost the FDA’s resources and speed up the review. “This is the kind of thing when the public gets wind of it, they think you’re paying the wolves to guard the sheep” said Joyce, but on the other hand, she pointed out, “the review is extensive and PDUFA dates [she means the 12-month review period promised in exchange for the fee] have been honored even during government shutdowns.”

The details of the FDA’s decision about a drug are outlined in a letter and if they choose to—and I imagine this mainly happens when a drug is not approved--the company can meet with agency officials to discuss the outcome, ask for a hearing, correct any problems, and plan to submit new information.

But let’s say the drug is approved. For a psychedelic, currently a Schedule 1 controlled substance, the next step is review by the Controlled Substances Staff for rescheduling since now the drug is no longer considered to be of no medical use. Once a Schedule 1 drug is approved for some kind of medical use, it won’t belong on Schedule 1 any longer. However, it will need to be rescheduled, though to where is another question. That process may take another 3 to 6 months before the drug can be marketed, usually with a requirement for a Risk Evaluation and Mitigation Strategy (REMS), which is in essence a risk prevention, monitoring, and management plan that FDA requires with drugs that might have serious safety concerns. But regardless of the company’s preferred scheduling, the folks at Controlled Substances will make up their own minds in the end.

Indeed, a newly approved psychedelic will probably have a very high level of safety surveillance during its first several years on the market. That’s what Phase 4 trials seek to do as well.

Phase 4

Yep, there’s another phase. Phase 4 trials occur after the drug has been approved. These trials might be additional treatment studies required of companies as a condition of FDA approval or based on data collected as part of the REMS or based on volunteers reporting their real-world experiences on the drug. Not only rare adverse reactions may be discovered but also reactions from prolonged use. In the case of psychedelics which are used once or a few times, those adverse reactions may be late onset mental problems associated with the trip. Or in the case of regular, small dosages like microdosing, issues like tolerance may appear, or cardiac issues which can occur with structurally similar drugs. Additionally, the public may submit their comments about the drug and how it has affected them to the FDA’s MedWatch program. The FDA also uses data from electronic health records and social media. In some cases, post-market surveillance will provide ongoing monitoring of a drug until FDA is satisfied that they fully understand the safety of the drug, or they conclude that the ongoing monitoring isn’t providing any meaningful new info. For example, the antipsychotic Clozapine had a REMS from 2015. It was just lifted.

Many folks in the scene argue that decades, even centuries of Psilocybe mushroom consumption illustrates the drug’s safety and that should be good enough, but the fact is FDA doesn’t approve drugs for the market based on anecdotal evidence. There’s no doubt psychedelics face a lot of challenges in our current system which is designed around assumptions based on conventional drugs. But I think it is important to remember that the challenges psychedelics face is not the fault of the FDA system. Psychedelics are being developed for treatment of various diseases, and FDA is grappling with how to regulate them. FDA isn’t perfect, but accusing the institution of subversion when it is acting within its legal mandate strikes me as vandalism.