We've updated our Privacy Policy to make it clearer how we use your personal data.

We use cookies to provide you with a better experience. You can read our Cookie Policy here.

Advertisement

LSD for Anxiety: A Deep Dive Into a New Clinical Trial

Four tabs of LSD, with a heart logo, are pictured on foil.
Read time:
 

A Phase II clinical trial investigating the efficacy of the psychedelic compound lysergic acid diethylamide (LSD)-assisted psychotherapy for the treatment of anxiety has reported positive results in the journal Biological Psychiatry. Participants given LSD showed greater reductions in their self-reported anxiety than those given a placebo. However, mixed data from a second group in the trial somewhat complicates the picture, highlighting the complex challenges of psychedelic clinical trial design.


The study took place across two centers: the University Hospital Basel and the clinic of Dr. Peter Gasser, who was also one of the study’s co-authors. Gasser, alongside first author Dr. Friederike Holze, a postdoctoral researcher in the Liechti lab at the University of Basel, presented the trial’s results at the recent Interdisciplinary Conference on Psychedelic Research (ICPR) in Haarlem, The Netherlands.

Crossover conundrums

Holze and Gasser’s study aimed to recruit people with anxiety disorders. Of the 42 participants enrolled, half of the group also had a life-threatening illness, such as cancer. The original goal of the trial was to examine how LSD-assisted psychotherapy affected anxiety symptoms with and without the context of a potentially fatal illness. The study used a “crossover” design, where participants were split into two groups, one of which received LSD and the other a non-psychedelic placebo. After a 24-week study period, the two groups were then brought back in, but this time their treatments were reversed. This approach, Gasser explained in his presentation, was an “ethical choice” – to make sure that everyone in the trial had a chance to access the LSD-assisted psychotherapy.


However, it was this well-intentioned design that ultimately necessitated a major rethink for the authors.


The trial involved two dosing sessions where patients were given an oral solution of 200 µg of LSD mixed in ethanol, or an ethanol-only placebo. These two sessions, six weeks apart, were supplemented by multiple non-psychedelic therapy sessions to help patients comprehend and integrate their experiences. The patients’ anxiety levels were measured using a patient-reported scoring system called the Spielberger’s State-Trait Anxiety Inventory–Global (STAI-G). Their anxiety was measured at baseline, 2, 8 and 16 weeks after their second dose of LSD.


After this period, Holze and Gasser intended to swap their study arms over to complete their crossover design.


The snag: their LSD patients were having too good a time.

Carryover effects and powerful placeboes

“We underestimated the duration of the LSD effect. Patients essentially stayed well after the first dose until they switched over into the placebo group,” Holze explains to Technology Networks. This meant that this group, termed the “LSD-first” group, had much lower anxiety scores when they began receiving placebo alongside psychotherapy in the study’s second half. This type of clinical trial issue is termed a carryover effect. On top of this, the LSD-first group continued to improve their scores in the study’s second half, even without any additional LSD being given.


A figure showing the desing of a study assessing the efficacy of LSD-assisted psychotherapy for anxiety.

The clinical trial design. One group is given two doses of LSD, while the other is given a placebo. The two groups are then swapped over after week 16. Participants are rated using various psychometric tests. Credit: Holze et al.


While great for the patients, these issues were problematic for Holze and Gasser’s data, as the effects of the LSD in the first half of the trial could influence the results of the second half of the trial. The authors took the decision to analyze only the first half results.


This data showed that patients given LSD had significantly greater reductions in their anxiety than those given placebo. The LSD group reported lower anxiety scores, on average 14.9 points lower on the STAI-G than at baseline, whereas the mean change for placebo was a 1.3-point increase. The difference between the two groups was statistically significant.


Additionally, the patients given LSD showed improvements in their scores on secondary outcomes, including measures of depression, as compared to placebo. The improvements patients experienced to their anxiety corresponded with the intensity of some of the psychedelic experiences they felt during LSD dosing, including the sensation of oceanic boundlessness (a commonly used metric in subjective measurement of psychedelic trips).


The patients’ anxiety reduction was first detectable at the two-week mark post treatment, write the authors. Patients with a life-threatening illness showed similar improvements to those with only anxiety, Holze says: “The study was also not statistically powered to assess such a difference validly. You would need more patients for a valid sub-group analysis. But we can say that LSD worked in both groups.”


Reassuringly, there was just one treatment-related adverse event encountered during the trial – a short-lived delusional state that was successfully treated with a combination of benzodiazepine and antipsychotic drugs.

A note of caution

This data all sits nicely together: LSD was extremely effective in reducing anxiety and depression, this benefit was linked to the psychedelic experience and there were limited side effects. But one observation in the paper’s discussion adds a note of caution to the findings.


The participants who were given placebo first, before waiting 16 weeks and then being given LSD, benefited less from the drug than those given LSD in the first half of the trial. This isn’t a psychedelic carryover effect, as these volunteers didn’t initially receive any active compound for there to be an effect from. Holze admits that this data confused the team: “We do not have an answer. Just speculations,” she says. “We were quite surprised to see a smaller response in those subjects who first had placebo and then LSD. These subjects mostly knew they would get LSD and still the response was somewhat smaller (still robust). Not clear why. They had more preparation and knew the therapist better, but this did not seem to enhance the response.”


Holze suggests that the effect could be due to chance – crossover study designs are rare in psychedelic research, so there isn’t data to suggest otherwise – but the finding does raise questions about how much of the powerful anti-anxiety effect experienced by the first group of participants receiving LSD was due to the drug and how much was due to the context of the trial, says Dr. Michiel van Elk, an assistant professor in the Institute of Psychology at the University of Leiden, who was not involved with the study.


van Elk, who presented at ICPR on the topic of improving clinical trials in psychedelics, points out that if the authors had pooled data from both groups receiving LSD, rather than only using data from the first group, the effects of the drug would have been markedly reduced.  


van Elk notes that the first group receiving a placebo showed very little improvement to their symptoms. “This looks very uncommon. Typically, you see an improvement, both for the placebo and for the psychedelic condition. People always improve because receiving something rather than nothing helps.”


van Elk suggests that this absence of an effect could be due to the trial’s crossover design. Placebo-receiving participants would likely have become unblinded, realizing what they had received during the first half of the study, although Holze and colleagues did not report blinding efficacy in the placebo group. Unblinding is an ongoing problem in a field where the active treatment usually induces intense hallucinations that are somewhat hard to miss. Given that participants were aware of the structure of the trial, placebo participants “would know the best is yet to come,” says van Elk. This knowledge may have contributed to a reduction in the placebo effect in the first half of the trial, which in turn makes the performance of the LSD arm look even more impressive.


The final consideration in the trial structure was the “ethical” decision to give LSD to all participants. This discussion was a focus at ICPR, where proponents of this approach insisted that withholding such a treatment from patients would be morally suspect. van Elk sympathizes with this view but points out that LSD’s efficacy for anxiety is not yet a sure thing: “We need to [withhold treatment from the placebo group] in the first place to establish whether a treatment is effective.” The assumption that LSD should be given to everyone because it is going to work adds a bias of expectancy to both patients and the trial team that could further enhance the drug’s effect, an issue discussed at length in a recent review of psychedelic clinical trials. van Elk notes a compromise approach that withholds the active drug from the placebo group but makes it freely available to that same cohort after the trial has concluded, which may strike a balance in this complex ethical dilemma.


As with many psychedelic trials, Holze and colleagues’ work remains at an early stage. The going is long and painstaking: Gasser mentioned in his presentation that the project had taken nearly 15 years to complete. Promising initial results, however, don’t disguise the need to improve and establish robust protocols for how psychedelic clinical trials should be designed.


Dr. Michiel van Elk and Dr. Friederike Holze were speaking to Ruairi J Mackenzie, Senior Science Writer for Technology Networks

Reference:

Holze F, Gasser P, Müller F, Dolder PC, Liechti ME. Lysergic acid diethylamide-assisted therapy in patients with anxiety with and without a life-threatening illness A randomized, double-blind, placebo-controlled Phase II study. Biological Psychiatry. 2022. doi:10.1016/j.biopsych.2022.08.025

Meet the Author
Ruairi J Mackenzie
Ruairi J Mackenzie
Senior Science Writer
Advertisement