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

BACKGROUND
This study aimed to investigate the efficacy and safety of lysergic acid diethylamide (LSD)-assisted therapy in patients who experienced anxiety with or without association with a life-threatening illness.


METHODS
The study is an investigator-initiated 2-center trial that used a double-blind, placebo-controlled, 2-period, random-order, crossover design with 2 sessions with either oral LSD (200 μg) or placebo per period. The primary end point was anxiety symptoms 16 weeks after the last treatment session, assessed by the Spielberger State-Trait Anxiety Inventory-Global score in 42 patients. Further outcome measures included ratings for depression symptoms (Beck Depression Inventory and Hamilton Depression Rating Scale, 21-item version) and ratings for acute subjective drug effects. The outcomes for the first period (between-subjects analysis) are primarily shown due to carryover effects.


RESULTS
LSD treatment resulted in significant reductions of State-Trait Anxiety Inventory-Global scores up to 16 weeks after treatment (least-square mean [standard error] change from baseline difference = -16.2 [5.8], 95% CI, -27.8 to -4.5, d = -1.18, p = .007). Similar effects were observed for ratings of comorbid depression on the Hamilton Depression Rating Scale, 21-item version (-7.0 [1.9], 95% CI, -10.8 to -3.2, d = -1.1, p = .0004) and the Beck Depression Inventory (-6.1 [2.6], 95% CI, -11.4 to -0.9, d = -0.72, p = .02). Positive acute subjective drug effects and mystical-type experiences correlated with the long-term reductions in anxiety symptoms. Transient, mild, acute untoward effects of LSD treatment were reported by 8 patients (19%). One treatment-related serious adverse event (acute transient anxiety) occurred (2%).


CONCLUSIONS
LSD produced long-lasting and notable reductions in anxiety and comorbid depression symptoms up to 16 weeks.

Anxiety is a common symptom of several mental illnesses and the leading symptom of anxiety disorders (1). In addition, cancer and other life-threatening diseases are commonly associated with anxiety (2). Treatment options include mainly selective serotonin reuptake inhibitors and psychotherapy. However, antidepressants need to be administered daily and have limited efficacy. Therefore, alternative treatment options are needed.
Lysergic acid diethylamide (LSD) is a well-known classic psychedelic substance. Psychedelics, including LSD and psilocybin, induce profound acute alterations of mind and mystical-type experiences primarily by interacting with the serotonin 5-HT 2A receptor (3)(4)(5). LSD has a more turbulent history than psilocybin. In addition, acute effects of LSD last longer than those of psilocybin. However, a recent study documented otherwise comparable acute subjective effects and tolerability of LSD and psilocybin in healthy subjects (6). In the 1950s to 1970s, LSD was used in patients with end-stage cancer to improve mood and reduce anxiety (7)(8)(9). However, these studies did not use rigorous trial methods. Modern research on the therapeutic potential of psychedelics mostly used psilocybin. A small pilot study reported the safety of a moderate dose of psilocybin in patients with advancedstage cancer (10). Two randomized, placebo-controlled trials in patients with life-threatening cancer reported significant reductions in anxiety and depression up to 5 and 6 weeks after a single dose of psilocybin (11,12).
A small pilot study for this trial used 2 high doses of LSD (200 mg) or an active placebo (20 mg LSD) in patients with lifethreatening illness (LTI). The study found trends toward reductions of anxiety up to 2 months compared with placebo (13,14). Therefore, the aim of this larger study was to corroborate these findings in patients with LTI. In addition, psychedelics have not yet been investigated in patients with psychiatric anxiety disorders in the absence of an LTI. Therefore, this trial investigated effects of LSD on anxiety, depression, and overall psychiatric symptomatology in patients with anxiety with and also without an LTI using the Spielberger State-Trait Anxiety Inventory (STAI), the Hamilton Depression Scale, 21-item version (HAM-D-21), the Beck Depression Inventory (BDI), and the Symptom-Checklist-90-R (SCL-90-R). Previous studies showed that acute effects of psilocybin on the 5 Dimensions of Altered States of Consciousness scale (5D-ASC; Oceanic Boundlessness but not Anxious Ego Dissolution ratings) (15) and Mystical Experience Questionnaire, 30-item version (MEQ30, total scores) were associated with long-term therapeutic effects on anxiety and depression (11,12,15). Therefore, we hypothesized that acute effects of LSD on these 5D-ASC and MEQ30 scores would similarly correlate with long-term reductions in anxiety.

Study Design and Participants
The study used a double-blind, placebo-controlled, 2-period, Patients were recruited through an advertisement that was placed on the homepages of the University Hospital Basel and Swiss Medical Society for Psycholytic Therapy trial registries or by word of mouth. All patients provided written informed consent before study inclusion. Written informed consent was obtained by the study psychiatrist who conducted the screening visit. Both patients and investigators were aware of the 2-period, placebo-controlled study design. Patients were informed that the study would investigate possible therapeutic benefits. The goal was to include patients with anxiety disorders or significant anxiety that was associated with a lifethreatening somatic illness. LTI was defined as any severe somatic disease, such as a diagnosis of cancer or another advanced-stage potentially fatal illness. Patients with an LTI had to meet the DSM-IV criteria for an anxiety disorder, including generalized anxiety disorder, social phobia, and panic disorder, as indicated by the Structured Clinical Interview for DSM-IV or have a score $ 40 on the state or trait STAI scale at study inclusion. Patients without an LTI had to meet DSM-IV criteria for at least 1 anxiety disorder. Thus, in patients without an LTI, elevated STAI scores were not sufficient for inclusion.
The exclusion criteria included age ,25 years, past or present diagnosis of a primary psychotic disorder, patients with a first-degree relative with psychotic disorders, past or present bipolar disorder (according to the DSM-IV), current substance use disorder (within the past 2 months according to the DSM-IV, except nicotine), suicide risk, likely to require psychiatric hospitalization during the course of the study, and central nervous system involvement of the cancer (Supplemental Methods). Patients were allowed to be in ongoing psychotherapy and were asked not to change therapists, not to increase or decrease the frequency of their therapy, not to commence any new type of therapy during the study, and generally to keep their psychiatric medications. Psychiatric medications, such as antidepressants, neuroleptics, and anxiolytic medications, on a fixed regimen, were tapered off typically 2 weeks before each LSD/placebo treatment session (i.e., 5 times the duration of a particular drug's half-life). Patients were instructed to refrain from using any psychoactive drugs, with the exception of long-term pain medication, within 24 hours of each LSD/placebo treatment session.
After study inclusion, the patients were randomly assigned to LSD or placebo in the first treatment period and vice versa in the second treatment period by order of enrollment and group. LSD-free base (. 99% purity; Lipomed AG) was administered as an oral solution in units that contained 100 mg LSD in 1 mL of 96% ethanol (16). Inactive placebo consisted of identical units that were filled with ethanol only. Randomization and production according to good manufacturing practice were performed by a licensed good manufacturing practice facility (Apotheke Dr. Hysek, Biel, Switzerland).

Procedures
The study included a screening visit and two 24-week treatment periods per participant. Each treatment period consisted of 2 treatment sessions and 5 study visits. Treatment sessions were separated by 6 weeks (62 weeks). Study visits were conducted at baseline (bl visit), between the sessions (btw visit), and 2, 8, and 16 weeks after the second treatment session (week-2, week-8, and week-16 visits). The week-16 visit in the second period also served as the end-of-study visit (Figures 1 and 2). Patients were generally called by phone by the therapists within the next few days after treatment sessions for a brief follow-up.
Screening consisted of written informed consent, an evaluation of the patient's physical and mental health background, a psychiatric interview (Structured Clinical Interview for DSM-IV), an assessment of anxiety severity (STAI), an assessment of depression (HAM-D-21 and BDI), and an assessment of further psychiatric symptomology (SCL-90-R) and a physical checkup. After successful screening, each patient was assigned to one investigator/therapist (Table S1) who conducted all treatment sessions and study visits for the whole duration of the study. Treatment sessions lasted approximately 12 hours, and study visits lasted approximately 1 hour (Supplemental Methods).
Study visits consisted of talking psychotherapy (Supplemental Methods) followed by an assessment of adverse events, changes in general medication, and administration of the STAI, HAM-D-21, BDI, and SCL-90-R.
Treatment sessions were conducted in a calm hospital room (University Hospital Basel) or in a calm practice room (Clinic Dr. Peter Gasser). Only 1 patient and 1 investigator/therapist were present during the treatment sessions (exceptions of more than 1 therapist being present were made upon request by the therapist or patient) (Supplemental Methods).

Outcome Assessments
The predefined primary end point of the study per protocol was the change in STAI-Global (STAI-G) score from baseline at week 16 after LSD compared with placebo within subjects. Because of the long-lasting effects of LSD and resulting carryover effects ( Figure 3) from the first to the second treatment period, the primary outcome analysis of the crossover (i.e., every patient served as his/her own control) was changed to week 16 after LSD compared with placebo in the first period (between subjects). Secondary end points were STAI scores at the btw, 2-, and 8-week visits and HAM-D-21, BDI, and SCL-90-R scores at the btw, 2-, 8-, and 16-week visits. Clinical response was defined as a STAI-G score reduction $30%. Further secondary end points were acute subjective drug effects during treatment sessions, assessed by the 5D-ASC and the MEQ30, acute autonomic drug effects during the treatment session (systolic and diastolic blood pressure and heart rate), adverse events (AEs), and serious AEs (SAEs) (during treatment sessions and during the entire study duration). In addition, correlations between acute LSD effects (MEQ30 total score, Oceanic Boundlessness, Anxious Ego Dissolution, and heart rate) and long-lasting therapeutic effects (16 weeks) on the STAI-G were assessed (Supplemental Methods).

Statistical Analyses
The sample size was calculated based on the pilot study (14) and using PASS (NCSS, Statisical Software; https://www.ncss. com/software/pass/), which required 80% power at a 2-sided alpha of 0.05 using a 10% difference in anxiety scores with a standard deviation of 15%, and was estimated to be 40 patients, allowing for 10 nonevaluable patients. Outcomes were analyzed using a mixed model for repeated measures (SAS 9.4 software; SAS Institute), including treatment, visit, treatment 3 visit interaction, period, and sequence as fixed factors and patient as the random factor. Baseline values of the periods were considered covariates (change from baseline values for the respective period). LTI was used as a stratification factor. The crossover analysis was planned, but in case of being not fully valid because of carryover effects and withdrawals, a betweensubjects analysis that was restricted to the first-period data was foreseen to substitute for the primary analysis of the full crossover data using the mixed model for repeated measures without the period and sequence factors. Outcomes were predefined and preregistered in Clinicaltrials.gov, and analyses were predefined a priori in the protocol. Pearson correlations were used to describe relations between acute effects of LSD on the 5D-ASC and MEQ30 and change in the STAI-G score from baseline at week 16. The criterion for significance was p , .05.  LSD produced strong reductions in anxiety, depression, and general psychiatric symptomatology compared with placebo in the first treatment period (Table 2; Figure S1). The least-square mean (6 standard error) changes from baseline in the STAI-G score at 16 weeks after the last session were 214.9 (64.1) in the LSD group and 1.3 (64.1) in the placebo group (difference = 216.2, 95% CI, 227.8 to 24.5, p = .007), indicating a significant difference between treatment groups (Table 2; Figure S1). Effects were maximal 2 weeks after the second LSD session and were sustained up to 16 weeks. Thirteen of 20 patients (65%) in the LSD group and 2 of 22 patients (9%) in the placebo group showed a clinical response at any outcome visit ($30% reduction of STAI-G scores, p = .003) ( Table S10). The secondary end point measurements, including the STAI-State, STAI-Trait, HAM-D-21, BDI, and SCL-90-R scores, all showed similarly rapid and sustained responses (Table 2; Figure S1) that were consistent with a lasting treatment effect on anxiety, depression, and general psychiatric symptomatology. Outcomes of the crossover analysis and the subgroups can be found in Figures S4-S8 and Tables S2-S7. Proportions of patients with the remission of depression are shown in Table S10. Individual responses and the time course can be found in Figures S11 and S13.

Enrollment began on
Acute subjective effects of LSD during the sessions of the first treatment period were significantly associated with the long-term outcome of a reduction in anxiety ( Figures S9, S10, S14-S16). Specifically, Oceanic Boundlessness and MEQ30 total score but not Anxious Ego Dissolution (average scores of both sessions) correlated with changes in STAI-G scores from baseline at week 16 in the between-subjects analysis Reporting Trials) profile. A total of 243 people contacted the study team and were prescreened. Of these, 62 attended a screening visit, and 46 were enrolled in the study. Two patients withdrew consent after enrollment and before randomization, resulting in 44 randomized patients. Twenty-one patients were assigned to the treatment group that received lysergic acid diethylamide (LSD) in the first period, and 23 patients were assigned to the group that received a placebo in the first period. There was a total of 7 dropouts. LSD-Assisted Therapy in Patients with Anxiety (r = 20.67, p = .001, r = 20.62, p = .003, and r = 0.049, p = .83, respectively; n = 20). Heart rate (E max ) during the acute LSD experience also correlated with long-term therapeutic outcome (r = 20.49, p = .026, n = 20) when analyzed between subjects and for the first treatment period. There were no associations if only the data from the second treatment period were analyzed.
During the entire study duration, a total of 9 (21%) SAEs occurred. Six SAEs occurred during the LSD period, and 3 occurred during the placebo period. Only 1 SAE (2%) was considered related to treatment and consisted of acute transient anxiety and delusions during an LSD session. The patient was successfully treated with lorazepam and olanzapine. A single dose of olanzapine was administered because lorazepam alone was not effective enough to fully block all symptoms. The patient was kept overnight and discharged in the morning and experienced no further long-term symptoms.
Consequently, the second LSD dose was reduced to 100 mg for this patient. Additional non-treatment-related SAEs in the LSD period that occurred outside study visits are described in ). In a total of 3 patients (7%), LSD in the second session was reduced to 100 mg (1 patient mentioned above and 2 other patients) because effects during the first session were considered too strong by the patient and therapist. There were no untoward effects during placebo sessions. A total of 229 additional AEs were reported during the entire trial duration (105 during the LSD period and 124 during the placebo period). None of these AEs were considered related to treatment. A detailed listing of all AEs is presented in Table S8. Autonomic effects are shown in Figure S15.
Outcome measures did not differ in patients with or without antidepressant cotreatment. The use of anxiolytic and antidepressant medications did not change substantially during the study (Supplemental Results).

DISCUSSION
In this double-blind, placebo-controlled, randomized trial, LSD administration during 2 treatment sessions induced rapid and lasting reductions in anxiety, depression, and general psychiatric symptomatology for up to 16 weeks. There was a significant reduction in the primary outcome measure of anxiety  (13,14). These findings are also consistent with a pilot study that used psilocybin in patients with advanced-stage cancer (10) and documented trends toward improvements in mood and anxiety. In addition, the findings of our study align with 2 randomized, placebo-controlled trials that used single administrations of psilocybin (22 or 30 mg/70 kg or 0.3 mg/kg) in patients with life-threatening cancer. These studies reported comparable reductions in anxiety and depression 5-6 weeks after the psilocybin session compared with an active placebo (1 or 3 mg/70 kg psilocybin or 250 mg niacin) (11,12). Effect sizes were comparable to this study, whereas very high and higher effect sizes than those herein were reported for reductions in depression in studies that used psilocybin without a placebo control group (17,18). However, trials that include a placebo condition likely more realistically reflect the benefits of psychedelic treatments. Reduction in the SCL-90-R overall scores indicates that psychedelic treatment might be transdiagnostically effective (19).
LSD was generally well tolerated this study. However, one SAE was related to LSD (i.e., transient anxiety and delusion reaction during the session). LSD may produce acute anxiety in some people, particularly at high doses (4,20). Specifically, effects of LSD at the dose and formulation that were used in this study have been characterized very well in healthy subjects, and the dose could be considered high and have a greater risk of acute anxiety (4,6). Transient anxiety and confrontation with a fear of death were also reported in the pilot study. Similar challenging emotions were also reported with psilocybin in patients with cancer (11,12) and patients with  LSD-Assisted Therapy in Patients with Anxiety Values are score changes from baseline shown as differences between LSD and placebo reported as least-square mean (95% confidence interval) in 42 patients. Btw visit indicates 5 weeks, and W2, W8, and W16 sessions indicate 10, 16, and 24 weeks, respectively. BDI

LSD-Assisted Therapy in Patients with Anxiety
Biological Psychiatry February 1, 2023; 93:215-223 www.sobp.org/journal major depressive disorder (18). According to our clinical impression, the 200-mg dose of LSD might be too high for some patients, especially if they are not experienced with the effects of psychedelic drugs. Thus, a first dose of 100 or 150 mg LSD may be more adequate in future studies, with an optional increase to 150 to 200 mg for further doses. The 200mg dose of LSD that was used in this study can be considered equivalent to 40 mg of psilocybin (6) and higher than the psilocybin doses of 15 to 30 mg that have been used in clinical trials to date. In addition, the high LSD dose was administered twice. The use of 2 doses has been strongly recommended (10,13), allowing patients to become more familiar with the effects of psychedelics and potentially have different experiences, particularly in cases in which the first dose produced a challenging experience.
The safety of LSD has been documented in healthy subjects (21). In this study, LSD moderately increased blood pressure and heart rate during the sessions compared with placebo and similar to studies in healthy subjects (4,6,21). In this study, no perceptual alterations were reported beyond the acute effects of LSD, consistent with the absence of clinically relevant flashback phenomena in controlled studies with LSD in healthy subjects (22). No participants reported acute suicidality at any time. One patient with a diagnosis of major depressive disorder reported transient feelings of depression, including suicidal thoughts (but no acutely increased suicidality, including preparation), 8 weeks after the last LSD treatment. AEs were comparable in the LSD and placebo periods.
In this study, LSD produced marked alterations of mind and overall very positive acute experiences, indicated by high Oceanic Boundlessness ratings in the upper range of the patients in limited LSD use programs (23) or in healthy subjects (4,6,20). Importantly, this study found that LSD-induced acute positive subjective effects (Oceanic Boundlessness score) and mystical-type experiences (MEQ30 total score) were associated with the long-term therapeutic outcome as similarly shown for psilocybin (11,12,15), indicating that acute positive effects of psychedelics may serve as a treatment response predictor and biomarker. Similarly, acute mystical-type effects of LSD (200 mg) correlated with changes in wellbeing/life satisfaction in healthy subjects (24). However, heart rate (E max ) as an autonomic measure of the acute drug response also correlated with the therapeutic effects, although this association was weaker.
In this study, there were differences in the treatment response between the group that received LSD in the first period and the group that received LSD in the second period. The LSD-first group showed an overall stronger response with a partial relapse after week 10, whereas the LSD-second group showed an overall weaker response with less relapse. In addition, the correlations between acute effects and therapy outcomes were seen only in the group that received LSD in the first period. This raises questions about the interactions between expectancies, psychotherapy, and pharmacological treatments that require more investigation.
This study has several strengths. It was relatively large. A total of 161 full-day sessions (80 LSD and 81 placebo) were conducted, and the study duration per patient was 12 months. The study used a randomized, double-blind, placebocontrolled design with a long 16-week follow-up for both treatments. In addition, all patients were intended to be treated with both treatments within the crossover design. Previous studies with psychedelics in patients with anxiety or depression used no or not a sufficiently large placebo control group (14,17,18) or used relatively short follow-up times after the last treatment session of only 2 weeks (10), 3 weeks (25), 4 weeks (18), 5 weeks (11), and 6 weeks (12) during the phase of the trial when a placebo group was run in parallel. Thus, the time course of effects and sustained effects up to 16 weeks after a psychedelic-assisted treatment session could be documented for the first time in a clinical trial compared with a placebo control group. In addition, symptom scores remained low in the LSD-first group until the end of the trial, potentially indicating persisting effects up to 12 months (without a control). Finally, to our knowledge, it was the first study with a psychedelic to include patients with anxiety without an LTI.
This study also has limitations. We used an inactive placebo as the control condition. The characteristic effects of LSD unblinded the treatment order assignment in most patients once the effects of LSD were perceived. Only 1 patient in the LSD-first group mistook LSD as a placebo and realized that he had had LSD the first time only when he received placebo during the second study phase. Measures of subjective expectancy were not included. Other studies with psychedelics used an active placebo, such as niacin, or a low dose of the psychedelic (e.g., 25 mg LSD) as the control condition. It remains to be shown whether blinding can be improved with these active placebo approaches, and valid blinding remains a challenge in any trials that use psychedelics as well as in many other clinical trials. The use of a crossover design can also be problematic. Patients in the LSD-first group showed therapeutic effects that persisted into the second period, thereby lowering baseline ratings of this period. In addition, a placebo in the second period tended to further improve outcome ratings. This finding may be explained by the intensive entire-day one-to-one patient-therapist interactions, potentially including partial re-experience of the LSD sessions and integrating and consolidating its effects. Such a consolidation response was not possible in the placebo-first group. In the placebo-first group, the patients and therapist spent more time together before the LSD sessions, but this seemed not to positively influence the responses to LSD treatment. Thus, carryover and order effects in the crossover occurred. A possible advantage of the crossover design was that each patient was sure to receive LSD at some point during the study. Notably, retention in the study was also high in patients who received LSD first. Another limitation is related to the setting and experience of the investigators and therapists. The therapists were mostly highly motivated and also very experienced with psychedelicassisted therapy and had previously treated patients with LSD within limited use programs (23) and the pilot study (13,14). The therapist treated the patients one-to-one within a highly intensive treatment setting that may not be generally available in other trials and future settings of psychedelic-assisted therapy. Patients were also motivated, reflected by the relatively low dropout when considering the long study duration and inclusion of a placebo period. Only motivated patients who initiated contact with the study team on their own were included. Patients were allowed to continue their treatments, including psychotherapy and drug therapy, with the exception of pausing serotonergic agents before treatment sessions for LSD-Assisted Therapy in Patients with Anxiety safety reasons. Thus, although many patients were already receiving treatment for their anxiety disorder, LSD-assisted treatment further improved their symptomatology.
In conclusion, this study suggests long-term benefits of LSD-assisted therapy in patients with anxiety disorders. Further, larger trials are needed to confirm these findings.