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

Anxiety is common across psychiatric disorders and frequently accompanies life‑threatening illnesses (LTIs) such as cancer. Current treatments, principally selective serotonin reuptake inhibitors and psychotherapy, require ongoing administration and show limited efficacy for many patients, motivating investigation of alternative approaches. Classic psychedelics act primarily at the serotonin 5‑HT2A receptor and can produce profound acute alterations of consciousness and mystical‑type experiences; modern clinical research has focused largely on psilocybin, whereas LSD has been less rigorously studied in contemporary controlled trials despite historical use in patients with terminal illness. Holze and colleagues designed this trial to evaluate whether LSD‑assisted therapy produces clinically meaningful and sustained reductions in anxiety and related symptoms in patients with anxiety both with and without an associated life‑threatening illness. The investigators hypothesised that two high‑dose LSD sessions would reduce anxiety at 16 weeks after treatment and that acute subjective effects measured with established scales (5D‑ASC and MEQ30) would correlate with longer‑term therapeutic benefit.

The trial was an investigator‑initiated, two‑centre, randomized, double‑blind, placebo‑controlled, two‑period crossover study, with each period containing two full‑day treatment sessions and five study visits. Treatment sessions within a period were separated by about six weeks. Participants were randomised to receive either two LSD sessions (200 µg oral solution) in the first period and placebo in the second, or vice versa, with order counterbalanced. The inactive placebo matched the vehicle solvent (ethanol) used to deliver LSD. Study medication was prepared under good manufacturing practice. Eligibility criteria excluded people aged under 25, those with current or past primary psychotic disorders or bipolar disorder, first‑degree relatives with psychotic disorders, active substance use disorder in the prior two months (except nicotine), elevated suicide risk, likely need for psychiatric hospitalisation during the study, and central nervous system involvement of cancer. Participants could remain in ongoing psychotherapy and continue stable psychiatric medications, but serotonergic drugs and other specified psychotropics were tapered off prior to each session using a washout that equalled about five times the drug half‑life. Patients were asked to avoid other psychoactive substances within 24 hours of sessions. Each treatment session lasted approximately 12 hours, conducted in a calm clinical room with one therapist assigned for the study duration, and included talking psychotherapy before and after drug administration. Outcome assessments occurred at baseline, between sessions, and at 2, 8 and 16 weeks after the second session of each period. The predefined primary endpoint was change in global anxiety (Spielberger State‑Trait Anxiety Inventory‑Global, STAI‑G) from baseline to 16 weeks after treatment. Secondary endpoints included other anxiety subscales (STAI‑S, STAI‑T), depression scales (HAM‑D‑21, BDI), general psychiatric symptoms (SCL‑90‑R), acute subjective effects (5D‑ASC and MEQ30), autonomic measures (blood pressure, heart rate), adverse events, and correlations between acute responses and long‑term outcomes. Sample size planning targeted about 40 patients to achieve 80% power for a prespecified effect, allowing for non‑evaluable participants. Analyses used a mixed model for repeated measures (MMRM) with fixed effects for treatment, visit, treatment by visit interaction, period and sequence where appropriate, and patient as a random factor; baseline values were included as covariates and life‑threatening illness status was a stratification factor. Because of observed carryover from the first to second period, the preplanned crossover analysis was superseded for the primary outcome by a between‑subjects analysis restricted to the first‑period data using the MMRM without period and sequence factors. Pearson correlations described associations between acute measures and change in STAI‑G. The significance threshold was p < 0.05.

Enrollment ran from June 2017 to February 2021 with the last patient visit in December 2021. Forty‑two patients who completed both treatment sessions and at least one outcome visit in the first period were included in the primary between‑subjects analysis; 20 had a life‑threatening illness and 22 had anxiety disorders without a somatic illness. LSD produced rapid and sustained reductions in anxiety compared with placebo in the first treatment period. At 16 weeks after the last session, the least‑square mean change from baseline in STAI‑G was −14.9 (±4.1) in the LSD group and 1.3 (±4.1) in the placebo group, yielding a between‑group difference of −16.2 (95% CI −27.8 to −4.5; p = 0.007). Effects were largest at 2 weeks after the second LSD session and were maintained through week 16. Using a clinical response threshold of ≥ 30% reduction in STAI‑G, 13 of 20 patients (65%) in the LSD group versus 2 of 22 patients (9%) in the placebo group met response criteria at any outcome visit (p = 0.003). Secondary measures including STAI‑S, STAI‑T, HAM‑D‑21, BDI and SCL‑90‑R showed similarly rapid and sustained improvements consistent with reductions in anxiety, depressive symptoms and overall psychiatric symptomatology, as reported in the first period analyses. Acute subjective effects measured during LSD sessions correlated with long‑term anxiety reductions: Oceanic Boundlessness (a 5D‑ASC subscale) and the MEQ30 total score correlated negatively with change in STAI‑G at week 16 (r = −0.67, p = 0.001; r = −0.62, p = 0.003; n = 20), whereas Anxious Ego Dissolution did not (r = 0.049, p = 0.83). Peak heart rate during the LSD experience also showed a modest correlation with longer‑term outcome (Emax r = −0.49, p = 0.026, n = 20). These acute–long‑term associations were observed in the group that received LSD in the first period but not when only second‑period data were analysed. Safety and tolerability data were reported over the entire trial: nine serious adverse events (SAEs; 21%) occurred overall, six during the LSD period and three during placebo; only one SAE (2%) was judged treatment‑related (an acute transient episode of anxiety with delusional features during an LSD session), successfully managed with lorazepam and a single dose of olanzapine, followed by overnight observation and recovery. During LSD sessions, eight patients (19%) reported transient untoward acute effects: anxiety (3 patients, 7%), nausea (4 patients, 10%) and headache (1 patient, 2%). Three patients (7%) had their second LSD dose reduced to 100 µg because the first session was judged too intense. Across the full study there were 229 additional adverse events (105 during LSD periods, 124 during placebo periods) that were not considered treatment‑related. LSD produced moderate, transient increases in blood pressure and heart rate during sessions; no clinically relevant persistent perceptual disturbances or flashback phenomena were reported. Outcome measures did not differ according to concurrent antidepressant cotreatment, and concomitant anxiolytic/antidepressant use did not change substantially during the trial.

Holze and colleagues interpret their findings as evidence that two high‑dose LSD sessions within a psychotherapeutic framework produced rapid and durable reductions in anxiety, depressive symptoms and general psychiatric distress up to 16 weeks compared with inactive placebo in the first treatment period. The pattern of results aligns with prior smaller pilot studies of LSD in patients with life‑threatening illness and with randomised trials of psilocybin in cancer patients, supporting a view that psychedelic‑assisted therapy can yield sustained symptom improvement beyond the acute drug effects. The investigators note that acute positive subjective experiences—particularly Oceanic Boundlessness and mystical‑type experiences measured by the MEQ30—were associated with better long‑term outcomes, a finding consistent with prior psilocybin research. Autonomic responses (peak heart rate) also correlated with outcome but to a lesser extent. The authors suggest that such acute response measures might serve as predictors or biomarkers of therapeutic response. Key limitations acknowledged include the use of an inactive placebo that made blinding difficult, with most participants becoming aware of treatment assignment once LSD effects were experienced. Carryover and order effects in the crossover design were substantial: patients who received LSD in the first period showed persistent benefit into the second period, complicating within‑subject comparisons and prompting reliance on a first‑period between‑group analysis for the primary endpoint. The intensive one‑to‑one treatment setting and high therapist expertise may limit generalisability to less specialised clinical contexts, and participants were self‑referred and motivated, which may also affect external validity. The authors further discuss dose considerations, noting that the 200 µg LSD dose is high and may provoke acute anxiety in some patients; they propose that starting doses of 100–150 µg with optional escalation may be preferable in future trials. Strengths highlighted are the relatively large sample for a psychedelic trial, the randomized double‑blind placebo‑controlled design, the long 16‑week follow‑up, and the inclusion of patients with anxiety both with and without LTIs. The authors conclude that the results support potential long‑term benefits of LSD‑assisted therapy for anxiety but state that larger trials are needed to confirm these findings and to address issues such as optimal dosing, blinding strategies, and broader generalisability.