Pharmacokinetics and subjective effects of a novel oral LSD formulation in healthy subjects

Earlier research has produced only limited and fragmentary pharmacokinetic (PK) data for lysergic acid diethylamide (LSD) in humans, with small intravenous studies and short, sparsely sampled oral studies leaving uncertainty about standardised exposure–time profiles. Holze and colleagues note that previous oral PK work by their group used a capsule formulation that lacked documented long-term stability, creating a need for well characterised, stable formulations for both experimental and clinical use. This study set out to characterise the PK of a newly manufactured, analytically confirmed oral LSD solution intended for clinical research and patient use. Secondary aims were to describe acute subjective effects, link those effects to plasma concentrations using pharmacokinetic–pharmacodynamic (PK/PD) modelling to derive EC50 values (the concentration producing half-maximal effect), to quantify the main metabolites 2-oxo-3-hydroxy-LSD (O-H-LSD) and N-desmethyl-LSD (nor-LSD) in plasma, and to compare exposure from the novel solution with previously tested capsule data.

The investigators conducted a double-blind, placebo-controlled, balanced cross-over study with four 12-hour experimental sessions per participant (100 μg LSD solution, 125 mg MDMA, 40 mg D-amphetamine and placebo); washout periods between sessions were at least 7 days. Only the LSD and placebo data are reported here. The study was performed in a hospital setting with one participant and one investigator present in a testing room; participants received standardised small meals and were supervised throughout the 12-hour sessions. The protocol was approved by the local ethics committee and registered at ClinicalTrials.gov (NCT03019822). Twenty-seven healthy volunteers participated (the extracted text reports 27 subjects in PK/PD analyses). Participants were asked to limit alcohol and avoid xanthine-containing drinks before sessions; regular medications likely to interact with LSD were exclusionary. Five participants had prior limited hallucinogen exposure (one to four lifetime uses); 22 were hallucinogen-naïve. Seven participants smoked (mean 4 ± 3 cigarettes/day) and 20 were non-smokers. The investigators noted that tobacco induces CYP1A2 in vitro, which is implicated in LSD metabolism, but human interaction data are lacking. LSD was administered as a single oral dose of 100 μg LSD base formulated as a 1 mL solution in 96% ethanol under GMP conditions; batch content was analytically confirmed at 96.2 ± 0.3 μg. Stability testing for the formulation and information about iso-LSD formation over time and with temperature were reported. Blood sampling for PK was intensive up to 11.5 hours: pre-dose and at 1, 1.5, 2, 3, 3.5, 4.5, 5.5, 6.5, 7.5, 9.5 and 11.5 hours post-dose; plasma was centrifuged immediately and stored at -20°C short term or -80°C for long-term storage. Plasma concentrations of LSD, O-H-LSD, nor-LSD and iso-LSD were measured by ultra-high-performance liquid chromatography tandem mass spectrometry (UHPLC-MS/MS) with calibration lines from 25 to 10 000 pg/mL and quality-control samples at several concentrations; the lower limit of quantification (LLOQ) was 25 pg/mL. The assay used isotope-labelled internal standards and provided accepted accuracy and precision criteria. Iso-LSD concentrations were measured primarily in samples collected around 2 hours. For pharmacodynamic assessment, repeated Visual Analog Scales (VASs) (0–100 mm) assessed "any drug effect," "good drug effect," "bad drug effect" and "ego dissolution" at the same post-dose time points as PK sampling. Pharmacokinetic analyses included both compartmental and noncompartmental approaches, and within-subject PK/PD modelling was performed to link plasma concentrations to VAS responses and to estimate EC50 and Emax parameters. Previously published capsule data were reanalysed using the same PK methods to enable comparisons between the new solution and the earlier capsule formulation.

LSD plasma concentrations followed first-order elimination kinetics. The modelled geometric mean Cmax for the oral solution was 1.7 ng/mL (reported range 1.0–2.9 ng/mL) reached at a median tmax of 1.7 hours (range 1.0–3.4 h). The geometric mean terminal half-life was 3.6 hours (range 2.4–7.3 h) and the AUC∞ was 13 ng·h/mL (range 7.1–28 ng·h/mL). No sex differences were observed in PK parameters, and no significant differences were found between light tobacco smokers and non-smokers. Of the metabolites measured, O-H-LSD was quantifiable in all subjects with a geometric mean Cmax of 0.11 ng/mL (range 0.07–0.19 ng/mL) at a tmax of 5 hours (range 3.2–8 h); O-H-LSD t1/2 and AUC∞ were 5.2 hours (range 2.6–21 h) and 1.7 ng·h/mL (range 0.85–4.3 ng·h/mL), respectively. Nor-LSD levels were below the level of quantification (reported as <0.01 ng/mL or otherwise not quantifiable). Subjective effects paralleled plasma concentrations. Mean onset of subjective effects was reported as approximately 0.7 hours after dosing, peak subjective effects at 2.5 ± 0.6 hours (range 1.6–4.3 h), and mean effect duration 8.5 ± 2.0 hours (range 5.3–12.8 h). The drug produced robust increases in VAS measures of "any drug effect" and "good drug effect," while "bad drug effect" increased only moderately at the group level due to transient anxiety in some participants. PK/PD model estimates (mean ± SD) were: for "any drug effect" EC50 1.1 ± 0.4 ng/mL with Emax 91% ± 17%; for "good drug effect" EC50 1.0 ± 0.5 ng/mL with Emax 83% ± 20%; for "bad drug effect" EC50 1.9 ± 1.0 ng/mL with Emax 40% ± 41%; and for "ego dissolution" EC50 1.4 ± 0.5 ng/mL with Emax 80% ± 34%. Across subjects, Cmax did not correlate with individual Emax values for the VAS measures. When compared with previously published capsule data reanalysed in the same way, the model-predicted geometric mean Cmax for the solution (1.7 ng/mL; CV 27%, 95% CI 1.6–2.0) was higher but not statistically significantly different from the capsule (1.3 ng/mL; CV 60%, 95% CI 1.2–1.9). Subjective-effect PD parameters were broadly similar between formulations, except that the EC50 for "any drug effect" was reported lower for the capsule. Times of onset, offset, duration, time to peak effect and area under the effect–time curve did not differ significantly, but the time to effect onset showed greater variance with the capsule than with the solution. The investigators also noted detection of iso-LSD in some earlier capsule-based samples, raising the possibility that older formulations contained inactive decomposition products.

Holze and colleagues interpret the data as providing a detailed characterisation of the plasma concentration–time profile and acute subjective effects of a novel, stability-documented oral LSD solution intended for clinical research. The investigators highlight that subjective effects rose and fell in parallel with plasma concentrations and that first-order elimination kinetics and a terminal half-life of about 3.6 hours were confirmed. Mean subjective onset, peak time and duration were consistent with a relatively protracted time to maximal response compared with many other psychoactive drugs. The authors position these results alongside prior work using capsule formulations, noting generally similar PK and PD parameter estimates but higher Cmax and AUC with the present solution. They offer two explanations: greater oral bioavailability of the solution, and potential underestimation of true content in prior capsule batches owing to limited stability data or pre-existing iso-LSD in those capsules. The relatively long tmax observed after both oral and intravenous dosing is discussed; possible mechanistic reasons proposed include dilution/distribution dynamics, slow blood–brain barrier passage or delays in the response mechanism itself. The authors also note that a threshold plasma concentration (0.3 ng/mL) was reached faster with the solution, which could reflect faster absorption or higher overall concentrations. Acknowledged limitations include the use of different subject samples for the capsule comparison (so between-study differences could confound direct comparisons), relatively sparse sampling during the early absorption/distribution phase (limiting characterisation of absorption kinetics), and assay improvements compared with earlier studies. Strengths cited are the relatively large sample size for a PK study, inclusion of both sexes, the use of a well characterised and stability-documented formulation, and provision of quality-assurance data. The investigators conclude that PK investigations such as this are essential to confirm exposure in experimental and clinical studies and that the presented data can serve as a reference for interpreting plasma concentrations in research and intoxication contexts.