Pharmacokinetics, Pharmacodynamics, and Urinary Recovery of Oral Mescaline Hydrochloride in Healthy Participants

Mescaline (3,4,5-trimethoxyphenethylamine) is a classic serotonergic psychedelic found in several cacti and acts as a 5-HT2A receptor agonist. Earlier research has established its psychedelic action in humans, but modern, detailed pharmacological characterisations are scarce compared with LSD and psilocybin. Mescaline has relatively low 5-HT2A receptor affinity and limited blood–brain barrier permeability, which helps explain why substantially higher oral doses (commonly 300–800 mg of mescaline hydrochloride) are required to produce full psychedelic effects. Prior human metabolism work was limited, with a single older radiolabel study in 12 men; contemporary compartmental PK and PK-PD modelling and urinary recovery data in diverse volunteers were lacking. Mueller and colleagues set out to provide a comprehensive pharmacokinetic, pharmacodynamic, and urinary recovery characterisation of oral mescaline HCl in healthy adults. Using dense sampling from two Phase I, randomized, double-blind crossover trials covering single doses from 100 to 800 mg, the investigators aimed to describe mescaline plasma kinetics with compartmental models, link plasma concentrations to acute subjective effects using PK-PD modelling, and quantify urinary recovery of mescaline and principal metabolites to infer extent of first-pass metabolism and renal elimination. This work was intended to inform dosing and safety considerations for future experimental and clinical use.

The analysis pooled data from two previously published Phase I studies conducted at the University Hospital Basel. Study 1 included two cohorts (each n=16) receiving single oral mescaline HCl doses of 300 or 500 mg within a larger crossover that also tested LSD, psilocybin and placebo. Study 2 was a crossover in 17 participants who received 100, 200, 400 and 800 mg mescaline HCl and an 800 mg plus ketanserin (40 mg) condition; the ketanserin arm was included in PK analyses but excluded from PK–PD modelling. Overall, 49 healthy volunteers (25 female, 24 male; mean age 30 ± 7 years, BMI 18–29 kg/m2) contributed 113 administrations, of which 105 administrations remained after excluding profiles affected by early vomiting or incomplete urine collection. Eligibility required physical and mental health, age 25–65 years and BMI 18–29 kg/m2; key exclusions included pregnancy, breastfeeding, major psychiatric history in the participant or a first-degree relative, chronic somatic disease and use of interfering medications. Dosing sessions began at 9 a.m. after a standardised small breakfast; participants were monitored continuously and outcome measures were recorded repeatedly for 24 h (Study 1) or 30 h (Study 2). Blood sampling was frequent during the acute phase (multiple samples from predose up to 24 or 30 h depending on the study). Urine was collected in 8-h intervals in Study 1 (0–8, 8–16, 16–24 h) and cumulatively to 30 h in Study 2. Mescaline, 3,4,5-trimethoxyphenylacetic acid (TMPAA) and N-acetylmescaline (NAM) were quantified by a validated HPLC–tandem mass spectrometry method. Pharmacokinetic analyses combined non-compartmental analyses (NCA) and sequential compartmental modelling using Phoenix WinNonlin 8.4. NCAs provided observed Cmax, tmax and AUC metrics; terminal half-life (t1/2) was estimated by log-linear regression and AUC∞ reported with percentage extrapolation. Urinary recovery was expressed as total µmol and percent of dose; renal clearance was calculated as urinary recovery/AUC∞. For compartmental PK modelling, a one-compartment model with first-order absorption and elimination was initially fitted; models explored inclusion of a lag time (tlag) and a second compartment. Parameter estimation used a naïve pooled engine with model selection guided by visual fit, -2 log likelihood (-2LL), AIC and likelihood ratio tests (drop of 3.84 in -2LL considered significant for adding one parameter). Values below the lower limit of quantification (BLQ) were handled with the M3-type censored likelihood approach. PK–PD modelling linked plasma mescaline concentrations to repeated single-item visual analogue scale (VAS) ratings (0–100%) that measured “any drug effect” (primary outcome), “good drug effect”, “bad drug effect”, “nausea” and “ego dissolution”. A sigmoid Emax model (Emax fixed at 100% to match VAS range) with EC50 (concentration producing 50% of maximal effect), a steepness parameter (γ) and an effect-compartment first-order equilibrium rate constant (ke0) was evaluated to account for counterclockwise hysteresis. Model diagnostics and derived PD metrics (e.g. area under the effect-time curve, effect onset, maxima and durations) were obtained by NCA on model-predicted individual effect-time profiles. Profiles with negligible observed effects produced extreme EC50 estimates and were treated specially in summarising PD parameters as described by the investigators.

Forty-nine participants (25 female, 24 male; mean age 30 ± 7 years) contributed data after exclusions. Thirty-four participants (69%) reported prior psychedelic experience. Early vomiting occurred in eight administrations (most often at 800 mg), and after excluding profiles with emesis shortly after dosing or incomplete urinary sampling, 105 administrations were retained for analysis. Urinary recovery and NCA findings: Across all doses and conditions, a geometric mean of 53% of the administered dose was recovered in urine as unchanged mescaline and 31% as TMPAA over 24–30 h. In many cases the combined recovery of mescaline and TMPAA approached nearly 100% by 24–30 h, and more than 50% of recovered analytes appeared in the first 8 h after dosing. Mescaline and TMPAA showed dose-proportional increases (dose linearity) between 100 and 800 mg. NAM concentrations were low, with peak plasma exposure <10% of mescaline and an elimination half-life near 2 h. Compartmental PK modelling: A one-compartment model with first-order absorption and elimination generally described mescaline plasma concentration–time data. Adding a second compartment did not materially improve fits and was discarded because of strong parameter correlations. Inclusion of a lag time improved fit around Cmax in many profiles and produced a significant reduction in -2LL in 59 of 105 profiles; the mean lag time was small (about 0.2 h). Geometric mean tmax was approximately 2.0 h and estimated terminal half-lives were about 3.5 h across doses. Model-derived parameters aligned well with NCA results. PK–PD modelling and subjective effects: The VAS item “any drug effect” was modelled using a sigmoid Emax model with a steepness parameter and an effect-compartment rate constant (ke0) to account for a counterclockwise hysteresis between plasma concentration and effect. Adding γ and ke0 significantly improved model fit across most profiles. Model-predicted onset of any drug effect was around 1 hour after dosing, with tmax of predicted effects ranging roughly from 1.9 h at 100 mg to 4.2 h at 500 mg. Predicted maximal intensity and effect duration increased with dose: for “any drug effect” maximal predicted intensity rose from about 13% with a duration of 2.8 h at 100 mg to about 89% with a duration of 15 h at 800 mg. Profiles with very low or zero observed effects produced extreme EC50 estimates and were excluded from summarised primary PD parameter statistics but retained in secondary analyses of predicted effect-time curves. Additional derived metrics and interpretation: The parallel concentration–time profiles of mescaline and TMPAA, with similar tmax and Cmax, were interpreted as consistent with substantial first-pass metabolism of mescaline to TMPAA. Based on urinary recovery, the investigators estimated oral bioavailability of at least 53%. Using that bioavailability, a calculated renal clearance of approximately 22 L/h (366 mL/min) was proposed, a value higher than typical glomerular filtration rates and suggestive of possible active renal secretion. The authors noted some inter-study differences: Study 2 (sampling to 30 h) tended to show more complete recovery than Study 1 (24 h sampling), and Study 1 had higher dose-corrected AUCs and lower recoveries for some doses, potentially reflecting cohort or sampling-duration differences.

Mueller and colleagues interpret their findings as the first detailed compartmental PK and PK–PD characterisation of oral mescaline hydrochloride in modern controlled human studies across a broad dose range (100–800 mg). They report dose-proportional pharmacokinetics from 100 to 800 mg, rapid oral absorption with median peak plasma concentrations at about 2.0 h, and plasma half-lives near 3.5 h. A one-compartment model with first-order absorption and elimination plus a small lag time adequately described plasma concentrations. The urinary recovery data showed substantial excretion of unchanged mescaline and the main metabolite TMPAA, supporting the view that first-pass metabolism converts a sizeable fraction of the oral dose into TMPAA and that renal elimination is the principal route for both analytes. From these urinary data the investigators infer an oral bioavailability of at least 53%. They further note that the estimated renal clearance (about 22 L/h) exceeds normal glomerular filtration rates, which may indicate active tubular secretion. The PK–PD relationship showed a counterclockwise hysteresis, modelled with an effect compartment and ke0, indicating a short delay between plasma concentrations and subjective effects; the dose-dependent increases in predicted maximal intensity and duration are consistent with mescaline producing subjective effects when present in plasma/brain and occupying 5-HT2A receptors. Strengths highlighted by the authors include the relatively large number of richly sampled profiles over a wide dosing range, analytically confirmed dosing, controlled experimental conditions and the combination of PK and urinary recovery data that allowed an indirect estimate of oral bioavailability. Several limitations are acknowledged: the healthy, mostly young volunteer sample limits generalisability to older people or patients with impaired renal function; prior psychedelic experience in many participants could have influenced PD responses; methodological differences between the two contributing studies (notably different urine sampling durations and study designs) may have affected recovery estimates; the small estimated lag time did not significantly improve fit in many profiles perhaps because absorption was rapid and early sampling was limited; and some PD profiles with minimal effects yielded implausible EC50 estimates, a problem the authors note might be better addressed by population-based modelling. Finally, they emphasise that confirming the oral bioavailability estimate would require an intravenous mescaline administration in a future study.

Mescaline HCl displayed dose-proportional pharmacokinetics between 100 and 800 mg and was well described by a one-compartment model with first-order absorption and elimination. Subjective acute effects tracked plasma concentrations with a short response lag and were captured by a sigmoid Emax PK–PD model without evidence of acute tolerance. Urinary recovery data indicate substantial first-pass metabolism to TMPAA and support an oral bioavailability of at least 53%, with renal excretion as the main elimination route for both mescaline and TMPAA. The authors note that these data can inform dose selection in future studies and highlight the need to consider renal function when translating findings to clinical populations.