Pharmacokinetics of Hoasca alkaloids in healthy humans

Hoasca (also called ayahuasca, yagé, daime and other names) is an Amazonian decoction prepared from Banisteriopsis caapi and Psychotria viridis that contains harmala alkaloids (harmine, harmaline, tetrahydroharmine, THH) together with N,N-dimethyltryptamine (DMT). Earlier pharmacology has identified harmine and harmaline as reversible inhibitors of monoamine oxidase A (MAO-A) and THH as a weak serotonin uptake inhibitor; these actions permit DMT to be orally active and increase serotonergic signalling. Despite a long history of ritual and medicinal use across the Amazon basin and an emerging research interest, the human pharmacokinetics of the tea’s characteristic alkaloids had been little described prior to this study. Callaway and colleagues set out to measure plasma concentrations and pharmacokinetic parameters of DMT, harmine, harmaline and THH following ingestion of hoasca in experienced ritual users, and to relate these concentrations to subjective, autonomic, cardiovascular and neuroendocrine responses. The study was conducted in a ceremonial context with volunteers from the União do Vegetal (UDV); analytical methods and psychometric evaluations were integrated into a prospective investigation of the tea’s acute pharmacology and pharmacokinetics.

This was a prospective, observational pharmacokinetic study conducted at a UDV temple near Manaus, Brazil. Plant material (B. caapi and P. viridis) was collected on site, authenticated, and prepared into approximately 120 l of hoasca by UDV members using their customary procedures. The final tea was analysed for alkaloid content by high-pressure liquid chromatography (HPLC) with fluorescence detection; DMT in plasma was measured by gas chromatography with nitrogen-phosphorus detection (GC-NPD). Fifteen male UDV members aged 26–48 years (mean 35.9 ± 6.9 years) who had used hoasca regularly for at least 10 years were randomly selected from a larger group of 24 and underwent medical screening including blood chemistry and ECG. An age-matched group of 15 hoasca-naive males was also assessed for baseline health measures but was not included in the acute pharmacokinetic phase. On the study day each experimental volunteer ingested 2 ml/kg of hoasca (the administered volume was rounded up in practice; the extraction does not clearly report the exact rounding rule) after baseline blood sampling. All doses were administered by a mestre and consumed rapidly. Blood was collected through an indwelling catheter into EDTA-containing tubes, centrifuged and plasma frozen at −80°C until assay. Samples were taken prior to ingestion and at serial time points thereafter, with a final 24 h sample on the next day; the extracted text does not clearly report the full set of intermediate sampling times. Harmine, harmaline and THH were quantified by HPLC-fluorescence and DMT by GC-NPD. Pharmacokinetic parameters and area-under-the-curve (AUC) values were estimated using PCNONLIN (Version 4.0) with non-linear regression. Neuroendocrine hormones (growth hormone, prolactin, cortisol) were measured by radioimmunoassay and autonomic/cardiovascular variables (heart rate, blood pressure, respiration, oral temperature, pupil diameter) were recorded repeatedly through 240 min using standard techniques. The hallucinogenic rating scale (HRS) was used to capture subjective effects; those psychological results are reported elsewhere.

Fifteen volunteers participated; one vomited about 45 min after ingestion and was excluded from pharmacokinetic analyses, leaving 14 subjects for most measures. For DMT, peak concentrations were measurable in all volunteers but only 12 had sufficient data for full pharmacokinetic calculations. The measured alkaloid concentrations in the tea were: harmine 1.70 mg/ml, harmaline 0.20 mg/ml, THH 1.07 mg/ml and DMT 0.24 mg/ml. Pharmacokinetic parameters (means ± SD) reported in plasma were: DMT (n = 12) Cmax 15.89 ± 4.4 ng/ml with Tmax 107.59 ± 32.5 min; harmine (n = 14) Cmax 114.8 ± 61.7 ng/ml with Tmax 102.0 ± 58.3 min; THH (n = 14) Cmax 91.0 ± 22.0 ng/ml with Tmax 174.0 ± 39.6 min. Harmaline concentrations were low in the beverage, and pharmacokinetics were reported for only five volunteers (Cmax 6.39 ± 3.1 ng/ml, Tmax 145.0 ± 66.9 min). None of the alkaloids were detected in pre-ingestion samples; THH was still detectable at low levels in three volunteers at 24 h. Subjective effects were reported by all 15 volunteers at the administered dose, with the most intense visionary experiences occurring between about 60 and 120 min, temporally coincident with the DMT Tmax. One volunteer vomited and was excluded from PK analyses. Neuroendocrine responses rose sharply above baseline: mean growth hormone increased to a maximum of 9.44 ± 2.41 ng/ml at 90 min before returning to baseline by 360 min; prolactin peaked at 33.90 ± 8.86 ng/ml at 120 min and returned to baseline by 360 min; cortisol rose to a maximum of 132.6 ± 10.7 ng/ml at 60 min, dipped below baseline by 360 min and showed a small but statistically significant increase vs baseline in the 24 h samples (73.90 ± 11.40 ng/ml, P < 0.05 as reported). Autonomic measures increased from baseline: pupil diameter rose from 3.79 ± 0.2 mm to 4.9 ± 0.2 mm at 180 min and remained dilated at 240 min (pupils typically returned to normal by ~6 h at this dose); respiration increased modestly from 18.4 ± 0.7 to 21.5 ± 1.0 breaths/min at 90 min; oral temperature rose slightly from 37.0 ± 0.1°C to 37.3 ± 0.1°C by 240 min, with ambient room temperature also rising over study sessions. Cardiovascular effects included an early heart rate increase from 71.9 ± 2.9 bpm to 79.3 ± 0.3 bpm at 20 min, a nadir at 120 min (64.5 ± 2.2 bpm) and return toward baseline by 240 min. Systolic and diastolic blood pressure rose to maxima at 40 min (137.3 ± 3.2 and 92.0 ± 3.0 mmHg, respectively) from baselines of 126.3 ± 3.9 and 82.7 ± 2.9 mmHg, and generally returned toward baseline by 180 min. The authors note these cardiovascular increases were not hypertensive; four individuals had heart rates <60 bpm at 120 min.

Callaway and colleagues interpret their findings as consistent with the known mechanism by which harmala alkaloids enable oral activity of DMT through MAO-A inhibition. Peak subjective effects coincided with DMT plasma Tmax, and neuroendocrine changes (growth hormone, prolactin, cortisol) and autonomic/cardiovascular responses tracked the alkaloid concentration-time profiles. The authors contrast the oral time-course to intravenous DMT, noting injected DMT produces a much faster onset and shorter duration; they suggest qualitative differences between oral and injected DMT may arise because elevated extracellular serotonin (from MAO inhibition and THH-mediated uptake blockade) competes with DMT at central serotonergic sites and thus alters the phenomenology of the experience. The discussion addresses side effects and mechanistic considerations: nausea, vomiting and diarrhoea were attributed to increased peripheral and central serotonin after MAO-A inhibition, and transient tremor or nystagmus may reflect tryptamine-receptor interactions. Harmine concentrations measured in plasma were several orders of magnitude above reported EC50 values for MAO-A inhibition, leading the investigators to suggest that typical hoasca doses could inhibit both MAO-A and, at higher levels, MAO-B. THH showed a pharmacokinetic profile relatively independent of harmine and may contribute indirectly via serotonin uptake inhibition rather than MAO inhibition. The neuroendocrine responses were interpreted as indicators of increased serotonergic activity, and elevations in cortisol and prolactin were similar to those reported after injected DMT but were prolonged here, consistent with the extended pharmacokinetic profile due to oral absorption and MAO inhibition. The authors acknowledge limitations in the present extraction: the control (hoasca-naive) group was used only for baseline health comparisons and was not included in the acute pharmacokinetic phase; the administered dose was considered mild by experienced users; and one subject’s data were excluded after vomiting. They present the study as preliminary but informative, providing direction for further research into the complex psychopharmacology of hoasca and related substances in healthy humans.

A long history of ritual use and the present observations led the investigators to conclude that, in their sample of experienced users, hoasca produced measurable plasma levels of DMT and harmala alkaloids that corresponded with subjective, autonomic and neuroendocrine effects without evidence of acute physical or psychological deterioration. They suggest the pharmacokinetic and neuroendocrine data from this prospective study offer a foundation for further clinical and pharmacological investigation of hoasca and similar preparations.