Hofmann vs. Paracelsus: Do Psychedelics Defy the Basics of Toxicology?-A Systematic Review of the Main Ergolamines, Simple Tryptamines, and Phenylethylamines

The investigators performed a systematic search of major medical and scientific databases from the early 20th century up to 31 August 2022. Search terms and MeSH headings included combinations such as psychedelic overdose, LD50 for specific compounds (LSD, MDMA, mescaline, psilocybin), psychedelic poisoning, and related phrases. The search was conducted between 1 July and 31 July 2022 and the review followed PRISMA 2020 guidelines. Inclusion criteria were original articles, case reports, reviews, English-language articles, and in vivo studies; excluded items were in vitro studies, abstracts, posters, and conference communications. Two non-English articles were handled by translating abstracts to extract key information. After screening 3032 records by title and abstract, full-text review left 67 eligible manuscripts, of which 33 provided direct data on high or lethal psychedelic doses; the remaining 34 were retained to contextualise discussion. Both human clinical reports and animal experimental LD50 data were included, and the authors report that many older clinical case reports predate modern analytical methods, limiting interpretation in some instances. The extraction appears to combine animal LD50 studies, human pharmacokinetic reports, clinical case reports, and case-series/reviews. The paper does not report conducting a formal meta-analysis or pooled quantitative synthesis; no specific risk-of-bias tool or statistical heterogeneity assessment is described in the extracted text. Where tables or figures are mentioned as summarising study characteristics and animal LD50s, the extracted text reproduces key numerical results within the narrative.

Sixty-seven articles met the authors' eligibility criteria; 33 of these contained information directly relevant to high or lethal doses of psychedelics. Of the 33, 18 (54.5%) were animal studies and 15 (45.5%) related to humans (including two experimental studies in humans, four reviews of clinical cases, and nine individual case reports). An additional 34 papers were used to inform the discussion. Ergolamines (LSD): Animal LD50 values varied substantially by species and route: reported figures include mice 50–60 mg/kg, rats 16.5 mg/kg, and rabbits 0.3 mg/kg after parenteral administration, while an isolated report suggested an elephant died after 0.06 mg/kg but a later repetition produced no fatality. Human pharmacokinetic data cited include an oral 160 µg dose producing 4.16 ng/mL in blood at two hours. Clinical case reports and series showed blood or tissue LSD concentrations in varied ranges (for example 0.0021–0.026 µg/mL in one eight-person series). The authors report two probable human deaths attributed to LSD in the literature, one with 31.2 µg/mL in liver tissue (authors of that case extrapolated a very large intravenous dose) and another with pre-mortem/post-mortem concentrations in the low ng/mL range; interpretation is complicated by analytical-method limitations and co‑ingestants in several reports. The review highlights multiple accounts of very large apparent exposures (including one person who reportedly snorted 55,000 µg) without lasting harm; historical LD50 estimates for humans have varied widely in the literature (e.g. 14,000 µg to 100,000 µg), but the authors note some of these values lack empirical support. Across the LSD literature assembled, the pattern the authors emphasise is that classic LSD toxicity in humans appears rare and that many reported harms involve contextual factors (set and setting) or external causes (e.g. traumatic injury), rather than straightforward pharmacological lethality. MDMA: The review describes MDMA as pharmacokinetically distinct from classic psychedelics, with non-linear kinetics and metabolism largely mediated by CYP2D6; metabolites can inhibit CYP2D6 and thus potentiate toxicity after repeated doses. Animal LD50s varied by species and route (examples: intraperitoneal LD50 approximately 97, 49, and 98 mg/kg for mouse, rat and guinea pig; intravenous LD50 22 mg/kg in monkey and 14 mg/kg in dog; oral LD50 in rats 160–325 mg/kg). Human forensic series report average post-mortem blood concentrations of ~0.45 mg/L (Australian series) and ~0.73 mg/L (Norwegian series), with a lethal concentration reported around 3 mg/L in a subset of 27 deaths; many MDMA-associated deaths involved other drugs or environmental factors. Using literature synthesis, the authors offer an approximate acute LD50 for a healthy 70 kg person of about 2 g (noting underlying assumptions), which they say corresponds to roughly 15–16 times a single recreational dose as reported in the sources. The review stresses that MDMA carries greater acute health risks than classical psychedelics (examples: hyperthermia, hyponatraemia, coagulopathy), and that the pattern of MDMA-related fatalities is influenced by route of administration, polydrug use, impurities, environmental factors and genetic variability in metabolism. Phenylethylamines (mescaline) and tryptamines (psilocybin): For mescaline (typically administered as cactus preparations), animal LD50s reported across studies were broad (mice 157–880 mg/kg depending on route; other reports included much higher figures for oral LD50 approaching ~1 g/kg in some species), and early human experimental work (1920s onward) reported no deaths at doses used in controlled settings. Two fatal human cases associated with mescaline ingestion are described in the literature: one death attributed to a fall while intoxicated and another due to aspiration during a ceremonial context. Typical recreational mescaline doses are around 200–450 mg (commonly cited 350 mg), and the authors point to pharmacokinetic properties—low lipid solubility, slow brain penetration, renal excretion—that may confer relative safety. For psilocybin, animal LD50s reported include mice 275–420 mg/kg (route-dependent), rabbits as low as 13 mg/kg intravenously in one report, and rats ~280 mg/kg. Human high-dose ranges are described qualitatively (30–40 mg considered high), and isolated fatal cases are noted but criticised for methodological problems in some instances. The review estimates that a very large mass of fresh mushrooms would be required to approach animal-derived LD50s in humans and concludes that psilocybin has relatively low toxicity, with cardiovascular risks relevant in individuals with pre-existing disease. Across the tryptamine literature, the pattern presented is a wide safety margin for classic tryptamines under controlled conditions.

Henríquez-Hernández and colleagues interpret the assembled evidence as broadly consistent with the view that classical psychedelics (LSD, psilocybin, mescaline) possess large safety margins under typical routes of administration and controlled settings. They argue that, for these compounds, reported fatalities are uncommon and often involve contextual factors—such as unsafe environments, polydrug use, or traumatic events—rather than clear, dose-dependent pharmacological lethality. The authors thereby question the universal applicability of Paracelsus' maxim in the context of these agents, while acknowledging that extreme exposures or atypical circumstances can still cause harm. The discussion contrasts the above pattern with MDMA, which the authors present as more hazardous in practice due to non-linear kinetics, active metabolites that inhibit CYP2D6, a higher incidence of acute medical complications (for example hyperthermia and hyponatraemia), and frequent involvement of impurities or environmental contributors in recreational settings. They highlight the potential importance of pharmacogenetics—particularly CYP2D6 variability—to explain some fatal outcomes and recommend targeted research to clarify these mechanisms. The authors also acknowledge methodological caveats: many historical reports predate modern analytical methods, animal-to-human extrapolations are imperfect, the review excluded non-indexed literature, and no quantitative meta-analysis was performed. These limitations, they say, constrain definitive statements about human LD50 values for many compounds. In terms of implications, the authors call for renewed investment in basic pharmacology and toxicology research (metabolism, pharmacokinetics, pharmacogenetics) to underpin safe therapeutic deployment and forensic interpretation. They further emphasise the role of set and setting in preventing harm and note that, when administered in guided therapeutic contexts, few significant adverse events or deaths have been reported for the classic psychedelics. The discussion thus frames a research agenda prioritising mechanistic studies, improved toxicological data, and careful attention to environmental and genetic moderators of risk.

The authors conclude that classic psychedelics demonstrate high potency at low doses but generally low addiction potential and substantial safety margins, with many substances tolerating large multiples of typical therapeutic doses without clear evidence of direct lethality. MDMA appears to be an outlier with higher associated acute risks, although many reported fatalities involve contextual or metabolic factors rather than simple dose–response relationships. Henríquez-Hernández and colleagues emphasise that understanding the unique toxicological behaviour of these agents will require investment in basic research—especially pharmacokinetics and pharmacogenetics—and that the set and setting of use remain critical determinants of harm.