Return to Hemp Reports

Derivation of THC Limits for Food

Pharmacological and Toxicological Basis of THC Limits for Food

Part II

1 Introduction

Excursus I: THC and Alcohol in food

2 Methodology
2.1 Extrapolation of different routes of administration to oral ingestion
2.2 Extrapolation of animal data to man
2.2.1 Interspecies comparison based on body weight
2.2.2 Interspecies comparison based on body surface
2.2.3 Interspecies comparison based on pharmacokinetics
2.2.4 Interspecies comparison based on precise toxicological data
2.3 Methodical basis for the determination of thresholds

3 Pharmacology and pharmacokinetics
3.1 Resorption and plasma level

Excursus II: THC in cosmetics and dermatics

3.2 Transport and metabolism

Excursus III: Detection of THC after ingestion of hemp-containing food

3.3 Influence of physical factors on THC content
3.4 Mode of action
3.5 Development of tolerance
3.6 THC effects and total toxicity

4 THC thresholds for psychotropic effects

5 Discussion of physical effects
5.1 Genetic material and cell metabolism
5.1.1 Cell studies
5.1.2 Studies with Cannabis users
5.1.3 Conclusion
5.2 Pregnancy
5.2.1 Birth complications
5.2.2 Birth defects
5.2.3 Pregnancy outcome
5.2.4 Brain development
5.2.5 Summary
5.3 Hormonal system and reproduction
5.3.1 Sex hormones
5.3.2 Glucocorticoids
5.3.3 Thyroid hormones
5.3.4 Glucose metabolism
5.3.5 Summary
5.4 Immune system
5.4.1 Cell-mediated immunity
5.4.2 Humoral immunity
5.4.3 AIDS infection
5.4.4 Summary

1 Introduction

This Part II of the THC study deals with the methodical, and especially the biological basis for THC limits in food. The essential ideas and results of Part II are summarized in Chapter 3 of Part I and provide the basis for the proposed THC limits. Part II addresses specifically the scientifically interested, who would like to explore individual aspects of the matter-or the whole topic-in more detail, in order to obtain a better understanding of the fundamentals. Here, the reader also finds numerous references.

The majority of the extensive data on the pharmacological effects of THC were collected in animal and cellular experiments, mostly using very high THC doses, often not by oral but by other routes of administration. Chapter 2 tackles the transferability of these results to oral THC administration to humans. A further section of this chapter reviews the methodical basis commonly used in the establishment of limits for other substances and hazards. In the course of this, some aspects are examined that are of special importance in connection with THC.

Chapter 3 contains a survey of the pharmacological effects of THC and of its overall toxicity. Furthermore, matters of bioavailability, metabolism, mode of action, as well as development of tolerance are dealt with. A separate section deals with physical parameters, such as effects of temperature on the biological efficacy of THC.

Chapter 4 examines the matter of the psychotropic threshold, i.e. the placebo threshold for the oral administration of THC. For the purpose of fixing a limit, this serves for the determination of the maximum daily quantity of THC that will not lead to undesired acute or chronic psychological reactions.

Chapter 5 deals with possible physical effects below the psychotropic threshold. It serves for the determination of the maximal daily dose that does not lead to undesirable acute or chronic physical effects.

Some Excursuses discuss special questions that may arise in the context of the application of industrial hemp products to man: Among them is a short comparison of THC and alcohol in food, matters of assimilation of THC by the skin, for instance through cosmetics, as well as the detection of THC in body tissues after the ingestion of hemp-based food in comparison to the impact of marijuana ingestion.

Excursus I: THC and Alcohol in Food

Alcohol-containing products such as beer, wine or spirits that are used for intoxication typically contain 3-50% alcohol (ethanol). Chronic heavy use of alcohol may irreversibly damage brain cells and other organs (liver, heart). There is evidence of serious alcohol-induced malformations in the fetus (alcohol-embryopathy). Alcohol-containing food products such as fruit drinks, sweets, and meat dishes, generally contain less than 0.3% ethanol, this being an amount ten times smaller than what is contained in intoxicating products. This alcohol content, which is generally considered low and innocuous, needs not be declared in this context. According to the WHO (World Health Organization), a regular ingestion of 7 g of ethanol per day is harmless (Verbraucherzentrale 1998).

THC-containing drug products such as marijuana and hashish, applied for drug use, typically contain 1-20% by weight of THC, sometimes more. THC shows some potential for physical impairment, though this is much weaker than in alcohol. (Hall et al. 1994b, WHO 1997).

Cannabis-based food products, such as edible oil based on seeds from low-THC hemp and hemp-seed-based cereal bars, contain less than 0.005% THC (50 ppm) on average. Thus their THC content is less than 1/200 of the amount found in drug products. In Switzerland, precisely this value (i.e. 0.005%) was set as the limit for the concentration of THC in edible oil derived from hemp seeds. In other products of which greater amounts can be consumed, lower limits were set: for pastries it is 0.0005% and for alcoholic beverages it is 0.00002%. Thus the difference between these concentrations and the THC concentration in marijuana amounts to a factor of between 2,000-50,000.

Table 1: Weight units and their abbreviations

Table 2: THC concentrations and their abbreviations

2 Methodology

2.1 Extrapolation of different routes of administration to oral ingestion

A large portion of data on the toxicology of THC for humans and animals were not obtained following oral administration but after inhalative or parenteral (intravenous, subcutaneous, intraperitoneal) application. Different routes of administration result in a different bioavailability and in different pharmacokinetics (surveys: Wall et al. 1983, Maykut 1985, Agurell et al. 1986, Harvey 1991). This has to be taken into account when dosage and plasma concentrations are translated to the situation of an oral administration, i.e. the relevant exposure route for food considered in the following.

Table 3: Comparison of the effectiveness of THC application to man via relevant routes (Agurell et al. 1986, Frytak et al. 1984, Harvey 1991, Stefanis 1978)

Intravenous administration: Intravenous administration results in a 100% bioavailability and an immediate rise of THC concentration in the blood. The onset of THC effects occurs within a few minutes. In order to achieve minimum psychotropic effects, 0.02 mg THC/kg BW (body weight) are necessary; as a function of body weight this typically corresponds to 1 mg THC. In the case of doses that lead to psychotropic effects, short-term peak plasma levels are reached that markedly decrease within minutes. The intravenous administration of 5 mg THC led to a plasma concentration of about 200 ng/ml THC, which decreased rapidly to 15 ng/ml after an hour and 3 ng/ml after 4 hours. The psychotropic effects had vanished after three hours. Smith and Asch (1984) intravenously administered monkeys with 2.5 mg THC/kg BW three times a week. This produced an average maximum plasma concentration of 300 ng/ml THC and an average long-term concentration of 15 ng/ml.

Inhalative administration: After inhalative administration, THC is quickly absorbed and the time course of plasma concentration is similar to the situation after intravenous administration. Bioavailability only reaches 10-30%, so that about five times the dose of intravenous administration is required to achieve the same effects. About 0.7 mg/kg, that is about 5 mg THC, are necessary to produce minimum psychotropic effects. An intoxication desired by cannabis consumers requires inhalation of at least 10-15 mg THC, which would lead to a maximum plasma concentration of 100 ng/ml after about 5 minutes. The concentration decreases rapidly, so that little THC will be detected after 2-3 hours. Chronic consumers need higher doses because of their development of tolerance to the active agent. This was shown in a study where 47 chronic Cannabis users tolerated inhaled THC doses of up to 180 mg without undesired side effects or nausea (Stefanis 1978).

Oral administration: The systemic bioavailability of THC reaches 10-20% after oral administration in a lipophilic vehicle. To achieve minimum psychotropic effects in man, 0.2-0.3 mg/kg are required, which equals 10-20 mg depending on the body weight. This is 10-15 times the dose of intravenous administration. The maximum plasma level after oral administration of this dose is of the order of 5 ng/ml and is reached after 2-4 hours. The psychotropic effect sets in after 30-60 minutes, peaks after 1-3 hours and lasts for 6-8 hours. The average maximum plasma concentration of THC in six cancer patients after oral ingestion of 15 mg THC was 3.9 ng/ml and was typically attained after two hours (Frytak et al. 1984). With the exception of one patient, the plasma level of THC in all cancer patients had dropped below 1 ng/ml or, respectively, no THC could be detected in the plasma any more after 6 hours. Three patients received three doses of 15 mg THC a day. The maximum plasma level ranged between 3.6 and 6.3 ng/ml; thus, it did not differ much from that following a single administration. Ohlsson et al. (1980) observed a maximum THC concentration of 5 to 6 ng/ml between 1 and 1.5 h after experienced marijuana smokers had ingested a chocolate cookie containing 20 mg THC. Similar results were found by Brenneisen et al. (1996) and Frytak et al. (1979), though they occasionally found higher plasma concentrations of more than 10 ng/ml.

The above comparison suggests that, in order extrapolate data from intravenous and inhalative administration to the ingestion route, different conversion factors for acute single application and chronic effects must be applied. The systematic bioavailability is relevant to chronic effects, whereas in relation to acute effects further aspects must be considered, such as the faster resorption and the considerably higher peak plasma concentrations of THC after smoking and intravenous intake relative to those after oral administration.

2.2 Extrapolation of animal data to man

One advantage of animal studies is that they allow a thorough control of the conditions of THC exposure, such as time and duration, as well as the control of possible confounding factors. Therefore they constitute an important element of toxicological research. However, for several reasons, caution is still required when using "data produced by those that continue to extrapolate animal data to humans without some attempt to discuss in detail the validity of their assumptions" (Campbell 1996).

The principle of phylogenetic continuity of species, including similarity of cell structure and energy metabolisms, is the common denominator for the cross-species extrapolation of toxicological data. Nevertheless there are some significant pharmacokinetic and other relevant differences between species that render extrapolation more difficult.

There are various methods for the extrapolation of animal data to man (Voisin et al. 1990, Winneke and Lilienthal 1992, Ings 1990). These shall only briefly be reviewed here:

2.2.1 Interspecies comparison based on body weight

The first approach to comparing toxicological data from various animal species with that of man is based on the body weight. Many biologic parameters such as water intake, creatinine clearance and synthesis of hemoglobin can be expressed as mathematical functions of body weight. These relations are fairly consistent over a wide range of species. The corresponding unit is milligram per kilogram (mg/kg). Results obtained from an application of 10 mg/kg THC to rats would be translated to man in the ratio of one to one if such a toxicological comparison based on body weight was employed. Hence 2 mg THC in a rat of 0.2 kg would correspond to 700 mg THC in a man weighing 70 kg.

However, many biological processes such as metabolic rate, cardiac function, and renal function are not directly proportional to body weight. Thus toxicological comparisons on the basis of body weight are likely to be inaccurate.

2.2.2 Interspecies comparison based on body surface

The body surface is another potential basis for the comparison of toxic effects between species. For instance, the heat loss from a warm-blooded animal is approximately proportional to its body surface. In studies on anti-cancer drugs it was demonstrated that the maximum tolerated dose in different animal species (mouse, rat, dog, monkey) correlated quite well with that in man, if adjusted to body surface. The corresponding unit is milligram per square meter (mg/m2) or milligram per square centimeter, respectively (mg/ cm2).

Table 4: Extrapolation of a dose of 1 mg/kg in a mouse and other animal species to man on the basis of body weight and body surface (according to Ings 1990)

Table 4 shows that the bodies of small animals have a comparatively large surface-to-weight ratio. Thus one employs conversion factors in order to adjust the disproportionate development of body weight and body surface to the growing size of the species (Mordenti 1986). The conversion of those parameters from rat to man requires a multiplication of the applied dose by a factor of 1/7 (see table 5). In the aforementioned example, where 10 mg/kg THC were administered to rats, the results thus obtained would be transferred to man at a rate of 7:1. Thus, 2 mg THC in a rat of 0.2 kg would be equivalent to 100 mg THC in a man of 70 kg. This result is more in line with human consumption patterns than a dose of 700 mg.

However, though extrapolation on the basis of body surface might be appropriate for some toxic agents, this procedure does not suit all cases. It is but a makeshift solution for the situation that no better data on specific toxicity in different species and on pharmacokinetics are available.

Table 5: Dosage conversion factors based on equal body surface (according to Voisin 1990)

2.2.3 Interspecies comparison based on pharmacokinetics

If the specific toxicity of a compound is unknown, the best correlations can be achieved on the basis of pharmacokinetic data, such as absorption (body intake), distribution, metabolism and excretion. Classical pharmacokinetic models are based on plasma concentration and the AUC (area under the curve) as plasma concentration over time, produced by known doses. Such data allow the determination of extent and duration of systemic exposure to the analyzed substance.

Table 6: Plasma concentration in man and rat (Frytak et al. 1984, Brenneisen et al. 1996, Hutchings et al. 1991, Scallet 1991)

Pharmacokinetics in man and in animals show some parallels and a number of conspicuous differences (Agurell 1986, Harvey and Brown 1991).

Unfortunately the available data on plasma concentrations after oral administration of THC in man and rat are available for different doses: the human subjects received lower doses whereas in animals the applied concentrations were higher (see Table 6). An oral intake of 0.2-0.3 mg/kg produced a peak plasma concentration of 3-10 ng/ml in man (Frytak et al. 1984, Brenneisen et al. 1996). In animal studies with rats, plasma concentrations peaked at about 100 ng/ml after oral administration of 15 mg/kg THC (Hutchings et al. 1991) and at about 150 ng/ml after 20 mg/kg THC (Scallet 1991). The concentrations maintained a high level for 6 hours. In another study where rats had received THC at doses of 15, 25 or 50 mg every day for two years, the THC concentration had finally reached a level of about 400, 1300 or 3000 ng/ml at the end of the test period (Chan et al. 1996). This is the result of the accumulation of THC, as known in man as well, that causes an augmented plasma level after chronic application.

Extrapolation of pharmacological data to different doses is problematic even if corresponding mathematical models exist (Inges 1990). It is desirable to have more animal data that would involve doses corresponding to human consumption patterns.

2.2.4 Interspecies comparison based on precise toxicological data

Unfortunately the results achieved by extrapolation on the grounds of the models presented above can be quite astonishing. Often toxicity in different species does not correspond to the toxicity determined by means of pharmacokinetic data. One example is the lethal dose of THC.

In a study with rats, the median lethal dose (LD50)was established to range between 800 and 1,900 mg/kg oral THC, depending on sex and strain (Thompson et al. 1973). When this dose is extrapolated on the basis of body surface, the oral LD50 in dogs would be a quarter of this dose (200-500 mg/kg) per kg BW and half of this dose (400-950 mg/kg) per kg BW in monkeys. However, the experimental studies revealed contrary results. No deaths were observed when orally administering the maximum THC doses either to dogs (up to 3,000 mg/kg THC) or to monkeys (up to 9,000 mg/kg THC) (Thompson et al. 1973). Instead of being 50% more sensitive, primates turned out to be at least five to ten times more resistant to THC. It has not yet become quite clear how, at low doses, the relevant target parameters (effects on hormonal system, immune system and fetus) would compare between species. Probably mice are especially predisposed to fetal malformations (Abel 1985). Also regarding the sensitivity of the hormonal system to THC, clear species-related distinctions were found that diminish the extrapolation of those results to man (Mendelson and Mello 1984).

Other examples of chemical substances demonstrate the wide range of possible conversion factors. Alcohol, methyl mercury and polychlorinated biphenyls cause embryonic malformations in man. The toxic doses per kg body weight in animal studies ranged from 0.2 to 8.0 times the toxic dose in man (Hemminki and Vineis 1985).

Conclusion: Animal data can provide an indication of the toxicity of THC to humans. However, the extrapolation of animal data to man is problematic, not only because of the use of high dosing regimens but also because of the lack of reliable conversion factors for the target parameter under examination. Smaller animals possibly experience a stronger THC toxicity compared to larger animals and primates. Thus, reliable data on the toxicity of THC in man should be based on studies with humans.

2.3 Methodical basis for the determination of thresholds

For those chemical substances whose undesirable effects are either known or suspected, certain concepts for the protection of health against impairments, such as the LOAEL ("lowest observed adverse effect level") and the NOAEL ("no observed adverse effect level"), have been established. In order to provide a safe margin of protection against potential harm, safety factors of 10 are typically applied. For each particular case, the precise value of a safety factor may vary with the reliability of the determined threshold, the relevance of the observed effects in animals and cells for man, the severity of effects above the threshold, and the understanding of the causal relationship between drug intake and observed effects. With regard to THC limits in food, the data situation is fortunate, as the daily THC intake of chronic Cannabis consumers (50-200 mg or 1-3 mg/kg BW, respectively) already clearly exceeds the daily doses of relevance to this study (0.1-0.2 mg/kg) by a factor of 10-15. Acute effects in man were mostly examined at THC doses that cause psychotropic effects. Thus, a NOAEL in this dose range constitutes a proper margin of protection for the ingestion of THC with food. Dosage and concentration in animal and cell studies mostly exceed this level by another one to three orders of magnitude.

Table 7: Plasma concentration after different THC doses (Frytak et al. 1984, Brenneisen et al. 1996, Chesher et al. 1990, Hutchings et al. 1991, Smith and Asch 1984, Chan et al. 1996, different cell studies: see Chapter 5)

Appropriate limits can be based only upon sufficient or at least probable evidence of effects and not on hypothetical effects. Early toxicological research with THC in the seventies often discovered health-impairing effects that could not be confirmed in the following years. Such inconsistent experimental findings cannot be utilized for the determination of limits for detrimental effects. Early studies were often short of well-designed methods concerning procedure, choice of controls and the consideration of possible confounding factors. For instance, insufficient imaging techniques (pneumoencephalography) suggested a cerebral atrophy as consequence of chronic Cannabis use, which later was refuted. Also, when "pair-feed controls" were employed in the eighties in animal studies, it was realized that even after high THC doses fetal impairments were not induced by THC toxicity, but rather resulted from the mother's reduced intake of food and water (Abel 1984). (Pair-feed controls are those that receive the same amount of food and water as the examined animals, who took in less food and water because of the deprivation caused by the administration of THC.)

At this point, two particularities of the toxicological evaluation of THC shall be emphasized:

1. Development of tolerance and accumulation: For most toxic agents, the toxicity increases with the duration of exposure and the NAOEL decreases correspondingly (Voisin et al. 1990). Such an increase of toxicity can be expected, especially for a substance which possesses a comparatively long half life (THC: 20-30 hours) and which accumulates with chronic administration. However, opposite effects were observed with THC, since tolerance develops for most effects, and this overcompensates for any accumulation that may occur. With an augmenting administration of THC, the psychological as well as most-if not all-physical effects mediated by specific receptors decrease (see Chapter 3.5). For instance, in THC studies of female rhesus monkeys, hormonal changes and a disruption of the menstrual cycle occurred (Smith et al. 1983). However, after six months of a persistent high-dosing schedule, the hormonal values and the menstrual cycle had returned to normal. Development of tolerance has also been established for most of the other effects (mood changes, cardiovascular effects, etc.).

2. Damage to children: Children generally respond more severely to chemical toxins and are rated as "sensitive persons," requiring larger margins of protection (Winnecke and Lilienthal 1992). A child's brain is, for instance, more susceptible to impairments than the brain of an adult. Contrary to most noxious substances, however, THC in relevant concentrations does not operate in an unspecified manner but acts on specific receptors on body cells (especially brain cells and immune cells). The number of cannabinoid receptors in adults is several times the number in children (see Chapter 5.2). This is also beneficial to the therapeutic uses of THC. Given the aforementioned reasons, children who were given THC (in this case delta-8-THC) in the course of a chemotherapy tolerated considerably higher doses (18 mg/m2 body surface) than adults would presumably have tolerated (Abrahamov 1995). Elderly people are more responsive to THC concerning psychotropic effects, even though the aging process leads to a slight reduction of THC receptors (Romero et al. 1998). However, caloric intake also usually decreases with age, causing a reduction in THC intake with food.

3 Pharmacology and pharmacokinetics

For a proper understanding of Cannabis effects, and for the sake of comparability of different routes of administration, a basic understanding of the pharmacokinetics of THC is needed (reviews: Agurell et al. 1986, Harvey 1991). It will be provided in the following sections.

3.1 Resorption and plasma level

Cannabis drug products (marijuana, hashish) are preferably inhaled (cigarettes, pipes) and only seldom orally ingested (tea, pastries, tincture). Besides these routes, the intravenous (injection into the blood vessel), the subcutaneous (injection under the skin), and the intraperitoneal (into the abdominal region) routes are widely applied in animal studies. This present study focuses on the toxicity from oral administration with food, as well as the transferability to the ingestion route of biological effects observed with other routes of administration.

Several studies determined the systemic bioavailability of THC after the smoking of a marijuana cigarette as ranging between 2 and 56% of the total amount of THC present in the cigarette. Generally, bioavailability appears to range between 10-30%, with inexperienced smokers achieving lower rates. Hence Lindgren et al. (1981) established a systemic bioavailability in heavy smokers of 23% (+/-16%) compared to 10% (+/-7%) in occasional users. The smoking of a marijuana cigarette containing 10-20 mg THC produces a maximum plasma level of about 100 ng/ml after three minutes. Subsequently, the plasma concentration drops rapidly. The subjective effect sets in after only a few puffs, and the maximum high is reached after 15-30 minutes, when the plasma level has already begun to decline. The psychotropic effects typically cease after three hours.

When orally ingested, the systemic bioavailability of the lipophilic THC molecule is typically 5-10%. However, it may be doubled when simultaneously applying a lipophilic carrier (fat, oil), thereby improving THC resorption. Thus the bioavailability is typically somewhat less than that via inhalation. The rate of intake also depends on additional factors, such as the fullness of the stomach, and varies between individuals. After oral administration of 15-20 mg THC, plasma levels peak at approximately 5 ng/ml THC, typically after 1-3 hours, with a large inter- and intra-individual variability. The subjective psychotropic effect has its onset after 30-90 minutes and lasts for about 6 hours. In order to achieve acute effects, considerably higher oral doses are required than with inhalation because of the slower intestinal resorption of THC and the somewhat lower bioavailability.

Excursus II: THC in cosmetics and dermatics

Substances that are applied to the skin can be systemically absorbed to an unknown extent. However, there have not yet been any quantitative studies of the dermal absorption of THC that would allow quantification. Nevertheless, this question is vital for the use of THC-containing products that are externally applied (cosmetics, dermatics for the treatment of neurodermitis). The physico-chemical characteristics of THC, however, allow a rough estimation of the amount of THC assimilated (Kalbitz et al. 1996, 1997).

Generally, the human skin is well protected against penetration by external substances. Many topically applied substances attain a systemic bioavailability of only a few percent (Hadgraft 1996). The main barrier to penetration is the cornea (stratum corneum), or more accurately, the cornified layer of the stratum corneum. In principle, substances can penetrate the space between the cells of the stratum corneum (intercellular), the cells themselves (intracellular), or the sebaceous and perspiratory glands and hair follicles. Only the first pathways of penetration generally play a relevant role (Hadgraft 1996, Kalbitz et al. 1996, Berti et al. 1995). For example, the route through the hair follicles and glands is of importance for polar molecules only. As a lipophilic molecule, THC does not belong to that group.

The permeation coefficient or, respectively, the permeability constant (Kp) constitute a quantitative expression for the ability of a substance to permeate the skin. The flux or absorption rate of a chemical results from a multiplication of the concentration (C) of a chemical on the skin surface with the permeability constant: flux = KpC (Mattie et al. 1994). The basic principles of absorption through the skin correspond to those of diffusion through semi-permeable membranes (Berti et al. 1995). Factors that influence the penetration through the skin are the thickness and condition of the skin, as well as the size of the penetrating substance and the carrier.

Molecules that enter and diffuse through the skin have to penetrate a number of lipid bilayers in the intercellular space, thereby repeatedly alternating from lipophilic to hydrophilic areas. Those molecules that are sufficiently lipophilic, such as glucocorticoids, easily cross those lipophilic phases. Thus, most publications still maintain as a rule that "highly lipophilic compounds with low molecular weights demonstrate the greatest flow rate through the stratum corneum" (Berti et al. 1995).

Occasionally, a direct relation between the coefficient of permeation and the octanol/water distribution coefficient is postulated (Guy 1995). The latter is a measure of a chemical's lipophilic and hydrophilic properties, respectively. A higher coefficient indicates stronger lipophilic characteristics. However, such a correlation could not be verified in experimental studies. Mattie et al. (1994) examined 13 substances with octanol/water coefficients ranging from zero to 1,400. The constant of permeability correlated only weakly with the octanol/water distribution coefficient (r2 = 0.04).

Instead, there is evidence that only a small fraction of strongly lipophilic substances, such as THC, overcomes the hydrophilic phases of the intercellular space. Gabriele Bast (1997) carried out a large number of experiments that involved different substances of differing lipophilic characteristics in different carriers, and stated: "When substances are applied in a lipophilic carrier the permeation coefficient Kp is notably decreased when the distribution coefficient (n-octanol/perfusion buffer (pH 7.4)) Poct exceeds 2000."

Conclusion: It may validly be assumed that, with an octanol/water distribution coefficient of 6,000 (Agurell et al. 1986), i.e. a strong lipophilic tendency, only a small amount of THC permeates the skin-and is systemically absorbed only on a small scale-when administered in an oily base, such as in cosmetics containing hemp oil. Based on experimental evidence obtained for other chemicals with known physico-chemical properties, the transdermal systemic bioavailability of THC thus is likely considerably less than the oral systemic bioavailability. Corresponding experimental studies should be conducted that quantify the exact rate of skin permeation.

3.2 Transport and metabolism

After absorption and infusion into the blood, more than 97% of THC and its metabolites are bound to plasma proteins. The octanol/water distribution coefficient is 6,000, the apparent volume of distribution in the body comes to 10 l/kg, typical of a lipophilic drug (Agurell 1986). THC crosses the blood-brain barrier comparatively easily and accumulates in fatty tissues from where it is re-released only slowly into other tissues, such as the blood. It is exponentially eliminated from the plasma, consistent with a multi-compartment model (2-4 compartments) with a terminal (ß-) plasma half-life of about 20-30 hours.

Ninety-five percent of THC is metabolized in the liver. Through microsomal hydroxylation, THC is converted to the equally pharmacologically effective 11-hydroxy-delta-9-THC, which in turn is converted into 11-nor-9-carboxy-delta-9-THC (THC-COOH) by the operation of alcohol dehydrogenase enzymes. Especially after oral administration, the 11-hydroxy-metabolite contributes considerably to the pharmacological effects, equaling those of THC. Besides the main metabolites, more than 20 other decomposition products are formed. The end products are 11-nor-acids and similar, more polar acids. About one third of those metabolites is excreted through the kidneys and about two thirds of them through the feces. The non-metabolized THC (5%) is defecated as well. The excretion through the urine is limited to acid metabolites only. THC-COOH has an elimination half life of 4-5 days. The complete elimination of a single THC dose may take up to 2-5 weeks. In chronic, heavy marijuana users, THC metabolites were still detected in the urine after 2-3 months after cessation of consumption.

Excursus III: Detection of THC after ingestion of hemp-containing food

The intake of a single oral dose of 16 or 33 mg THC by different experimentees caused a urine concentration of the THC metabolite THC-COOH of 170-240 ng/ml in an immunoassay. According to a GC/MS analysis (gas chromatography/mass spectrometry), these concentrations reached about 400 ng/ml (Lehman et al. 1997). With a chronic intake of such THC quantities, the metabolites accumulate. Consequently, a urine assay of heavy chronic Cannabis users might find THC-COOH concentrations of 500 to 1,000 ng/ml (Solowij et al. 1995). Because of the long elimination half-life of the THC metabolites and an accumulation in body tissues, they might remain detectable in the urine for weeks or even months after the last use (Bell et al. 1989, Ellis et al. 1985). In 86 chronic Cannabis users, THC metabolites were detected in an immunoassay up to 77 days after the last intake.

The ingestion of THC with food may also result in the detection of THC metabolites in the urine. A single oral intake of 40 ml of a hemp oil with a comparatively high THC content of 151 mg/ml, corresponding to about 6 mg THC, produced a maximum urinary THC-COOH concentration of about 100 ng/ml in an immunoassay (Alt and Reinhardt 1996). The intake of 135 ml of hemp oil containing an unknown THC concentration over a period of 4.5 days led to a THC-COOH concentration of 55 ng/ml (Struempler et al. 1997). It took two days after the last ingestion before the immunoassay tested negative. In another study, different patients tested positive for THC-COOH in the immunoassay after the intake of 15 ml hemp oil (Constantino et al. 1997). The daily intake of 10 ml hemp oil of an unknown concentration for 25 days caused a THC-COOH-concentration of 36 ng/ml, measured by GC/MS (Callaway et al. 1997).

Energy bars containing hemp seeds show only low THC concentrations. For example, a THC concentration of 4.4 mg/ml was found in the "Hempy-bar" energy bar of Green Machine Ltd. (Alt and Reinhardt 1997). After ingestion of one or two hemp seed bars, the urinary immunoassay tested positive (cutoff of 20 ng/ml). However, if measured quantitatively by CG/MS, only low concentrations, typically 1-2 ng/ml were detected (Fortner et al. 1997).

Conclusion: After the ingestion of hemp-based food, THC metabolites may be detectable in the urine. The observed THC-COOH concentration corresponds to the to the amount of THC ingested with food. Relevant THC-COOH concentrations are detected in the urine especially after intake of hemp oil, whereas other hemp products produce only low levels of metabolites as a result of their low THC content. Because of the accumulation of THC metabolites in body tissues, a long-term intake of small THC quantities may cause significant urine concentrations. In general, those results cannot be distinguished from those obtained after low-level drug consumption. However, the high concentrations that are found after chronic heavy marijuana use cannot be detected following ingestion of hemp seed foods.

3.3 Influence of physical factors on THC content

Ninety-five percent of the THC present in the Cannabis plant is found in a pharmacologically inactive form, i.e. one of two delta-9-tetrahydrocannabinolic acids (THCA) (Turner 1980), while the majority of biological effects are caused by the corresponding neutral phenolic forms of THC (Dewey 1986). Thus the question arises to what extent the total amount of THC species detected in food is pharmacologically active.

A conversion of THC acids into the pharmacologically active phenols-chemically a decarboxylation process (separation and release of carbon dioxide)-is accomplished most effectively by heating. A heating for five minutes to 200 to 210°C was found to be optimal for decarboxylation (Brenneisen 1984). Under these conditions, THCA was completely converted into neutral THC, while avoiding the subsequent oxidation to cannabinol (CBN). When marijuana is smoked and temperatures of 600° C are reached, obviously only a few seconds are sufficient for decarboxylation.

A much more gradual decarboxylation occurs at room temperature. Hence, after marijuana has been stored for a year, about 50% of its THCA had been converted into the active, neutral THC (Brenneisen 1984). However, storage also leads to a decrease in total THC content as THC oxidizes to neutral CBN, a non-psychotropic cannabinoid (Fairbairn 1976). The total THC content of marijuana dropped to 87% after 47 weeks of storage in the dark (20° C) and to 64% with exposure to light (Fairbairn 1976). Since THC in food is not protected by the plant's glands, as it is in marijuana, THC present in food is converted much faster into CBN. This is suggested by experiments with powdery Cannabis resin and alcoholic extracts (Fairbairn 1976).

Baker et al. (1981) analyzed 64 marijuana samples (Cannabis herb) and 26 hashish samples (Cannabis resin) for their relative amounts of THCA and THC, and found a wide range of ratios, especially in marijuana. In Cannabis resin, the ratio ranged between 0.5 to 1 and 6.1 to 1. Lower rates, corresponding to a low THCA fraction, were found in Cannabis samples from the Indian subcontinent, whereas samples originating from Mediterranean countries displayed higher rates. It may be assumed that much of the THC in cold-pressed hemp oil and other hemp based food is present in the form of the biologically inactive tetrahydrocannabinolic acid, as long as heating during the production process was insufficient for effective decarboxylation. Thus, for a realistic toxicological assessment of food, the concentration of phenolic THC must be determined. Only few studies have examined the percentage of pharmacologically inactive THC-acids in hemp based food. In a study of 10 commercially available hemp oils in Switzerland, THC-acids generally constituted less than 10% of the total THC-content (Lehmann et al. 1997). Thus, in this case the by far larger fraction was biologically active.

3.4 Mode of action

Most specific THC effects are mediated through cannabinoid receptors (reviews: Howlett 1995, Pertwee 1995, Matsuda 1997). At very high doses, non-specific effects on membrane fluidity and other non-specific effects may also become relevant (Martin 1986).

So far two types of THC receptors, CB1 and CB2, each with additional subtypes, have been identified and cloned. The CB1 receptor is found predominantly in brain cells, with a particularly high receptor density in motor, limbic, associative, cognitive, sensory and autonomic brain structures (basal ganglia, cerebellum, limbic system, hypothalamus, cerebral cortex). In addition, it was also found in the testes and other peripheral tissues.

The CB2 receptor has so far only been found outside the brain, particularly in cells of the immune system, such as in the spleen, tonsils, thymus, mast cells, and blood cells. Presumably it is involved modulating the operation of immune cells. Often CB1 and CB2 receptors are expressed from the same immune cells.

The endogenous ligands first discovered for the cannabinoid receptors were arachidonic acid derivatives (arachidonylethanolamides), which differ greatly from plant cannabinoids in their molecular structure (Devane 1992). They are called anandamides (reviews: Di Marzo and Fontana 1995, Mechoulam et al. 1996). Only recently another ligand, 2-arachidonylglycerol (2-AG) was identified (Stella et al. 1997).

3.5 Development of tolerance

Tolerance develops to most THC effects (Romero et al. 1997). This applies also to undesirable effects, such as effects on the psyche, the cardiovascular system and the hormonal system. An acute use of THC, for example, leads to a marked increase of the heart rate, whereas with chronic administration this effect no longer occurs, or at least not to the same extent. This tolerance phenomenon can be explained by two factors, i.e. enhancement of THC metabolism and down-regulation of brain cannabinoid receptors. The number of receptors decreases and their response to THC declines, such that only higher doses can produce the known THC effects. Thus, rats that had been administered THC over a period of five days exhibited a decreased specific binding in different receptor sites of the brain, ranging from 20 to 60% of that measured in controls. (Romero et al. 1997).

3.6 THC effects and total toxicity

THC has acute effects on almost every body system (reviews: Hall et al. 1994, Hollister 1986, Dewey 1986, Maykut 1985). The most conspicuous effects are those on the central nervous and cardiovascular systems. With regard to physical effects, THC produces an increased heart rate, reddened eyes and a dry mouth. As for psychotropic effects, a mild euphoria, an enhanced sensory perception, fatigue and eventually dysphoria together with anxiety have been observed.

As a function of dose, the following effects were observed in clinical studies in vivo (in living organisms) or in vitro (i.e. in laboratory dishes), respectively:

_ Psyche and perception: fatigue, euphoria, enhanced well-being, dysphoria, anxiety, disturbed orientation, increased sensory perception, heightened sexual experience, hallucinations, psychotic states.

_ Cognition and psychomotor performance: fragmented thinking, enhanced creativity, disturbed memory, unsteady walk, slurred speech.

_ Nervous system: attenuation of pain, muscle relaxation, appetite enhancement, decrease in body temperature, vomiting, anti-emetic effects, neuroprotective effects in brain schema.

_ Cardiovascular system: increased heart rate, enhanced heart activity and increase in oxygen demand, vasodilation, drop in blood pressure, collapse.

_ Eye: reddened conjunctivae, reduced tear flow, reduced intraocular pressure.

_ Respiratory system: bronchodilation, dry mouth.

_ Gastrointestinal tract: reduced bowel movements.

_ Hormonal system: effects on LH, FSH, testosterone, prolactin, somatotropin, TSH, reduced sperm count and sperm mobility and quality, suppressed ovulation and suppressed menstruation.

_ Immune system: impairment of cell-mediated and humoral immunity, anti-inflammatory and immune-stimulating effects.

_ Fetal development: fetal malformations, fetal growth retardation, impairment to fetal and postnatal cerebral development, improved postnatal development.

Despite these observed effects, the overall physical human toxicity of Cannabis is low. Human fatalities following acute Cannabis intoxication have not been reported. Chronic heavy marijuana use is also not related to mortality (Sidney et al. 1997). Long-term heavy marijuana users did not show any abnormal health features that distinguished them from the population as a whole (Gruber et al. 1997).

Strong side effects are rare, even with high THC doses. The median lethal dose (LD50) for rats is in the range of 800 to 1900 mg/kg (Thompson et al. 1973). No toxic deaths were observed in experiments of rhesus monkeys with an acute oral application of 9000 mg/kg (Thompson et al. 1973). For illustration purposes: 9000 mg/kg THC in a man weighing 70 kg corresponds to the consumption of 630 grams THC or 3 kg of high-percentage Cannabis resin (hashish), or 15 kg of marijuana of a medium quality.

Even a long-term high-dosing regimen of THC is tolerated relatively well. This was suggested by the example of rats that ingested 50 mg/kg THC every day for two years (Chan et al. 1996). At the end of the two-year period a mean of 45% of the controls and 70% of the dosed animals had survived. The higher survival in the THC group was primarily due to a decreased incidence of cancer.

4 THC thresholds for psychotropic effects

Some experimental and clinical studies report experiences with threshold values for psychotropic THC doses. Acute effects below the psychotropic threshold cannot be distinguished from placebo effects.

Lucas and Laszlo (1980) found pronounced psychotropic reactions (anxiety, marked visual distortions) in patients undergoing cancer chemotherapy who had received 15 mg THC/m2 (square meter of body surface), which corresponds to 25 mg THC in an average adult person (body surface: 1,7 m2). A reduction to 5 mg THC/m2, about 7.5-10 mg THC, produced only mild reactions.

No mood changes were observed in six cancer patients after administration of a single oral dose of 15 mg THC for antiemetic treatment (Frytak et al. 1984). Brenneisen et al. (1996) administered single oral doses of 10 or 15 mg THC to two patients. Physiologic parameters (heart rate) and psychological parameters (concentration, mood) were not modified by the administration. In a study by Chesher et al. (1990) of a healthy population dosed orally with 5 mg THC, no difference in the subjective level of intoxication was found compared to placebo controls. Doses of 10 and 15 mg THC caused slight differences compared to the placebo, and a dose of 20 mg, finally, caused marked differences in subjective perception.

In light of these findings, one may validly assume the psychotropic threshold to be in the range of 0.2-0.3 mg THC per kg body weight for a single oral dose taken in a lipophilic carrier, corresponding to an administration of 10-20 mg THC to an adult person. A single dose of 5 mg THC can be regarded as a placebo dose. In various clinical studies, psychotropic reactions were also observed following single doses of 5 mg THC. However, these cannot be distinguished from effects that occur after administration of placebos. As the duration of action of THC in therapeutic dosage ranges between 4 and 12 hours, a daily intake of 2 x 5 mg which equals 10 mg THC, administered orally in a lipophilic carrier, will not have any effects that could be distinguished from placebo effects.

Return to Hemp Reports