Geoffrey W. Guy and Colin G. Stott
GW Pharmaceuticals plc, Porton Down Science Park, Salisbury, Wiltshire SP4 OJQ, UK
Cannabis has been used medicinally for 4000 years [1-4] in a variety of cultures and was re-introduced into British medicine in 1842 by W. O'Shaughnessy . It remained in the British pharmacopaeia until 1932, when cannabis, extract of cannabis and tincture of cannabis were among 400 medicines removed, though all three remained in the British Pharmaceutical Codex of 1949 .
However, following the 1961 UN Single Convention on Narcotic Drugs, cannabis and cannabis derivatives became scheduled products and were subject to special measures of control and parties could ban their use altogether. Following the 1971 UN Convention on Psychotropic Substances, the UK enacted the Misuse of Drugs Act 1971. Cannabinol and its derivatives, including A9-tetrahydrocannabinol (A9-THC), appeared in Schedule I to the Convention, and their regular medical use was prohibited. The introduction of the Misuse of Drugs Regulations in the UK in 1973 listed cannabis and cannabis products in Schedule 4 (now Schedule I in current legislation), thereby prohibiting medical use altogether .
Although the medicinal properties of cannabis had been well documented for a number of years, the constituent(s) responsible for therapeutic efficacy had, until recently, not been identified. The discovery, isolation (and subsequent synthesis) of the principal cannabinoid present in cannabis, A9-THC, by Raphael Mechoulam and Yehiel Gaoni in 1964  ensured that interest in cannabinoid chemistry remained and led to an expansion of cannabinoid research.
Despite the scheduling and prohibition of cannabis and the ban on medical use of cannabis-based products in the 1970s, research into the pharmacology and toxicology of A9-THC continued through the 1970s and 1980s, mainly by the National Institute of Health (NIH) in the USA.
However, much of the work concentrated solely on A9-THC (NTP program, NIH) . In many cases, the investigation of the pharmacokinetics of cannabis components involved the delivery of smoked marijuana, and the measurement of A9-THC levels and its primary metabolite, 11-hydroxy-tetrahydrocannabi-nol (11-OH-THC).
Recent research and development of a cannabis-based medicine
In January 1997, the White House Office of National Drug Control Policy (ONDCP) asked the Institute of Medicine (IOM) to conduct a review of the scientific evidence to assess the potential health benefits and risks of marijuana and its constituent cannabinoids. That review began in August 1997 and resulted in the report published in 1999 . Reports were also published in August 1997 by the US NIH  and in December 1997 by the American Medical Association (AMA) .
In parallel with the timing of the IOM review, a number of expert bodies in the UK were asked to review the medical and scientific evidence for and against the use of cannabis as a medicine. The British Medical Association (BMA) published a report on the topic in 1997 . The UK Department of Health commissioned three literature reviews on cannabis, at the request of the Advisory Council on the Misuse of Drugs (ACMD); and these were reviewed by the House of Lords Select Committee on Science and Technology in 1998. The authors of the report all gave evidence to the House of Lords inquiry [12-14].
Dr Geoffrey Guy was also invited to submit evidence to the House of Lords enquiry, and subsequently GW Pharmaceuticals Ltd was founded in the UK in early 1998. As GW's Executive Chairman, Dr Guy successfully floated the company (GW Pharmaceuticals plc) on the Alternative Investment Market (AIM) of the UK Stock Exchange in June 2001. The first UK Home Office licenses received by GW were to cultivate, possess and supply cannabis for research purposes were received in June 1998 and cultivation began in August 1998.
In November 1998, the House of Lords Select Committee on Science and Technology published its report Cannabis: The Scientific and Medical Evidence , which recommended that clinical trials of cannabis medicines should be carried out as a matter of urgency. The Committee warmly welcomed GW's research programme.
September 1999 saw the start of GW's first phase I clinical trials in healthy volunteers and in March 2000 GW received authorization from the Medicines Control Agency (MCA; now the Medicines and Healthcare Products Regulatory Agency, MHRA) to start phase II clinical trials in patients.
In March 2001, the same House of Lords Select Committee published a follow-up report, Therapeutic Uses of Cannabis , which confirmed the UK Government's intention to permit the prescription of cannabis-based medicines (CBMs) subject to the approval of the MHRA.
GW entered into its pivotal phase III clinical trials programme in March 2001. The initial phase III studies involved patients with multiple sclerosis (MS), neuropathic pain and cancer pain. The results of the first four phase III studies were reported in November 2002, and six of the trials have now been completed, yielding positive results, and a further three are due to report in 2005.
In March 2003 GW submitted an application to the MHRA for its first product, Sativex®.
In May 2003 GW entered into an exclusive UK marketing agreement for Sativex® with the German pharmaceutical company Bayer AG. This agreement was extended in November 2003, to add the Canadian market.
In May 2004 GW submitted a New Drug Submission for Sativex® to the Canadian regulatory authorities, Health Canada.
The discovery and chemical synthesis of A9-THC initiated the modern era of cannabis research because it enabled investigation of the effects and mode of action of individual cannabinoids in laboratory models . The production of synthetic analogues of A9-THC enabled structure - activity relationships of A9-THC to be established. Further, pharmacological investigation of A9-THC indicated that it might exert its effects by interacting with a specific receptor protein in the brain [18, 19]. The conclusion from this work was that the so-called cannabinoid receptor was a G-protein-coupled receptor. Once a CB receptor agonist, CP-55,940, was synthesized, radiolabelled binding studies were performed , and the distribution of CP-55,940-binding sites were found to be similar to those coded for by cDNA for another G-protein-coupled receptor, SKR6, a receptor without a known ligand (an orphan receptor). Further investigation using cannabinoid-binding assays revealed that SKR6 was indeed a cannabinoid receptor identified in rat brain . Soon afterwards a human G-protein receptor was identified that had an amino acid sequence 98% identical to the SKR6 receptor in rat brain.
In 1993, a second G-protein-coupled cannabinoid receptor sequence (CX5) was identified among cDNAs from the human promyelocytic leukaemic cell line HL60 .
Munro et al.  suggested that the brain receptor be referred to as CB1 and that the second receptor, which is expressed by cells of the immune system, be referred to as CB2.
It has since become widely accepted that CB1 receptors are widely distributed but are particularly abundant in some areas of the brain, including those concerned with movement and postural control, pain and sensory perception, memory, cognition and emotion, and autonomic and endocrine functions [23, 24]. They are also prevalent in the gut, testes and uterus. The role of the second type of receptor, CB2 receptor, is still under investigation but it is believed to mediate the immunological effects of cannabinoids [23, 24].
In the meantime, Mechoulam and Devane isolated and elucidated the structure of a brain constituent that bound to the cannabinoid receptor : arachi-donylethanolamide (AEA, anandamide). During subsequent investigation of several lipid fractions collected from rat brain, it was discovered that the fractions also contained materials that bound to cannabinoid receptors . Characterization of these fractions revealed that some contained polyunsaturated acid ethanolamides (similar to AEA), but others contained a distinct lipid component, 2-arachidonoyl glycerol (2-AG).
AEA is found to be a partial agonist at CBj receptors; whereas 2-AG binds to CBi and CB2 with similar affinities, and is a full agonist at CBj. 2-AG occurs in concentrations in the brain that are 170 times higher than those of AEA .
The role of these endogenous cannabinoids (so-called endocannabinoids) is currently unclear, and others have subsequently been identified: noladin ether , virodhamine , N-arachidonoyl-dopamine (NADA)  and arachi-donoyl-serine (ARA-S) . The identification of AEA and 2-AG has led to a resurgence of interest in the field of cannabinoid medicine, especially within the pharmaceutical industry, as they may represent potential molecular targets for the treatment of a number of disorders.
In the wake of widespread availability of synthetic CB receptor-specific ligands, research into the identification of potential sites of action of cannabinoids has increased around the world. However, until recently, the lack of significant available quantities of pure cannabinoids other than A9-THC and cannabidiol (CBD) has been a constant source of frustration for researchers.
To date, of the synthetic research receptor ligands, only SR-141716A (CB1 receptor antagonist) has shown sufficient potential to be developed into a pharmaceutical product (Rimonabant). A number of other synthetic cannabinoids have been developed into pharmaceuticals including Marinol®, Synhexyl, Nabilone and Levonantradol. However, regulatory approval of these products varies between territories and, as a result, they are not currently widely used or accepted.
The existence of the various types of cannabinoid molecule available and their source has led to the proposal of four distinct classes of cannabinoids:
1. phytocannabinoids: those which occur naturally in the plant;
2. endocannabinoids: those that occur naturally in the body (AEA, 2-AG, etc.);
3. synthetic cannabinoids: cannabinomimetic compounds resulting from chemical synthesis (e.g. dronabinol, nabilone, HU-210, CP-55,940, SR-141716A);
4. fatty acid amide hydrolase (FAAH) inhibitors: compounds that affect AEA production, release, metabolism and re-uptake.
Production of cannabis-based medicines
Cannabis-based medicines may be produced according to the regulatory requirements in a variety of ways:
• isolation and purification of individual molecules from plant sources;
• chemical synthesis of required molecular components;
• extraction of required plant components;
• selective delivery of required components.
Rationale for the development of a cannabis-based medicine as a whole-plant extract
The cannabinoids that are currently of most interest and have received the most scientific interest to date are the principal components of cannabis, A9-THC and CBD. Both have important pharmacology [31, 32]. A9-THC has analgesic, anti-spasmodic, anti-tremor, anti-inflammatory, appetite-stimulant and anti-emetic properties; CBD has anti-inflammatory, anti-convulsant, anti-psychotic, anti-oxidant, neuroprotective and immunomodulatory effects. CBD is not intoxicating and indeed it has been postulated that the presence of CBD in cannabis may alleviate some of the potentially unwanted side effects of A9-THC.
It is postulated that the beneficial therapeutic effects of cannabis result from the interaction of different cannabinoids . This may explain why cannabis-based medicines made from whole-plant extracts may be more effective than single cannabinoid products, as the extracts consist of multiple cannabinoids in defined, specific ratios. Different ratios of cannabinoids may be effective in treating different diseases or conditions across a number of therapeutic areas.
Although research has focused primarily on the two principal cannabinoids, A9-THC and CBD, it is possible that other components within the plant are also important, which is why GW Pharmaceuticals' medicines are made from whole-plant extracts. McPartland and Russo  cite a number of literature reports, which support this theory. Mechoulam et al.  suggested that other compounds present in herbal cannabis might influence A9-THC activity. Carlini et al.  determined that cannabis extracts produced effects "two or four times greater than that expected from their THC content." Similarly, Fairbairn and Pickens  detected the presence of unidentified "powerful synergists" in cannabis extracts causing 330% greater activity in mice than A9-THC alone.
Other compounds in cannabis may ameliorate the side effects of A9-THC . Whole cannabis causes fewer psychological side effects than synthetic
A9-THC, seen as symptoms of dysphoria, depersonalization, anxiety, panic reactions and paranoia .
It is possible that the observed difference in side-effect profiles may also be due, in part, to differences in routes of administration: orally administered A9-THC undergoes 'first-pass metabolism' in the small intestine and liver, to 11-OH-THC; and the metabolite has been reported to be psychoactive, albeit on the basis of limited evidence . Inhaled A9-THC undergoes little first-pass metabolism, so less 11-OH-THC is formed [38,39]. The effect of the route of administration on tolerability has been known for years. Walton, in 1938, remarked that "smoking cannabis is a satisfactory expedient in combating fatigue, headache and exhaustion, whereas the oral ingestion of cannabis results chiefly in a narcotic effect which may cause serious alarm" .
The other classes of compounds present in cannabis also have their own pharmacology (e.g. terpenoids, flavonoids) [31, 32]. The potential for interaction and synergy between compounds within the plant may play a role in the therapeutic potential of cannabis as a medicine. This may explain why a cannabis-based medicine using extracts containing multiple cannabinoids, in defined ratios, and other non-cannabinoid fractions, may provide better therapeutic success and be better tolerated than the single synthetic cannabinoid medicines currently available.
CBD, as a non-psychoactive cannabinoid, is currently the cannabinoid of considerable interest. CBD, along with A9-THC, has been demonstrated to have a wide range of pharmacological activity, with the potential to be developed for a number of therapeutic areas . It is likely that other cannabinoids, present in small amounts in Cannabis sativa L., may also have interesting pharmacological properties, for example tetrahydrocannabivarin (THC-V), cannabichromene (CBC) and cannabigerol (CBG) [31, 32, 39].
The pharmaceutical development of cannabis-based medicines is well documented [42, 43]. For cannabinoids to be made into pharmaceuticals, licensed by the regulatory bodies around the world, they must reach strict requirements laid down in terms of the product's quality, safety and efficacy and increasingly the healthcare industry requirement of cost-effectiveness. Such standards are achieved by adhering to the industry and regulatory standards of Good Laboratory Practice (GLP), Good Manufacturing Practice (GMP) and Good Clinical Practice (GCP), according to the guidance documents provided by the International Conference on Harmonisation . All requirements are now implemented through European Union and national legislation. In the case of plant-based medicines they must also adhere to Good Agricultural Practice (GAP) standards.
As a result, quality control is required throughout the whole of the manufacturing chain, including the production of raw materials. For pharmaceuti-
cals produced from plants, the regulatory authorities have produced their own guidelines on the production of botanical drug products (BDPs) . As botanical pharmaceuticals have more than a single chemical entity present, their control is paramount, and hence detailed characterization and specification is required.
Breeding of cannabis plants for generation of cannabis extracts
Cannabis is in most cases a dioecious plant; that is to say, the species produces separate male (staminate) and female (pistillate) plants .
Analysis of the various parts of the plant confirms that the major source of cannabinoids is the female flower. Cannabinoids are not detected in the roots. The richest sources of the principal cannabinoids A9-THC and CBD are the leaves and flowers and hence these plant components are selected for the production of A9-THC- and CBD-based medicines.
In the wild, Cannabis is a short-day-length plant. This means that the plant grows vegetatively through the long days of summer. Only when the day length falls, signalling the end of summer, does the female plant start to flower and hence the cannabinoids are produced. As an annual herb in the field, normally only one crop per year would be produced.
It is during the last few weeks of life that the female plant is most active in the production of cannabinoids and terpenes. The plant will produce variable inflorescences, these being complex clusters of flowers and bracts. Each flower consists of a furled specialized single leaf - the calyx - within which is housed the ovary. Each calyx is covered in minute sticky organelles - the stalked glandular trichomes. When viewed through a hand lens, each trichome resembles a golf ball (the resin head, also known as the glandular head) sitting on a tee (the trichome's stalk; Fig. 1)
The particular day length that induces flowering is termed the 'critical day length' . This will differ according to the geographical and genetic origin of the plant in question. Thus, flowering in response to exposure to a defined amount of light may be achieved through selective breeding.
Cannabinoid content varies in different varieties but the high cannabinoid content of modern varieties is purely due to plant breeding.
However, by growing under glass in controlled conditions, a succession of crops can be planned to meet production requirements. To be suitable for long-term commercial use, plants must have selected characteristics. Plants that are selectively bred for their characteristics are termed chemovars. In order to be commercially useful, they must possess the following characteristics:
• high rate of cannabinoid production;
• high yield of cannabinoid per unit area;
• high level of purity of the desired cannabinoid (purity as used here defines the consistency of cannabinoid content as a ratio);
• high inflorescence-to-leaf ratio (the harvest index);
• natural resistance to pests and diseases;
• sturdy growth capable of bulk plant handling;
• minimal production of anthers on female plants.
The production of uniform high-quality botanical raw material (BRM) of defined composition is dependent upon the bulk production of cloned plants; that is to say, all plants are derived from cuttings taken from a few select mother plants. Being genetically identical, all the cloned plants have the potential to replicate exactly the characteristics of the mother plant.
BRM is obtained from distinct varieties of C. sativa plant hybrids to maximize the output of specific cannabinoids. The chemovars used are the result of an extensive breeding programme spanning more than 15 years.
GW's cannabis-based medicines are pharmaceutically formulated whole-plant extracts of chemovars of C. sativa produced by selective breeding to give a high content of defined cannabinoids, optimum habit and early flowering. A wide range of chemovars of C. sativa has been selectively bred by GW Pharmaceuticals. Each of these chemovars has a different cannabinoid profile, and the chemovars have been specifically bred to produce the required level of specified cannabinoids. From this range, two separate chemovars, one that produces A9-THC as the principal cannabinoid and one that produces CBD as the principal cannabinoid, have been selected for production of Sativex®.
Cultivation of chemovars for generation of cannabis extracts
Crops are produced from cuttings, which ensures that the genotype is fixed, giving a constant ratio of cannabinoid content. Cannabinoid content may be selectively bred to produce defined ratios of principal and other minor cannabinoids. By further careful, selective breeding, it is possible to cultivate chemovars which produce minor cannabinoids (CBC, CBG, THC-V, etc.) in greater amounts than have been observed to date in wild-type cannabis plants or in varieties produces by recreational growers. The pharmacology of the minor cannabinoids has yet to be clearly established, but may yet provide a whole new range of therapeutic options for both patient and clinician.
Potter  has described the use of "mother plants" to maintain the genotype for each subsequent generation of plants (rooted cuttings, termed "clones"). Once potted up and grown in continuous bright light [75 W/m2 PAR (photo-synthetically active radiation)] at 25 °C in optimized compost, a rooted cutting will reach a height of 2 m in 12 weeks. This plant is then capable of being heavily pruned; the removed branches being cut up to produce up to 80 cuttings per mother plant. If well kept, over the next 10-15 weeks the trimmed mother plant will regrow to produce at least two more flushes of cuttings. The vigour of the mother plant then wanes, and the plant is destroyed to make way for younger mothers.
Branches of the mother plant are removed where there are sufficient numbers of axial buds developing, these being the new growths that eventually develop into mature plants. Each branch is then cut into sections, each supporting only one axial bud. The cutting is then placed in rooting powder and immediately transferred into a very moist peat plug. In the correct environment, roots begin to appear after 7 days, and the cuttings allowed to acclimatize to their surroundings before they are potted up.
Rooted cuttings are transferred into large pots, filled with a proprietary growing media, which contains sufficient fertilizer to stimulate vegetative growth and flower production.
For the first 3 weeks after potting, plants are grown in continuous bright light. With no night-time breaks during this period the plant grows to around 50 cm and establishes a healthy root system.
After 3 weeks the lighting is switched to a 12-h light/12-h dark cycle. Having established themselves in a 24-h daylight environment in subtropical temperatures, the plants suddenly detect the change in light exposure, as if they had experienced the immediate arrival of the autumn equinox. For a short-day plant (i.e. late summer/autumn flowering) like cannabis, the response is dramatic. The GW chemovars flower within 5 days of the photoperiod switch. The inflorescences (flowers) increase in size over the next 6 weeks, becoming white with myriad receptive stigmas. The unfertilized stigmas then start to senesce to an orange/brown colour. After 8 weeks in flower, the bulk of stigmas have senesced and the rate of cannabinoid biosynthesis in the selected varieties slows rapidly. At this point, the crop is harvested.
Mother plants, seedlings and mature clones are produced under glass, which allows a very high degree of control of growing conditions to be exercised. The controls significantly exceed the controls possible for field-grown crops. In particular:
• proprietary compost is used, warranted free of artificial pesticides and herbicides by the supplier;
• the compost contains sufficient fertilizer to ensure optimum vegetative growth and eventual flowering;
• stringent hygiene conditions reduce ingressive pests and diseases - adventitious infestation is controlled biologically with predatory mites;
• fresh potable water, rather than stored or untreated water, is used for the irrigation of the plants; this reduces the potential for contamination with water-borne organisms;
• during growing, the plants are inspected regularly, and plants showing male characteristics are removed to avoid fertilization of plants;
• growing conditions are strictly controlled via computer technology to ensure that optimal cultivation conditions are maintained at all times in terms of light, temperature, humidity, airflow, etc.
At harvest, the entire plant is cut and dried in a temperature- and humidity-controlled environment until it meets the specification for loss on drying. Leaves and flowers are stripped from the larger stems to provide the BRM, which is stored in suitable containers protected from light under controlled conditions.
Drying the crop as quickly as possible reduces the cannabinoid losses, and this is achieved by keeping the plants in a stream of dehumidified air. Plants are crisp to the touch in less than 7 days.
As part of GAP and GMP, the BRM must conform to a specification. The specification for BRM includes tests for identification, extraneous matter and identification and assay for cannabinoids and cannabinoic acids, confirmatory thin-layer chromatography (TLC) and loss on drying. Additionally, BRM is tested for aflatoxins and microbial bioburden. The growing parameters employed have been selected to minimize the conditions that would be expected to result in microbial and fungal spoilage.
Cannabinoids are present in the plant as the corresponding carboxylic acid and it is necessary to decarboxylate material before extraction. The conditions for efficient decarboxylation have been optimized to maximize decarboxylation and minimize oxidation. The process is time- and temperature-dependent and a criterion of not less than 95% efficiency was adopted for BRM used in subsequent manufacture of botanical drug substance (BDS; whole-plant extract).
Development work has shown that efficient extraction can be carried out using patented extraction technology. The conditions of the extraction have been carefully assessed during development and are essential to ensure the optimum conditions and hence the correct composition of the extract produced. The extraction produces a whole-plant extract, from which the BDS is prepared.
The whole-plant extract is subject to further processing (covered by intellectual proprietary rights) to remove unwanted materials from the extract. The exact content of the BDS is defined by a specific BDS specification. BDS is transferred to sealed, stainless steel containers and stored at -20 ± 5 °C to maintain stability.
A schematic diagram of the process flow from cultivation to final processing and quality-control release of the pharmaceutical product is detailed in Figure 2.
Using any defined BRM, a corresponding BDS may be created using the above GW proprietary process. The contents of the BDS will depend on the genetically defined content of the BRM, and the technology used to extract the active constituents. Thus, BDSs may be produced which have defined levels of principal cannabinoids, other cannabinoids and other non-cannabinoid constituents. Thus a series of individual BDSs may be described.
Each BDS contains a cannabinoid fraction and a non-cannabinoid fraction. GW describes its BDSs individually as each BDS generated has a unique composition. The two BDSs used to generate Sativex® are Tetranabinex®, an extract of a chemically and genetically characterized cannabis plant, contain-
Propagation from mother plants
Was this article helpful?