WO2024165731A1 - A cannabinoid-based oil-in-water emulsion for sublingual and buccal administration - Google Patents

Abstract

The present disclosure relates to a formulation which is an oil-in-water emulsion comprising; a water phase; and an oil phase which comprises one or more dissolved cannabinoid compounds; and a surfactant system; wherein the D50 particle size of the oil phase droplets as measured by dynamic light scattering is in the range 100nm to 300nm. The emulsion can be used to treat various diseases or disorders, e.g. pain. It can be administered via a spray for sublingual or buccal absorption.

Description

TITLE A cannabinoid-based oil-in-water emulsion for sublingual and buccal administration TECHNICAL FIELD The present disclosure relates to a cannabinoid-containing oil-in-water emulsion suitable for sublingual and buccal administration to patients suffering from conditions related to imbalances in their endocannabinoid system (ECS). The emulsion can for example be used to manage pain. BACKGROUND Cannabinoids have been known for many centuries to be useful active compounds, which can be used to treat various diseases. For example, these compounds can be used for pain relief. Cannabis plants or extracts from the cannabis plant have been historically used as a natural medicine. More recently, there have been efforts to develop medicaments including cannabinoids, including their incorporation into modern conventional products for oral use such as solutions, tablets, lozenges etc. Such formulation efforts are complicated by the fact that the cannabinoids are highly lipophilic and thus difficult to formulate for oral administration by conventional drug delivery technologies. Conventional tablets require dissolution of the active pharmaceutical ingredient to facilitate permeation across the gastro-intestinal tract and ultimately transport the drug to therapeutic receptor targets via the blood stream. A well- characterized cannabinoid like cannabidiol (CBD) has a reported water solubility of 0.7 µg/mL and an intestinal permeability that collectively render the cannabinoid a Class II drug of the Biopharmaceutical Classification System (BCS). Its poor water solubility presents a significant hurdle to the preparation of an oral delivery formulation which results in acceptable bioavailability of the active ingredients. In addition, first-pass metabolism imposed by the liver after entering the portal vein is known also to eliminate a significant fraction of the drug further restricting bioavailability. An attractive alternative is administration via mucosa in the oral cavity, either as sublingual (under the tongue) or buccal (between gum and cheek) administration. In practice administration by either route often involves a combination of both, due to distribution of the drug in the oral cavity, which is difficult to avoid. With such administration, a fraction of the drug is unintentionally swallowed and thus subject to the above-mentioned fate in the gastro-intestinal tract. To minimize the fraction of drug reaching the GI tract in sublingual/buccal administration, it is desirable to formulate a product that rapidly penetrates the epithelium in the oral cavity. The faster the rate, the less a fraction of dose is swallowed, ultimately improving bioavailability, i.e. the fraction of dose reaching the systemic circulation where it is transported to receptor sites for therapeutic action. In order to increase sublingual and buccal absorption rate, two major aspects of the drug delivery technology must be addressed: 1) Distribution of a dose onto the mucosa and penetration, and 2) Organoleptic properties of the formulation. Some known cannabis-based medicines rely on sublingual administration via a spray. Some such products use ethanol as means to dissolve the cannabinoids. While the low viscosity of ethanol does ensure a distributed spray cone from the spray, a major drawback of such a formulation is its poor organoleptic properties owing to the burning sensation of ethanol on the mucosa. Mouthfeel and taste are essential contributors to patient compliance, both of which can cause the patient to supplement the administration with e.g. a flush of water. This will result in substantial loss of dose to the GI tract and thus poor absorption and hepatic clearance. In addition, for drug delivery systems targeting the endocannabinoid system (ECS), high drug loads are needed to increase the likelihood of a therapeutic response for a given disease indication. The ECS is an endogenous multifunctional pro-homeostatic signaling system being almost ubiquitously distributed within the body. The system is generally recognized to consist of three main parts 1) receptors: G protein-coupled receptors (GPCRs); cannabinoid receptor 1 (CB1R) and 2 (CB2R), 2) endocannabinoids: the body´s own signaling molecules regulating the ECS through the cannabinoid receptors, including N-arachidonylethanolamine (anandamide i.e., AEA) and 2-arachidonoylglycerol (2-AG) and 3) the enzymes: responsible for the metabolism and regulation of endocannabinoids available at a given time. In line with the development of research in the field, additional components have been discovered being part of the ECS. This including 1) receptors: GPCRs (e.g., GPR18, GPR55 and GPR119), ion channels (e.g., Transient Receptor Potential Vanilloid 1 (TRPV1) and nuclear receptors (e.g., Peroxisome Proliferator-activated receptor gamma (PPAR-y), 2) endocannabinoid-like compounds: e.g., Palmitoylethanolamide (PEA) and Oleoylethanolamine (OEA) and 3) synthesizing and degradative enzymes and transport proteins of the endocannabinoids and alike ligands. The ECS acts locally at the various body parts it is localized within, where it responds to and is activated by disturbances occurring within these systems, with the aim to recover and maintain homeostasis. Use of exogenous cannabinoids to target a specific disease caused by an imbalance in ECS homeostasis, is complicated by a dose-dependent effect on the receptor response observed in vivo. That is, the level of cannabinoid available at the receptor site induces either advantageous or deleterious effects, for example for cognitive functions where low doses of the CB1 agonist, Δ9-THC, in mice improved cognitive performance whereas high doses impaired these functions (DOI:10.3390/ijms21082778). Finally, a drug delivery system intended for delivery of multiple cannabinoids that collectively aim to establish homeostasis in the ECS, must incorporate a broad dose range to accommodate the difference in receptor binding (and ultimately modulation) of the included cannabinoids. WO2016/147186 A1 claims up to a 50% loading of oil phase in the water phase with up to 50% cannabinoids in the oil phase. The feasibility of such a claim is highly doubtful, considering a concomitant claim of droplet sizes as small as 50 to 400 nm. The above problems are addressed by the present invention, which seeks to develop a cannabinoid-containing mouth spray which simultaneously allows good penetration / absorption and which has good organoleptic properties to ensure compliance. Other objectives include the ability to allow an adequate amount of the active to be contained and delivered, and that the composition must have good chemical and physical stability, in particular avoiding chemical degradation of the active and physical separation of components of the composition over time. It is also an objective for the compositions to be safe and for side-effects to be minimized, and ideally avoided. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 shows a flow diagram illustrating the product of a cannabinoid-based emulsification system in accordance with the embodiments of the present disclosure. Figure 2 shows the DLS particle size distribution of Example emulsion A1. Figure 3 shows stability data of emulsions of the present invention. Figure 4 shows pictures of the emulsions of the present invention. Figure 5 shows the PAMPA membrane used in in vitro testing. Figure 6 shows in vitro data relating to flux of cannabinoids in the emulsions of the present invention through the PAMPA membrane. Figure 7 shows in vivo pharmacokinetic (PK) data relating to bioavailability of cannabinoids in the emulsions of the present invention following GI administration. DESCRIPTION Summary of the Invention The present invention relates to a formulation which is an oil-in-water emulsion comprising; a water phase; an oil phase which comprises one or more dissolved cannabinoid compounds; and a surfactant system; wherein the D50 particle size of the oil phase droplets as measured by dynamic light scattering is in the range 100nm to 300nm. The Oil-in-water Dispersion The present invention formulates cannabinoids in an oil-in- water emulsion. This emulsion is primarily designed for use as an oral spray, wherein the cannabinoid is primarily intended to be absorbed through the oral mucosa – however, the emulsions may also be used for other delivery routes, e.g. oral administration. For an oral spray, the absorption may specifically be sublingual (under the tongue) and/or buccal (between gum and cheek). The formulation is designed to allow the droplets of oil in the emulsion to pass through the oral mucosa. There are two specific formulation aspects which enable this to work well: Firstly, the oil droplets are associated with the surfactant system to form a micelle. As is typically the case in oil-in- water emulsions, the surfactant will be placed on the outside of the oil droplets, i.e. at the interface between the oil and water phases. Thus, the oil droplets are effectively coated with the surfactant molecules so that the droplets can be absorbed readily – the surfactant turns an otherwise hydrophobic droplet into a hydrophilic micelle, which can pass through the oral mucosa much more easily. Secondly, the present inventors have found that the ability of the micelle to pass through the oral mucosa is dependent on the particle size of the micelle. The present inventors have found that the use of an emulsion leads to improved sublingual and/or buccal absorption compared to e.g. products based on cannabinoid solutions, where a solvent is used to dissolve the cannabinoid. In particular, when a cannabinoid solution in ethanol is sprayed into the mouth, it comes into contact with saliva. It is believed that once the composition mixes with saliva, the hydrophilic saliva effectively acts as anti-solvent to the cannabinoid, causing the cannabinoid to precipitate out of solution. Effectively, the hydrophobic cannabinoids are not absorbed in the oral cavity and instead swallowed followed by delayed absorption in the gut. Similarly, it has previously been proposed to dissolve cannabinoids in coconut oil or similar medium chain triglycerides, and to administer these compositions directly. Also relative to such approaches, the present invention offers significant advantages. A cannabinoid dissolved in such oils is not effectively absorbed by the human body due to hepatic clearance; this applies to the fraction of dose entering the GI tract. For this reason, the use of an oil-in-water emulsion (and not a solution in e.g. ethanol or coconut oil) is an important aspect of the present invention. According to the present invention, the oil phase preferably constitutes 2-20 wt.% of the emulsion. The remainder is preferably made up by the water phase and the surfactant system, i.e. there are preferably no other components in the emulsion than the oil phase, water phase and the surfactant system. More preferably, the oil phase constitutes 5-15 wt.% of the emulsion, most preferably 8.5 wt.%. Droplet Size / Particle Size In the present disclosure, the expression “particle size” is used to characterize the size of the oil droplets / micelles, even though these are not solid particles. This is because the techniques for characterizing the droplet sizes is the same as that used in measuring particle sizes in dispersions. The present invention does not relate to dispersions. According to the present invention, it is important that the emulsion has an appropriate particle size. The present inventors have found that the absorption of the cannabinoids varies with the particle size. The present inventors have identified that there is particularly good absorption when the particle size is around 100-300nm, more preferably around 200nm. When the particles are within the range, the absorption is maximized compared to similar formulations which use larger or smaller oil droplets / micelles. According to the present invention, the D50 value is in the range 100-300nm. The D50 value is the particle size value where 50% of the particles are larger than said value, and 50% are smaller, i.e. it represents the median particle size value. In addition, it is preferred that the D10 value is in the range 50-100nm. It is also preferred that the D90 value is smaller than 600nm. Preferably, all of the above D50, D90 and D10 values are simultaneously fulfilled. As is conventional, the D10 value is the particle size where 10% of the particles are smaller than this threshold value. Similarly, the D90 value is the threshold value where 90% of the particles are smaller than this value. It is preferred for as many droplets as possible to be as close to 200nm as possible. In the presently claimed emulsions, it is reasonable to assume that the oil droplets / micelles are spherical. The D10, D50 and D90 values can be determined directly using well-known test methods. According to the present invention, the particle size of the oil droplets / micelles are determined by Dynamic Light Scattering (DLS), which is considered the best way to measure particles in the nano- range. Measurement of droplet sizes is carried out in accordance with ISO 22412:2017. For testing, 1 mL of emulsion sample is used. It is diluted 60 times in water to maximize dispersion of particles thereby minimizing risk of overlapping particles that cannot be separated. The measurement is repeated three times, and the average taken to yield particle size results. DLS measurements can e.g. be performed on a Malvern Zetasizer Nano ZS. The measurements are made at 25°C and using an equilibration time of 1 minute. Prior to measurement, measures are taken to remove bubbles, e.g. inversion and redispersion of the sample. Such equipment and methods are suitable to measure particle sizes in the range 1 nm to 10 µm. Zeta Potential The zeta potential of the droplets in an emulsion is related to the overall stability because particles that are highly charged (either positively or negatively) will repel each other in solution and thus be less likely to aggregate and eventually separate out of dispersion. On the other hand, droplets that are relatively uncharged or neutral are more likely to interact with each other rather than the solvent leading to larger particles or complete separation back into two phases. Thus, ensuring optimal droplet zeta potential is helpful to ensure that the emulsion will be stable over time. According to the present invention, the emulsion preferably has a zeta potential which is close to zero but slightly negative, e.g. in the range -40 mV to -1 mV. Measurement of zeta potential is carried out in accordance with ISO 13099-1:2012. The zeta potential can be measured using e.g. the Malvern Zetasizer Nano ZS, i.e. the same device useable for measuring particle size. For testing, 1 mL of emulsion sample is used. It is diluted 60 times in water to maximize dispersion of particles Zeta potential measurements can e.g. be performed on a Malvern Zetasizer Nano ZS. The measurements are made at 25°C and using an equilibration time of 420 seconds. Prior to measurement, measures are taken to remove bubbles, e.g. inversion and redispersion of the sample. The Water Phase According to the present invention, the emulsion comprises a water phase. The water phase preferably contains no other solvent than water, although trace amounts of solvents may be present. The water phase may contain at least 90 wt.% water, preferably at least 95 wt.% water and more preferably at least 98 wt.% water. The addition of solvents which may help to dissolve cannabinoids in the water phase are preferably absent. In particular, ethanol is preferably absent (although trace amounts may be tolerated). This is because it is desired to keep the cannabinoid in the oil phase. It is a Pharmacopeial requirement (Ph. Eur. 5.4) to test for residual organic solvents used in the production of said emulsion. Ethanol is often used as extraction solvent for phytocannabinoids as well as cleaning detergent in the emulsion production process, which justifies the test for residual ethanol. The maximum amount of ethanol present in the emulsions of the present invention is preferably set to 0.5 wt.% based on the entire emulsion. In addition to water, the water phase may contain various additives. These are discussed below. It is also possible that surfactant is contained in the water phase. However, for the purposes of the present invention, the surfactant system is not included as part of either the water phase or the oil phase when calculating relative amounts. The Oil Phase According to the present invention, the emulsion comprises an oil phase. The cannabinoid is present in this phase. The oil phase acts as a carrier for the cannabinoid, and the cannabinoid is soluble in it. Any oil can in principle be used, as long as the cannabinoid can be dissolved in it. The oil phase preferably comprises one or more long chain triglycerides (LCT) or very long chain triglycerides (VLCT). Triglycerides are tri-esters derived from the condensation reaction of glycerol with three fatty acids. The three fatty acids may be the same or different, and they may each be saturated or unsaturated. An LCT is defined as a triglyceride in which the average carbon number of the fatty acid-derived groups is in the range C12-C15. For example, all three of the fatty acid- derived groups can have a carbon number in the range C12-C15. Analogously, a VLCT has an average carbon number of the fatty acid-derived groups in the range C16-C22. For example, all three of the fatty acid-derived groups can have a carbon number in the range C16-C22. LCT/VLCT oils have been found by the present inventors to result in greater absorption of the cannabinoid(s) in the emulsion, at least for any portion of the emulsion which is swallowed. The emulsions of the present invention may be designed for administration to the oral cavity, in which case it is inevitable that a portion will be swallowed and enter the GI tract, even if this is not the primary intention. Alternatively, the emulsion may be expressly designed for oral administration, i.e. to be deliberately swallowed. The inclusion of LCTs and/or VLCTs is believed to result in a higher fraction of the dose reaching the systemic circulation, when compared to a similar emulsion based on medium-chain triglycerides (MCTs). LCTs and VLCTs are significantly more viscous than MCTs. Perhaps for this reason, it is conventional to use MCTs for dissolving cannabinoids, especially when preparing a sprayable formulation. The present inventors have surprisingly found that the advantages of using LCTs and VLCTs outweigh any predicted disadvantages. Preferably, the total amount of LCT and VLCT in the oil phase is in the range 45-75 wt.%, more preferably 55-65 wt.% based on the total weight of the oil phase. Examples of triglycerides which can be comprised in the oil phase include: OLL: 1-oleyl-2-linoleyl-3-linolenoylglycerol; OOL: 1,2-dioleyl-3-linolenoylglycerol; PLL: 2,3-dilinoleyl-1palmitoylglycerol; POL: 1-palmitoyl-2-oleyl-3-linoleylglycerol; OOO: 1,2,3-trioleylglycerol; POO: 2,3-dioleyl-1-palmitoylglycerol; SOO: 2,3-dioleyl-1-stearoylglycerol. The use of MCTs is preferably substantially avoided, or at least minimized, in the present invention. Preferably, the oil phase comprises no more than 10 wt.% MCTs, preferably no more than 5 wt.% MCTs. In a particularly preferred embodiment, the oil phase comprises olive oil and/or rapeseed oil. Preferably, both of these are contained, preferably in a weight ratio of olive oil:rapeseed oil which is in the range 1:6 to 1:2, preferably around 1:4. The fatty acids present in the triglycerides in olive oil and rapeseed oil are set out in Table 1 below:

Table 1: Overview of fatty acid composition of olive oil and rapeseed oil listed in their respective monograph in Ph. Eur. (European Pharmacopoeia). S = Saturated, US = Unsaturated. Composition Fatty acid Type IUPAC name Formula O_{live oil} ^Rapeseed o_il Palmitic Hexadecanoic a_cid ^S a_cid ^{C16H32O2 7.5-20.0% 2.5-6.0%} Palmitoleic (Z)-hexadec-9- a_cid ^US e_{noic acid} ^{C H O ≤ 3.5% N/A} Stearic Octadecanoic a_cid ^S a_cid ^{C H O 0.5-5.0% ≤ 3.0% (Z)-octadec-} 56.0- 50.0- O_{leic acid US} ^9- C _{H O enoic acid} 85.0% 67.0% (9Z,12Z)- Linoleic 16.0- U^S octadeca-9,12- C H O 3.5-20.0% a_cid 30.0% dienoic acid (9Z,12Z,15Z)- Linolenic octadeca- 6.0- U^S C H O ≤ 1.2% a_cid 9,12,15- 14.0% trienoic acid Arachidic a_cid ^{S Icosanoic acid C H O ≤ 0.7% N/A} Eicosenic (Z)-icos-11- a_cid ^US e_{noic acid} ^{C H O ≤ 0.4% ≤ 5.0%} Behenic a_cid ^{S Docosanoic acid C H O ≤ 0.2% N/A} E_{rucic acid US} ^{(Z)-docos-13-} C _{H O N/A ≤ 2.0% enoic acid} Lignoceric Tetracosanoic a_cid ^S a_cid ^{C H O ≤ 0.2% N/A} In addition, the oil phase may contain various additives. These are discussed below. It is also possible that surfactant is contained in the oil phase. However, for the purposes of the present invention, the surfactant system is not included as part of either the water phase or the oil phase when calculating relative amounts. The Cannabinoid The presently claimed emulsion comprises one or more cannabinoid compounds, which is/are present in the oil phase. According to the present invention, the cannabinoid is dissolved in the oil. The cannabinoid compound(s) may be provided in the form of a cannabinoid concentrate, e.g. an extract from the Cannabis sativa L. plant (phytocannabinoids). The cannabinoid(s) may also be synthetically derived by means of organic chemistry for example to resemble the body’s own cannabinoids (endocannabinoids), phytocannabinoids or structural analogues thereof. Cannabinoids may also be produced by biosynthesis for example by fermentation technology. Cannabis sativa L. is an annual herbaceous flowering plant. Indigenous to Eastern Asia, the plant is now of cosmopolitan distribution due to widespread cultivation. It has been cultivated throughout recorded history and used as a source of industrial fiber, seed oil, food, and medicine. Although the main psychoactive constituent of Cannabis is tetrahydrocannabinol (THC), the plant is known to contain more than 500 compounds, among them at least 113 cannabinoids; however, most of these "minor" cannabinoids are only produced in trace amounts. Besides THC, another cannabinoid produced in high concentrations by some plants is cannabidiol (CBD). CBD is a phytocannabinoid discovered in 1940. It accounts for up to 40% of the plant's extract. CBD is a herbal dietary supplement. CBD does not have the same psychoactivity as THC, and can modulate the psychoactive effects of THC on the body if both are present. Unlike THC, which acts on the cannabinoid receptor type 1 (CB1) as a partial agonist, CBD instead is a negative allosteric modulator of CB1 receptors. Material types from the plant are distinguished from their degree of purification and thus concentration of the major cannabinoid: extracts (cannabinoid 45-75% w/w), distillates (60-95 % w/w) and isolates (≥95% w/w). The use of an plant-based extract as model compound for the invention does not restrict its applicability to synthetic and biosynthetic cannabinoids. It is also possible to use e.g. synthetic CBD and/or synthetic THC. As for their phytocannabinoids counterparts, these are poorly water- soluble and they can therefore be dissolved in a carrier oil in the same way as a cannabinoid plant-based extract. When a cannabinoid plant-based extract is used, this is preferably included in the oil phase in an amount of 15-55 wt.%, preferably 25-45 wt.% based on the weight of the oil phase. The amount of cannabinoid plant-based extract in the formulation is preferably 1-10 wt.%, preferably 2-8 wt.%. The same preferred amounts apply also to the use of synthetic cannabinoids such as CBD and THC; the total amounts of cannabinoids in the oil phase should in that case preferably be 15-55 wt.%, more preferably 25-45 wt.% based on the weight of the oil phase. Process for Extracting Cannabinoids from Cannabis Sativa L. Various solvents have been used for the extraction of cannabinoids as the main bioactive compounds of Cannabis sativa L. Common organic solvents can be used, although supercritical CO₂ is preferred for larger-scale preparative applications. Alcohols such as ethanol (EtOH) and methanol (MeOH) are widely employed, and their usage has been warranted by previous research findings. EtOH presents the additional advantage of low toxicity, besides the reported efficiency for cannabinoids. Production of a cannabis extraction involves the following steps: The cannabis plant material is harvested, i.e. some or all of the stalks, stems and leaves of the plant. The plant material is then dried and chopped. The chopped material is placed in a vessel together with the solvent. The solvent (contained dissolved cannabinoids taken up from the plant material) is then separated from the solid plant material, e.g. by filtration, and the filtrate collected. The solvent is then removed from the filtrate e.g. by evaporation. The resulting composition is the cannabinoid extract. Further steps may be included, such as removal of waxes, removal of terpenes and decarboxylation. Cannabinoid extraction processes are described further in US9987567B1. The composition of the cannabinoid extract, i.e. which cannabinoids are contained in which relative amounts, is dependent on various factors, including the choice of plant material and the solvent used. Conditions applied in the extraction process may also play a role, for example temperature and extraction time. According to the present invention, the cannabinoid(s) is/are dissolved in the oil phase. It is not contemplated that the cannabinoids in the claimed emulsions are present in the form of solid particles of cannabis plant material which contain the cannabinoid(s). In particular, the present inventors have realized that the use of dissolved cannabinoids in the oil phase leads to advantages relative to the inclusion of cannabis plant material. Firstly, when the cannabinoid(s) is/are dissolved in the oil phase, bioavailability is improved as compared to the use of solid cannabis material. Secondly, it is believed that safety is improved because it is uncertain how the body may react to the penetration of solid, insoluble material across the buccal membranes. The Surfactant System A surfactant system is needed to reduce the interfacial surface tension between oil and water and ultimately improve the physical stability of the emulsion during its shelf-life. Furthermore, the surfactant system is needed in order to render the outside surface of the oil droplets much more hydrophilic, in order to allow these droplets (or micelles, as they are in the presence of the surfactant system) to pass through the oral mucosa. As outlined above, the oral mucosa can be much more easily penetrated by hydrophilic entities than by hydrophobic ones. Thus, the surfactant system performs a dual role in the emulsions of the present invention, and the constituent components of the surfactant system are selected accordingly. The HLB number (hydrophilic-lipophilic-balance) of the selected surfactant system is preferably selected to approximately match that of the disperse phase, i.e. the oil phase in the present invention. Most preferably, the surfactant system comprises non-ionic surfactants. It is preferably to use a blend of hydrophilic and hydrophobic non-ionic surfactants. Preferred surfactants which can be included in the surfactant system in the emulsion of the present invention include non- ionic surfactants such as: Sorbitan esters, such as sorbitan monooleates, sorbitan monolaurates, sorbitan monopalmitates, sorbitan monostearates, sorbitan tristearates, sorbitan trioleates. Preferably, the sorbitan monooleate is used, most preferably the specific surfactant Span® 80. This is a lipophilic surfactant. Polysorbates including polyoxyethylene sorbitane monolaurates, monopalmitates, monostearates, monooleates, tristearates, trioleates and preferably mono-oleates. A particularly preferred surfactant is polyoxyethylene sorbitan mono-oleate (Tween® 80). This is a hydrophilic surfactant. Preferably, the surfactant system comprises both Span® 80 and Tween® 80. If both Span® 80 and Tween® 80 are contained, the weight ratio between them is preferably in the range 1:3 to 1:0,50. The selected ratio may depend on how much oil phase is contained in the emulsion. For large amounts of oil, say ca. 25 wt.%, then the ratio of Span® 80 and Tween® 80 may be in the range 1:0,50 to 1:1, preferably around 1:0.75. For more moderate amounts of oil, say ca. 10 wt.%, then the ratio of Span® 80 and Tween® 80 may be in the range 1:3 to 1:1,50, preferably around 1:2. If both Span® 80 and Tween® 80 are contained, they preferably together make up at least 50 wt.% of the surfactant system, preferably 60-100 wt.% of the surfactant system, preferably at least 80 wt.% or at least 90 wt.%. In a preferred embodiment, the surfactant system consists of Span® 80 and Tween® 80. The amount of the surfactant system relative to the weight of the emulsion as a whole used can be calculated in view of the desire to form micelles. The concentration therefore has to be selected to be above the critical micelle concentration (CMC). For example, when the oil phase constitutes 10 wt.% of the emulsion, then the total amount of surfactant would typically be 6 wt.%, with the remainder of the emulsion being the water phase. At least some surfactant will be contained at the interface between the oil phase and the water phase. Additionally, surfactant may also be contained in the water phase and/or the oil phase. For the purposes of the present invention, the surfactant system is considered to be a separate part of the composition to the oil phase and the water phase, so that it is not included for the purposes of calculating the total amount of either of these phases, which is needed e.g. for calculating relative amounts of the two phases in the emulsion, and for calculating relative amounts of ingredients in each of these two phases. Antioxidants It is preferred that the emulsions of the present invention comprise one or more antioxidants. The antioxidant primary serves to prevent oxidation of the cannabinoid. It can also help to prevent oxidation of triglycerides which are contained in the oil phase – in particular, unsaturated triglycerides are prone to oxidation. In a preferred embodiment of the present invention, the emulsion comprises antioxidant components in both the oil phase and in the water phase. In that case, the antioxidant in the oil phase has to be soluble therein, and the antioxidant in the water phase has to be soluble therein. The antioxidant in the oil phase may e.g. be Vitamin E. The antioxidant in the water phase may e.g. be Vitamin C. Suitable amounts of antioxidant are e.g. 0.05 to 2 wt.% based on the emulsion, preferably 0.50 to 1 wt.%. Further Ingredients Further ingredients may be contained in the emulsion. For example, a preservative may be included. If included, the preservative may be potassium sorbate and/or sodium benzoate. Alternatively, preservatives such as benzalkonium chloride, cetylpyridinium chloride, thiomersal, benzoic acid, and propylene glycol. Less preferred preservatives include biguanides (e.g. chlorhexidine), phenols, benzyl alcohols, methyl parabens, ethyl parabens, and propyl parabens. The emulsion may also contain a pH-modifier. The preferred pH of the emulsions of the present invention is in the range 4- 6, preferably around 5. Addition of a pH-modifier may be necessary if the pH of the emulsion at the end of the emulsification process is outside this preferred range. Such pH modifying agents may be e.g. known buffers, e.g. citric acid. In another preferred embodiment, the formulation may contain a pH-modifier such as lactic acid or a lactate, phosphoric acid or a phosphate, hydrochloric acid, or a hydroxide salt. The amount added is determined based on the pH of the emulsion at the end of the emulsification process and the target pH. If the emulsion resulting immediately from the emulsification process has a pH within the preferred range, then there may be no need to add a pH-modifier. A taste-masking agent may also be included. A preferred taste-masking agent is menthol. Menthol is poorly water- soluble, so that it will primarily be found in the oil phase of the emulsions of the present invention. The formulation of the present disclosure may further comprise an osmotic agent, preferably selected from the group consisting of glycerin, glucose, sucrose, sorbitol, sodium phosphate and any combination thereof. Processing Technology Formation of nano-sized droplets requires introduction of energy and/or application of mechanical force to break apart the oil phase into droplets. Ultrasonication and high- pressure homogenization are preferred ways to achieve this in present invention. Ultrasonication, as the name implies, applies ultrasonic waves (frequency greater than 20 kHz). The mechanical vibration creates acoustic cavitation resulting in the formation of droplets. High-pressure homogenization uses a high-pressure piston pump to mechanically reduce the size of dispersed oil droplets. Several cycles of homogenization may be introduced to further reduce the size of the droplets so as to arrive at an emulsion which has the recited D50 value of 100nm to 300nm. Medical Uses The emulsions of the present invention have many potential uses. The emulsions can be used in palliative care, and/or in the treatment or alleviation of a disease, preferably wherein the disease is selected from the group consisting of pain, in particular acute or chronic pain, somatic pain, visceral pain, neuropathic pain, cancer pain, chronic back pain, chronic central nervous pain; neurological disorders, neurodegenerative diseases, insomnia, psychiatric disorders, nausea, anorexia, vomiting and nausea caused by chemotherapy, diabetic polyneuropathy, fibromyalgia, Tourette-Syndrome, multiple sclerosis, spasm at multiple sclerosis, anxiety disorders, schizophrenia, social phobia, sleep disorder, skin related diseases like psoriasis and neurodermatitis, glaucoma, restless leg syndrome, epilepsy, Alzheimer disease, movement disease like Dystonias, Huntington‘s disease, Parkinson’s disease, bipolar diseases, as well as other medical indications which are affected by the endocannabinoid system and which are affected by any other receptors affected by cannabinoids (e.g. GPR18, GPR119, GPR55). Uses for further processing The oil-in-water emulsion may be diluted with pure water or aqueous mixtures containing, for example, antioxidants, preservatives, pH modifiers and taste masking agents solvents. There is full miscibility between the emulsion and an aqueous solvent due to the emulsion being water-based. Administration The emulsions of the present invention are designed to be optimized for buccal and/or sublingual application, preferably by spraying the composition into the mouth of a user. As discussed above, the emulsions may also be formulated to optimize cannabinoid absorption when the emulsion is swallowed. For example, when preferred oils are used, the uptake is improved and elimination of the cannabinoid in the liver is reduced. When intended for the gastro-intestinal tract, the emulsions may also be administered to the user as an oral solution. In such a case, they can be swallowed directly, or swallowed following a dilution with either pure water or water mixed with, for example, antioxidants, preservatives, pH modifiers and taste masking agents solvents. Containers and Dispensers The emulsions of the present invention are preferably provided in a dispenser with a spray nozzle it is intended to be use these for buccal and/or sublingual administration. Examples of such dispensers include a standard buccal pump comprising a container (e.g. a glass or plastic bottle) and a spray pump which connects to the container. Such containers are available from e.g. Aero Pump GmbH and Frapak Packaging. In one embodiment of the invention, the dispenser is substantially air-tight both prior to use. It is preferred that the dispenser include means for protection against microbiological contamination. For example, the 3K and COMFORT systems produced by UrsaTec can be used. These dispensers are designed to prevent microbiological contamination. When such containers are used, it is not necessary to include a preservative in the emulsion formulation. Examples Example 1 – Preparation of the Formulation Oil-in-water emulsions were prepared. The oil phase consists of a cannabis extract dissolved in carrier oils (olive and rapeseed oil), lipophilic antioxidant, and taste masking agent (menthol). A lipophilic surfactant (low-HLB) is mixed with the ingredients of the oil phase. The water phase consists of preservative and hydrophilic antioxidant dissolved in WFI water. A hydrophilic surfactant (high-HLB) is mixed with the ingredients of the oil phase. A blank emulsion was also created for the lowest oil phase concentration (10.0 wt%). The blank emulsion is prepared in the same manner as the cannabinoid emulsions (see step-wise description below), except that no cannabinoid extract was added. The following manufacturing steps were used: A. Water phase premix • The water phase ingredients are mixed. B. Oil phase premix • The cannabinoid extract is weighed in a container of suitable size and liquified by the application of heat (up to 60°C). using either a water bath or heat plate. • All other ingredients are weighed separately and added to the extract-container under mixing. C. Addition of oil phase to the water phase under ultrasonication (1. emulsification) • Water phase is weighed in a container of suitable size. • The tip of an ultrasonic probe is installed into the water phase and turned on. • The oil phase is added in a controlled manner using a pump while mixing and under ultrasonication. • Mixing and ultrasonication continues after addition of oil phase. The total ultrasonication time is 60 min at a batch size of 25 L (25 kg). (approx. 700 W). D. High-pressure homogenization (2. emulsification) • The ultrasonicated emulsion is subjected to final emulsification using a high-pressure homogenizer. • The batch is processed through the homogenizer at a pressure of at least 100 MPa. • Two cycles of homogenization are performed. E. pH adjustment • The pH is measured, and if necessary, the emulsion is pH adjusted to 4.9 - 5.1 using HCl/NaOH. F. Filtration to single use storage bags • The finished emulsion is filtered through a 0.45 mm filter to a single use bag (closed system). • The pressure under filtration is continuously monitored and the filter is replaced if the pressure exceeds 1 bar. Using the above preparation method, the emulsions of Table 2 below were produced. Refined olive oil was chosen as LCT and refined rapeseed oil as VLCT and mixed together in a 1:4 ratio in the oil phase. A lipophilic antioxidant, all-rac- ^- Tocopheryl acetate (Vitamin E acetate), was added in 1.2 wt.% concentration relative to the oil phase. A hydrophilic antioxidant, ascorbic acid (Vitamin C), and a preservative, potassium sorbate, were added to the water phase in 1.2 wt.% and 0.1 wt.% concentration relative to the water phase, respectively. All emulsifications, except D1, were successfully completed. D1 did not result in homogenous emulsion and remained phase separated, confirming the edge of failure of the formulation design. Pictures of emulsions B1, C2 and the failed D1 are provided in Fig. 2A, 2B, and 2C, respectively: Table 2: Formulation design for emulsions. Lab scale is less than 5 L emulsion; pilot scale is between 5 and 30 L emulsion * = Extract from CBD chemovar; ** = Extract from THC chemovar ; *** = 1:1 mix of CBD and THC emulsions O^il Water Surfactant system (relative to E^xtract p_hase phase emulsion) (wt.% (wt.% of ID Scale (wt.% of Tween of oil water ^Spa HLB e_mulsion) ^{n 80} 8₀ ^Total phase) phase) surfactant (_{wt.%) (wt.%) (wt.%)} ^blend A1 Pilot 31* 10.0 85.4 2.1 4.0 6.1 11.3 A2 Lab 31* 10.0 85.4 2.1 4.0 6.1 11.3 A3 Pilot 31** 10.0 85.4 2.1 4.0 6.1 11.3 A4 Lab 31** 10.0 85.4 2.1 4.0 6.1 11.3 A5 Pilot 31*** 10.0 85.4 2.1 4.0 6.1 11.3 A6 Pilot 0 10.0 85.4 2.1 4.0 6.1 11.3 A7 Lab 0 10.0 85.4 2.1 4.0 6.1 11.3 B1 Lab 42* 20.7 76.4 4.2 3.6 7.8 9.2 C1 Lab 40* 27.8 71.0 5.3 4.0 9.3 8.9 C2 Lab 42* 27.6 71.7 5.3 3.4 8.6 8.5 D1 Lab 40* 35.4 66.9 6.3 3.1 9.4 7.9 Example 2 – Particle size measurements and zeta potential DLS measurements were performed on the emulsions A1-A4, A6-A7 and C1 which were produced in Example 1. D10, D50 and D90 values were measured, as well as the zeta potential. The size measurements were performed using a Malvern Zetasizer Nano ZS (Table 3) in accordance with ISO 22412:2017. Measurements of zeta potential were carried out on the same equipment in accordance with ISO 13099-1:2012. The test parameters were as shown in Tables 3 and 4: Table 3: DLS size method details Item Description Cell Disposable cuvette Refractive index of dispersant 1.34 Mixture of water & oil emulsion Material (RI/absorbance) 1.450/0.001 Viscosity (cP) 1.0366 Display range 0.6 – 6000 nm Temperature (°C) 25 Equilibration time (min) 1 Number of measurements 3 Number of sample preparations per 2 sample Table 4: DLS zeta potential method details Item Description Cell Disposable folded capillary cells Refractive index 1.450/0.001 (RI/absorbance) Viscosity 1.0366 Dielectric constant 78.5 Model value 1.50 Temperature (°C) 25.0 Equilibration time (s) 420 Number of measurements 1 Number of sample 2 preparations The results are shown in Table 5 below: Table 5: Measured particle size distributions and zeta potential of the emulsions Surfactant system (relative to emulsion) ID D10 D50 D90 ) _(n ^{ZP (mV)} (nm) (nm _m) A1 67.1 117 214 -10.80 A2 68.4 128 240 -6.60 A3 82.8 160 360 -15.20 A4 68.3 146 389 -5.10 A6 58.1 107 194 -9.44 A7 63.1 119 219 -7.30 C1 113 175 279 -35.00 The droplet size distribution is robust towards changes in the extract load of the oil phase (A1-A4 compared to the blank emulsions, A6-A7). The droplet size increases slightly for the highest loaded formulation, C1, but it is still well within the target ranges for droplet size. The droplet size distribution of sample A1 is shown in Figure 3. Example 3 – Stability evaluation 12 month stability testing was performed on emulsion A5 which was produced in Example 1. Stability was tested under ICH Q1 conditions: Refrigerated, intermediate (30°C / 65 %RH) and accelerated (40°C / 75 %RH). The samples were packed in 30 mL HDPE bottles with screw caps and stored in qualified climate chambers. Analyses are summarized in below table 6:

Table 6: Analyses conducted per condition and time point. Condition C^ompendial 5 ^C Analysis 30 ^C/65 %RH 40 ^C/75 %RH r_eference (refriger (intermediate) (accelerated) ated) Cannabinoid T0,T3,T6, a_ssay ^{DAB All time points} (T9,T12) P^{h. Eur.} T0,T3,T6, Microbiology T0,T6,T12 T0,T6,T12 5_.1.4-1 (T9) Residual Ph. Eur. 2_.4.24 ^{T0 T0 T0} solvents p_H ^{Ph. Eur.} A_{ll time points 2.2.3} Peroxide Ph. Eur. A^{ll time} ue ₂ ^points val _.5.5 Visual All time points a_ppearance ^N/A The results were: Visual appearance All samples - at any time point and condition - appeared yellow and milky and showed no sign of phase separation. Cannabinoids Cannabinoid stability data are illustrated in Figure 4A-C for the three conditions. The following is observed: ^ Under refrigerated conditions, the emulsion remains fully stable throughout the 12 months test period. No incipient signs of cannabinoid degradation are observed (Figure 4A). ^ Degradation of THC is observed at T12 under intermediate conditions (Figure 4B) and at T9 under accelerated conditions (Figure 4C), i.e. degradation starts somewhere between T9 and T12 at 30 ^C and somewhere between T6 and T9 at 40 ^C. ^ No critical degradation products were detected, as CBN remains below its specification limit (<0.1 %; not shown) and with no quantifiable ^8-THC formation assessed for 40 ^C at T9 (not shown). Microbiology & pH Microbiological purity is within specification at any condition and at all time points. This is backed-up by the pH which remains in the range 4.8-5.8, considering that bacterial growth (or chemical degradation) can impact the pH of the product. Table 7: pH measurements from QC. The emulsion is adjusted to a pH of approx. 5 during production. 5 ^{^C} 30 ^C/65 %RH 40 ^C/75 %RH C^onditions (_{refrigerated)} (intermediate) (accelerated) T0 5.2 5.2 5.2 T3 5.1 5.6 5.8 T6 5.3 5.7 5.5 T9 5.3 5.5 5.0 T12 5.4 5.1 4.8 Peroxide value (“oil rancidity”) Peroxide values are provided in Table 8. The peroxide value I_P is the number that expresses in milliequivalents of active oxygen the quantity of peroxide contained in 1 kg of the substance. It is used to evaluate the rancidity of the oil phase in the emulsion. Rancidity can impact the taste and smell of the product and thus patient compliance. The peroxide value at which oxidation of oils can be detected as an off-flavour varies widely depending on the nature of the oil. Samples of olive oil may not be perceived as rancid till the peroxide value reaches 20 meq/kg according to literature. It is not surprising to see an increase in the peroxide value at the harshest condition after nine months, 2.0 meq/kg, increasing very slightly to 2.1 after 12 months. Peroxide levels are slightly elevated to 1.3 after 12 months for the 30/65 condition. However, peroxide values in this range (< 2 meq/kg) are well below the 20 meq/kg threshold of rancidity for pure oils and is therefore not expected to impact taste and smell of the product. Table 8: Peroxide values. 5 ^{^C} 30 ^C/65 %RH 40 ^C/75 %RH C^onditions (_{refrigerated)} (intermediate) (accelerated) T0 < 0.04 < 0.04 < 0.04 T3 < 0.04 < 0.04 < 0.04 T6 < 0.04 < 0.04 0.8 T9 < 0.08 < 0.08 2 T12 < 0.08 1.3 2.1 Example 4 – In vivo and in vitro experiments on bioavailability Drug delivery performance of the emulsion technology was tested at a 10% oil phase relative to water phase, loaded with THC and CBD extracts. The tests were performed using the emulsion A5 of example 1 designated Formulation #1 in the experiment (see tables 9 and 10). Formulation #2 of tables 9 and 10) was a simple mixture of extract and MCT oil. Formulation #3 of table 9 was only included in the in vitro experiment. It was prepared from a commercially available EtOH-PG (ethanol and propylene glycol) based formulation with CBD and THC, which was diluted with EtOH-PG to provide a final cannabinoid concentration similar to Formulation #1 and #2 (0.85 wt.% of each cannabinoid). A comparison was made relative to other types of cannabinoid formulation technologies. This is covered by an in vitro and in vivo part with the following aims: • In vitro: To evaluate the transport of CBD and THC from three different formulations over an artificial membrane, designed to mimic the oral mucosa, using the µFLUX setup. • In vivo: To elucidate the pharmacokinetic (PK) profile of CBD and THC after oral administration in rats. Table 9: Summary of experimental setup of the in vitro flux experiments.

(1) o/w emulsion (A5) Formulations (2) MCT-based (3) EtOH-PG based (1): THC and CBD extracts standardized in the final delivery system to approx. 0.85 wt.% of each cannabinoid in the emulsion. (2) THC and CBD extracts standardized in the final delivery system to approx. 0.85 wt.% of each cannabinoid using MCT as Active diluent compounds (3): THC and CBD of an intermediate product diluted in the final delivery system to approx. 0.85 wt.% of each cannabinoid using a 0.43:0.57 mixture of ethanol (EtOH) and propylene glycol (PG) by weight as diluent - Concentrations of all three equalized (see above), to eliminate any bias from difference in concentration gradients. - Flux through an artificial membrane and Parameters artificial saliva (PAMPA = Parallel Artificial Membrane Permeability assay) - Receptor media assayed at specific timepoint using UV-HPLC In vitro: µFLUX experiments The in vitro permeation experiments were performed using a µFLUX™ setup from Pion Inc. (Billerica, MA, USA) consisting of two 20 mL compartments separated by a horizontal barrier with an absorptive area of 1.0 cm². A schematic overview of the µFLUX setup is shown in Figure 5. The barrier of the setup comprised a PAMPA-membrane prepared by the addition of 25 µL phosphatidylcholine solution (20% w/v in dodecane) to a hydrophobic 0.45 µm PVDF filter. An additional hydrophilic 0.45 µm PVDF filter was introduced to act as a physical boundary between the PAMPA-membrane and the donor compartment. To mimic the in vivo situation, artificial saliva (Orthana Saliva) 12 mL was added to the donor side along with 3 mL formulation and 20 mL acceptor sink buffer (ASB); 20 mM HEPES, 1% w/v SLS adjusted to pH 7.4 was added to the receiver side. Both donor and receptor compartments were maintained at 37 °C and stirred using 20 mm cross-shaped magnetic stirrers operating at 250 rpm. Samples of 100 μL were taken from the receptor compartments at t = 15, 30, 60, 90, and 120 min and replaced with 100 μL ASB. The samples were diluted with 100 μL methanol and centrifuged before the resulting supernatant was assayed for CBD and THC using HPLC- UV. All treatments were done in triplicate. In vivo: GI tract absorption in rodents Table 10: Summary of experimental setup of the in vivo PK experiments. F_ormulations ^{(1) o/w emulsion} (_{2) MCT-based} (1): THC and CBD extracts standardized in the final delivery system to approx. 0.85 wt% of each cannabinoid in the emulsion. Active (2) THC and CBD extracts standardized in compounds the final delivery system to approx. 0.85 wt% of each cannabinoid using MCT as diluent - Fasted rats (n=4 per cohort) dosed by oral cavage - Blood samples taken from the tail vein Parameters at timepoints: 0.5, 1, 1.5, 2, 3, 4, 6 and 24 h - THC/CBD in supernatant of blood samples assayed by LC-MS Detailed description of rat studies Male Sprague-Dawley rats from Janvier Labs (Le Genest-Saint- Isle, France) were used for the study which was carried out in agreement with the Danish law on animal experiments as approved by the Danish Animal Experiments Inspectorate in accordance with the EU directive 2010/63/EU under license number 2019-15- 0201-00262. THC and CBD strengths were determined for both formulations prior to the in vivo study. The formulations were diluted 1000x in MeOH and analyzed by HPLC-UV relative to standards diluted in MeOH. The rats were fasted overnight with ad libitum access to water during their inactive period, by adjusting the day/night rhythm, so that the rats had their inactive period during the night. Four rats were dosed with the emulsion and four rats were dosed with the MCT-based formulation. All rats were dosed by oral gavage with a volume of 3 mL/kg (51 mg/kg, CBD+THC). Blood samples (100-150 µL) were taken from the tail vein at the timepoints: 0.5, 1, 1.5, 2, 3, 4, 6 and 24 h (8 h sample was excluded due to insufficient blood flow from the tail vein) and the plasma collected by centrifugation at 10,000 × g at 4 °C for 10 min. The rats were killed prior to the 24 h sample which was taken from the heart prior to harvesting of the brains. Both plasma samples and brains were stored at -20 °C until the day of LC-MS analysis. Precipitation of plasma proteins was done prior to analysis by mixing 50 µL plasma with 200 µL methanol followed by vortex, 10 min storage at -20 °C and centrifugation. The supernatant was analyzed by LC-MS and quantified relative to standards prepared in the same way, yet by mixing 50 µL of blank plasma with 200 µL of CBD/THC solutions in methanol. The difference in dilution factor between samples and standards was accounted for during the data treatment. The brains were each homogenized in 10 mL methanol and the homogenate centrifuged at 5500 × g for 5 min. The supernatant was evaporated to dryness and resuspended in 500 µL MeOH followed by centrifugation and LC-MS analysis. Results Results from the µFLUX studies are shown in Figure 6A and 6B, for CBD and THC flux, respectively. Results from the in vivo study in rats are shown in Figure 7A and 7B, for CBD and THC, respectively. • Permeation through the PAMPA membrane is superior for the emulsion and almost non-existent for other formulations --> high likelihood of sublingual absorption. The poor flux from the EtOH-PG based formulation is believed to be explained partially by precipitation of the poorly soluble cannabinoids from the EtOH-PG solution upon contact with artificial saliva, the latter acting as an anti-solvent to the cannabinoids. Precipitation was visually observed when dispersing the EtOH-PG formulation with artificial saliva. • Significantly faster onset of absorption and high exposure of emulsion following administration to the GI tract compared to the MCT-based formulation. • The combined pre-clinical package establishes the effectiveness of the emulsion as an enabling drug delivery technology. In addition to sublingual permeation, the fraction of dose, which is inevitably lost to the GI tract, is still absorbed to a high extent.

Claims

CLAIMS 1. A formulation which is an oil-in-water emulsion comprising; a water phase; and an oil phase which comprises one or more dissolved cannabinoid compounds; and a surfactant system; wherein the D50 particle size of the oil phase droplets as measured by dynamic light scattering is in the range 100nm to 300nm. 2. A formulation according to claim 1, wherein (i) the D10 value is in the range 50-100nm and/or (ii) the D90 value is smaller than 600nm, wherein the D10 and D90 values are measured by dynamic light scattering. 3. A formulation according to any preceding claim, wherein the oil phase makes up 2-20 wt.% of the formulation, preferably 5-15 wt.% of the formulation. 4. A formulation according to any preceding claim, wherein the oil phase comprises long-chain triglycerides (LCT) and/or very long chain triglycerides (VLCT). 5. A formulation according to claim 4, wherein the total amount of LCT and VLCT in the oil phase is in the range 45 to 75 wt.%, preferably 55 to 65 wt.%, based on the total weight of the oil phase., 6. A formulation according to any preceding claim, wherein the surfactant system comprises a lipophilic surfactant and a hydrophilic surfactant, which are preferably a sorbitan ester and a polysorbate, more preferably Span 80 and Tween 80. 7. A formulation according to any preceding claim, wherein the formulation comprises antioxidants in both the water phase and in the oil phase, which are preferably Vitamin C and Vitamin E. 8. A formulation according to any preceding claim, wherein the total amounts of cannabinoids in the oil phase is 15-55 wt.%, preferably 25-45 wt.%, based on the weight of the oil phase. 9. A formulation according to any preceding claim, wherein the emulsion is substantially free of solid particles, preferably free of solid particles. 10. A formulation according to any preceding claim, wherein the formulation is substantially free of ethanol, preferably free of ethanol. 11. A formulation according to any preceding claim, wherein the formulation is contained in a spray dispenser. 12. A formulation according to any preceding claim, for use in treating one or more of: ^ pain (preferably acute or chronic pain, somatic pain, visceral pain, neuropathic pain, cancer pain, chronic back pain, chronic central nervous pain); ^ neurological disorders; ^ neurodegenerative diseases; ^ insomnia; ^ psychiatric disorders; ^ nausea; ^ anorexia; ^ obesity; ^ vomiting and/or nausea caused by chemotherapy; ^ diabetic polyneuropathy; ^ fibromyalgia; ^ Tourette-Syndrome; ^ multiple sclerosis; ^ spasms related to multiple sclerosis; ^ anxiety disorders; ^ schizophrenia; ^ social phobias; ^ sleep disorders; ^ skin related diseases (preferably psoriasis and/or neurodermatitis); ^ glaucoma; ^ restless leg syndrome; ^ epilepsy; ^ Alzheimer disease; and ^ movement diseases (preferably Dystonias or Huntington‘s disease). 13. An emulsion for use according to claim 12, wherein pain is treated. 14. An emulsion for use according to claim 12 or claim 13, wherein the treatment comprises oral administration, sublingual administration and/or buccal administration of the emulsion. 15. A method of producing a formulation according to any preceding claim, comprising the steps of: (i) Preparing an oil phase by dissolving one or more cannabinoid compounds in a carrier oil; (ii) Preparing a water phase; and (iii) Mixing the oil phase and the water phase by ultrasonication and/or high-pressure homogenization so that the D50 particle size of the oil phase droplets as measured by dynamic light scattering is in the range 100nm to 300nm, wherein the formulation also comprises a surfactant system, whose components are present in (a) the oil phase and/or (b) the water phase and/or (c) the mixture of the oil phase and the water phase.