Cannabis-Based Oral Formulations for Medical Purposes: Preparation, Quality and Stability

Current legislation in Italy provides that medical Cannabis may be administered orally or by inhalation. One of the fundamental criteria for the administration of oral formulations is that they deliver a known consistent quantity of the active ingredients to ensure uniform therapies leading to the optimisation of the risks/benefits. In 2018, our group developed an improved Cannabis oil extraction technique. The objective of the present work was to carry out a stability study for the oil extracts obtained by this method. Furthermore, in order to facilitate the consumption of the prescribed medical Cannabis therapy by patients, a standard procedure was defined for the preparation of a single-dose preparation for oral use (hard capsules) containing the oil extract; thereafter, the quality and stability were evaluated. The hard capsules loaded with the oil extract were analysed and found to be uniform in content. The encapsulation process did not alter the quantity of the active molecule present in the oil. The stability tests yielded excellent results. Since the capsule dosage form is easily transported and administered, has pleasant organoleptic properties and is stable at room temperature for extended periods of time, this would facilitate the adherence to therapy by patients in treatment.


Introduction
Before the last century, when it became illegal in most Countries, Cannabis was widely used in medicine. The reason for the restriction in its availability as a therapeutic agent was its growing notoriety as a psychotropic agent and its consequent abuse [1,2]. In recent years, however, there has been a resurgence in support for the legalisation of cannabinoids for medical use as a result of media attention as well as expectations of their efficacy, albeit, this is not always supported by scientific evidence [3][4][5].
The phytocomplex of the Cannabis plant contains over 500 different molecules, of which approximately a hundred belong to the cannabinoid chemical class; among these, small differences in molecular structure may induce widely different effects [6]. The molecules of greatest pharmacological interest from the point of view of their effects are the decarboxylated forms of delta-9-tetrahydrocannabinol (THC) and cannabidiol (CBD) as these are easily absorbed in the intestine [7]. Hence, the determination of the quantities of these compounds present in medications to be administered to patients is a fundamental prerequisite.
In the last few years, reforms in Italy have opened the door to the use of medical Cannabis in tightly regulated cases. Consequently, Cannabis is now available for this purpose. In Italy medical Cannabis is produced by the Stabilimento Chimico Farmaceutico Militare (Pharmaceutical Chemical Military Facility) in Florence.
Since 2016, a variety of Cannabis, FM2, has been available. This is supplied as dried, ground Cannabis inflorescences containing delta-9-tetrahydrocannabinol in quantities ranging from 5% to 8% and cannabidiol in percentages from 7.5% to 12%. Since 2018, another variety of Cannabis, FM1, has also been made available; this contains delta-9tetrahydrocannabinol in quantities from 13% to 20% while the cannabidiol content is lower than 1%. Note that the percentages reported refer to the "total" content: that is, the sum of the molecule in both acid form (delta-9-tetrahydrocannabinolic acid-THCA-and cannabidiolic acid-CBDA) and decarboxylated form (delta-9-tetrahydrocannabinol-THC and cannabidiol-CBD) [8][9][10].
Currently in Italy, the law states that medical Cannabis may be administered orally or by inhalation. The administration by inhaler is to be considered the second-choice option and must only be selected when oral administration does not produce the desired pharmacological effects or when the physician considers it opportune [9,11].
Concerning oral administration, in accordance with Minister of Health directives, decoctions represent the first-choice pharmaceutical form. The decoction must be prepared in compliance with the official procedure reported in "Recommendations for doctors prescribing FM2 Cannabis inflorescence derivatives" [11]. In a previous study, our research group demonstrated that the prescription of decoction-based formulations, considering the low yields of THC and CBD and, consequently, the high volume that a patient would have to consume as well as the high costs of processing the raw material to obtain the required quantity of active molecule, should not be recommended [12].
As well as the decoction for oral administration, the legislation in effect in Italy specifies that medical Cannabis may also be administered as an oil extract (hereafter, oil) on condition that this has been previously titrated for the active molecule using the proper instrumentation as set out by the current regulations (gas or liquid chromatography coupled with mass spectroscopy) [9,11]. The administration of formulations containing known quantities of active molecule is essential to ensure the uniformity of therapies leading to the consequent optimisation of the risks/benefits. On this point, it is important to note that although a number of preparation methods have been reported in scientific literature [13][14][15][16], the situation for oils, in particular, was that an exhaustive comparative study was lacking which investigated the technical aspects of preparation procedures for Cannabis-based formulations for medical purposes.
In light of this, in 2018, our research group developed a novel preparation method (denominated β-4) that allowed us to obtain a significantly higher amount of THC and CBD than those for water extraction (decoction) or an oil extraction using the previously known methods most widely used in Italy [12].
Having optimised the extraction procedure, the objective of the present study was to conduct stability studies on the oils obtained through the β-4 method. In addition, considering that oils have received considerable attention due to their easier dose management during the treatment period, but their organoleptic characteristics are particularly unpleasant, in order to facilitate the consumption of the prescribed medical Cannabis therapy by a patient in treatment, a standard procedure was defined for the preparation of a single-dose preparation for oral use (hard capsules) using the oil-based formulation. Both the oil and the capsules were then evaluated for quality and stability. Furthermore, the preparation of a pharmaceutical form that masks the organoleptic characteristics of the oil has the advantage of allowing the establishment of a control/placebo group in a clinical trial.

Capsules Preparation
Using the oils obtained with the β-4 procedure, rigid capsules were prepared.
The optimal order in which the components should be added was experimentally evaluated. The most efficacious method was adding the olive oil first and then, the silica. Indeed, when we attempted to dispense oil into a capsule already containing a layer of silica, the presence of the excipient made the operation extremely difficult in that each drop of oil disturbed the powder and caused a small puff of the powder to drift out of the capsule given its fine composition. It was more practical to add the oil first; cover it with a layer of silica and then, mix the two by inverting 180 • the capsule filling machine to accelerate the mixing process.

Capsules' Quality
The capsules, prepared with oils obtained with the β-4 method employing the technique described in Section 4.3, analysed employing the method described in Section 4.4, were uniform in mass and content and complied with the directives set out in the European Pharmacopeia [17] In particular, the weight of the individual dosage units and the amount of the lipophilic liquid phase present in the single capsules varied by less than 10% from the average value.
The titration of the active molecules for the oil extracted from the capsules varied by less than 10% from the expected result. Table 1 reports the titrated quantity of active molecules in the oil extracted from capsules prepared using two different volumes of oil (156 µL/cps and 312 µL/cps) deriving from the β-4 extraction procedure. As the results show, the quantity of oil extracted from the capsule is consistent with the expected result and independent of the quantity of oil dispensed into the capsule.
The encapsulation process, and, hence, the contact with the silica and the capsule gelatin, did not alter the quantity of the tested active molecules present in the oil independently of the quantity used to fill the capsule. Furthermore, no alteration of the envelope occurred: the use of different volumes of oil had also the purpose of evaluating whether alteration of the gelatin envelope occurred as the quantity of oil increased.

Stability Tests
The stability tests conducted employing the method described in Section 4.5 both on the oils obtained from the β-4 procedure and the capsules containing the same oil, yielded excellent results: the variation in the content of active molecules was less than 10% both after refrigerated storage for 180 days and after storage at room temperature for 180 days. This result was valid for both the oils and the oil in capsules. The maximum variation for oils was 9.29%, for capsules was 9.10%. Tables 2 and 3 respectively report the results of the analysis of three 100 mL batches of oil prepared using the β-4 procedure and the capsules filled with the same oils. Figures 1  and 2 show the active molecules concentrations in oils and capsules during 180 days: the variations were always less than 10.        Table 2.   Table 3.

Discussion
In 2018 our research group developed a novel preparation method (denominated β-4) and the results obtained using this procedure were compared with the three established techniques most widely used in Italy [12]. In detail, two of these techniques instruct that the Cannabis be, first, ground and then, mixed with olive oil. The resulting mixture is heated (for two hours in a water bath at boiling point for one method, and for two hours at 110 • C for the other) and then, filtered to obtain the oil extract. The third method, instead, directs that the Cannabis is chopped and then, pre-heated at 115 • C for 40 min. Subsequently, the Cannabis is mixed with olive oil and further ground with a turbo-emulsifier for three minutes. The mix of Cannabis and olive oil is then heated in a water bath at boiling point for 40 min, before being filtered and added with butylhydroxytoluene (BHT) 0.02%. The three examined methods specify that the weight (mg) to volume (mL) ratio between plant material and solvent is 100:1. Using type FM2 Cannabis, the most effective of the three established methods described above yields oil with a maximum concentration of THC equal to 0.37% ± 0.08% (3.38 mg/mL) and that of CBD is equal to 0.70% ± 0.19% (6.40 mg/mL) [12][13][14][15][16]. The β-4 method revealed itself to be the most effective: using a quantity of vegetable material twice that of the other methods (weight/volume ratio of plant material to solvent of 200:1), led to an average concentration of active molecules in the decarboxylated form more than double of the other methods: 8.04 mg/mL for THC and 13.05 mg/mL of CBD. These values represent significantly higher returns than those for water extraction (decoction) or an oil extraction using the previously known methods [12].
Considering the good results, we performed stability studies on the oils obtained through the β-4 method. In contrast to previous studies on oil preparation [14,[18][19][20][21], the β-4 oils were stable up to 180 days not only if stored in the refrigerator, but also if stored at room temperature, regarding both the two major components -THC and CBD-and the others have been tested.
The stability has been the same for the capsules fitted with different amounts of the β-4 oils: the results revealed that the encapsulation process has no effect on the oils.
We did not performed tests and there is no evidence in the literature describing the losses of the phytocomplex components during the various stages of preparation [18]. The deepening of the losses occurred during the different preparation phases, which until now has been presented only in terms of concentrations of active compounds obtained in final oils, need to be investigated, in particular focused on decarboxylation.
The choice of the hard capsule as the dosage form for the oral route was tied to the fact that this is a relatively easy form to prepare and, in addition, the equipment required for this process is readily available in the majority of Italian hospital and community pharmacies.
The optimal encapsulation procedure was selected after a variety of tests. In the initial tests, the oil in its pure form without excipient was used to fill the capsule as it was presumed that this would not affect the capsule casing. However, while the casings appeared to be intact, on closer examination, the lack of an excipient to provide internal support caused it to be particularly fragile. Furthermore, the oil could leak from around the joint between the two parts of the capsule. Hence, it was decided to use an excipient such as silica able to gel the oil thus preventing the oil from leaking and solidifying the contents of the capsule.
Successively, the optimal order in which the components should be added was evaluated and a variety of preparation tests were conducted. These allowed us to understand that the most efficacious method was adding the olive oil first and then, the silica.
The development of dosage forms for oral use based on oil formulations aims to facilitate the consumption of the prescribed therapy by the patient in treatment with medical Cannabis. The encapsulation process offers numerous advantages: firstly, it masks the organoleptic properties of the formulation and this favours adherence to therapy by the patient. The oils in particular have particularly unpleasant organoleptic properties, which may have negative consequences for proper adherence. In addition, encapsulation makes handling and transportation easier; furthermore, considering the fact that storage at low temperature is not necessary for at least six months, the management of the prescribed therapy is certainly simplified for the patient even in terms of domestic storage of the dosage units.
The preparation of a single-dose unit to be administered orally, such as that developed in the course of the present study, allows the administration of a placebo in a clinical trial; this is an essential step considering the fact that, to date, the efficacy of medical Cannabis has not been demonstrated definitively by scientific literature and requires further study. One of the particular organoleptic properties of Cannabis-based formulations is the odour, which is intense and unmistakable; therefore, a subject or the experimenter would easily realize in which group the subject has been included. This would nullify the randomness of subjects thus rendering the entire experiment void.
The method β-4 allows to obtain oils whose average concentration of active molecules in the decarboxylated form is 8.04 mg/mL for THC and 13.05 mg/mL of CBD [12].

Materials for the Galenic Preparation and Reagents for Quantitative Analysis
The other materials used for producing the galenic preparation described below (olive oil, distilled water, micronized silica anhydride, type 0 rigid capsules), were purchased from a pharmaceutical supplies company (Farmalabor s.r.l, Canosa di Puglia, Bari, Italy) and complied with the relevant monograph of the European Pharmacopoeia (Eur.Ph). The preparation of the capsules was performed using a 100-hole manual capsule filling machine (Farmalabor, Optima Aluminium ® ) and a precision pipette (Gilson, Microman ® ).

Oils and Capsules Preparation
For the preparation of the Cannabis oil, a precise quantity of FM2 Cannabis inflorescences was immersed in a precise quantity of extra-virgin olive oil in a weight to volume ratio of 200:1 (mg/mL). The oil containing the inflorescences was then placed in a water bath at boiling point with stirrer for 60 min. Subsequently, the oil was filtered using cotton gauze or hydrophilic cotton gauze in a manual press. Before the oil extraction phase, the Cannabis flowers had been ground for 60 s to produce a uniformly sized batch; spread in a thin layer (5 mm max; optimal thickness 1-2 mm) and placed in an oven at 140 • C for 30 min. The temperature applied was based on the fact that this is close to the evaporation point of THC (145 • C) [22]. The volume of the batches prepared was between 5 and 100 mL [12].
Using the oils obtained with the described procedure, rigid capsules were prepared. In detail, after opening the casing using a manual capsule filler, the oil was dispensed into the capsules in quantities from 0.1 mL to 0.4 mL using a precision pipette. The remaining space in each capsule was filled with micronized anhydrous silica in quantities ranging from 5 mg to 20 mg. Thereafter, the capsules were closed. The capsules were not removed from the capsule filling machine, but it was inverted 180 • so that the oil would mix with the silica. The capsules were allowed to rest in this position for 12 h: after that, they were removed from the capsule filling machine.

Analytical Method
Chromatographic analysis [23] was performed by Acquity ® UPLC system coupled with a TQD mass spectrometer (Waters, Milan, Italy). The chromatographic separation was carried out using an Acquity UPLC HSS T3 column (2.1 × 30 mm, 1.8 µm) (Waters, Milan, Italy) at a constant 30 • C. The chromatographic separation was obtained by a gradient of mobile phases A (acetonitrile and water in a ratio of 70:30 + 0.05% Formic acid) and B (isopropanol and acetonitrile in a ratio of 80:30 + 0.05% Formic acid) at a flow rate of 0.4 mL/min. The initial condition of the gradient was 100% solution A; after 3.5 min the mobile phase was brought to 100% solution B and kept there for 1.5 min. Then the column was re-equilibrated to the initial condition for 1 min (total run time 6 min). The autosampler was kept at 10 • C, the injection volume was 10 µL. Data acquisition, data processing and system control were managed by MassLynx software (Waters, Milan, Italy). The mass spectrometer coupled to the UPLC system was set in positive ionisation mode (ESI+) with a capillary voltage of 3.5 kV, a source temperature of 150 • C and a desolvation temperature of 400 • C. The flow rate of the Nitrogen was 800 L/h for the desolvation and the cone flow rate was 60 L/h. Ion monitoring was performed in multiple reaction mode, with the mass transitions and collision energies (CE) as reported here: CBD 315. 14 [23][24][25].
All the standard cannabinoid solutions necessary to create the calibration curve were diluted to concentrations between 1250 ng/mL and 5 ng/mL. CBD-d3 and THC-d3 were used as internal standards.
All the samples to be analysed were diluted with isopropanol to obtain a final concentration suitable for the range of the calibration curve.
In order to analyse the oils dispensed in the capsules, the capsules were opened by separating the two halves. Both parts were then immersed in a sufficient quantity of isopropanol (10 mL) so that the oil from the capsules could mix with the solvent. The solid residue was thereafter separated from the liquid by centrifugation. It was subsequently diluted and analysed as described above.
Each time the stability of the oils was tested, the analysis was carried out in triplicate. As for the capsules, each test was performed on 10 capsules of the same lot.

Stability Test
The oils deriving from the β-4 procedure and the same oils contained in the capsules were tested at regular intervals (at least every 30 days) in order to evaluate the stability after refrigerated storage (2-8 • C) and storage at room temperature (15-25 • C). The analytical method employed is that described in Section 4.4.

Statistical Evaluation
For each of the active molecules of interest, the average content, the corresponding standard deviation as well as the maximum and minimum concentrations in a determined quantity of the finished product were evaluated.

Conclusions
The objective of the present work was to carry out a stability study for the oil extract obtained by a specific method developed by our research group. Furthermore, in order to facilitate the consumption of the prescribed medical Cannabis therapy by patients, a standard procedure was defined for the preparation of hard capsules containing the oil extract; thereafter, the quality and stability were evaluated. The encapsulation process did not alter the quantity of the active molecule present in the oil and the stability tests yielded excellent results.
The capsule form, thanks to its features (ease of transportation, neutral organoleptic properties and stability at room temperature for extended periods) would facilitate and ensure adherence to therapy on the part of the patients in treatment.
As widely discussed in literature, oral administration has a lower bioavailability (5-20%) than inhalation. Pharmacological effects range from 30 min to 3 h and the maximum concentration of cannabinoids in the blood is usually reached within 2 h. Despite this, oral administration is generally preferred, as it is easy to administer [26][27][28][29]. The use of the formulation developed during our studies, could therefore represent a promising option for the use in therapy.
Given that gastric juices may alter some of the components of the oil, such as CBD [26][27][28][29][30], further studies will address the issue of how to make the capsules containing oils prepared by the β-4 procedure gastro-resistant. Subsequently, an assessment will be performed of whether this type of formulation improves the bioavailability of the cannabinoids of interest.

Patent
An Italian patent was granted by the Italian Office for Patents and Brands for the procedure for Cannabis oil production (patent number 102018000011128, 17 November 2020). and Marco Parente for the support during the development of the research project. The authors also wish to thank Tom O Byrne for the linguistic revision of the text.

Conflicts of Interest:
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.