Stability Study of Cannabidiol in the Form of Solid Powder and Sunflower Oil Solution

pharmaceutics

Article Stability Study of Cannabidiol in the Form of Solid Powder and Sunﬂower Oil Solution

Ema Kosovi´c 1,2 , David Sýkora 2 and Martin Kuchaˇr 3,*

1 Institute of Chemical Process Fundamentals of CAS v.v.i., Rozvojová 135, 16502 Prague, Czech Republic; [email protected] 2 Department of Analytical Chemistry, University of Chemistry and Technology Prague, Technická 5, 16628 Prague, Czech Republic; [email protected] 3 Forensic Laboratory of Biologically Active Substances, Department of Chemistry of Natural Compounds, University of Chemistry and Technology Prague, Technická 5, 16628 Prague, Czech Republic * Correspondence: [email protected]

Abstract: Stability studies represent an essential component of pharmaceutical development, en- abling critical evaluation of the therapeutic potential of an active pharmaceutical ingredient (API) or a final pharmaceutical product under the influence of various environmental factors. The aim of the present study was to investigate the chemical stability of cannabidiol (CBD) in the form of a solid powder (hereinafter referred to as CBD powder) and also dissolved in sunflower oil. We performed stress studies in accordance with the International Conference on Harmonization (ICH) guidelines, where 5 mg of marketed CBD in the form of a solid powder and in form of oil solution were exposed for 7 and 14, 30, 60, 90, 180, 270, and 365 days to precisely defined temperature and humidity conditions, 25 ◦C ± 2 ◦C/60% RH ± 5% and 40 ◦C ± 2 ◦C/75% RH ± 5% in both open and closed vials in the dark. CBD powder was significantly more stable than CBD in oil solution. Such finding is important because CBD is often administered dissolved in oil matrix in practice due to Citation: Kosović,E.; Sýkora, D.; very good bioavailability. Thus, the knowledge on admissible shelf time is of paramount importance. Kuchaˇr,M. Stability Study of Cannabidiol in the Form of Solid Keywords: cannabidiol; degradation; oil matrix; stability study; tetrahydrocannabinol; cannabinol Powder and Sunflower Oil Solution. Pharmaceutics 2021, 13, 412. https:// doi.org/10.3390/pharmaceutics 13030412 1. Introduction Academic Editor: Anne Marie Healy It has been known for thousands years that the cannabis plant interacts with the human body on many levels: to relieve neuropathic pain [1], to lower intraocular pres- Received: 2 February 2021 sure [2], to increase appetite [3], and to decrease nausea and vomiting [4]. The substances Accepted: 16 March 2021 responsible for these effects are collectively known as cannabinoids, among which the most Published: 19 March 2021 prominent are ∆9-tetrahydrocannabinol (THC), cannabidiol (CBD), and cannabinol (CBN) (Figure1). Due to the fact that a combination of psychoactive THC with non-psychoactive Publisher’s Note: MDPI stays neutral cannabinoids shows a higher activity than THC alone, a lot of effort and resources have with regard to jurisdictional claims in been invested in studying and isolating non-psychoactive alkaloids and their application published maps and institutional affil- in medicine [5]. iations. CBD is the second most prevalent active ingredient of cannabis, which does not possess any psychoactive effects. According to the report from World Health Organiza- tion, to-date, there is no evidence of public health related problems associated with the use of pure CBD [6]. Only a few publications can be found in the literature concerning Copyright: © 2021 by the authors. storage conditions and degradation of cannabinoids in general, where authors clearly Licensee MDPI, Basel, Switzerland. specify the storage conditions and individual forms of cannabis (oil, resin, etc.) [7,8]. The This article is an open access article mentioned papers emphasized a promising medical potential of CBD. Many clinical trials distributed under the terms and and government approvals have been completed or are currently underway. One of these conditions of the Creative Commons trials provided lung cancer treatment to patients with adenocarcinoma [9]. Earlier works Attribution (CC BY) license (https:// have shown that CBD has probably anti-neoplastic properties, but more recent data have creativecommons.org/licenses/by/ indicated that CBD may also provide a significant response in patients with this kind of 4.0/).

Pharmaceutics 2021, 13, 412. https://doi.org/10.3390/pharmaceutics13030412 https://www.mdpi.com/journal/pharmaceutics Pharmaceutics 2021, 13, 412 2 of 10

shown that CBD has probably anti-neoplastic properties, but more recent data have indicated that CBD may also provide a significant response in patients with this kind of cancer. However, on the other hand, there are some indications that CBD oil accelerates cer-

Pharmaceuticstain2021 types, 13, 412 of cancers; therefore, cancer should be treated with great care and distinction2 of 10 as to what type of cancer is being treated [10]. Due to its promising results in early trials, there is an increasing interest in research of CBD as a dietary supplement with potential medicinal benefits. cancer. However, on the other hand, there are some indications that CBD oil accelerates To assure an adequatecertain types use of of cancers; CBD therefore, in medicine, cancer it should is necessary be treated to with know great careits physical, and distinction chemical, and biologicalas to what properties, type of cancer and is info beingrmation treated [ 10regarding]. Due to its stability promising and results shelf-life. in early trials, Lindholst [11] describedthere is cannabis an increasing resin interest and in ex researchtract as of CBDvery as susceptible a dietary supplement to degradation with potential medicinal benefits. caused by light, temperature,To assure and an adequateoxygen, use whil ofe CBD Carbone in medicine, et al. it [12] is necessary provided to know an almost its physical, complete review ofchemical, THC degradation and biological prod properties,ucts. and A informationsimilar degradation regarding stability mechanism and shelf-life. was Lind- proposed by Laytonholst et [al.11] [13], described who cannabis focused resin more and extracton the as identification very susceptible of to degradation caused products caused byby the light, above-mentioned temperature, and oxygen,factors, while but Carbonethe results et al. are [12 ]not provided fully pharmaceu- an almost complete review of THC degradation products. A similar degradation mechanism was proposed by tically acceptable dueLayton to the et al. used [13], methanolic who focused morematrices on the and identification length of of the degradation study [14]. products caused The aim of ourby work the above-mentioned was to investigate factors, the but chemical the results stability are not fully of pharmaceuticallycommercially mar- acceptable keted formulations dueof CBD to the forms used methanolic under defin matricesed International and length of theConference study [14]. on Harmoniza- tion (ICH) guideline requirements.The aim of our The work stability was to investigate of CBD thewas chemical evaluated stability in ofthe commercially form of a mar- solid powder and anketed oil formulations solution for of CBDa one- formsyear under time defined period, International under ICH Conference stability on condi- Harmoniza- tion (ICH) guideline requirements. The stability of CBD was evaluated in the form of a tions. The study wassolid adapted powder andto the an oilclimate solution conditions for a one-year of time the period,country under where ICH stabilitythe experi- conditions. ment was conducted,The i.e., study the was Czech adapted Repub to thelic. climate High-performance conditions of the country liquidwhere chromatography the experiment was (HPLC) in combinationconducted, with i.e.,ultraviolet-visible the Czech Republic. (UV-Vis) High-performance spectroscopic liquid chromatographyand mass spectro- (HPLC) in metric (MS) detectioncombination was used with as ultraviolet-visible a suitable method (UV-Vis) for spectroscopicthe stability and measurements. mass spectrometric To (MS) detection was used as a suitable method for the stability measurements. To the best of our the best of our knowledge,knowledge, no no other other long-termlong-term stability stability study study of CBD of in CBD both thein both above-mentioned the above- forms mentioned forms hashas been been performed to-date. to-date.

Figure 1. Chemical structures of active ingredient cannabidiol (CBD) and its possible degradation products

(∆9-tetrahydrocannabinol (∆9-THC) and cannabinol (CBN)). Figure 1. Chemical structures of active ingredient cannabidiol (CBD) and its possible degradation products (Δ9-tetrahydrocannabinol2. Materials and (Δ Methods9-THC) and cannabinol (CBN)). 2.1. Chemicals and Reagents 2. Materials and MethodsCBD with a purity of 99.9% was purchased from Lipomed AG (Arlesheim, Switzer- land), CBN (>95%) from CBDepot s.r.o. (Teplice, Czech Republic), and ∆9-THC (>95%) 2.1. Chemicals and Reagentsfrom PharmaCan s.r.o. (Prague, Czech Republic). Methanol, acetonitrile, and formic acid CBD with a puritywere allof LC-MS99.9% gradewas andpurchased purchased from from Lipomed Merck (Prague, AG Czech(Arlesheim, Republic). Switzer- Flower Gold land), CBN (>95%) sunflowerfrom CBDepot oil, with s.r.o. saponification (Teplice, value Czech of 189–195 Republic), mg KOH/g and Δ and9-THC ratio (>95%) of 12:88 sat- urated:unsaturated fatty acids, was procured from Eseltix (Bratislava, Slovakia). Ultra- from PharmaCan s.r.o.pure (Prague, water used Czech for the Republic). LC-MS analyses Methanol, was obtained acetonitrile, in-house and from formic a PureLab acid Ultra were all LC-MS gradesystem and (Elga, purchased UK). from Merck (Prague, Czech Republic). Flower Gold sunflower oil, with saponification value of 189–195 mg KOH/g and ratio of 12:88 satu- 2.2. Stability Study rated:unsaturated fatty acids, was procured from Eseltix (Bratislava, Slovakia). Ultrapure Sample preparation for stability studies was conducted by weighing 5 ± 0.1 mg of water used for the LC-MS analyses was obtained in-house from a PureLab Ultra system the CBD powder into 2 mL glass vials, which were then placed in the dark within stability (Elga, UK). chambers and under appropriate conditions without any further modification. In the case of the CBD oil samples, 1 mL of sunflower oil, whose exact mass was determined 2.2. Stability Study by weighing (KERN ABP 200-5DM, Balingen, Germany), was added to the CBD powder. Sample preparation for stability studies was conducted by weighing 5 ± 0.1 mg of the CBD powder into 2 mL glass vials, which were then placed in the dark within stability chambers and under appropriate conditions without any further modification. In the case

Pharmaceutics 2021, 13, 412 3 of 10

An ultrasonic bath (Bandelin Sonorex, Berlin, Germany) at 35 kHz for 3 min was used to completely dissolve CBD in sunflower oil. All of the samples were prepared in duplicate for each condition (temperature and humidity) and time point in open and closed vials to address the possible influence of free access of air to the CBD samples. Closed vials (screw caps) were wrapped with Parafilm (Pechiney Plastic Packaging, Ohio, USA). The prepared samples were placed in the stability chambers at Quinta Analytica, s.r.o. (Prague, Czech Republic). According to ICH guidelines [15], the recommended storage conditions for both forms of CBD were 25 ◦C/60% RH and 40 ◦C/75% RH for 7 and 14, 30, 60, 90, 180, 270, and 365 days. Photostability tests included the preparation of four samples and their exposition to UV and Vis light, respectively. Those samples were prepared by the same procedure as mentioned above. The required conditions were total exposure in duration of 5 days at least 1.2 million lux hours to the Vis and another 5 days to 200 W hours/m2 to the UV light source. The source used to illuminate Vis radiation was equipped with TL D W18/33 640 fluorescent tubes and the UV intensity measurements were performed with a standard UV spectroradiometer in the spectral range of 350–405 nm. After exposure to stability conditions, the viscosity of samples was measured with a HAAKE RheoStress 600 instrument (Waltham, Thermo Scientific, USA) using a PP35 sensor at a constant temperature of 25 ◦C.

2.3. Preparation of Stability Samples and Reference Standard Samples before LC-MS At definite time intervals, the stability samples were taken from the stability chambers and stored in a freezer in the dark at −50 ◦C until LC-MS measurement. Samples for LC-MS analysis were prepared by dilution in methanol to a final concentration of 0.5 mg/mL. The CBD oil samples required a more complex sample preparation protocol. In the first step, the oil samples were diluted 100 times with methanol. Due to the limited miscibility of oil with methanol, the samples were intensively treated in an ultrasonic bath at 35 kHz for 3 min and then further diluted 10 times with methanol. The resulting homogenous solution was injected directly into the LC-MS system. Reference samples which were not exposed to the environmental influences were diluted with methanol and injected in LC-MS as described above. The area under the peak of reference samples was considered as the standard form determining the degradation in the stressed samples.

2.4. LC-UV-MS Analysis of Stability Samples The experiments were performed using a Dionex Ultimate 3000 HPLC system (Waltham, Thermo Scientific, USA) composed of a dual pump, an autosampler, a column thermostat compartment, and a diode array detector (DAD). An Eclipse C18 Plus column (Agilent Technologies, Santa Clara, CA, USA) with dimensions of 50 mm length and 2.1 mm internal diameter with particle size of 5 µm was used for was used for separation. The column was equipped with a guard column with dimensions 5 mm length and 2.1 mm internal diameter, packed with the same sorbent. The column oven and autosampler were thermostated at a temperature of 25 ◦C. Detection was performed simultaneously with the DAD and a triple quadrupole mass spectrometer (3200 Q TRAP, AB Sciex, Ontario, Canada) with electrospray ionization (ESI) in a positive mode. Analyst 1.6 (AB Sciex) software was used to acquire and evaluate data from the LC-UV-MS system. The mobile phase consisted of ultrapure water containing 0.1% formic acid as phase A and acetonitrile containing 0.1% formic acid as phase B at a flow rate of 0.2 mL/min. The LC gradient had the following time profile: 0.0–10.0 min linearly from 5% to 100% B; 100% B was maintained from 10.0 to 12.0 min; then from 12.0 to 12.5 min 100% to 5% B. Finally, 5% B was maintained to 20.0 min. The injection volume was 5 µL. Serial detection was performed using the DAD, where the UV spectra were scanned in the range of 200–400 nm and MS detection took place with a mass range of 80–1550 m/z. Quantification of the powder and CBD oil samples was based on area normalization of automatically integrated peaks with appropriate weight correction of samples using UV detection at 225 nm and 210 nm, respectively. Pharmaceutics 2021, 13, 412 4 of 10

the DAD, where the UV spectra were scanned in the range of 200–400 nm and MS detection took place with a mass range of 80–1550 m/z. Quantification of the powder and CBD oil samples was based on area normalization of automatically integrated peaks with appropriate weight correction of samples using UV detection at 225 nm and 210 nm, respectively.

Pharmaceutics 2021, 13, 412 4 of 10 2.5. Statistical Analysis All statistical data presented in the following text and p values were calculated by applying Student’s2.5. significant Statistical Analysis test (MS Excel 2010, USA). All statistical data presented in the following text and p values were calculated by 3. Results applying Student’s signiﬁcant test (MS Excel 2010, USA). 3.1. HPLC-MS Analysis3. Results Cannabidiol3.1. and HPLC-MS its potential Analysis degradation products were separated on a C18 column by gradient elution, asCannabidiol described and in itsSection potential 2.4., degradation utilizing products simultaneous were separated UV and on MS a C18 detec- column by gradient elution, as described in Section 2.4, utilizing simultaneous UV and MS detection. tion. The additionThe of addition formic ofacid formic to acidthe tomo thebile mobile phase phase improved improved chromatographic chromatographic resolution resolu- and tion and providedprovided better betterpeak peak shapes, shapes, while while working working at at a ﬂowa flow rate rate of 0.2 of mL/min 0.2 mL/min provided pro- good vided good chromatographicchromatographic performance performance (Figure(Figure2a,b). 2a,b).

FigureFigure 2. HPLC-UVchromatograms 2. HPLC-UVchromatograms of CBD reference of CBD samples reference (black line)samples and CBD (black samples line) stored and forCBD 365 samples days (red line) in thestored stability for chambers. 365 days The (red CBD line) powder in the was stability measured chambers. at 225 nm (a )The and CBD the CBD powder oil solution was at measured 210 nm (b). at 225 nm (a) and the CBD oil solution at 210 nm (b). Each sample exposed to specified conditions was measured four times. The test used to detect a possible decrease in the amount of CBD over time was based on a statistical Each samplecomparison exposed to of multiplespecified samples conditions to determine was measured whether the four average times. peak The area test values used of the to detect a possiblereference decrease standard in the sample amount statistically of CBD differed over from time the was average based areas on of thea statistical real stressed comparison of multiplesamples samples taken from to the determine stability chambers whether at the a specific average time. peak For thisarea purpose, values Student’sof the reference standardt-test sample was applied statistically with a significance differed from level oftheα =average 0.05. areas of the real stressed samples taken from3.2. the Evaluation stability of CBD chambers Stability Samplesat a specific time. For this purpose, Student’s t- test was applied withThe a significance results of the level CBD of powder α = 0.05. stability study are presented in Figure3. As can be seen, CBD in the form of powder is stable through almost the whole of the testing ◦ ◦ 3.2. Evaluation of CBDperiod. Stability The integrity Samples of the CBD exposed to 25 C ± 2 C/60% RH ± 5% remained statistically unchanged for 270 days. The samples stored in open vials for one year showed The results ofa slight the CBD decrease powder of 10.37 stability± 0.51% instud totaly (arep = 9.43presented× 10−6). in The Figure samples 3. stored As can in closed be seen, CBD in the vialsform under of powder the same is conditionsstable through (25 ◦C ±almost2 ◦C/60% the RHwhole± 5%) of alsothe showedtesting statisticallyperiod. The integrity of thesignificant CBD exposed differences to after 25 one°C ± year 2 °C/60% of storage RH (p = ± 7.42 5%× remained10−7) with statistically a smaller decrease un- of changed for 270 days.8.01 ± 0.67%The samples in total. stored in open vials for one year showed a slight de- The stress conditions corresponding to 40 ◦C ± 2 ◦C/75% RH ± 5% led to a statistically −6 crease of 10.37 ± significant0.51% in decreasetotal (p in= the9.43 amount × 10 of). CBDThe insamples samples stored for in 180 closed days (vialsp = 7.91 under× 10− 7) the same conditionsin both (25 open °C ± and 2 °C/60% closed vials. RH The ± 5%) observed also averageshowed decrease statistically was 8.21 significant± 0.57%. differences after one yearIn additionof storage to temperature (p = 7.42 × and 10− humidity,7) with a another smaller important decrease parameter of 8.01 was ± 0.67% free ver- in total. sus restricted air access to the samples (open vs. closed vials). The results showed that the change in the amount of CBD stored as solid powder depended mainly on tempera- The stress conditions corresponding to 40 °C ± 2 °C/75% RH ± 5% led to a statistically ture. However, the differences between samples stored in open and closed vials seemed −7 significant decreaseslightly in the more amount pronounced of CBD (but in statistically samples insignificant) stored for 180 at 25 days◦C ± (2p◦ C/60%= 7.91 × RH 10± )5% in both open andafter closed 365 days.vials. The observed average decrease was 8.21 ± 0.57%. In addition to temperature and humidity, another important parameter was free ver- sus restricted air access to the samples (open vs. closed vials). The results showed that the change in the amount of CBD stored as solid powder depended mainly on temperature. However, the differences between samples stored in open and closed vials seemed slightly

Pharmaceutics 2021, 13, 412 5 of 10

Pharmaceutics 2021, 13, 412 more pronounced (but statistically insignificant) at 25 °C ± 2 °C/60% 5RH of 10 ± 5% after 365 days.

FigureFigure 3.3.Stability Stability of theof the CBD CBD powder powder samples samples expressed expressed as the amount as ofthe CBD amount over time of measuredCBD over time meas- uredby LC-UV by LC-UV (225 nm). (225 nm). When stored at 25 ◦C ± 2 ◦C/60% RH ± 5%, CBD in sunflower oil was stable for a periodWhen of at least stored 180 days.at 25 The °C first ± 2 statistically °C/60% RH significant ± 5%, difference CBD in between sunflower the samples oil was stable for a periodwas observed of at least at 270 180 days days. in open The vs. first closed statisti vials (pcally= 3.17 significant× 10−6), where differen in opence samples, between the samples wasa CBD observed decrease at of 270 11.41 days± 0.45% in open compared vs. closed to samples vials stored(p = 3.17 in the × 10 closed−6), where vials was in open samples, measured (Figure4). After a storage period of one year, the amount of CBD reached a value aof CBD 58.03% decrease of the original of 11.41 amount, ± 0.45% which compared corresponds to tosamples an average stored total in loss the of 41.97%closed ofvials was meas- uredCBD in(Figure open and 4). closedAfter vials.a storage Using period the t-test, of calculated one year,p valuethe amount showed of no CBD significant reached a value of 58.03%difference of between the original the amount amount, of CBD which in the samplescorresponds stored into open an average and closed total vials loss for of 41.97% of CBDthe whole in open testing and period. closed vials. Using the t-test, calculated p value showed no significant ◦ ± ◦ differenceUnlike between the CBD powder the amount samples, of the CBD CBD in oil the samples samples stored stored at 40 inC open2 C/75% and closed vials for RH ± 5% in open vials underwent significant degradation as early as after 90 days, with a theloss whole of 20.2% testing of CBD period. (Figure4). Moreover, there was a very significant difference between samplesUnlike stored the in CBD open andpowder closed samples, vials. The the analysis CBD ofoil samples samples stored stored in closed at 40 vials °C ± 2 °C/75% RH ±showed 5% in noopen statistical vials underwent decrease in the significant amount of degradation CBD after 90 days as early under as the after same 90 con- days, with a loss −12 ofditions. 20.2% After of CBD 180 days, (Figure however, 4). aMoreover, significant CBDthere loss was of 16.5%a very (p =significant 3.68 × 10 difference) was between determined. After a storage period of one year, the amount of CBD stored in the closed samplesvials reached stored a value in open of 24.09%, and withclosed total vials. degradation The analysis of 75.91%, of while samples the remaining stored in closed vials showedamount of no CBD statistical in open vials decrease showed in a value the amount of 1.03%, correspondingof CBD after to 90 a total days degradation under the same conditions.of 98.97% After CBD. 180 days, however, a significant CBD loss of 16.5% (p = 3.68 × 10−12) was determined.All ofAfter the oila storage samples wereperiod weighed of one before year, and th aftere amount the storage of CBD period. stored This elimi- in the closed vials nated any potential imperfections in the CBD amount determination caused by possible oil reached a value of 24.09%, with total degradation of 75.91%, while the remaining amount evaporation, and no oil evaporation was registered. of CBDOur in preliminary open vials experiment showed confirmed a value that of samples1.03%, ofcorresponding sunflower oil displayed to a total New- degradation of 98.97%tonian behavior, CBD. where their viscosity depended on the shear-strain rate. The viscosity of samplesAll storedof the in oil open samples vials at 40were◦C ± weighed2 ◦C/75% before RH ± 5%, and increased after the from storage 0.045 Pa.s period. to This elimi- nated3.62 Pa.s any after potential 365 days (Figureimperfections5). Although in athe statistically CBD amount significant determination increase in viscosity caused by possible was also observed for oil samples in the closed vials after 90 days (p = 0.001), the change in oilviscosity evaporation, of the samples and no in theoil closedevaporation vials was was much registered. slower. As the viscosity increased, the solubility of the oil samples in methanol decreased, and a longer ultrasonic mixing was necessary for the sample preparation preceding liquid chromatograph-ultraviolet-mass spectrometric (LC-UV-MS) analysis.

Pharmaceutics 2021, 13, 412 6 of 10

Pharmaceutics 2021, 13, 412 6 of 10 Pharmaceutics 2021, 13, 412 6 of 10

Figure 4. Stability of the CBD oil samples expressed as the amount of CBD over time measured by LC-UV (UV 210 nm).

Our preliminary experiment confirmed that samples of sunflower oil displayed Newtonian behavior, where their viscosity depended on the shear-strain rate. The viscosity of samples stored in open vials at 40 °C ± 2 °C/75% RH ± 5%, increased from 0.045 Pa.s to 3.62 Pa.s after 365 days (Figure 5). Although a statistically significant increase in viscosity was also observed for oil samples in the closed vials after 90 days (p = 0.001), the change in viscosity of the samples in the closed vials was much slower. As the viscosity increased, the solubility of the oil samples in methanol decreased, and a longer ultrasonic mixing was necessary for the sample preparation preceding liquid chromatograph-ultraviolet- FigureFigure 4.4.Stability Stability of of the the CBD CBD oil samplesoil samples expressed expressed as the amountas the amount of CBD over of CBD time over measured time by measured by LC-UVmassLC-UV spectrometric (UV(UV 210 210 nm). nm). (LC-UV-MS) analysis.

FigureFigure 5.5.Dependence Dependence of of the the viscosity viscosity of the of oil the samples oil samples stored stored at 40 ◦C at± 402 ◦ °CC/75% ± 2 °C/75% RH ± 5% RH in ± 5% in closedclosed andand open open vials. vials. The irradiated samples of CBD in both forms (powder and oil) for 5 days were subjectedThe toirradiated statistical analysissamples of of CBD CBD content in both after assessmentforms (powder for possible and oil) visible for changes 5 days were sub- jectedin physical to statistical properties analysis such as clarity,of CBD color, conten etc.t Figure after 6assessment along with thefor statistical possible testsvisible changes inshows physical that light properties exposure didsuch not as lead clarity, to a demonstrable color, etc. Figure degradation 6 along of the with CBD the samples statistical tests showsexposed that to 25light◦C ±exposure2 ◦C/60% did RH not± lead5%. However,to a demonstrable a statistically degradation signiﬁcant of decrease the CBD samples of approximately 4.5% was determined for the oil samples stored at 40 ◦C ± 2 ◦C/75% RH ± 5% in the open vials (p = 0.043).

Figure 5. Dependence of the viscosity of the oil samples stored at 40 °C ± 2 °C/75% RH ± 5% in closed and open vials.

The irradiated samples of CBD in both forms (powder and oil) for 5 days were sub- jected to statistical analysis of CBD content after assessment for possible visible changes in physical properties such as clarity, color, etc. Figure 6 along with the statistical tests shows that light exposure did not lead to a demonstrable degradation of the CBD samples

Pharmaceutics 2021, 13, 412 7 of 10

exposed to 25 °C ± 2 °C/60% RH ± 5%. However, a statistically significant decrease of ap- Pharmaceutics 2021, 13, 412 proximately 4.5% was determined for the oil samples stored at 40 °C ± 2 °C/75%7 of 10 RH ± 5% in the open vials (p = 0.043).

Figure 6.6.Photostability Photostability of of the the CBD CBD samples samples measured measured by LC-UV. by LC-UV. (LE, light (LE, exposure). light exposure).

Despite not being the main aim of the study, we tried to detect degradation products Despite not being the main aim of the study, we tried to detect degradation products of CBD during the course of the stability study using LC-UV-MS. Two important poten- oftial CBD CBD during degradation the course products of the were stability available study for ususing in the LC-UV-MS. form of authentic Two important standards, potential CBDspecifically, degradation∆9-THC products and cannabinol were available (CBN) (Figure for 1us). in the form of authentic standards, specifically,Because Δ9-THC all of theand CBD cannabinol stability measurements(CBN) (Figure were 1). performed utilizing UV in series withBecause MS detection all of inthe the CBD MS stability scanning measurements mode, it was possible were performed to register utilizing the potential UV in series withpresence MS ofdetection∆9-THC in and the CBN MS in scanning partially mode, degraded it was CBD possible samples. to A register detailed the analysis potential of pres- enceall of theof Δ obtained9-THC and LC-UV-MS CBN in chromatograms partially degraded provided CBD evidence samples. of CBN A detailed in many analysis partially of all of 9 thedegraded obtained samples, LC-UV-MS whereas chromatograms∆ -THC was not provided found. However, evidence it should of CBN be in noted many that partially the de- detection limits of the quadrupole mass spectrometer used in the scanning mode is rather graded samples, whereas Δ9-THC was not found. However, it should be noted that the limited and a possible low amount of ∆9-THC may have been left undetected. In addition, detection limits of the quadrupole mass spectrometer used in the scanning mode is rather we found several oxidation products of CBD eluted at various retention times in partially limiteddegraded and samples, a possible as demonstrated low amount by of the Δ9-THC presence may of [M+H] have +been331.1 leftm/ undetected.z and 347.3 peaks In addition, wein the found MS chromatograms several oxidation [12]. products of CBD eluted at various retention times in partially degraded samples, as demonstrated by the presence of [M+H]+ 331.1 m/z and 347.3 peaks in4. Discussionthe MS chromatograms [12]. An integral part of the systematic approach to evaluating a drug stability is to verify 4.its Discussion behavior under the influence of various external factors. The stability assessment, the length of the stability study, and the storage conditions are usually the main focus of a pharmaceuticalAn integral scientist’s part of attentionthe systematic in the development approach to of evaluating all dosage forms. a drug In stability general, theis to verify itsdrug behavior should beunder evaluated the influence under storage of various conditions, external with appropriatefactors. The tolerances, stability toassessment, verify the lengthits thermal of the stability stability and moisture study, and sensitivity. the storage Stability conditions test design are must usually take in considerationthe main focus of a pharmaceuticalthe nature of the substancesscientist’s andattention geographic in the area development in which the drugsof all willdosage be used forms. [15]. In general, the drugIn order should to evaluate be evaluated the effect ofunder heat, humidity,storage conditions, oxygen access, with matrix, appropriate and light, tolerances, it was to verifynecessary its thermal to find an stability effective, and sensitive, moisture and sensitivity. selective analytical Stability method test design for the must detection take in con- siderationand determination the nature of CBD of the and substances its possible degradationand geographic products. area The in literaturewhich the refers drugs to will be a few studies where cannabinoids were analyzed in various matrices using isocratic and used [15]. gradient elution profiles, while a combination of UV and MS/MS was used to measure cannabidiol-richIn order to productsevaluate [ 16the–19 effect]. Separation of heat, of hu cannabinoidsmidity, oxygen under access, isocratic matrix, conditions and light, it wasis a challenging necessary to task find due an to effective, their physical sensitive, and chemical and selective properties analytical [17,20]. method Therefore, for we the detec- tionchose and the determination gradient profile, of using CBD acetonitrile and its possible and water. degradation All samples products. were measured The literature by the refers tosame a few LC method.studies where cannabinoids were analyzed in various matrices using isocratic and gradientThe results elution of the CBDprofiles, powder while stability a combination study are of in correspondenceUV and MS/MS with was a used stability to measure cannabidiol-richexperiment performed products by Turner [16–19]. et al. Separation [21], who studied of cannabinoids the influence under of higher isocratic tempera- conditions istures a challenging on dried cannabis task due plant to material. their physical They showed and chemical that a significant properties loss of[1 cannabinoids7,20]. Therefore, we

Pharmaceutics 2021, 13, 412 8 of 10

is caused by thermal exposure at 37 ◦C and 50 ◦C in the first 10 weeks, while the content of CBD in all of the stored materials remained relatively constant without any detectable decomposition for 100 weeks. Considering the fact that our experiments were performed in stability chambers, light played no part in the decomposition of CBD in the above-mentioned stability tests. In this regard, the results cannot be in total correspondence with data published by Trofin et al. [22], who showed that the content of CBD in cannabis oil, obtained by extraction from herbal cannabis decreased with storage time after light exposure. They observed a steady decrease of ∆9-THC and CBD over the entire storage period of four years, which was more pronounced in samples exposed to light at 22 ◦C than in samples stored in the dark at 4 ◦C. The average loss of CBD recorded during each year was 4–5.5%. At the same time, the content of CBN continuously increased during the storage period with an average gain of 4.5%. Nevertheless, it may be concluded that the behavior of CBD does not deviate much depending on the used oil matrix. In addition to a statistically proven decrease in the amount of CBD after a specific time under given conditions, certain other physical and chemical changes to the oil samples were observed during the course of our stability test. Specifically, a slight change in color to a darker yellow was observed in the case of the samples stored for 180 + days. This was accompanied by an increase in viscosity of samples stored for 180 + days at 40 ◦C ± 2 ◦C/75% RH ± 5% (Figure5) (vide supra). A rapid color development of heated oil mentioned by Maskan may be correlated to the chemical processes taking place in the sunflower oil samples [23]. A change of color of the oil from orange to brown indicates the processes of oxidation, polymerization, and other chemical changes, which also lead to an increase in the viscosity of heated oil [24]. Therefore, the observed change in color and the increase in viscosity mainly for samples exposed to 40 ◦C ± 2 ◦C/75% RH ± 5% is likely attributed to the higher temperature and the influence of oxygen in open samples, resulting in the formation of complex structures in the oil [25]. There is an apparent link between the change of oil viscosity after 90 days at higher temperatures and the incipient CBD degradation under such conditions. In their work, Pavlovic et al. [26] compared olive oil, hemp seed oil, and medium chain triglycerides (MCT), all three containing the same amount of CBD. They determined that MCT oil was more stable and more likely to maintain a consistent flavor and visual characteristics over time, which improved the bioavailability of CBD. Photostability studies, necessary to evaluate the overall photosensitivity of CBD, are very important due to a possible change that may occur when a product is stored under light exposure. According to some authors, cannabinoids are less stable in light, but this also depends on other conditions such as the solvent in which they are stored, as well as temperature, oxygen access, and many other factors. Fairbairn et al. [27] showed that light-exposed samples stored in various solvents degraded faster than samples stored in the same solvents in the dark, while Turner and Hanry [28] reported that THC and CBD either in solution or in crude extract form were stable for six days when exposed to both natural and artificial light. The main conclusion that may be derived is that the exposure to light alone for the given time period does not lead to significant changes in the CBD content, but in combination with other factors, such as the solvent used, increased temperature, and presence of oxygen, light may accelerate the degradation process.

5. Conclusions While CBD stored in the form of a solid powder remained mostly intact over a one- year time period under the studied stability conditions with an approximately 10% decrease in CBD, the influence of temperature, humidity, and air oxygen on CBD oil samples was much more pronounced. Very significant degradation of CBD took place between the 90th and 180th day in the case of CBD oil samples in open vials at 40 ◦C ± 2 ◦C/75% RH ± 5%, with a complete degradation of CBD after approximately 270 days. Exposure to artificial Pharmaceutics 2021, 13, 412 9 of 10

light showed that the light itself does not have strong adverse effects on CBD. Nevertheless, in combination with other factors, it may accelerate the degradation process. The experimental results of our CBD stability study showed that under deﬁned stability conditions, various degradation products, including CBN and several oxidation products, were formed. However, ∆9-THC was not detected by our method.

Author Contributions: Conceptualization, E.K. and D.S.; methodology, M.K., D.S., E.K.; formal analysis, D.S.; resources, E.K.; data curation, E.K.; writing—original draft preparation, E.K.; writing— review and editing, D.S.; supervision, D.S. and M.K.; project administration, M.K. and D.S.; funding acquisition, M.K. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by Ministry of the Interior of the Czech Republic, grant number MV0/VI20172020056. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: The source data available upon a reasonable request. Acknowledgments: We would like to express our thanks to Jakub Rudovský from Quinta Analytica, s.r.o. for his valuable suggestions in the planning and developing of the stability study, and Vˇera Pˇenkavová from the Institute of Chemical Process Fundamentals of CAS for providing ideas and overall assistance in the viscosity experiments of the CBD oil samples. We also thank Quinta Analytica, s.r.o. for allowing us to use their stability chambers. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References 1. Lee, G.; Grovey, B.; Furnish, T.; Wallace, M. Medical Cannabis for Neuropathic Pain. Curr. Pain Headache Rep. 2018, 22, 8. [CrossRef][PubMed] 2. Tomida, I.; Pertwee, R.G.; Azuara-Blanco, A. Cannabinoids and glaucoma. Br. J. Ophthalmol. 2004, 88, 708–713. [CrossRef] [PubMed] 3. Sansone, R.A.; Sansone, L.A. Marijuana and body weight. Innov. Clin. Neurosci. 2014, 11, 50–54. [PubMed] 4. Parker, L.A.; Rock, E.M.; Limebeer, C.L. Regulation of nausea and vomiting by cannabinoids. Br. J. Pharmacol. 2011, 163, 1411–1422. [CrossRef][PubMed] 5. Pacifici, R.; Marchei, E.; Salvatore, F.; Guandalini, L.; Busardo, F.P.; Pichini, S. Evaluation of cannabinoids concentration and stability in standardized preparations of cannabis tea and cannabis oil by ultra-high performance liquid chromatography tandem mass spectrometry. Clin. Chem. Lab. Med. 2017, 55, 1555–1563. [CrossRef][PubMed] 6. Beardsley, P.M.; Brands, B.; Ekwere, I.T.; Elliott, S.; Gysling, K.; Jain, R.; Kaduri, P.; Kitanaka, J.; Rahimi-Movaghar, A.; Nudmamud- Thanoi, S.; et al. WHO Expert Committee on Drug Dependance: No. 1013, Cannabidiol; WHO Press: Geneva, Switzerland, 2018; pp. 1–27. 7. Szaflarski, J.P.; Bebin, E.M.; Comi, A.M.; Patel, A.D.; Joshi, C.; Checketts, D.; Beal, J.C.; Laux, L.C.; De Boer, L.M.; Wong, M.H.; et al. Long-term safety and treatment effects of cannabidiol in children and adults with treatment-resistant epilepsies: Expanded access program results. Epilepsia 2018, 59, 1540–1548. [CrossRef] 8. Leweke, F.M.; Piomelli, D.; Pahlisch, F.; Muhl, D.; Gerth, C.W.; Hoyer, C.; Klosterkotter, J.; Hellmich, M.R.; Koethe, D. Cannabidiol enhances anandamide signaling and alleviates psychotic symptoms of schizophrenia. Transl. Psychiatry 2012, 2, 94. [CrossRef] 9. Sulé-Suso, J.; Watson, N.A.; Van Pittius, D.G.; Jegannathen, A. Striking lung cancer response to self-administration of cannabidiol: A case report and literature review. SAGE Open Med. Case Rep. 2019, 7, 2050313X19832160. [CrossRef] 10. Martinez-Martinez, E.; Martin-Ruiz, A.; Martin, P.; Calvo, V.; Provencio, M.; Garcia, J.M. CB2 cannabinoid receptor activation promotes colon cancer progression via AKT/GSK3beta signaling pathway. Oncotarget 2016, 7, 68781–68791. [CrossRef] 11. Lindholst, C. Long term stability of cannabis resin and cannabis extracts. Aust. J. Forensic Sci. 2010, 42, 181–190. [CrossRef] 12. Carbone, M.; Castelluccio, F.; Daniele, A.; Sutton, A.; Ligresti, A.; Di Marzo, V.; Gavagnin, M. Chemical characterisation of oxidative degradation products of ∆9-THC. Tetrahedron 2010, 66, 9497–9501. [CrossRef] 13. Layton, C.; Runco, J.; Aubin, A. Forced Degradation of Cannabidiol; Waters Corporation: Milford, MA, USA, 2016; Volume 6, p. 720005766EN. 14. International Conference on Harmonization Guideline: Impurities: Guideline for Residual Solvents Q3C (R5); ICH Steering Committee: London, UK, 2005; Volume 4, pp. 1–25. Available online: https://www.ema.europa.eu/en/documents/scientific-guideline/inter national-conference-harmonisation-technical-requirements-registration-pharmaceuticals-human-use_en-33.pdf (accessed on 25 July 2020). Pharmaceutics 2021, 13, 412 10 of 10

15. Guidance for Industry Q1A(R2): Stability Testing of New Drug Substances and Products; US. Food and Drug Administration: Rockville, MD, USA, 2003. Available online: https://www.fda.gov/regulatory-information/search-fda-guidance-documents/q1ar2-stabili ty-testing-new-drug-substances-and-products (accessed on 25 July 2020). 16. Palazzoli, F.; Citti, C.; Licata, M.; Vilella, A.; Manca, L.; Zoli, M.; Vandelli, M.A.; Forni, F.; Cannazza, G. Development of a simple and sensitive liquid chromatography triple quadrupole mass spectrometry (LC–MS/MS) method for the determination of cannabidiol (CBD), ∆9-tetrahydrocannabinol (THC) and its metabolites in rat whole blood after oral administration of a single high dose of CBD. J. Pharm. Biomed. Anal. 2018, 150, 25–32. [PubMed] 17. Citti, C.; Ciccarella, G.; Braghiroli, D.; Parenti, C.; Vandelli, M.A.; Cannazza, G. Medicinal cannabis: Principal cannabinoids concentration and their stability evaluated by a high performance liquid chromatography coupled to diode array and quadrupole time of flight mass spectrometry method. J. Pharm. Biomed. Anal. 2016, 128, 201–209. [CrossRef][PubMed] 18. Deidda, R.; Avohou, H.T.; Baronti, R.; Davolio, P.L.; Pasquini, B.; Del Bubba, M.; Hubert, C.; Hubert, P.; Orlandini, S.; Furlanetto, S. Analytical quality by design: Development and control strategy for a LC method to evaluate the cannabinoids content in cannabis olive oil extracts. J. Pharm. Biomed. Anal. 2019, 166, 326–335. [CrossRef][PubMed] 19. Nemeškalová, A.; Hajkova, K.; Mikulu, L.; Sykora, D.; Kuchar, M. Combination of UV and MS/MS detection for the LC analysis of cannabidiol-rich products. Talanta 2020, 219, 121250. [CrossRef] 20. Zgair, A.; Wong, J.C.; Sabri, A.; Fischer, P.M.; Barrett, D.A.; Constantinescu, C.S.; Gershkovich, P. Development of a simple and sensitive HPLC-UV method for the simultaneous determination of cannabidiol and Delta(9)-tetrahydrocannabinol in rat plasma. J. Pharm. Biomed. Anal. 2015, 114, 145–151. [CrossRef] 21. Turner, C.E.; Hadley, K.W.; Fetterman, P.S.; Doorenbos, N.J.; Quimby, M.W.; Waller, C. Constituents of Cannabis sativa L. IV: Stability of cannabinoids in stored plant material. J. Pharm. Sci. 1973, 62, 1601–1605. [CrossRef][PubMed] 22. Trofin, I.; Dabija, G.; Vaireanu, D.I.; Laurentiu, F. The influence of long-term storage conditions on the stability of cannabinoids derived from cannabis resin. Rev. Chim. 2012, 63, 422–427. 23. Maskan, M. Change in colour and rheological behaviour of sunflower seed oil during frying after adsorbent treatment of used oil. Eur. Food Res. Technol. 2003, 218, 20–25. [CrossRef] 24. Tan, Y.A.; Ong, S.H.; Berger, K.G.; Oon, H.H.; Poh, L.B. A study of the cause of rapid color development of heated refined palm oil. J. Am. Oil Chem. Soc. 1985, 62, 999–1006. [CrossRef] 25. Stevenson, S.G.; Vaisey-Genser, M.; Eskin, N.A.M. Quality control in the use of deep frying oils. J. Am. Oil Chem. Soc. 1984, 61, 1102–1108. [CrossRef] 26. Pavlovic, R.; Nenna, G.; Calvi, L.; Panseri, S.; Borgonovo, G.; Giupponi, L.; Cannazza, G.; Giorgi, A. Quality traits of “Cannabidiol oils”: Cannabinoids tontent, terpene fingerprint and oxidation stability of European commercially available preparations. Molecules 2018, 23, 1230. [CrossRef][PubMed] 27. Fairbairn, J.W.; Liebmann, J.A.; Rowan, M.G. The stability of cannabis and its preparations on storage. J. Pharm. Pharmacol. 1976, 28, 1–7. [CrossRef][PubMed] 28. Turner, C.E.; Henry, J.T. Constituents of Cannabis sativa L. IX: Stability of synthetic and naturally occurring cannabinoids in chloroform. J. Pharm. Sci. 1975, 64, 357–359. [CrossRef][PubMed]