Pharmacokinetic Profile of ∆⁹-Tetrahydrocannabinol, Cannabidiol and Metabolites in Blood following Vaporization and Oral Ingestion of Cannabidiol Products

Abstract

There is limited data on the comparative pharmacokinetics of cannabidiol (CBD) across oral and vaporized formulations. This within-subject, double-blind, double-dummy, placebo-controlled laboratory study analyzed the pharmacokinetic profile of CBD, ∆⁹-tetrahydrocannabinol (∆⁹-THC) and related metabolites in blood and oral fluid (OF) after participants (n = 18) administered 100 mg of CBD in each of the following formulations: (1) oral CBD, (2) vaporized CBD and (3) vaporized CBD-dominant cannabis containing 10.5% CBD and 0.39% ∆⁹-THC (3.7 mg); all participants also completed a placebo condition. Oral CBD was administered in three formulations: (1) encapsulated CBD, (2) CBD suspended in pharmacy-grade syrup and (3) Epidiolex, allowing for pharmacokinetic comparisons across oral formulations (n = 6 per condition). An optional fifth experimental condition was completed for six participants in which they fasted from all food for 12 h prior to oral ingestion of 100 mg of CBD. Blood and OF samples were collected immediately before and for 57–58 h after each drug administration. Immunoassay screening and LC–MS-MS confirmatory tests were performed, the limit of quantitation was 0.5 ng/mL for ∆⁹-THC and 1 ng/mL for CBD. The mean C_max and range of CBD blood concentrations for each product were as follows: vaporized CBD-dominant cannabis, 171.1 ng/mL, 40.0–665.0 ng/mL, vaporized CBD 104.6 ng/mL, 19.0–312.0 ng/mL and oral CBD, 13.7 ng/mL, 0.0–50.0 ng/mL. Of the three oral formulations, Epidiolex produced the greatest peak concentration of CBD (20.5 ng/mL, 8.0–37.0 ng/mL) relative to the capsule (17.8 ng/mL, 2.0–50.0 ng/mL) and syrup (2.8 ng/mL, 0–7.0 ng/mL). ∆⁹-THC was detected in the blood of 12/18 participants after vaporized CBD-dominant cannabis use, but neither ∆⁹-THC nor its metabolite THC-COOH were detected in the blood of any participants after vaporized or oral CBD-only administration. These data demonstrate that different oral and vaporized formulations produce substantial variability in the pharmacokinetics of CBD and that CBD alone is unlikely to convert to ∆⁹-THC or produce positive drug tests for ∆⁹-THC or its metabolite.

Introduction

Cannabis is one of the most self-administered psychoactive drugs (1, 2). Although medicinal and non-medicinal cannabis use is becoming increasingly permissible through state-level legalization, under US federal law, cannabis containing >0.3% delta-9-tetrahydrocannabinol (∆⁹-THC) remains illegal and a Schedule I drug (i.e., deemed to have no medicinal use and high abuse potential). In December 2018, the passing of the Agriculture Improvement Act of 2018 (aka the “Farm Bill”) removed hemp (cannabis with ≤0.3% ∆⁹-THC) and hemp-derived products from the controlled substances list in the USA (3). Several other countries similarly differentiate cannabis from hemp from a regulatory perspective (e.g., <0.05%–<1% ∆⁹-THC), although the amount of ∆⁹-THC allowed varies.

∆⁹-THC is the primary psychoactive phytocannabinoid in cannabis that produces feelings of “high” and other hallmark subjective and physiological effects commonly associated with acute cannabis use. Cannabidiol (CBD) is another phytocannabinoid, and, in contrast to ∆⁹-THC, is not typically associated with robust drug effects or impairment (4, 5). Retail CBD products, typically derived from legal hemp, include those intended to be smoked, vaporized, orally ingested and topically applied, and there are a variety of formulations for products within each of those intended routes of administration.

Presently, inadequate regulation enforcement of retail CBD product limits poses risks of label inaccuracies and contamination with ∆⁹-THC, potentially leading to positive drug tests for cannabis (6, 7). Further, commercial CBD products in the US can legally contain up to 0.3% ∆⁹-THC (this threshold varies in other countries). It is unclear what impact the presence of low levels of ∆⁹-THC in CBD products may have on drug testing. Inaccurate test results could have serious consequences in several contexts in which drug-testing commonly is done, including the workplace, the criminal justice system and drug treatment centers. Typically, cannabis drug tests probe for the presence of ∆⁹-THC or the acid metabolite of ∆⁹-THC (e.g., THC-COOH) in one of several biological matrices (e.g., blood, oral fluid (OF), urine, hair). Recently, we reported that acute administration of CBD-dominant cannabis with low levels of ∆⁹-THC (0.39%) can produce a positive drug test in urine (8, 9). More research is needed to discern how unique CBD products can influence drug testing results and the pharmacokinetics of various cannabinoids in blood.

To date, the majority of such pharmacokinetic comparisons for CBD across formulations have been made across studies (reported in systematic reviews or meta-analyses) and have focused primarily on oral formulations (10, 11). Generally, these reviews have found that, compared with oral capsules, oral solutions elicit a slower absorption rate of CBD (measured in blood), and individuals in a fasted state demonstrate slower absorption rates and lower bioavailability of oral CBD compared to individuals under a fed state (12). Given that these comparisons were often made across studies, it is critical to replicate these findings in the same individuals using a within-subject experimental design, as the pharmacokinetics of CBD vary widely across individuals. Moreover, given the growing popularity of vaporization as a route of CBD administration, within-subject comparisons of vaporized formulations of CBD relative to oral formulations are also needed. This is especially important considering recent evidence that vaporized and oral CBD-dominant cannabis products with low levels of ∆⁹-THC may produce positive urine drug test results (8, 13).

Another factor that has been scarcely studied in humans is how the co-administration of ∆⁹-THC may influence the pharmacokinetic properties of CBD. One study found that CBD may partially inhibit hydroxylation of ∆⁹-THC to 11-OH-THC (14, 15) and produce small increases in ∆⁹-THC C_max, suggesting that the two cannabinoids may reciprocally influence metabolism. Importantly, low levels of ∆⁹-THC are permissible under the Farm Bill, and, even when not specified, CBD products can contain ∆⁹-THC (6). Therefore, it is important to understand if the presence of ∆⁹-THC alters the pharmacokinetics of CBD and its metabolites, especially because CBD and ∆⁹-THC are both hepatically metabolized and share a number of enzymatic pathways, which can reciprocally slow absorption, metabolism or elimination of either of these cannabinoids (16, 17).

In this report, we present analyses of ∆⁹-THC and its metabolites in blood following the vaporized and oral administration of CBD products and comment on a critical analytical method issue that compromised interpretation of OF data.

Methods

Participants

Eighteen individuals (nine men; nine women) completed all study procedures not including the fasting condition; demographic and substance use characteristics for study participants can be found in Table I. Participants were recruited for the study using online/print media advertisements and word-of-mouth communication. Volunteers first completed a brief telephone screening followed by a laboratory screening visit where they provided informed consent and underwent procedures to confirm study eligibility. The Institutional Review Board of JHU School of Medicine approved the study, which was conducted in accordance with ethical standards established in the Helsinki Declaration. Participants were compensated for their time after each study visit.

Table I.

Demographics and Substance Use History

ID	Gender	Age	Race	Ethnicity (Hispanic: Y/N)	Height (ft, in)	Weight (lbs)	BMI (kg/m2)	Average # Alcoholic Drinks/Week	Last Cannabis Use (days)	Cigarette smoker (Y/N)	Session Order
038	M	27	White	N	5′8′′	182	28	3	365	N	c, b, d, a
053	F	41	White	Y	5′2′′	194	36	7	32	N	c, b, d, a
054	F	35	Black	N	5′11′′	204	29	1	150	N	a, c, b, d
063	F	31	White	N	5′3′′	128	23	11	36	N	b, d, a, c
066	M	29	White	N	5′6′′	159	26	0	61	N	d, a, c, b
068	M	38	White	N	6′5′′	203	24	0	120	N	d, a, c, b
070	F	29	White	N	5′1′′	132	25	4	52	N	a, c, b, d, e
079	M	23	White	N	6′4′′	188	23	2	74	N	b, d, a, c, e
087	F	27	Black	N	5′10′′	185	27	2	39	N	d, a, c, b
097	F	31	White	N	5′3′′	133	24	8	35	N	c, b, d, a
105	M	36	White	Y	5′4′′	160	28	0	180	N	c, d, a, c, e
106	M	39	White	N	5′11′′	211	29	11	117	N	c, b, d, a, e
112	F	23	White	N	5′4′′	141	24	2	120	N	d, a, c, b, e
113	F	40	White	N	5′6′′	151	24	3	7	N	c, b, d, a, e
114	M	35	Black	N	6′3′′	169	21	1	60	N	b, d, a, c
116	M	24	White	N	5′8′′	130	20	2	62	N	a, c, b, d
118	M	30	White	N	6′3′′	210	27	0	31	N	a, c, b, d
119	F	21	Black	N	5′5′′	160	27	2	1095	N	a, c, b, d

a, placebo; b, oral CBD (capsule, syrup, or Epidiolex); c, vaporized CBD; d, vaporized CBD-dominant cannabis (3.7 mg of THC); e, oral CBD-containing syrup following overnight fasting (optional fifth session).

Principal study inclusion criteria included: 1) good health (determined via medical history, 12-lead electrocardiogram, blood chemistry, hematology, and serology analysis, vital signs and a physical examination); 2) no self-reported use of cannabis or other psychoactive drugs (aside from alcohol, nicotine/tobacco, or caffeine) for one month prior to the screening visit; 3) experience inhaling cannabis; 4) negative urine test for cannabis (<50 ng/mL THC-COOH) and other illicit drugs and negative alcohol breath test (both at the screening visit and on arrival for each study session); 5) 18–45 years old; 6) body mass index (BMI) between 19 and 36 kg/m²; 7) negative pregnancy test (via serum at the screening visit and via urine before each session) or not currently breastfeeding (for females only); and 8) demonstrated ability to expectorate 3–5 mL of “native” OF over a 5 min period at the screening visit. Beyond failure to meet these inclusion criteria, reasons for exclusion included: 1) current use of medications (either prescription or OTC) or other drug products (e.g., herbal supplements) that may have interfered with study outcomes and/or the safety of study participants (e.g., drugs metabolized through CYP2D6, CYP2C9, CYP2C19 and CYP2B10 enzymes or drugs that inhibit CYP3A4 enzymes (16–19); 2) use of dronabinol in the 6 months prior to screening or use of hemp seeds/hemp oil in the 3 months prior to screening; 3) history of xerostomia (dry mouth) or the presence of mucositis, gum infection or bleeding or other significant oral cavity disease or disorder that may have affected the collection of OF samples; 4) presence of anemia; or 5) enrollment in another clinical trial or receiving of any drug as part of a research study within 30 days prior to study intake.

Study design

A within-subject crossover study design was used, and drug administration was double blind and double dummy (i.e., participants inhaled from a vaporizer and orally ingested study drugs or placebo in all test sessions). All participants completed four experimental test sessions and 6 of the 18 participants completed an optional fifth test session. The four primary test sessions were 1) oral CBD (100 mg) with vaporized placebo; 2) oral placebo with vaporized CBD (100 mg); 3) oral placebo with vaporized CBD-dominant cannabis (100 mg of CBD; 3.7 mg of ∆⁹-THC); and 4) oral placebo with vaporized placebo. These drug conditions were administered in a randomized order that was counterbalanced using a modified Latin-square procedure to minimize order effects. A manipulation of formulation was included in the oral CBD dose condition; participants were exposed to 1 of 3 possible drug formulations: 1) six received encapsulated CBD (100 mg); 2) six received 100 mg of CBD suspended in pharmacy-grade syrup; 3) six received 1 mL of Epidiolex (an FDA-approved oral CBD product at 100 mg of CBD/mL of liquid); see the Study Drug section below for further details. An optional fifth experimental condition was completed for six participants in which they fasted from all food for 12 h prior to oral ingestion of 100 mg of CBD in pharmacy-grade syrup and inhalation of vaporized placebo cannabis. This fifth condition always occurred after completion of the first four test sessions.

Experimental session procedures

Study conditions entailed two phases, which occurred over 3 days (58 h total): 1) an acute drug administration session that lasted approximately 8 h and 2) a two-day/three-night stay on a closed residential unit (Johns Hopkins Bayview Clinical Research Unit). Acute drug administration sessions were separated by ≥ 1 week to ensure sufficient washout of the prior study dose.

On Day 1 of each experimental condition, participants came to the Johns Hopkins Behavioral Pharmacology Research Unit at approximately 0730 h. Participants consumed a standard low-fat breakfast consisting of toast and jam (this step was skipped during the optional fifth dosing session conducted after overnight fasting), provided a urine sample to test for recent drug use and pregnancy (for females) and completed an alcohol breathalyzer (these tests all needed to be negative to proceed with the session). In addition, participants had an intravenous catheter inserted into a forearm vein to allow for repeated blood sampling. Participants who completed the test session after overnight fasting stayed on the residential research unit the night before their experimental drug administration session where they were monitored by nursing staff to ensure compliance with fasting requirements. Also, before the start of each test session, participants reported their use of cannabis, alcohol and tobacco since their last study visit, via the Timeline Follow-Back questionnaire (20), and use of any medications, vitamins or dietary supplements. Next, participants completed baseline pharmacodynamic outcomes (e.g., subjective questionnaires and cognitive performance tasks) as described elsewhere (4) and provided baseline biological specimens that included urine, blood and OF; urine results have been previously reported (8, 9):.

After baseline procedures, participants swallowed their oral CBD dose (i.e., capsule, syrup or Epidiolex) or corresponding oral placebo (see Study Drug section). Exactly 1 h after oral dosing, participants inhaled the vaporized study dose, which was either CBD-only, CBD-dominant cannabis or placebo cannabis. Vaporized doses were administered with the Volcano Medic^® (Storz and Bickel, Tuttelingen, Germany). The Volcano Medic vaporizer heated each study dose at 204°C (400°F) and the resulting aerosol (i.e., “vapor”) was captured in a balloon. Participants inhaled this vapor using a one-way valve attached to the balloon; participants were given 10 min to inhale three full balloons ad libitum. New balloons were used for each session/participant to prevent contamination from prior drug doses. Balloons were covered with an opaque bag to reduce the visibility of vapor and preserve the drug blind for participants and study staff. After drug administration, participants continued to provide biospecimens and complete pharmacodynamic measures for 8 h. After the conclusion of the drug administration session, participants were escorted to the nearby residential research unit where they resided for the next two days (up to 58 h post-drug administration).

Study drug

Two batches of cannabis (one active, one placebo) were obtained for the study from the National Institute on Drug Abuse (NIDA) Drug Supply Program. The active batch was a CBD-dominant cannabis chemotype that contained: 10.5% CBD, 0.39% ∆⁹-THC, 0.02% ∆⁸-THC and 0.05% cannabinol (CBN); all percentages based on dry weight. The placebo batch of cannabis contained 0.001% ∆⁹-THC, 0.003% CBD and 0.005% CBN and had no detectable ∆⁸-THC. To assist with drug blinding, the same amount of cannabis (953 mg) was used in placebo and active sessions; this quantity was selected to yield a total CBD dose of 100 mg in the CBD-dominant cannabis condition. Placebo and active cannabis were administered via vaporization with the Volcano Medic (see Session Procedures).

CBD (a crystalline powder) was obtained from Albany Molecular Research, Inc. (AMRI) for this study (∆⁹-THC was not detected in this product by independent testing). In the CBD-only vapor dose condition, 100 mg of CBD powder was placed on a steel wool dosing pad, heated and vaporized by the Volcano Medic, and participants inhaled the resulting vapor using the procedures described above for placebo/active cannabis. The CBD powder was also used to make 2 of the 3 oral drug formulations. Six of 18 participants ingested gelcaps (size 00) that contained 100 mg of CBD and filled with inert cellulose; for placebo oral conditions, these six participants ingested identical capsules, which were only filled with cellulose. For the next set of six participants, 100 mg of CBD powder was suspended in a pharmacy-grade cherry syrup (10 mL) that participants orally ingested; for placebo oral conditions, these six participants ingested a solution of an equal volume which only contained the pharmacy-grade cherry syrup. The remaining six participants ingested 1 mL (i.e., 100 mg of CBD) of the FDA-approved oral CBD product Epidiolex (GW Pharmaceuticals, Greenwich, England), which was mixed with 9 mL of the aforementioned pharmacy-grade cherry-flavored syrup to preserve the blind of study medications (placebo Epidiolex was not available for this study). Placebo oral conditions for these six participants was oral ingestion of 10 mL of pharmacy-grade cherry syrup only. Rationale for dose selection is described in our prior reports on this study (4, 8).

Outcome measures

Blood and OF specimens were collected at baseline (before drug administration) and 0.5, 1, 1.5, 2, 3, 4, 5, 6, 8, 12, 22, 26, 30, 34, 46, 50, 54 and 58 h post-oral drug. Because oral and vaporized dosing were staggered by 1 h, the timeline of sample collection differed by 1 h. between routes of administration: for vaporized doses, samples were collected at baseline and 0 (immediately after), 0.5, 1, 2, 3, 4, 5, 7, 11, 21, 25, 29, 33, 45, 49, 53 and 57 h post-inhalation of vaporized study drugs. Whole blood specimens (∼ 6 mL each) were collected, mixed by inversion and stored at −80°C until shipped on dry ice for quantitative analysis by the Immunalysis Corporation (Pomona, CA). Both qualitative (ELISA) and quantitative (LC–MS-MS) methods were used to analyze blood and OF samples.

The Cannabinoids Direct ELISA Kit (THC-A/C-THC) was used for qualitative analysis of THC-COOH; the manufacturer-recommended cutoff concentration of 10 ng/mL was used. Cross-reactivity for this assay (as described by the manufacturer) was THC-COOH (100%); 11-nor-9-carboxy-∆⁸-THC (110%); ∆⁹-THC (21%); ∆⁸-THC (45%); 11-OH-THC (<5%); 8–11-dihydroxy-∆⁹-THC (<5%); CBN (<5%); and CBD (<5%). Quantitative whole blood concentrations of ∆⁹-THC, ∆⁸-THC, 11-OH-THC, THC-COOH, CBN, CBD and 7-COOH-CBD were determined for all participants using liquid chromatography with tandem mass spectral detection (LC–MS-MS) (21). A solid-phase extraction technique was used for these analyses followed by MS detection in positive electrospray ionization mode. The limit of quantitation (LOQ) was 0.5 ng/mL for ∆⁹-THC, 2 ng/mL for 11-OH-THC and 1 ng/mL for ∆⁸-THC, CBD, 7-COOH-CBD and CBN. The upper limit of linearity (ULOL) was 100 ng/mL.

For OF collection, participants expectorated into 8 mL glass, screwcap culture tubes (Thermo Fisher Scientific, Waltham, MA, 16 × 100 mm, #14-959-35AA). These tubes contained a PTFE-liner (Thermo Fisher Scientific, #4506615), and before being used for specimen collection in the study, they were salinized with Sylon-CT^TM (Sigma-Aldrich, St Louis, MO, USA, #33065-U), rinsed with methanol and dried. Participants were asked to produce each sample within a 5 min period and they were not permitted to consume food or drink for at least 10 min before providing each specimen. After collection, the capped OF samples were wrapped with para-film and stored in a refrigerator until shipped overnight (in refrigerated containers) for analysis.

The Saliva/OF Cannabinoids Direct ELISA Kit was used for qualitative analysis of ∆⁹-THC in OF; the manufacturer-recommended cutoff concentration of 4 ng/mL was used. Cross-reactivities for this assay (as described by the manufacturer) were as follows: ∆⁹-THC (100%); ∆⁸-THC (66.7%); CBN (4%); CBD (50%); and conjugated-THC (25%). The Ultra-Sensitive Cannabinoids Direct ELISA Kit was used for qualitative analysis of THC-COOH for blood testing; a manufacturer-recommended cutoff concentration of 0.05 ng/mL was used. Cross-reactivities for this assay (as described by the manufacturer) were as follows: THC-COOH (100%); 11-nor-9-carboxy-∆⁸-THC (125%); ∆⁹-THC (10%); 11-OH-THC (33%); CBN (<0.25%); and CBD (<0.25%). Quantitative OF concentrations of CBD, ∆⁹-THC, ∆⁸-THC and THC-COOH were determined for all participants using LC–MS-MS analysis (see (19) for details). For ∆⁹-THC, the LOQ was 1 ng/mL and the ULOL was 100 ng/mL. For THC-COOH, the LOQ was 0.02 ng/mL and the ULOL was 0.1 ng/mL.

Data presentation and analysis

Data are largely presented using descriptive statistics to show the relative differences in analytes across formulation and time points.

Pharmacokinetic outcomes (C_max, T_max) for analytes in whole blood (CBD, 7-COOH-CBD and ∆⁹-THC) were compared between male and female participants while controlling for weight using ANOVAs for each formulation. All analytes were evaluated for the oral and vaporized CBD-only formulations and the CBD-dominant cannabis formulation. Those same analytes were compared across all time points between male and female participants using repeated-measure ANOVAs for each formulation. Statistically significant differences are determined by P < 0.05. The data underlying this article will be shared on reasonable request to the corresponding author.

Results

All baseline whole blood samples tested negative for CBD, 7-COOH-CBD, ∆⁹-THC and several ∆⁹-THC metabolites by LC–MS-MS, suggesting compliance with cannabis abstinence requirements in between experimental test sessions. Supplemental Table I displays the full ELISA and LC–MS-MS results for each participant at all time points. Note that ∆⁸-THC, 11-OH-THC and CBN were measured, but are not included in the table for parsimony because only a few sporadic samples for the entire study were positive and concentrations were ≤ 1 ng/mL. Unless otherwise noted, oral CBD data include all three formulations (i.e., Epidiolex, capsule and syrup) collapsed together, but excludes the overnight fasting condition.

During this protocol we collected blood and OF samples to test the pharmacokinetics of CBD, ∆⁹-THC and several ∆⁹-THC metabolites. Unfortunately, it was recognized post hoc that the exposure of OF samples to an acidic buffer during extraction produced up to 6% conversion of CBD to ∆⁹-THC and up to 1.1% conversion to ∆⁸-THC (22). Subsequent experiments determined that the small batch sizes used for validation experiments showed no conversion but larger batch sizes and longer extraction times used for production samples, resulted in more conversion as a consequence of longer exposure to acid for some samples. Because of this, we only present individual-level OF data as supplemental material to this report (Supplemental Table II). No additional analyses or results are included on the OF data.

CBD detection by route of administration

Vaporized CBD and CBD-dominant cannabis produced greater mean peak concentrations (C_max) of CBD in whole blood compared with orally-administered CBD (Figure 1, Panel A). On average, peak concentrations of CBD were observed in blood immediately after vaporization and 4 hours after oral ingestion. Vaporized CBD-dominant cannabis produced the largest C_max for CBD, including when compared with the same dose of vaporized CBD (Table II). The window of detection for CBD in blood was shortest following oral administration and longest following vaporized CBD-dominant cannabis (Table III).

Figure 1.

Quantitative blood concentrations (ng/mL, mean + SEM) of CBD, 7-COOH-CBD and ∆⁹-THC and THC-COOH across routes of administration. Note that data are collapsed across oral formulation conditions.

Table II.

Mean ∆⁹-THC, THC-COOH, CBD and 7-COOH-CBD Blood Maximum Concentration (C_max), Time to Maximum Concentration (T_max) and Individual Ranges after Administration of 100 mg of CBD across Three Formulations

	CBD (ng/mL)		7-COOH-CBD		∆9-THC		THC-COOH
	C max	T max	C max	T max	C max	T max	C max	T max
Oral	13.7 (0.0–50.0)	2.9 (1.0–5.0)	108.9 (2.0–624.0)	4.5 (1.0–26.0)	0.1 (ND-2.0)a	0.2 (ND-0.2)a	ND	ND
Vape	104.6 (19.0–312.0)	0.0 (0.0–0.0)	15.2 (0.0–68.0)	1.3 (0.0–5.0)	0.1 (ND-2.0)a	0.2 (ND-0.2)a	ND	ND
Vape + THC	171.1 (40.0–665.0)	0.1 (0.0–0.5)	22.4 (0.0–68.0)	0.7 (0.0–2.0)	5.9 (ND-32.0)	0.2 (ND-0.5)	0.8 (ND-4)	0.66 (ND-2.0)

Note: Specimens were collected at 0.5, 1, 1.5, 2, 3, 4, 5, 6, 8, 12, 22, 26, 30, 34, 46, 50, 54 and 58 h post-oral active study drug and 0 (immediately after), 0.5, 1, 2, 3, 4, 5, 7, 11, 21, 25, 39, 33, 45, 49, 53 and 57 h post-inhalation of vaporized active study drugs. Oral data is collapsed across the three formulations (capsule, Epidiolex, syrup). “Vape” refers to CBD-only condition, while “Vape+THC” refers to CBD-dominant cannabis condition. The limit of quantitation (LOQ) was 0.5 ng/mL for ∆⁹-THC and 1 ng/mL for CBD, 7-COOH-CBD and THC-COOH.

^aOne blood sample collected at a single timepoint evidenced ≤ 2 ng/mL of ∆⁹-TH.

Table III.

Mean (Range) Time at First and Last Detection for CBD and 7-COOH-CBD LC–MS-MS Testing in Blood across Formulations

	CBD (ng/mL)		7-COOH-CBD (ng/mL)
	LC–MS-MS		LC–MS-MS
	Time at First Detection	Time at Last Detection	Time at First Detection	Time at Last Detection
Oral 100 mg of CBD	1.5 (0.5–3.0)	9.7 (0.5–30.0)	0.5 (0.5–12.0)	47.9 (5.0–58.0)
Vape 100 mg of CBD	0.0 (0.0–0.0)	9.9 (2.0–53.0)	0.3 (0.0–4.0)	51.1 (7.0–57.0)
Vape + THC 100 mg of CBD, 3.7 mg of THC	0.1 (0.0–0.5)	16.8 (1.0–57.0)	0.1 (0.0–0.5)	48.2 (0.1–57.0)

Note: Specimens were collected at 0.5, 1, 1.5, 2, 3, 4, 5, 6, 8, 12, 22, 26, 30, 34, 46, 50, 54 and 58 h post-oral active study drug and 0 (immediately after), 0.5, 1, 2, 3, 4, 5, 7, 11, 21, 25, 39, 33, 45, 49, 53 and 57 h post-inhalation of vaporized active study drugs. LC–MS-MS = liquid chromatography with tandem mass spectrometry. The limit of quantitation (LOQ) was 0.5 ng/mL for ∆⁹-THC and 1 ng/mL for CBD, 7-COOH-CBD and THC-COOH and the upper limit of linearity (ULOL) was 100 ng/mL. Oral data are collapsed across the 3 formulations (capsule, Epidiolex, syrup). “Vape” refers to CBD-only condition, while “Vape+THC” refers to CBD-dominant cannabis condition.

7-COOH-CBD detection by route of administration

Oral CBD produced the largest 7-COOH-CBD C_max and the most delayed T_max compared with vaporized CBD-dominant cannabis and vaporized CBD-only (Table II). Vaporized CBD-dominant cannabis produced slightly greater concentrations of 7-COOH-CBD in blood compared to the same dose of the vaporized CBD-only formulation across all timepoints (Figure 1, Panel B). On average, 7-COOH-CBD was detected in blood 2 days after the initial use across both routes of administration (i.e., vaporized and oral). Following oral CBD administration, 7-COOH-CBD remained detectable in blood for 14/18 participants at the final testing time point (58 h post-administration) (Table III).

∆⁹-THC detection by route of administration

∆⁹-THC was detected in blood most frequently following inhalation of CBD-dominant cannabis. For CBD-dominant cannabis, blood ∆⁹-THC concentrations peaked immediately after administration (all samples < 4 ng/mL), but ∆⁹-THC was no longer detectable 3 h after administration for any participant. ∆⁹-THC was not detected in any blood specimen following acute administration of CBD-dominant cannabis (3.7 mg of ∆⁹-THC) for 2 of 18 study participants. This is consistent with prior studies in which a subset of participants had no detectable ∆⁹-THC nor its metabolites in blood after low doses of oral ∆⁹-THC (23, 24).

∆⁹-THC was detected following both oral and vaporized CBD in a single participant (Supplemental Table I, ID: 97). In both conditions, ∆⁹-THC was detected at ≤ 2 ng/mL at isolated, non-systematic, timepoints after exposure, which suggests that this was residual ∆⁹-THC from prior exposure rather than the result of in vivo conversion of CBD to ∆⁹-THC.

CBD detection by oral formulation

As described above, three different oral formulations (each containing 100 mg of CBD) were administered: capsule, Epidiolex and syrup. CBD and 7-COOH-CBD blood concentrations across the three oral formulations are depicted in Figure 2. Epidiolex produced the greatest CBD C_max relative to capsule and syrup (Table IV). Across the three formulations, CBD peaked approximately 3 h after administration. On average, CBD was detected for the shortest period after administration of the pharmacy-grade syrup (Table IV).

Figure 2.

Quantitative blood concentration (ng/mL, mean + SEM) of CBD and 7-CBD-COOH across oral formulations (six participants per group).

Table IV.

Mean (Range) C_max, T_max and Time to First and Last Detection for CBD and 7-COOH-CBD LC–MS-MS Testing after Administration of 100 mg of CBD across Oral Formulations

	CBD (ng/mL)				7-COOH-CBD (ng/mL)
	LC–MS-MS				LC–MS-MS
	C max	T max	Time to First Detection	Time to Last Detection	C max	T max	Time to First Detection	Time to Last Detection
Capsule n = 6	17.8 (2.0–50.0)	2.5 (1.0–4.0)	1.3 (1.0–3.0)	16.8 (8.0–34.0)	36.1 (2.0–159.0)	6.0 (1.0-9.0)	2.3 (0.5–4.0)	29.2 (5.0–58.0)
Epidiolex n = 6	20.5 (8.0–37.0)	3.3 (1.0–5.0)	1.7 (0.5–3.0)	13.0 (8.0–30.0)	249.0 (81.0–624.0)	4.3 (3.0–6.0)	0.8 (0.5–1.5)	58.0 (58.0–58.0)
Syrup n = 6	2.8 (0.0–7.0)	3.2 (1.0–4.0)	1.4 (0.5–3.0)	6.4 (0.5–12.0)	62.2 (12.0–189.0)	9.3 (3.0–34.0)	0.9 (0.5–1.5)	58.0 (58.0–58.0)

Note: Specimens were collected at 0.5, 1, 1.5, 2, 3, 4, 5, 6, 8, 12, 22, 26, 30, 34, 46, 50, 54 and 58 h post-oral active study drug and 0 (immediately after), 0.5, 1, 2, 3, 4, 5, 7, 11, 21, 25, 29, 33, 45, 49, 53 and 57 h post-inhalation of vaporized active study drugs. LC–MS-MS = liquid chromatography with tandem mass spectrometry. The limit of quantitation (LOQ) was 0.5 ng/mL for ∆⁹-THC and 1 ng/mL for CBD, 7-COOH-CBD and THC-COOH and the upper limit of linearity (ULOL) was 100 ng/mL.

7-COOH-CBD was tested and compared across all three oral formulations. Epidiolex produced greater blood 7-COOH-CBD C_max relative to capsule and syrup. Epidiolex produced the greatest C_max and earliest T_max of the three conditions for 7-COOH-CBD (Table IV). Interestingly, although CBD concentrations were greater after encapsulated CBD compared with CBD in pharmacy-grade syrup, the opposite was true for concentrations of the 7-COOH-CBD metabolite.

Fasting conditions

Six participants in this study completed an optional fifth test session in which they were orally administered 100 mg of CBD in pharmacy-grade syrup after fasting from all food for ≥ 12 h. CBD blood C_max was greater after participants ate a low-fat breakfast compared with fasting conditions (Figure 3).

Figure 3.

Quantitative blood concentration (ng/mL, mean + SEM) of CBD under fasting and non-fasting conditions before and after administration of CBD syrup (six total participants per group).

Sex differences

We analyzed pharmacokinetic sex differences for each formulation while controlling for weight. C_max and T_max did not differ as a function of sex across all analytes and formulations. Additionally, CBD and ∆⁹-THC across timepoints did not differ as a function of sex nor were there significant sex X time interactions for any formulation.

Oral fluid results

All baseline OF samples tested negative for CBD, CBN, ∆⁸-THC and ∆⁹-THC by LC–MS-MS, confirming cannabis abstinence requirements. The full ELISA and LC–MS-MS results for each participant are detailed in Supplemental Table II (note that THC-COOH and CBN were detected at sporadic concentrations <LOQ and are not listed). No additional analyses were performed using OF results because methods of extraction converted some CBD to ∆⁹- and ∆⁸-THC as described in the Methods section.

Discussion

This report describes notable variations in the pharmacokinetics of CBD, 7-COOH-CBD and ∆⁹-THC in blood that have important implications for dosing and drug testing. First, there are dietary and oral formulation factors, which influence CBD concentrations. Second, this work demonstrated the utility of the CBD metabolite, 7-COOH-CBD, to be a marker for CBD product exposure given its substantially longer window of detection relative to CBD. Third, we did not find any evidence suggesting that CBD converts to ∆⁹-THC in vivo due to acidic gastric environments. ∆⁹-THC was only detected at an isolated timepoint following ingestion and vaporization of CBD only (<2 ng/mL), which suggests that this was residual ∆⁹-THC from prior exposure rather than the result of in vivo conversion of CBD to ∆⁹-THC. However, inhalation of vaporized CBD-dominant cannabis containing 0.39% ∆⁹-THC produced detectable concentrations of ∆⁹-THC in blood for most study participants. Finally, exposure of OF samples containing CBD to an acidic reagent can lead to the in vitro formation of ∆⁹- and ∆⁸-THC. Similar observations have been reported for assays utilizing acidic derivatizing reagents (25). Therefore, ∆⁹-THC positive results can be an artifact of the testing method used. Together, these data provide important, novel information related to drug testing methods and interpretation, as well as product-related differences with respect to CBD dosing.

We observed great variability in cannabinoid pharmacokinetics across the three oral formulations. Epidiolex produced the greatest peak concentrations of CBD in blood, and the pharmacy-grade syrup containing CBD produced the lowest CBD concentration of the three different oral formulations tested. Previous studies have indicated that absorption is slowest among oral solutions compared to oral capsules (10, 11, 26). Those studies often included much larger doses (1000–6000 mg), which could be one reason our results partially differed. Differences in solution properties may explain the differences observed across the oral formulations (i.e., the lipophilicity or alcohol content of each solution). For example, preclinical studies have demonstrated that lipid-based solutions are associated with higher CBD levels relative to non-lipid solutions (27, 28). This finding is especially relevant as a growing number of liquid CBD products are available (e.g., sodas, teas, oils, tinctures). Future research should systematically study how different CBD solution vehicles and other product characteristics (e.g., temperature, vehicle, carbonation, use of nanotechnology) impact the pharmacokinetics of CBD.

Consistent with previous findings (11, 29), diet also impacted the concentrations of CBD in blood. Overnight fasting prior to acute oral CBD dosing resulted in notably lower concentrations of CBD in blood compared with conditions in which participants consumed a low-fat breakfast. These findings support providing education to individuals who are taking CBD with medicinal intent given that dietary factors significantly influence dose delivery. The CBD metabolite, 7-COOH-CBD, was present days after acute exposure of vaporized and oral CBD products. On average, 7-COOH-CBD was detected in blood 53–58 h after administration, depending on route of administration. This is in comparison to CBD, which was only detected in blood on average 9–18 h after acute exposure. These data suggest that CBD metabolism occurs over several days, and this means that the PK of other drugs may be impacted during this prolonged period, particularly those that share similar hepatic pathways. Future studies are needed to characterize the drug-drug interactions of CBD, 7-COOH-CBD and other metabolites.

These data should be considered in light of some limitations. First, this study only included one dose of CBD (100 mg) and ∆⁹-THC (3.7 mg) and two routes of administration (vaporized and oral). Given the wide range of products and preparations available, these data only reflect a small subset of the products currently being used. Second, these data were collected after acute dosing among individuals who were infrequent cannabis and/or CBD product users. Future studies should examine the pharmacokinetics of CBD and CBD-dominant cannabis exposure under chronic dosing conditions. Third, the sample sizes for each oral formulation condition were relatively small (n = 6 for each formulation). Despite this, there were notable differences in pharmacokinetics across oral dose formulations, which warrants further study, especially given the increasing prevalence and diversity of oral cannabis/CBD products. The sample size also limits any conclusions that can be made about sex differences on cannabinoid pharmacokinetics. Prior work has demonstrated that peak concentrations for 11-OH-THC (a primary metabolite of THC) is greater in females compared to males following administration of ∆⁹-THC-dominant cannabis (30). Additional studies in larger samples are needed to determine whether robust sex differences in CBD metabolism exist, and, if so, whether those differences are associated with clinically relevant outcomes. Finally, this study did not quantify ∆⁸-THC-COOH, which may interfere with quantification of ∆⁹-THC-COOH. Because the source cannabis used in this study contained only trace amounts of ∆⁸-THC (0.017%; 23 times lower concentration than ∆⁹-THC), it is unlikely that significant interference occurred.

In summary, this study examined within-subject variations in the pharmacokinetics of CBD products under acute oral and vaporized dose conditions. Furthermore, this study reiterated that exposure of biological specimens to an acidic environment can result in CBD conversion to ∆⁹-and ∆⁸-THC. We also demonstrated that CBD bioavailability is significantly impacted by route of administration, formulation and dietary condition.

Supplementary Data

Supplementary Data is available at Journal of Analytical Toxicology online.

Funding

This research was supported by the Substance Abuse and Mental Health Services Administration (SAMHSA) and the National Institute on Drug Abuse (NIDA; T32DA07209).