Cannabidiol: Pharmacology and Therapeutic Targets

Abstract

Cannabidiol (CBD) products lacking regulatory approval are being used to self-treat a myriad of conditions and for its unsubstantiated health benefits. The scientific evidence supporting these claims largely arises not from controlled clinical trials, but from the recognition that CBD has numerous biological targets. Yet, CBD is commonly consumed and often in over-the-counter products that are unapproved and of unknown composition. Epidiolex® is the only product that has undergone rigorous pharmacokinetic assessment and testing in clinical trials; it was approved as a non-scheduled drug by the U.S. Food and Drug Administration for the treatment of intractable childhood-onset seizures. However, studies investigating CBD for other medical conditions are limited in number and often lack the scientific rigor, controls, or sample sizes required to draw clinically meaningful conclusions. Although Epidiolex® is safe for human consumption, recent changes in regulation of commercially available CBD products has resulted in limited quality control and products marketed with unknown CBD bioavailability. Even scientifically rigorous studies have used different sources of CBD and different suspension vehicles for administration, making it difficult to compare results among studies and resolve mixed outcomes. This paper reviews the molecular targets, pharmacokinetics, and safety and abuse liability of CBD; additionally, the extant evidence on its potential therapeutic effects for neurological disorders, pain, inflammation, conditions related to immune function, psychiatric disorders, and substance use are described.

Introduction

The phytocannabinoid cannabidiol (CBD) is a 21-carbon terpenophenolic compound with numerous molecular targets. Evidence that CBD has therapeutic promise largely stems from preclinical cellular and rodent studies, which suggest that CBD could be neuroprotective, cardioprotective, and anti-inflammatory. However, few highly controlled clinical trials investigating CBD have been conducted to elucidate its therapeutic potential. Moreover, many studies have been conducted with CBD preparations that have no regulatory approval. Epidiolex® (GW Pharmaceuticals) is the only marketed CBD monotherapy with U.S. Food and Drug Administration (FDA) approval. This review discusses CBD’s molecular targets, its pharmacokinetic profile, evaluates its therapeutic potential, and highlights concerns regarding unregulated over-the-counter products. Only human studies (e.g., clinical trials and laboratory studies) investigating CBD alone (across various preparations and routes of administration [e.g., oral, inhaled/vaporized, topical, and sublingual]) are included; reports on nabiximols, such as Sativex® or cannabis material containing THC, are generally excluded. Additionally, review of the therapeutic evidence focuses largely on controlled clinical trials. See Pisanti et al. (2017) for a thorough review of the preclinical literature.

Evidence for CBD Molecular Targets

The two primary cannabinoid receptors [cannabinoid 1 (CB1) and cannabinoid 2 (CB2)] are G_i/o-coupled protein receptors. CB1 receptors are located throughout the central nervous system (Herkenham et al. 1991), but have also been found in cardiac, lung, small intestine, kidney, and liver tissues (Buchholz et al. 2017), and on immune cells (Galiegue et al. 1995). In contrast, CB2 receptors are primarily located on immune cells (Nunez et al. 2004), in the gastrointestinal tract (Galiazzo et al. 2018), and in low densities in the central nervous system (Van Sickle et al. 2005). Evidence suggests that CBD activity at cannabinoid receptors is limited. A systematic review concluded that CBD effects at CB1 receptors are primarily due to indirect effects (i.e., no direct interaction with the orthosteric CB1 receptor binding site) (McPartland et al. 2015). One proposed mechanism of indirect CBD action at CB1 receptors is negative allosteric modulation, which has been reported in several in vitro studies (Chung et al. 2019; Laprairie et al. 2015; Tham et al. 2019). CBD inhibition of fatty acid amide hydrolase (FAAH) with increased anandamide has also been reported (Bisogno et al. 2001; De Petrocellis et al. 2011) – this is another proposed indirect mechanism of action. However, another study reported CBD activation of FAAH (Massi et al. 2008), and these cross-study inconsistencies have been attributed to differences in in vitro physiological test environments (Bih et al. 2015). With regard to CB2 receptors, CBD was reported to act as a low affinity agonist in receptor binding preparations (for review see McPartland et al., 2015; McPartland et al. 2015). In vivo studies support a potential role for CB2 as both CBD-induced reductions in cocaine self-administration (Galaj et al. 2019) and CBD’s anti-seizure effects (Vilela et al. 2017) were blocked by CB2 antagonist pretreatment.

Although direct effects of CBD on cannabinoid receptors appear limited, over 65 molecular targets for CBD have been identified, including transient receptor potential vanilloid (TRPV) channels and serotonin (5-HT_1A) receptors, which have the most supporting evidence (Bih et al. 2015) and are at least partially responsible for CBD’s pharmacodynamic effects (Campos et al. 2012; Li et al. 2020; Nichols and Kaplan 2020; Soares and Campos 2017). Multiple studies have demonstrated that CBD acts as a low-potency, full agonist at TRPV1 and causes rapid desensitization of TRPV1 (Bisogno et al. 2001; De Petrocellis et al. 2011; Iannotti et al. 2014; Iannotti et al. 2019; Ligresti et al. 2006) (but also see Qin et al. 2008). In vivo studies have reported blockade of CBD effects by TRPV1 antagonists, including reductions in cocaine self-administration (Galaj et al. 2019), anti-seizure effects (Vilela et al. 2017), decreases in heart rate (in anesthetized rodents) (Kossakowski et al. 2019), and anti-inflammatory effects (Couch et al. 2017; Petrosino et al. 2018). CBD has also been shown to activate other TRPV receptors, including TRPV2, TRPV3, and TRPV4 (De Petrocellis et al. 2011; Morelli et al. 2014; Nabissi et al. 2015; Nabissi et al. 2013; Qin et al. 2008). Together, these findings suggest a role for TRPV receptors, particularly TRPV1, in mediating several potential therapeutic effects of CBD, such as neuroprotection and anti-convulsant effects (Gray and Whalley 2020; Lazarini-Lopes et al. 2020), anti-psychotic effects (Campos et al. 2012), and immunomodulatory effects (Nichols and Kaplan 2020).

CBD is an agonist at 5-HT_1A receptors both in vitro (Russo et al. 2005; Yang et al. 2010) and in vivo (Alves et al. 2010; Galaj et al. 2019; Gomes et al. 2013; Gomes et al. 2012; Gomes et al. 2011; Hartmann et al. 2019; Resstel et al. 2009; Sartim et al. 2016; Soares et al. 2010; Sonego et al. 2016; Zanelati et al. 2010). In vivo, the 5-HT_1A antagonist WAY100635 blocked CBD-induced panicolytic effects (Soares et al. 2010), antidepressant-like effects (Sartim et al. 2016; Zanelati et al. 2010), reversal of haloperidol-induced catalepsy (Sonego et al. 2016), anti-aggression (Hartmann et al. 2019), reductions in cocaine self-administration (Galaj et al. 2019), and reductions in autonomic stress responses (Resstel et al. 2009). WAY100635 also blocked CBD-induced anxiolytic effects (Gomes et al. 2011), stress-associated cardiovascular effects (Gomes et al. 2013), fear-associated freezing behavior (Gomes et al. 2012), and changes in baroreflex activity (Alves et al. 2010) when CBD was microinjected into the bed nucleus of the stria terminalis. These rodent studies suggest that many of CBD’s behavioral effects are due to actions at 5HT_1A and that CBD may potentially be therapeutic for certain psychiatric disorders (Campos et al. 2012; Schier et al. 2012; Soares and Campos 2017). While these preclinical findings are exciting, rigorous clinical trials of CBD for psychiatric disorders are needed.

While there is more evidence supporting a role for TRPV channels and 5-HT receptors in CBD’s mechanism of action, there is an emerging literature suggesting a multitude of other potential targets, including, but not limited to, adenosine, G-coupled protein receptor (GPR) 55, GPR18, GPR119, proliferator-activated receptor alpha, and glycine receptors (Atalay et al. 2019). Due, in part, to the multiplicity of CBD molecular targets (for review see Bih et al. 2015; Lee et al. 2017), the speculation regarding its therapeutic potential has been broad and includes applications for pain, inflammation, and psychiatric disorders, among others. However, evidence in support of its efficacy for these conditions is quite limited and only the efficacy of Epidiolex® for the treatment of epilepsy has been rigorously tested in humans resulting in FDA approval (see Therapeutic Implications section below for a review of clinical trials).

Pharmacokinetics

While numerous studies have investigated the pharmacokinetics of CBD-THC combinations (for review see Millar et al. 2018), few have examined CBD alone (Table 1). CBD pharmacokinetics have been investigated using oral (Birnbaum et al. 2019; Consroe et al. 1991a; Devinsky et al. 2018b; Haney et al. 2016; Schoedel et al. 2018; Taylor et al. 2019; Taylor et al. 2018; Wheless et al. 2019), sublingual (Guy and Flint 2004), intravenous, and smoked administration methods (Ohlsson et al. 1986) (Table 1) in humans; however, only four used Epidiolex® (Devinsky et al. 2018b; Schoedel et al. 2018; Taylor et al. 2019; Taylor et al. 2018), the only approved formulation with data ensuring consistent product concentration and stability.

Table 1.

Pharmacokinetic studies of cannabidiol (CBD) in humans.

Reference	Population	Route	Dose	Matrix	Analytical Method	Tmax	Cmax	t1/2
Birnbaum et al. (2019)	Adults w/ refractory epilepsy n=8	Oral (Viero Health Violet Formulation)	200–300 mg	Plasma	Negative-ion mode electro-spray-ionization liquid chromatography-tandem mass spectrometry	Fed: Mean: 2.4 h Range: 1–6 h Fasting: Mean: 3.2 h Range: 2–5 h	Fed: 0.45 (ng/ml)/mg Fasting: 0.03 (ng/ml)/mg	Fed: 24.3 h Fasting: 38.9 h
Consroe et al. (1991a)	Huntington’s patients n=14	Oral (National Institute on Drug Abuse)	10 mg/kg daily for 6 weeks	Plasma	Capillary gas chromatography/ion trap mass spectroscopy	--	--	Mean: 68.2 h Range: 41.4–113 h
Devinsky et al. (2018b)	4–10 year old children w/ Dravet syndrome n=34	Oral solution (Epidiolex®)	1.5 mg/kg once	Plasma	Ultra-performance liquid chromatography with tandem mass spectrometry	2.5 h	29.3–37.6 ng/ml	--
			5 mg/kg daily			Day 22: 2.5 h	Day 22: 130 ng/ml
			10 mg/kg daily			Day 22: 5 h	Day 22: 288 ng/ml
			20 mg/kg daily			Day 22: 2.5 h	Day 22: 380 ng/ml
Guy and Flint (2004)	Healthy adults w/ a history of cannabis use n=12	Sublingual (GW-3009–01)	20 mg	Plasma	--	2.17 h	2.05 ng/ml	--
Haney et al. (2016)	Adult regular cannabis users n=8	Oral (STI Pharmaceuticals)	800 mg	Plasma	Liquid/liquid extraction, derivatization, and gas chromatography-tandem mass spectrometry	Mean: 3 h Range: 2–6 h	Mean: 77.9 ng/ml Range: 1.6–271.9 ng/ml	--
Ohlsson et al. (1986)	Adult men w/ a history of cannabis use n=5	Intravenous	20 mg	Plasma	Gas chromatography mass spectrometry	--	Mean: 686 SD: 239	Mean: 24 h
		Smoking	19 mg				Mean: 110 SD: 55	Mean: 31 h
Schoedel et al. (2018)	Adult polydrug users n=41	Oral solution (Epidiolex®)	750 mg 1500 mg 4500 mg	Plasma	--	5.11 h	336.2 ng/ml	--
						6.13 h	524.5 ng/ml
						4.07 h	426.9 ng/ml
Taylor et al. (2018)	Healthy adults n=6/dose	Oral solution (Epidiolex®)	1500 mg	Plasma	High-performance liquid chromatography with tandem mass spectrometry	Median: 4 h Range: 3–5 h	Mean: 292.4 ng/ml	Mean: 14.43 h
			3000 mg			Median: 5 h Range: 3–5 h	Mean: 533.0 ng/ml	Mean: 14.39 h
			4500 mg			Median: 5 h Range: 5–5 h	Mean: 722.1 ng/ml	Mean: 16.61 h
			6000 mg			Median: 5 h Range: 3–5 h	Mean: 782.0 ng/ml	Mean: 15.42 h
	n=9/dose		750 mg twice daily for 6 days			Day 1 AM Median: 5 h Range: 2.5–5 h	Day 1 AM Mean: 290.8 ng/ml
						Day 7 AM Median: 3 h Range: 2.5–5 h	Day 7 AM Mean: 330.3 ng/ml	Day 7 AM Mean: 56.41 h
	n=12/dose		1500 mg twice daily for 6 days			Day 1 AM Median: 5 h Range: 2.5–5 h	Day 1 AM Mean: 361.8 ng/ml
						Day 7 AM Median: 3 h Range: 2–4 h	Day 7 AM Mean: 541.2 ng/ml	Day 7 AM Mean: 60.54 h
			1500 mg			Fed: Mean: 3 h Range: 1.5–5 h	Fed: 1628 ng/ml	Fed: 24.4 h
						Fasting: Mean: 3.5 h Range: 2.5–5.03 h	Fasting: 335.4 ng/ml	Fasting: 30.33 h
Taylor et al. (2019)	Adults w/ mild to severe hepatic impairment vs. healthy controls	Oral solution (Epidiolex®)	200 mg	Plasma	Liquid chromatography with tandem mass spectrometry
	Healthy adults n=8					Mean: 2.3 h	Mean: 148 ng/ml	Mean: 8.58 h
	Mild hepatic impairment n=8					Mean: 2.8 h	Mean: 233 ng/ml	Mean: 15.7 h
	Moderate hepatic impairment n=8					Mean: 2.0 h	Mean: 354 ng/ml	Mean: 20.5 h
	Severe hepatic impairment n=6					Mean: 2.5 h	Mean: 381 ng/ml	Mean: 22.1 h
Wheless et al. (2019)	Children w/ treatment-resistant epilepsy n=20	Oral solution (INSYS Manufacturing LLC)	5 mg/kg	Plasma	High-performance liquid chromatography with tandem mass spectrometry	Median: 2.6 h Range: 1–8 h	Mean: 59.03 ng/ml	Mean: 31.3 h
	n=20		10 mg/kg			Median 4.0 h Range: 1–8.1 h	Mean: 110.5 ng/ml	Mean: 33.5 h
	n=21		20 mg/kg			Median 3.2 h Range: 1–24 h	Mean: 256.9 ng/ml	Mean: 21.6 h
	n=20		10 mg/kg daily for 6 days			Day 6: Median: 3 h Range: 1–4.2 h	Day 6: Mean: 119.6 ng/ml
	n=20		20 mg/kg daily for 6 days			Day 6: Median: 2 h Range: 0–6 h	Day 6: Mean: 220 ng/ml
	n=21		40 mg/kg daily for 6 days			Day 6: Median: 3 h Range: 0–6 h	Day 6: Mean: 426.8 ng/ml

Oral administration has been the primary route used in controlled human studies with doses ranging from 20 to 6000 mg. Across studies the mean time to maximum concentration (T_max) for oral administration is highly variable and ranges from 1 to 6.13 h post-ingestion. One study examined the pharmacokinetics of oral CBD in healthy adults; Taylor and colleagues found that a single administration of Epidiolex® had a T_max range of 3–5 h (Taylor et al. 2018). In adult polydrug users (self-report of cannabinoid use in the past 12 weeks and lifetime exposure of other drug classes included opioids and stimulants in 80–90% of participants) the T_max of Epidiolex® was similar to that of healthy adults, 4.07 to 5.11 h (Schoedel et al. 2018), and in regular cannabis users oral CBD had a T_max of 3 h (Haney et al. 2016). Sublingual CBD (GW-3009–01) appears to have an earlier T_max than oral administration (T_max= 2.17 h) (Guy and Flint 2004), although no studies have done a direct comparison.

Several controlled studies suggest that CBD produces a dose-dependent, but not dose-proportional peak concentration (i.e., maximum concentration; C_max). The C_max of Epidiolex® was 292 and 782 ng/ml after 1500 and 6000 mg, respectively, in healthy adults (Taylor et al. 2018). Epidiolex® also produced a dose-dependent C_max after both a single dose and repeated dosing in children with Dravet syndrome (Devinsky et al. 2018b). Similar findings were observed in children with treatment resistant epilepsy receiving an unlicensed CBD solution (Wheless et al. 2019) (Table 1). In contrast, in polydrug users the relationship between Epidiolex® dose and C_max was not dose-dependent (Schoedel et al. 2018), and, in regular cannabis users, unlicensed CBD C_max was highly variable ranging from 1.6–271.9 ng/ml after a single dose of 800 mg oral CBD (Haney et al. 2016). It is unclear if past or recent experience with cannabinoids or other psychoactive substances can alter CBD pharmacokinetics; as no studies have directly compared the pharmacokinetics of CBD between healthy adults and regular cannabis users or people who use multiple psychoactive substances. One additional study investigated excretion of CBD in urine after both oral and vaporized administration in healthy adults and found that oral CBD (100 mg) resulted in a higher urine C_max than vaporized CBD-dominant plant material containing 100 mg CBD and 3.7 mg THC (776 and 261 ng/ml, respectively) (Spindle et al. 2020b).

Few studies have determined the half-life (t½) of CBD; after acute dosing it has been reported that Epidiolex® has a t½ of 14.39–16.61 h (Taylor et al. 2018), and between 21.6–33.5 h was reported for children with treatment-resistant epilepsy (Wheless et al. 2019). One study in adult men with a history of cannabis reported a t½ of 24 and 31 h, respectively, for intravenous CBD (20 mg) smoked CBD (19 mg) (Ohlsson et al. 1986). After repeated CBD administration, a t½ of 68 h for oral CBD was reported in Huntington’s patients (Consroe et al. 1991a). There is also emerging interest in CBD metabolites with at least one, 7-OH-CBD, reported to be an active metabolite (Taylor et al. 2019; Taylor et al. 2018; Wheless et al. 2019). Several studies have examined the pharmacokinetic profile of different metabolites and these findings are detailed in Table 2.

Table 2.

Pharmacokinetic studies of cannabidiol (CBD) metabolites 6-OH-CBD, 7-OH-CBD, and 7-COOH-CBD in humans.

Reference	Route & Population	CBD Dose	6-OH-CBD			7-OH-CBD			7-COOH-CBD
			Tmax (h)	Cmax (geometric mean, ng/ml)	t1/2 (mean h)	Tmax (h)	Cmax (geometric mean, ng/ml)	t1/2 (mean h)	Tmax (h)	Cmax (geometric mean, ng/ml)	t1/2 (mean h)
Taylor et al. (2018)	Epidiolex® in healthy adults	1500 mg	Median: 4 Range: 2.5–4	10.7	40.75	Median: 3.5 Range: 2.5–4	238.7	18.70	Median: 4 Range: 4–5	3060	25.98
		3000 mg	Median: 4.5 Range: 2.5–5	14.4	22.78	Median: 4.5 Range: 3–5	332.2	15.42	Median: 5 Range: 4–5	3557	23.88
		4500 mg	Median: 5 Range: 4–5	14.5	33.92	Median: 5 Range: 4–5	404.8	14.89	Median: 5 Range: 4–8	5120	25.18
		6000 mg	Median: 5 Range: 3–5	23.5	28.67	Median: 5 Range: 3–5	515.8	14.46	Median: 5 Range: 4–8	4591	30.24
		750 mg twice daily for 6 days	Day 1 AM Median: 5 Range: 2.5–6	Day 1 AM 8.2	Day 1 AM --	Day 1 AM Median: 4 Range: 25–6	Day 1 AM 123.0	Day 1 AM --	Day 1 AM Median: 5 Range: 4–5	Day 1 AM 2785	Day 1 AM --
			Day 7 AM Median: 2.5 Range: 2–5	Day 7 AM 12.8	Day 7 AM 21.54	Day 7 AM Median: 2.5 Range: 2–5	Day 7 AM 152.6	Day 7 AM 24.73	Day 7 AM Median: 4 Range: 3–5	Day 7 AM 9824	Day 7 AM 21.32
		1500 mg twice daily for 6 days	Day 1 AM Median: 4 Range: 3–5	Day 1 AM 9.0	Day 1 AM --	Day 1 AM Median: 4 Range: 3–5	Day 1 AM 139.5	Day 1 AM --	Day 1 AM Median: 5 Range: 4–12	Day 1 AM 2748	Day 1 AM --
			Day 7 AM Median: 3 Range: 2.5–5	Day 7 AM 16.3	Day 7 AM 82.21	Day 7 AM Median: 3 Range: 2.5–5	Day 7 AM 187.9	Day 7 AM 31.70	Day 7 AM Median: 3 Range: 0.5–5	Day 7 AM 16,306	Day 7 AM 22.00
Taylor et al. (2019)	Epidiolex® in adults w/ mild to severe hepatic impairment vs. healthy controls	200 mg
	Healthy adults		Mean: 2.5	3.19	13.2	Mean: 2.8	41.8	13.3	Mean: 4.5	823	19.8
	Mild hepatic impairment		Mean: 2.5	5.78	17.5	Mean: 3.5	54.9	14.8	Mean: 3.5	706	21.8
	Moderate hepatic impairment		Mean: 1.5	7.56	20.8	Mean: 2.0	76.4	15.6	Mean: 2.8	804	22.8
	Severe hepatic impairment		Mean: 2.0	5.35	20.3	Mean: 3.5	45.5	21.7	Mean: 4.0	221	Not calculatable
Wheless et al. (2019)	Oral CBD solution in children w/ treatment-resistant epilepsy	5 mg/kg	--	--	--	Median 2.6 Range: 1–6.1	22.03	18.4	--	--	--
		10 mg/kg				Median 4.0 Range: 1–12	34.56	25.6
		20 mg/kg				Median: 3.1 Range: 1–121	71.7	14.2
		10 mg/kg daily for 6 days				Day 6: Median: 2.1 Range: 1–4.1	Day 6: 65.6	--
		20 mg/kg daily for 6 days				Day 6: Median: 2 Range: 1–6	Day 6: 97.1	--
		40 mg/kg daily for 6 days				Day 6: Median: 2 Range: 0–5.9	Day 6: 217.7	--

It has been shown that the presence of food (i.e., a high fat meal) can significantly increase CBD exposure (i.e., 1500 mg, p.o.), with a 4-fold increase in exposure (i.e., area-under-the curve; AUC) compared to fasting in healthy normal volunteers (Taylor et al. 2018). A study in adults with refractory epilepsy also reported a food-induced increase in CBD exposure (i.e., 300 mg, p.o.) that was even greater (i.e., 15-fold increase in AUC) (Birnbaum et al. 2019). CBD t½ was also modified by food-intake from 30.33 h to 24.4 h (~20%) in the fasted versus fed condition in healthy adults receiving Epidiolex® (Taylor et al. 2018) and from 39 h to 24.3 (38%) in adults with refractory epilepsy (Birnbaum et al. 2019). It should also be noted that patients with hepatic impairment may require adjustment of their CBD dose; in one study AUC increased with severity of hepatic impairment, and those with severe impairment had a roughly 5-fold increased AUC (Table 1; Taylor et al. 2019).

Additionally, drug-drug interactions may result in changes in the pharmacokinetic effects of both CBD and other drugs. Documentation of drug-drug interactions with CBD have been reported in recent publications (Brown and Winterstein 2019; Fitzcharles et al. 2020). CBD acts as an inhibitor or inducer of several cytochrome P450 isoforms including 3A4, 2C19, 2C8, 2C9, 2D6, 1A2 and 2B6, and has minor activity at several others. (Highlights of Prescribing Information 2018; Stout and Cimino 2014). As CYP450 enzymes are involved in metabolism of the majority of pharmacotherapies (Zanger and Schwab 2013), CBD has the potential to cause interaction effects with many over-the-counter and prescription medications. For example, in a pediatric expanded-access study, 13 participants taking clobazam (anti-seizure treatment) also received oral CBD (titrated up to 20 mg/kg/day) for 4 weeks – concomitant CBD administration resulted in a 60% increase in serum clobazam and a 500% increase in the metabolite norclobazam (Geffrey et al. 2015). Two pediatric case reports identified drug-drug interactions with other pharmacological agents: 1) CBD was implicated in a 4-fold increase in serum concentrations of everolimus (an antineoplastic chemotherapeutic also used for seizure management) – the authors suggest that this was likely mediated through a CYP 3A4 interaction (Wiemer-Kruel et al. 2019); 2) one patient taking methadone experienced increased fatigue and somnolence and over a 2-fold increase in serum methadone concentrations after initiating CBD - these effects resolved upon removal of CBD (Madden et al. 2020). Additionally warnings in the Epidiolex® package insert suggest that when given in combination with the anticonvulsant valproate, liver enzymes can become elevated (Highlights of Prescribing Information 2018) Overall, the potential for drug-drug interaction should be considered when starting treatment on CBD treatment.

Safety and Abuse Liability

The World Health Organization’s report on CBD concluded that it has a good safety profile with limited side effects (Cannabidiol (CBD) Critical Review Report 2018). Several controlled human laboratory studies of oral (200–800 mg) and sublingual (20 mg) CBD reported limited effects on physiological outcomes, including heart rate and blood pressure (Babalonis et al. 2017; Guy and Flint 2004; Haney et al. 2016; Martin-Santos et al. 2012; Schoedel et al. 2018; Taylor et al. 2018). In contrast, two recent randomized, double-blind, placebo-controlled studies found a modest decrease in arterial pressure and systolic blood pressure after acute CBD administration (Jadoon et al. 2017; Sultan et al. 2020), but this effect dissipated when CBD was administered daily for 7 days (Sultan et al. 2020). The registration studies for Epidiolex® reported the most common side effects as diarrhea, headache, decreased appetite, and somnolence (Devinsky et al. 2014; Devinsky et al. 2017; Devinsky et al. 2016; Devinsky et al. 2018a; Devinsky et al. 2018b; Schoedel et al. 2018; Szaflarski et al. 2018; Taylor et al. 2018). Interestingly, a recent meta-analysis reported, in children with epilepsy, that CBD was associated with higher rates of pneumonia compared to placebo and that high doses of CBD (≥20 mg/kg) were associated with abnormal liver function tests (Chesney et al. 2020).

With regard to abuse liability, the vast majority of studies evaluating acute dosing concluded that there is no signal for abuse potential with CBD (Babalonis et al. 2017; Haney et al. 2016; Martin-Santos et al. 2012; Schoedel et al. 2018). This is consistent with the re-scheduling of Epidiolex® as a non-scheduled drug in the U.S. (Highlights of Prescribing Information 2018). The exceptions to this body of evidence include two randomized, double-blind, placebo-controlled studies: one examined vaporized CBD (100 mg) and reported increased ratings of “Pleasant Drug Effect” and “Like Drug Effect” (Spindle et al. 2020a) while another reported that vaporized CBD (400 mg) increased subjective ratings of intoxication on a visual analog scale (Solowij et al. 2019).

Of critical importance, the majority of CBD products being sold have not been approved by the FDA. Unregulated CBD is available in numerous formulations, including oral capsules or tinctures; sublingual oils; topical creams, balms, and salves; e-liquids or crystalized formations (wax) for vaporization; and dietary supplement forms. These products are sold online and in retail shops with advertising suggesting a vast array of unsubstantiated medical and psychiatric benefits, and to improve beauty, hygiene, and nutrition. Estimated sales of these products were between 600 million and 2 billion USD in 2018, and investment companies predict sales will reach 16 billion USD by 2025 (Azer et al. 2019). Whether or not these unregulated products contain CBD as advertised is unknown, and unsanctioned CBD products may contain hazardous chemicals (Bonn-Miller et al. 2017; FDA 2019; FDA 2020b; Poklis et al. 2019). Analysis of 48 products purchased online found that only 31% were accurately labeled in regard to CBD concentration and 21% contained THC (Bonn-Miller et al. 2017). Similarly, a study in the United Kingdom reported only 38% of over-the-counter products contained ±10% of the advertised quantity and 55% contained THC (Liebling et al. 2020). The FDA reported similar findings with inconsistent CBD concentrations and the presence of THC (FDA 2019). The FDA has sent a myriad of warning letters due to inaccurate labeling and false health claims, including, but not limited to, treatment for chronic pain, anxiety, and opioid use disorder (FDA 2019; FDA 2020b). Contamination with 5-fluoro MDMB-PINACA and dextromethorphan has also been reported (Poklis et al. 2019). CBD product contamination could lead to unanticipated psychoactive effects and positive urine drug screens in the case of THC. Several states with medical cannabis laws require laboratory testing of products, but uniform testing protocols have not been developed (e.g., array of tests conducted, assay technology, sensitivity parameters). There have been no systematic evaluations to determine the accuracy of laboratory reports, product labels, or product contamination rates of products available in the state-sponsored dispensaries and those sold elsewhere (i.e., online, unregulated retail shops) (Corroon et al. 2020).

There has been concern that oral CBD could transform into THC in the human gut, but this hypothesis has recently been refuted by empirical studies confirming that CBD does not transform to THC in humans, even at high doses (4500 mg acute oral dose; Crippa et al. 2019; Schoedel et al. 2018; Spindle et al. 2020b). Thus, intoxication or THC positive drug screens associated with CBD products are thought to be due to contamination.

Therapeutic Implications

Neurological Disorders

The most compelling evidence supporting the medicinal use of CBD is for the treatment of epilepsy (Table 3). An early study suggesting a benefit of CBD for seizure disorders was a very small (n=15) randomized, double-blind, placebo-controlled study of patients with secondary generalized epilepsy (Cunha et al. 1980). CBD (200–300 mg, p.o./daily for up to 18 weeks) was associated with reduced seizure episodes on a monthly seizure scale of 0–3. Open-label and expanded access studies with limited rigor have also examined CBD in patients with childhood onset seizures and febrile infection-related epilepsy (Devinsky et al. 2016; Gofshteyn et al. 2017; Szaflarski et al. 2018). An open-label trial of 137 patients (ages 1–30) with childhood onset seizures described as “severe, intractable, [and] treatment-resistant” explored oral CBD (up to 50 mg/kg/day) and reported a 36.5% reduction of median monthly seizures (Devinsky et al. 2016). In a similar open-label add-on study, adults and children with various types of epilepsy received 12–48 weeks of treatment with Epidiolex® (up to 50 mg/kg/day); after 12 weeks of treatment, Chalfont Seizure Severity Scores decreased by 51.3% (Szaflarski et al. 2018).

Table 3.

Clinical studies of CBD effects on neurological disorders.

Reference	Study Sample	CBD Dose and Route	CBD Formulation	Study Details	Summary Outcomes
Chagas et al. (2014)	Parkinson’s disease (n=21)	75 or 300 mg/day, p.o. for 6 weeks	Gelatin capsule containing 99.9% pure CBD (THC-Pharma) dissolved in corn oil	Randomized, double-blind, placebo-controlled	300 (but not 75) mg CBD was associated with better scores on the Parkinson’s Disease Questionnaire. Similar scores were found on the Unified Parkinson’s Disease rating scale with CBD and placebo treatment
Consroe et al. (1991b)	Huntington’s disease (n=15)	10 mg/kg/day, p.o., for 5 weeks	Gelatin capsule containing CBD (National Institute on Drug Abuse) dissolved in sesame oil	Randomized, double-blind, placebo-controlled, crossover	No improvement of Huntington’s disease-related symptoms
Cunha et al. (1980)	Generalized epilepsy (n=15)	200–300 mg, p.o./daily for up to 18 weeks	Gelatin capsule containing CBD or glucose (placebo)	Randomized, double-blind, placebo-controlled	Reduced reported monthly seizure episodes on scale of 0–3
Devinsky et al. (2016)	Patients ages 1–30 with severe, intractable, treatment-resistant childhood onset seizures (n=137)	Up to 50 mg/kg/day, p.o.	Epidiolex®	Open-label	36.5% decrease in median monthly seizures
Devinsky et al. (2017)	Dravet Syndrome patients ages 2–18 (n=120)	Escalated to 20 mg/kg/day, p.o., over two weeks, followed by 20 mg/kg/day for 12 weeks	Epidiolex®	Randomized, double-blind, placebo-controlled	CBD decreased median monthly motor seizers by 38.9% (placebo decrease was 13.3%)
Devinsky et al. (2018a)	Lennox-Gastaut patients, ages 2–55 (n=212)	10 or 20 mg/kg/day, p.o. for 14 weeks	Epidiolex®	Randomized, double-blind, placebo-controlled	10 or 20 mg/kg/day oral CBD had a 37.2% and 41.9% decrease in drop-seizure activity from baseline, respectively, compared to a 17.2% decrease in patients who received placebo
Gofshteyn et al. (2017)	Febrile infection-related epilepsy (n=7)	Titrated to 15–25 mg/kg/day	Epidiolex®	Open-label case series (emergency/expanded access)	6 out of 7 patients had decreases in seizure frequency
(Hess et al. 2016)	Tuberous Sclerosis Complex (n=18)	Titrated to 50 mg/kg/day	Epidiolex®	Open-label, expanded access	Decreased median monthly seizure frequency from 22 at baseline to 13.3 after 3 months of treatment (48.8% decrease).
Szaflarski et al. (2018)	Adults and children with various types of epilepsy (n=139)	Titrated up to 50 mg/kg/day, p.o., for 12–48 weeks	Epidiolex®	Open-label, add-on	Decreased mean Chalfont Seizure Severity Scores from 80.7 to 39.3 after 12 weeks of treatment. Scores were stable from week 12 to week 48.
Thiele et al. (2018)	Lennox-Gastaut patients ages 2–55 (n=171)	20 mg/kg/day, p.o., for 14 weeks	Epidiolex®	Randomized, double-blind, placebo-controlled	CBD decreased drop seizures by 43.9%, placebo decreased drop-seizures by 21.8%
Zuardi et al. (2009)	Parkinson’s disease (n=6)	150 mg/day, p.o. Dose increased by 150 mg/day weekly for 4 weeks.	Gelatin capsule containing 99.9% pure CBD (THC-Pharma) dissolved in corn oil	Open-label, pilot	CBD was associated with improved scores on the Brief Psychiatric Rating Scale and the Parkinson’s Psychosis Questionnaire

Rigorous studies have recently established CBD (Epidiolex®) as an effective treatment for Dravet and Lennox-Gastaut syndromes, and more recently seizures that are secondary to tuberous sclerosis complex, all forms of childhood epilepsy associated with severe, intractable seizures (Table 3). A randomized, double-blind trial of 120 patients ages 2–18, with Dravet Syndrome found that 20 mg/kg/day oral CBD for 14 weeks decreased median monthly motor seizures by 38.9% compared to 13.3% with placebo (Devinsky et al. 2017). Two double-blind randomized studies examined CBD for Lennox-Gastaut. In the first, patients (n=212; ages 2–55) who received 10 or 20 mg/kg/day oral CBD had a 37.2% and 41.9% decrease in drop-seizure activity from baseline, respectively, compared to a 17.2% decrease in patients who received placebo (Devinsky et al. 2018a). In the second, patients with treatment-resistant Lennox-Gastaut (n=171; ages 2–55) received 20 mg/kg/day oral CBD or placebo for 14 weeks. Those receiving CBD had a 43.9% reduction in drop-seizures, compared to a 21.8% decrease in patients who received placebo (Thiele et al. 2018). These studies provided the supportive evidence of efficacy and safety for CBD for Dravet and Lennox-Gastaut syndromes leading to its approval by the FDA in 2018. In July 2020 the FDA expanded Epidiolex® approval to include treatment of seizures due to tuberous sclerosis complex (FDA 2020a). An initial open-label add on study reported that 3 months of treatment with Epidiolex® titrated up to 50 mg/kg/day decreased median weekly seizures by 48.8% (Hess et al. 2016). The FDA reported that in a randomized, double blind, placebo-controlled trial of 224 patients, Epidiolex® decreased seizure frequency after 8 weeks of treatment (FDA 2020a).

CBD has been investigated in other neurological conditions, including Huntington’s and Parkinson’s disease (Table 3). A small randomized, double-blind, placebo-controlled, crossover study of 15 patients with mild to moderate Huntington’s disease received 10 mg/kg/day oral CBD (Consroe et al. 1991b). Patients who received CBD had no improvement of Huntington’s disease-related symptoms (e.g., chorea, tongue extension) and no benefit of CBD was found by physician nor patient assessment (Consroe et al. 1991b). Two studies have explored CBD for the treatment of Parkinson’s disease. A small, open-label pilot study (n=6) suggested that oral CBD (starting at 150 mg/day and increasing weekly for 4 weeks) might improve Parkinson’s-related psychotic symptoms (Zuardi et al. 2009). In a subsequent, randomized, double-blind study by the same group, 21 patients with Parkinson’s disease received 75 or 300 mg oral CBD or placebo daily for 6 weeks (Chagas et al. 2014). Patients who received CBD scored similar to those who received placebo on the Unified Parkinson’s Disease Rating Scale (a measure of Parkinson’s disease severity), but patients who received 300 mg CBD had better scores on the Parkinson’s Disease Questionnaire compared to those who received placebo (Chagas et al. 2014). The results of these two small trials are promising; however, larger randomized controlled trials are needed to determine the efficacy of CBD for the treatment of Parkinson’s disease. For a review of CBD effects on neurological disorders in animal models see Elsaid et al. (2019).

Pain, Inflammation, and Immune Function

A limited number of studies of varying rigor have investigated CBD for efficacy in pain conditions, including multiple sclerosis, fibromyalgia, Crohn’s disease, and neuropathic pain (Table 4). In a randomized, double-blind, placebo-controlled, crossover study 20 patients with multiple sclerosis, spinal injury, brachial plexus lesions, or amputation received CBD (2.5–120 mg/day of sublingual spray) treatment which decreased pain on a visual analog scale, but did not improve other symptoms such as spasms, coordination, bladder control, or emotional well-being (Wade et al. 2003). In contrast, a randomized, double-blind, placebo-controlled, crossover study in which 34 patients with chronic pain primarily due to multiple sclerosis received a relatively low dose (2.5 mg) of sublingual CBD spray daily for 8 weeks found little-to-no pain improvement (Notcutt et al. 2004). Other studies using low doses also found that CBD was not analgesic: in a randomized, double-blind, placebo-controlled, crossover study of 20 patients with chronic pain due to fibromyalgia, vaporized (inhaled) Bedrolite (18.4 mg CBD & < 1 mg THC) did not improve experimental pressure or electrical pain (van de Donk et al. 2019) and sublingual CBD (20 mg/day for 8 weeks) did not alter the Crohn’s disease activity index – a global measure of Crohn’s disease severity that includes pain – compared to placebo in a randomized, double-blind study of 19 patients (Naftali et al. 2017). Collectively these studies suggest that CBD is not an effective analgesic, but they are limited by a low dose of CBD. One exploratory study reported that 50 out of 94 patients on opioids for chronic pain who were treated with CBD-rich hemp extract (self-titrated to ~30 mg/day, p.o.) reduced their opioid intake (Capano et al. 2020). In a larger randomized, double-blind study, patients with peripheral neuropathy of various etiologies (n=29) received topical CBD (250 mg/3 fl. oz.) up to 4 times daily for 4 weeks or placebo, after 4 weeks of treatment the placebo group crossed-over to receiving CBD and all participants were treated for an additional 4 weeks (Xu et al. 2020). This study reported a larger decrease in “intense, sharp, cold, and itchy” on the Neuropathic Pain Scale in the CBD group compared to placebo. While this study is intriguing, further studies are required to confirm that CBD is an efficacious analgesic when applied topically.

Table 4.

Clinical studies of CBD effects on pain, inflammation, and immune function.

Reference	Study Sample	CBD Dose and Route	CBD Formulation	Study Details	Summary Outcomes
Capano et al. (2020)	Patients on opioids for chronic pain (n=94)	Self-titrated to 30 mg/day, p.o.	CBD-rich hemp extract	Exploratory	50 out of 94 patients reduced their opioid intake
Couch et al. (2019)	Male participants who received aspirin to increase absorption of lactulose and mannitol (n=38)	600 mg, p.o.	99.65% pure CBD in cellulose (Artelo Biosciences)	Randomized, double-blind, placebo-controlled	Mannitol and lactulose absorption were decreased after treatment with CBD
Irving et al. (2018)	Ulcerative colitis (n=6)	Titrated up to 250 mg/day, p.o. for 10 weeks	CBD-rich botanical extract in a gelatin capsule	Randomized, doublebind, placebo-controlled	CBD did not improve remission rates
Jadoon et al. (2016)	Type II diabetes (n=62)	200 mg/day (route unknown) for 13 weeks		Randomized, double-blind, placebo-controlled	CBD did not improve glycemic control
Naftali et al. (2017)	Crohn’s disease (n=19)	20 mg/day, sublingual, for 8 weeks	99.5% pure CBD extracted in house and dissolved in olive oil	Randomized, double-blind, placebo-controlled	No change in Crohn’s disease activity index
Notcutt et al. (2004)	Chronic pain (primarily due to multiple sclerosis) (n=34)	2.5 mg, sublingual spray for 8 weeks	Botanical CBD extract dissolved in tetrafluoroethane 80%, ethanol 20% for n=6 and in ethanol 50%, propylene glycol 50% for n=28	Randomized, double-blind, placebo-controlled, crossover	No improvement of chronic pain on a visual analog scale
Tomida et al. (2006)	Ocular hypertension or glaucoma (n=6)	20 or 40 mg, sublingual spray	provided by GW Pharmaceuticals	Randomized, double-blind, placebo-controlled, crossover	20 mg CBD did not alter intraocular pressure, 40 mg CBD increased intraocular pressure
van de Donk et al. (2019)	Chronic pain due to fibromyalgia (n=20)	18.4 mg, vaporized (inhaled)	Bedrolite (18.4 mg CBD & < 1 mg THC),	Randomized, double-blind, placebo-controlled, crossover	No improvement of experimental pressure or electrical pain
Wade et al. (2003)	Pain due to multiple sclerosis, spinal injury, brachial plexus lesions, or amputation (n=20)	2.5–120 mg/day, sublingual spray	CBD extract in unknown solvent (GW Pharmaceuticals)	Randomized, double-blind, placebo-clntrllled, crossover	CBD treatment was associated with decrease pain on a visual analog scale. CBD did not improve spasms, coordination, bladder control, or emotional well-being.
Xu et al. (2020)	Neuropathic pain in the lower extremities due to various etiologies (n=29)	250 mg/3 fl. oz. topical CBD applied up to 4 times a day for 4–8 weeks	Theramu Relieve CBD compound Cream (Theramu) or emu oil (placebo)	Randomized, double-blind, placebo controlled	CBD decreased ratings of “intense, sharp, cold, and itchy” on the Neuropathic Pain Scale
Yeshurun et al. (2015)	Patients undergoing hematopoietic cell transplantation (n=48)	300 mg/day, p.o. prophylactic starting 7 days before transplantation and continuing until day 30	CBD (STI pharmaceuticals) dissolved in olive oil at 2.5% concentration	Historical control	Patients that received prophylactic CBD had lower rates of graft-versus-host-disease compared to a historical control of 101 patients

There are also a multitude of conditions related to inflammation and immune function for which a single randomized controlled study has reported on the effects of CBD, including gastrointestinal inflammation, ulcerative colitis, ocular hypertension and glaucoma, graft-versus-host disease, and diabetes (Table 4). First, in a randomized, double-blind, placebo-controlled study 38 male participants received aspirin to increase gastrointestinal absorption of lactulose and mannitol and this effect was reduced in participants who received CBD (600 mg, p.o.) (Couch et al. 2019), suggesting CBD may decrease gastrointestinal inflammation. However, a randomized, double-blind, placebo-controlled, pilot study (n=6) found that CBD-rich extract did not improve remission rates in patients with ulcerative colitis (Irving et al. 2018). In a randomized, double-blind, placebo-controlled, crossover, pilot study, patients (n=6) with ocular hypertension or glaucoma received sublingual CBD (20 and 40 mg). Intraocular pressure did not change when patients received 20 mg CBD, but increased when they received 40 mg, suggesting CBD may worsen ocular-related disease (Tomida et al. 2006). A phase II clinical trial of 48 patients who received oral CBD solution (300 mg/day) in addition to prophylactic immunosuppressive treatment for the prevention graft-versus-host-disease due to allogeneic hematopoietic cell transplantation suggested CBD may decrease rates of graft-versus-hostdisease when compared to a historical control of 101 patients (Yeshurun et al. 2015). Lastly, in a randomized, double-blind, placebo-controlled study of 62 patients with type II diabetes, CBD (200 mg/day) produced minimal effects on glycemic control (Jadoon et al. 2016). While these reports from well controlled studies are intriguing, there is insufficient evidence to draw clinically meaningful conclusions. For a review of preclinical findings on CBD effects on pain and inflammation see Burstein (2015).

Psychiatric Disorders and Substance Use

Numerous studies of varying rigor have examined CBD for its anxiolytic effects (Table 5). This body of work is difficult to interpret because studies test different doses, several limit enrollment to men, and some enroll patients with anxiety disorders while others employ experimental anxiety paradigms. A small (n=10), randomized, double-blind, placebo-controlled, crossover study in men with generalized anxiety reported that CBD (400 mg, p.o.) reduced subjective anxiety on a visual analog mood scale (Crippa et al. 2011). A double-blind study of healthy adults (n=40) reported that CBD (300 mg, p.o.) decreased anxiety after a simulated public speaking test compared to placebo (and similar to other anxiolytics) (Zuardi et al. 1993). CBD 300 mg, p.o., but not 100 or 900 mg, also reduced subjective ratings of anxiety during a test of experimentally induced public speaking in a randomized, double-blind, placebo-controlled study (n=60) (Zuardi et al. 2017). Similarly, a randomized, double-blind, placebo-controlled study (n=57) in men reported that 300 mg oral CBD, but not 150 or 600 mg, decreased anxiety during a simulated public speaking test (Linares et al. 2019). In a randomized, double-blind, placebo-controlled, crossover study (n=16) CBD (600 mg, p.o.) was not anxiolytic (Martin-Santos et al. 2012). CBD (600 mg, p.o.) was also not anxiolytic in a randomized, double-blind, placebo-controlled study of 32 participants with high trait paranoia and persecutory ideation (Hundal et al. 2018). In two double-blind, placebo-controlled, within-subjects studies (n=15/study) CBD (600 mg, p.o.) did not reduce subjective ratings of anxiety on a visual analog mood scale or the Spielberger State Trait Anxiety Inventory (Borgwardt et al. 2008; Fusar-Poli et al. 2009). However, one study found that 600 mg oral CBD was anxiolytic in a randomized, double-blind, placebo-controlled study of 36 undergraduate students with social phobia (Bergamaschi et al. 2011). In regard to repeated CBD administration, in one randomized, double-blind study, patients (n=58) with a clinically high risk for psychosis received 600 mg CBD (p.o.) daily for 1 week (Appiah-Kusi et al. 2020). There was no significant difference between participants who received CBD versus placebo on the Tier Social Stress Test. Overall, these mixed results suggest that controlled studies to identify an effective dose range and dosing regimen (acute, repeated dosing) are needed, particularly in individuals with anxiety disorders. Despite this lack of controlled data, over the counter CBD products are being advertised as beneficial for such conditions. For review of pre-clinical studies examining CBD anxiolysis see Blessing et al. (2015).

Table 5.

Clinical studies of CBD effects on psychiatric disorders and substance use.

Reference	Study Sample	CBD Dose, and Route	Drug Formulation	Study Details	Summary Outcomes
Arkell et al. (2019)	Healthy adults (n=14)	125 mg plant material, inhaled (vaporized)	11% THC and <1% CBD 11% THC and 11% CBD <1% THC and <1% CBD (placebo) (from Tilray)	Randomized, double-blind, placebo-controlled crossover	No subjective difference between THC and the THC+CBD combination
Appiah-Kusi et al. (2020)	Adults with high risk for psychosis (n=58)	600 mg/day, p.o. for 1 weeks	Capsule (STI Pharmaceuticals)	Randomized, double-blind, placebo-controlled	No difference between patients that received CBD versus placebo on the Tier Social Stress Test
Bergamaschi et al. (2011)	Undergraduate students with social phobia (n=36) and healthy controls	600 mg, p.o.	Gelatin capsule containing 99.9% pure CBD (SIT Pharmaceuticals and THC-Pharm) dissolved in corn oil	Randomized, double-blind, placebo-controlled	CBD decreased anxiety during a simulated public speaking test
Bhattacharyya et al. (2010)	Adults who experienced THC-induced psychosis (n=3)	5 mg, I.V.	--	Pseudorandomized, double-blind, crossover	CBD prevented the acute psychotic symptoms to THC (1.25 mg, I.V.)
Boggs et al. (2018)	Schizophrenia (n=36)	600 mg/day, p.o. for 6 weeks	STI Pharmaceuticals	Randomized, double-blind, placebo-controlled	CBD did not affect scores on the MATRICS Consensus Cognitive Battery or the Positive and Negative Symptoms Scale
Borgwardt et al. (2008)	Healthy adults (n=15)	600 mg, p.o.	Capsule	Pseudorandomized, double-blind, placebo-controlled, crossover	No effect of CBD on the Visual Analog Mood Scale, Spielberger State Trait Anxiety Inventory test, or Positive and Negative Symptoms Scale
Consroe et al. (1979)	Healthy adults (n=10)	200 mg, p.o.	Gelatin capsule containing crystalline CBD (from Dr. R. Mechoulam)	Randomized, double-blind, placebo-controlled, crossover	Consumption of 1 g/kg alcohol with CBD resulted in lower blood alcohol concentration than only alcohol, but CBD did not alter the behavioral effects of alcohol
Crippa et al. (2011)	Men with generalized anxiety (n=10)	400 mg, p.o.	Gelatin capsule containing 99.9% pure CBD (THC-Pharma) dissolved in corn oil	Randomized, double-blind, placebo-controlled, crossover	CBD reduced anxiety on a visual analog mood scale
Englund et al. (2013)	Healthy adults (n=22)	600 mg, p.o.	Capsule (STI Pharmaceuticals)	Randomized, double-blind, placebo-controlled	CBD prevented THC (1.5 mg, I.V.) - induced paranoia via assessment with the State Social Paranoia Scale
Fusar-Poli et al. (2009)	Healthy adults (n=15)	600 mg, p.o.	Gelatin capsule containing 99.9% pure CBD (THC-Pharma)	Randomized, double-blind, placebo-controlled, crossover	No effect of CBD on the Visual Analog Mood Scale, Spielberger State Trait Anxiety Inventory test, or Positive and Negative Symptoms Scale
Hallak et al. (2010)	Schizophrenia (n=28)	300 or 600 mg, p.o.	Gelatin capsule containing CBD (from Dr. R. Mechoulam)	Double-blind, placebo-controlled	CBD had no effect on the Positive and Negative Symptom Scale, Brief Psychiatric Rating Scale, or performance on the Stroop Color Word Test
Haney et al. (2016)	Cannabis smokers (n=31)	200, 400, or 800 mg, p.o.	Size 00 opaque capsule containing CBD (STI Pharmaceuticals)	Randomized, double-blind, placebo-controlled, crossover	CBD did not change the subjective intoxicating effects of smoked cannabis containing THC (5.08–5.30%), nor did it alter cannabis self-administration
Hindocha et al. (2018)	Non-treatment seeking tobacco smokers undergoing abstinence (n=33)	800 mg, p.o.	Opaque capsule containing pure synthetic CBD (STI Pharmaceuticals)	Randomized, double-blind, placebo-controlled, crossover	CBD decreased attention bias toward cigarette cues, but did not alter craving or withdrawal ratings
Hundal et al. (2018)	Adults with high trait paranoia and persecutory ideation (n=32)	600 mg, p.o.	Gelatin capsule containing CBD (GW Pharmaceuticals) in Killophor EL and M1944CS	Randomized, double-blind, placebo-controlled	No significant difference between CBD and placebo on the State Social Paranoia Scale or the Community Assessment of Psychic Experiences questionnaire
Hurd et al. (2019)	Participants with heroin use disorder who were abstinent (n=42)	400 or 800 mg., p.o. acute, or daily for three days	Epidiolex®	Randomized, double-blind, placebo-controlled	Acutely, CBD decreased drug-cue induced craving on a visual analog scale, but not heroin. One week after their last treatment, participants that received CBD still had lower scores on a visual analog scale for drug-cue induced craving. Heroin craving (assessed by out of clinic questionnaire) was not altered by CBD. Those who received CBD had lower scores on a visual analog scale of anxiety.
Karniol et al. (1974)	Healthy adults (n=40)	15, 30, or 60 mg, p.o.	CBD (Makor Chemicals Ltd.) dissolved in 0.9 ml of ethanol and placed in 200 ml of orange juice	Double-blind, placebo-controlled	CBD inhibited the subjective effects of THC (30 mg, p.o.) on a scale of 0–4
Leweke et al. (2012)	Schizophrenia (n=33)	Increasing to 800 mg/day over the first week, then 800 mg/day for 3 additional weeks	--	Randomized, double-blind	CBD improved clinical symptoms compared to baseline.
Linares et al. (2019)	Healthy adult men (n=57)	150, 300, or 600 mg, p.o.	Gelatin capsule containing 99.9% pure CBD (STI Pharmaceuticals) dissolved in corn oil	Randomized, double-blind, placebo-controlled	300 mg (but not 150 or 600 mg) reduced anxiety during a simulated public speaking test.
Manini et al. (2015)	Healthy adults with experience with opioids (non-dependent) (n=17)	400 or 800 mg, p.o.	Gelatin capsule containing 99.9% pure CBD (GW Pharmaceuticals) dissolved in corn oil	Double-blind, placebo-controlled, crossover	CBD did not alter the pharmacokinetics or adverse effects of 0.5 or 1.0 μg/kg I.V. fentanyl
Martin-Santos et al. (2012)	Healthy adult men (n=16)	600 mg, p.o.	Opaque capsule with 99.9% pure CBD (THC-Pharm and STI Pharmaceuticals)	Randomized, double-blind, placebo-controlled, crossover	CBD did not alter scores on the Positive and Negative Symptoms Scale or on the Spielberger State Trait Anxiety Inventory
McGuire et al. (2018)	Schizophrenia (n=88)	1000 mg/day for 6 weeks	Oral solution (GW Pharmaceuticals)	Randomized, double-blind, placebo-controlled	CBD decreased positive psychotic symptoms on the Positive and Negative Symptoms Scale and those who received CBD were rated as better on the Clinical Global Impression Scale by a physician.
Morgan et al. (2013)	People seeking treatment for tobacco smoking (n=24)	Inhaler use, as needed, 400 μg administered in each depression of the inhaler	CBD (STI Pharmaceutical) dissolved in absolute ethanol ~5%	Randomized, double-blind, placebo-controlled	CBD decreased tobacco smoking by 40%
Morgan et al. (2018)	Light cannabis users (n=24)	16 mg, inhaled (vaporized)	CBD (STI Pharmaceuticals) dissolved in alcohol	Randomized, double-blind, placebo-controlled	CBD inhibited THC (8 mg, vaporized) - induced increases on the Psychomimetic State Inventory
Solowij et al. (2019)	Current cannabis users and non-naive non-users (n=36)	4 or 400 mg inhaled (vaporized)	CBD (STI Pharmaceuticals) dissolved in ethanol (10% solution)	Randomized, double-blind, placebo-controlled	CBD-THC combinations with relatively high CBD (400 mg) were less intoxicating than THC alone, however, when the CBD dose was reduced to 4 mg it increased THC-induced intoxication
Zuardi et al. (2017)	Healthy adults (n=60)	100, 300, or 900 mg, p.o.	Gelatin capsule containing 9.6% pure CBD dissolved in corn oil	Randomized, double-blind, placebo-controlled	300 mg (but not 100 or 900 mg) reduced subjective ratings of anxiety.
Zuardi et al. (1993)	Healthy adults (n=40)	300 mg, p.o.	Gelatin capsule containing CBD (from Dr. R. Mechoulam) dissolved in corn oil	Randomized, double-blind, placebo-controlled	CBD decreased anxiety after a simulated pubic speaking test.

In regard to other psychiatric disorders, a randomized, double-blind study in 33 patients reported that CBD (200–800 mg/daily) improved clinical symptoms of schizophrenia compared to baseline and similar to the antipsychotic amisulpride (Leweke et al. 2012). Another randomized, double-blind, placebo-controlled clinical trial in 88 patients with schizophrenia found oral CBD solution (1000 mg/day for 6 weeks) decreased positive psychotic symptoms (McGuire et al. 2018). Yet, CBD (600 mg/day, p.o.) did not improve psychotic symptoms in a randomized, double-blind, placebo-controlled study of 36 patients with schizophrenia (Boggs et al. 2018). Additionally, CBD (300 and 600 mg, p.o.) did not alter scores on the Positive and Negative Symptoms Scale or the Brief Psychiatric Rating Scale and did not change performance on the Stroop Color Word Test in a double-blind, placebo-controlled study of 28 patients with schizophrenia (Hallak et al. 2010). For a review of CBD effects on psychiatric disorders in animal models see Elsaid et al. (2019).

Although it is unclear if CBD can reduce psychiatric illness, it may reduce psychotic symptoms associated with THC usage (Table 5). In a double-blind, placebo-controlled, crossover study, CBD (5 mg, I.V.) prevented the acute psychotic symptoms of THC (1.25 mg, I.V.) in three out of three men that experienced THC-induced psychosis (Bhattacharyya et al. 2010). Similarly, oral CBD (600 mg) prevented THC (1.5 mg I.V.) -induced paranoia in a randomized, double-blind, placebo-controlled study (n=22) (Englund et al. 2013). Vaporized CBD (16 mg) inhibited THC (8 mg, vaporized) -induced increases on the Psychomimetic State Inventory in a randomized, double-blind, placebo-controlled study of light cannabis users (n=24) (Morgan et al. 2018). CBD may also alter THC intoxication, although results from these studies are inconsistent, possibly due to differences in CBD formulation and/or route of administration. In an early, double-blind, placebo-controlled study (n=40) oral CBD (15–60 mg) inhibited THC’s subjective effects (Karniol et al. 1974). In a randomized, double-blind study (n=36) vaporized CBD-THC combinations with relatively high CBD (400 mg) were less intoxicating than THC alone (8 mg), however, when the CBD dose was reduced to 4 mg it increased THC-induced intoxication (Solowij et al. 2019). Another randomized, double-blind, placebo-controlled, crossover study in 31 cannabis smokers found that CBD (200–800 mg, p.o.) did not change the subjective intoxicating effects of smoked cannabis containing THC (5.08–5.30%), nor did it alter cannabis self-administration (Haney et al. 2016). In a randomized, double-blind, placebo-controlled, within-subjects study participants (n=14) reported no subjective difference in drug effect between vaporized CBD+THC (11% CBD, 11% THC) and THC only (<1%CBD, 11%THC) (Arkell et al. 2019).

Despite popular belief, few human studies examining CBD treatment of substance use have been conducted and while some of these results are intriguing, not enough evidence exists to indicate CBD as a viable treatment option for substance use disorders (Table 5). Inhaled CBD decreased tobacco smoking by 40% in a randomized, double-blind, placebo-controlled study of 24 participants who wanted to quit smoking (Morgan et al. 2013). In a randomized, double-blind, placebo-controlled, crossover study of 33 non-treatment seeking smokers undergoing short-term tobacco abstinence, CBD (800 mg, p.o.) decreased attention bias toward cigarette cues, but did not alter craving or withdrawal ratings (Hindocha et al. 2018). In a randomized, double-blind, placebo-controlled, crossover study (n=10), CBD (200 mg, p.o.) decreased blood alcohol levels, but did not alter the behavioral effects of alcohol (Consroe et al. 1979). For a review of CBD effects on alcohol consumption in animal models see Turna et al. (2019). Regarding opioids, one double-blind, placebo-controlled, crossover study (n=17) reported that CBD (400 or 800 mg, p.o.) did not alter pharmacokinetics or adverse effects of I.V. fentanyl (Manini et al. 2015). Another randomized, double-blind, placebo-controlled study of participants (n=42) with heroin use disorder who were abstinent reported that Epidiolex® (400 or 800 mg) inhibited drug-cue induced craving and anxiety, but not heroin craving (Hurd et al. 2019). While these studies are intriguing, they do not provide substantive data to draw clinically meaningful conclusions.

Conclusions

CBD has diverse molecular targets, including indirect activity at cannabinoid receptors and agonism of TRPV and 5-HT_1A, and additional molecular targets are being investigated (Bih et al. 2015). This array of targets has resulted in claims that CBD is efficacious for a myriad of health conditions, but clinical data supporting CBD as a pharmacotherapy are limited to the anti-epileptic effects of Epidiolex®. Epidiolex® has been developed to produce accurate biological exposure and has undergone rigorous pharmacokinetic evaluations (Devinsky et al. 2018b; Schoedel et al. 2018; Taylor et al. 2019; Taylor et al. 2018). However, six studies have characterized the pharmacokinetic effects of other formulations of CBD, including oral preparations (Birnbaum et al. 2019; Consroe et al. 1991a; Haney et al. 2016; Wheless et al. 2019), sublingual CBD (Guy and Flint 2004), and smoked CBD (Ohlsson et al. 1986). There are variations in the reported pharmacokinetic profiles across these studies that may be attributable to the formulation, the chosen study sample (i.e. drug use history), drug-drug interactions (i.e., concomitant seizure medication), and other factors (e.g., food intake with high fat content).

While the consensus is that pharmaceutical-grade CBD has a favorable safety profile with limited side effects (Cannabidiol (CBD) Critical Review Report 2018), this may not generalize across all populations (children, elderly) (Chesney et al. 2020) or across all CBD formulations, as unregulated retail products carry inaccurate labels and have the potential to be contaminated with hazardous chemicals (Bonn-Miller et al. 2017; Corroon et al. 2020; FDA 2019; FDA 2020b; Poklis et al. 2019).

The FDA has approved the use of Epidiolex® for the treatment of seizures associated with Dravet Syndrome, Lennox-Gastaut Syndrome, and tuberous sclerosis complex. Clinical trials investigating the efficacy of CBD for the treatment of pain, autoimmune diseases, psychiatric disorders, substance use, and various other conditions often rely on a single acute dose, but effective doses may vary across disease states. Clinical trials with multiple doses given repeatedly for extended periods of time are needed before CBD can be recommended as a viable pharmacotherapy for these conditions. However, patients are self-treating with over-the-counter, untested CBD products for numerous psychiatric and medical conditions. The lack of evidence supporting the efficacy of CBD for these conditions and the absence of consistent manufacturing quality-control with these unapproved CBD products warrants concern for public health and patient safety.