Efficacy, pharmacokinetics, and safety of oral cannabinoids
Efficacy of oral cannabinoids
The first placebo-controlled study demonstrating the efficacy of THC for the treatment of CINV in patients with cancer was published in 1975 . Since publication of this initial report, numerous clinical studies have examined the antiemetic efficacy of oral dronabinol or nabilone capsules for the treatment of patients with CINV, with two meta-analyses reporting significant improvements with cannabinoids compared with conventional antiemetic therapy [23, 24], and one meta-analysis finding that, while the antiemetic efficacy of cannabinoids was favored, compared with the antiemetic prochlorperazine for the resolution of nausea, vomiting, or nausea and vomiting (Table 1), the findings did not achieve significance . However, studies included in these meta-analyses differed in methodology (e.g., crossover study, blinding), discontinuation rates, sample sizes, timing of drug administration, tumor type, and chemotherapeutic agent(s) used [23,24,25]. Further, meta-analyses differed not only in the specific outcomes analyzed (e.g., antiemetic efficacy vs resolution of nausea or vomiting, or both), but also in the specific studies included in the evaluation of a particular efficacy outcome (i.e., number of studies) and whether cannabinoids were evaluated individually or by drug class [23,24,25].
In addition, a pooled analysis of 14 studies indicated that a significantly greater percentage of patients preferred cannabinoids compared with conventional antiemetics for the treatment of CINV [61 vs 26%, respectively; relative risk (RR) 2.4; 95% CI 2.1–2.8; number needed to treat, 2.8] . These data were confirmed by a 2015 meta-analysis of nine studies that also reported patients preferred cannabinoids to conventional antiemetic agents (RR 2.8; 95% CI 1.9–4.0) . Reasons for these similar findings were not provided, but the results may be antiemetic-dependent, as there were no differences between patient preference for cannabinoids over metoclopramide or chlorpromazine in single studies with a small number of participants (N = 40 and N = 64, respectively) .
Medical use of marijuana is controversial and no clinical trials have been conducted to date to compare the antiemetic efficacy of medical marijuana with the conventional antiemetic agents recommended as first-line therapy by the NCCN Clinical Practice Guidelines in Oncology for Antiemesis [17, 26]. Medical marijuana is currently not recommended by the NCCN for antiemesis in CINV . However, given that approximately half of all states in the United States have approved the use of medical marijuana, health care providers are increasingly likely to encounter patients interested in receiving medical marijuana for antiemesis .
Variability in pharmacokinetics and pharmacodynamics of cannabinoids
Lack of antiemetic efficacy (i.e., failure to decrease incidence of CINV) of oral THC was initially reported in patients with sarcoma receiving chemotherapy. In these patients, the lack of antiemetic efficacy was thought to possibly be associated with the type of chemotherapeutic agent administered . However, absorption of THC was also highly variable, with decreased incidence of nausea and vomiting associated with higher drug plasma concentrations of THC (50% incidence at >5 ng/mL vs 83% at <5 ng/mL). Subsequent studies of orally administered THC have confirmed high PK variability in healthy individuals (Table 2) [28,29,30,31,32,33]. Oral absorption of dronabinol is high (90–95%), but slow and variable . Peak plasma concentrations of dronabinol and its metabolites have been observed at ~2 h postdose with dronabinol capsule in healthy individuals [31, 32]. Variability in peak plasma concentrations (C
max) was estimated between 150 and 200% . In healthy individuals currently reporting cannabis use, administration of supratherapeutic doses of THC (i.e., 75–90 mg) were associated with high interindividual variability [C
max range 9.0–127.1 ng/mL; time to C
max) range 1–12 h] .
A new tablet formulation of THC designed to improve drug uptake demonstrated a more rapid T
max in healthy individuals, with interindividual variability in C
max ranging from 42 to 62% . In one study, peak plasma levels of oral THC (e.g., tablet, capsule) were lower and were achieved over a longer duration compared with IV THC, which reached peak plasma concentrations within 20 min of administration (C
max, 62 ng/mL) . In this study, participants reported a maximal “high” feeling, a subjective psychological effect of THC use, 30–50 min after IV administration, compared with 2.5–3 h after oral THC administration. By further comparison, participants who smoked marijuana reported a maximal “high” feeling at 30–90 min postdose, long after C
max was achieved. These physiologic effects may be mediated by the major metabolites of THC, including 11-hydroxy-Δ9-tetrahydrocannabinol (11-OH-THC) and 11-nor-9-carboxy-Δ9-tetrahydrocannabinol (THC-COOH) [32, 35], which were shown to have a longer T
max than THC when marijuana was smoked (i.e., T
max for a marijuana cigarette containing 3.55% THC: THC, 0.14 h; 11-OH-THC, 0.2 h; THC-COOH, 1.35 h) . Systemic availability of oral THC is lower than that of smoked THC (4–12% vs 8–24%, respectively) [34, 37], as the systemic availability of oral THC is limited by extensive first-pass hepatic metabolism .
With regard to medical marijuana, dosing of smoked marijuana is variable, given interindividual differences in frequency and depth of inhalations, and the type of cannabis selected, as cannabinoid content varies by blend [28, 29]. High interindividual PK variability was demonstrated in a study of healthy individuals smoking low- and high-dose cannabis (i.e., containing 1.75 and 3.55% THC, respectively) . The smoking protocol of this study included a 2-s inhalation, 10-s hold period, and 72-s exhalation and rest period for a total of 8 puffs over 11.2 min. Peak plasma concentrations of THC were observed 8.4 min after the first puff, with a mean C
max of 84.3 ng/mL (range 50–129 ng/mL) for low-dose cannabis and 162.2 ng/mL (range 76–267 ng/mL) for high-dose cannabis; THC plasma concentrations then decreased rapidly to 17.3 and 29.7 ng/mL, respectively, after 30 min.
Inhaling vaporized THC reduced exposure to harmful byproducts produced by smoking cannabis [26, 38]. Peak plasma concentrations of THC after inhalation of low- or high-dose vaporized cannabis (2.9% THC, 46.5 µg/L; 6.7% THC, 62.1 µg/L) were achieved within 10 min of administration in healthy regular users of cannabis . Interindividual PK variability was observed with inhalation of vaporized THC, which was attributed, in part, to rate and depth of inhalation and time THC was held in the lungs of participants, as well as factors associated with delivery of vaporized THC, including heating temperature, number of balloon fills, and the amount and type of cannabis used .
The PD of oral THC is variable and differs from that of smoked or IV THC (Fig. 1) [30, 37, 39]. The PK/PD profile of orally administered THC (i.e., cookie) was shown to differ from that of smoked and IV administration of the drug in individuals with prior cannabis exposure . Maximal feeling of “high” was achieved 30 min after smoking or IV administration of THC, and declined to baseline levels after 4 h; in contrast, after oral administration of THC, maximal feeling of “high” was slower in onset (i.e., 2–4 h), with a decline to baseline levels after 6 h. Peak plasma concentrations were achieved within 3 min following smoking or IV administration, but within approximately 1 h with oral administration. Plasma THC concentrations and the degree of “high” experienced by participants had high intra- and interindividual variability. Further, clinical signs of cannabis intoxication (e.g., reddening of conjunctivae, increased pulse rate) differed between smoked and IV administration vs oral administration . Reddening of the conjunctivae reached a maximum effect by 10 min following smoking and IV administration, compared with a maximal effect observed 1–3 h after oral administration. In general, reddening of conjunctivae occurred with plasma THC concentrations >5 ng/mL, even in the absence of feeling “high.” The median increase from baseline in pulse rate was comparable between smoking and IV administration: an increase of 34 beats per minute (bpm) was observed with a median THC concentration of 45 ng/mL obtained via smoking, vs 40 bpm with a median plasma concentration of 100 ng/mL via IV administration. The effect of oral administration on pulse rate was lower compared with smoking and IV administration (26 bpm with a median THC concentration of 4.5 ng/mL). Pulse rate often returned to baseline or below while plasma concentrations remained >5 ng/mL and patients still reported feeling “high” . Thus, it is apparent that plasma THC concentrations >5 ng/mL correlate better with reddening of conjunctivae than pulse rate, although both are considered clinical indicators of cannabis intoxication.
Single-dose oral administration of THC tablets (3.0, 5.0, or 6.5 mg) in healthy adults ≥65 years of age indicated no association between plasma THC concentrations and eyes open-body sway scores [P = 0.1; determined by SwayStar™ (BESTec-etp Freiburg GmbH, Freiburg, Germany), a device used to measure body movement when standing with eyes open or closed]. However, the eyes open-body sway scores were associated with plasma concentrations of the THC metabolites 11-OH-THC and THC-COOH . This is a potentially clinically relevant finding in the context of falls, which are a primary cause of morbidity and mortality in the elderly . In the same study of THC tablets, alertness scores were not associated with plasma concentrations of THC (P = 0.5), or its metabolites 11-OH-THC (P = 0.7) and THC-COOH (P = 0.8) .
The high PK variability of oral THC tablets and capsules may compromise accurate and consistent dosing of dronabinol . The oral dronabinol solution formulation, approved by the US Food and Drug Administration (FDA) in July 2016, has been shown to have less variability, with drug detected in plasma in 15 min in 100% of individuals receiving this formulation compared with <25% of individuals receiving dronabinol capsule (Fig. 2) . Further, the intraindividual variability in the mean area under the concentration–time curve from time 0 extrapolated to infinity (AUC0–∞) was decreased with oral dronabinol solution compared with the capsule. These findings have important clinical implications for patients with CINV, as patients may derive therapeutic benefit faster with oral dronabinol solution than with dronabinol capsule. Further, decreased intraindividual variability with oral dronabinol solution vs capsule may minimize the need to individualize dosing to obtain optimal therapeutic effects . However, if needed, individualized dosing based on body surface area and titration of dosing to achieve clinical benefit is supported by current US labeling .
Safety of cannabinoids
Cannabinoids are associated with a number of potential adverse effects, including a “high” feeling, euphoria, disorientation, and depression . Adverse effects of cannabinoids on non–central nervous system functions (e.g., tachycardia, reddening of conjunctivae, decreased GI motility) are attributed to the ubiquitous localization of cannabinoid receptors throughout the body . Results of a meta-analysis showed that patients receiving oral dronabinol or nabilone capsules had greater incidence of adverse effects compared with those receiving conventional antiemetic therapy or placebo: dizziness, 49 vs 17%, respectively; hypotension, 25 vs 11%; dysphoria or depression, 13 vs 0.3%; hallucinations, 6 vs 0%; and paranoia, 5 vs 0% . Medical use of cannabinoids (including oral dronabinol and nabilone capsules) for various conditions, including CINV, chronic pain, spasticity related to multiple sclerosis or paraplegia, human immunodeficiency virus/AIDS, and sleep disorder, was associated with a greater risk of adverse effects compared with an active comparator or placebo in 62 studies . Overall, the most common adverse effects following administration of cannabinoids included disorientation (OR 5.4; 95% CI 2.6–11.2), dizziness (OR 5.1; 95% CI 4.1–6.3), euphoria (OR 4.1; 95% CI 2.2–7.6), confusion (OR 4.0; 95% CI 2.1–8.0), and drowsiness (OR 3.7; 95% CI 2.2–6.0). Although data are limited, single-dose dronabinol oral solution was shown to be generally well tolerated, with nausea, dizziness, somnolence, and headache the most common adverse events reported by healthy volunteers .
Considerations for oral cannabinoids compared with medical marijuana
Numerous routes of administration are available for patients with cancer receiving medical marijuana, including smoking, oral (e.g., cookie, candy, beverages), and mucosal [44,45,46]. In contrast with oral cannabinoids (i.e., dronabinol, nabilone), medical marijuana is not currently regulated by the FDA . Thus, there is currently a lack of standardization regarding dosing and potency across available medical marijuana formulations; additionally, the potential for food safety issues cannot be excluded for users of oral products (i.e., foodstuffs, beverages) [44,45,46, 48]. While edible medical marijuana products are required to have child-resistant packaging and labeling, unintentional pediatric exposures may still occur [47, 49]. Use of smoked marijuana for medical purposes by patients with cancer has several limitations, including a patient’s inability to tolerate smoked marijuana due to taste or the potential for airway obstruction, which may result from inflammation of the airway following smoking [50, 51]. Medical marijuana may also increase the risk for atrial fibrillation, myocardial infarct, and chronic bronchitis . Further, patients who are immunocompromised may risk additional immunosuppression (e.g., by suppressing lymphocyte proliferation) following use of medical marijuana [26, 52]. In addition, insurance will generally cover the costs associated with use of cannabinoids approved by the FDA, but not medical marijuana. Overall, additional studies comparing the safety and efficacy of oral cannabinoids with various formulations of medical marijuana are needed.