To date, cannabis is still the most widely used illicit drug worldwide, although meanwhile legalized for recreational purposes in several countries, with, in 2019, almost 4% of the global population (aged 15 to 64 years) having used cannabis at least once, the equivalent of about 200 million people (1). In Central and Western Europe as well as North America, the risk perception associated with cannabis use is on the decrease, while regular cannabis use increased in the long-term, with a prevalence of 7.8 and 14.5%, respectively, in the adult population (1).

This development poses a notable public health issue as recreational cannabis use is associated with considerable adverse health effects, including cognitive deficits, motor impairment and psychosis (2). In addition, about 20–30% of the regular cannabis users have been found to develop a cannabis use disorder (CUD) over time (3). The risk of CUD increases with daily cannabis use, earlier age at first use and higher potency of cannabis (4). According to the Diagnostic and Statistical Manual of Mental Disorders, 5^th Edition (DSM-V), CUD is defined by impaired control, social impairment, risky use, tolerance, and withdrawal (5). A recent meta-analysis reported that almost half of the regular or dependent cannabis users is affected by a cannabis withdrawal syndrome (6) which typically begins soon after cessation of use, peaks within a couple of days and lasts for up to 3 weeks (7). The cannabis withdrawal syndrome is characterized by craving, irritability, nervousness, sleep disorders, depressed mood and decreased appetite (8).

Although the number of people with CUD seeking treatment is increasing in Europe and North America, the overall utilization of CUD-specific treatments is relatively low and the majority of the affected individuals is untreated (9, 10). In the United States, CUD treatment seeking behavior and CUD treatment admissions among young adults (aged 18 to 24 years) have even declined (11). On the other hand, effective treatment options for CUD are limited and focus primarily on psychosocial interventions, including motivational enhancement therapy, cognitive behavioral therapy and contingency management (12). However, access to psychosocial interventions as well as coverage from insurance companies are often limited. Moreover, clinical trials showed that these treatments are cost-intensive and their abstinence rates are only modest and decline after treatment, raising the need for further therapeutic options, especially pharmacological treatments (13–16).

In the last two decades, numerous studies explored the potential of different medications with various pharmacological targets for the treatment of CUD. Human laboratory and clinical studies particularly aimed to identify pharmaceutical agents which are effective in the treatment of cannabis withdrawal syndromes, the maintenance of abstinence, the retention in treatment and the reduction of cannabis use. The most promising candidates included several antidepressants (e.g., bupropion, escitalopram, mirtazapine, nefazodone and venlafaxine), antipsychotics (e.g., quetiapine), anticonvulsants (e.g., valproic acid, gabapentin and topiramate) as well as lithium, buspirone and N-acetylcysteine (17, 18). Although some of these agents produced some benefits for distinct individual aspects in patients with CUD, none of these treatments has demonstrated sufficient empirical evidence to provide clear therapeutic recommendations and to achieve the approval for the treatment of CUD by the authorities, mainly due to insufficient study designs, sample sizes and outcome measures (19).

More recently, the endocannabinoid system and its components have been proposed to provide novel and unique systemic targets for the treatment of CUD. The primary constituent of cannabis, Δ⁹-tetrahydrocannabinol (THC), produces its acute psychoactive effects via partial agonism at the cannabinoid type 1 receptor (CB1) in the central nervous system (20). Regular cannabis use is associated with the development of craving, tolerance and withdrawal symptoms which was related to a dysregulation of the endocannabinoid system, particularly to CB1 receptor downregulation and desensitization and reduced levels of the endocannabinoids N-arachidonoylethanolamide (AEA) and 2-arachidonoylglycerol (2-AG) (21). It was therefore suggested that the potentiation of endocannabinoid function by medical cannabinoids might serve as a promising treatment strategy for CUD (22). In this context, the following cannabinoids and cannabinoid preparations are of particular interest: dronabinol (THC), nabilone, a synthetic derivative of THC, cannabidiol (CBD), a natural inhibitor of the hydrolysis and reuptake of endocannabinoids as well as a negative allosteric modulator of the CB1 receptor, nabiximols, which contains a combination of THC and CBD at a ratio of ~1:1, and PF-04457845, a synthetic inhibitor of the endocannabinoid-degrading enzyme fatty acid amide hydrolase (FAAH).

This systematic review aims to summarize and discuss the main findings of randomized controlled trials (RCTs) evaluating the efficacy, safety and tolerability of different medical cannabinoids in the treatment of CUD.

This study aimed to systematically review the current literature on the use of cannabinoids in the treatment of CUD and to summarize the main findings in terms of efficacy, safety and tolerability provided by randomized controlled trials. We identified eight studies which examined the effects of five cannabinoid preparations on specific clinical outcome variables, i.e., cannabis use, abstinence, withdrawal, craving, retention in treatment and adverse events. Regarding their, at least in part, different mechanisms of action, we could classify the cannabinoids to two therapeutic strategies: (A) agonist substitution and (B) endocannabinoid modulation.

The agonist substitution therapy, also known as replacement therapy, has been proven to be effective in various substance use disorders, particularly in nicotine and opioid dependence. It therefore appeared obvious to test cannabinoid receptor agonists in the treatment of CUD. In this respect, three cannabinoid preparations were of particular interest, dronabinol, nabilone, and nabiximols. Dronabinol and nabilone are currently used as a second line treatment for patients with AIDS/cancer cachexia and for chemotherapy patients experiencing nausea or vomiting (37, 38). Nabiximols is used in the treatment of central neuropathic pain in multiple sclerosis and as an adjuvant analgetic in adults with advanced malignancy (36). In several human laboratory studies, CB1 receptor agonists have been shown to alleviate symptoms of cannabis withdrawal and to reduce relapse in patients with CUD (39, 40).

As expected, dronabinol attenuated cannabis withdrawal symptoms and improved retention in treatment but failed to reduce cannabis use and to improve abstinence, which was the primary outcome of the two studies (26, 27). It was suggested that the CB1 receptor-agonistic properties of dronabinol successfully counteracted the development of withdrawal symptoms, but the low motivation to quit among the participants might have been responsible for the lack of an effect on cannabis use and abstinence. The authors therefore concluded that the participants would have benefited from a longer maintenance period in order to better promote a motivation for sustainable change. Similarly, nabiximols improved withdrawal symptoms, craving and treatment retention, but, in contrast to dronabinol, also reduced cannabis use and improved abstinence (29, 31, 32). This significant difference from dronabinol in the therapeutic profile might be explained by the additional presence of CBD for which “anti-addictive” action has been described recently, see further below (41). On the other hand, nabilone failed to demonstrate any beneficial effects on cannabis use, withdrawal or craving. In this context, the authors speculated whether the dose of nabilone might have been too low in order to display therapeutic efficacy in CUD (28). Moreover, the sample size of 18 participants of whom only 12 completed the overall treatment phase seems too small to draw any robust conclusion on nabilone's efficacy.

The modulation of the endocannabinoid system is a relatively novel approach in the treatment of CUD (22). The endocannabinoid system consists of specific cannabinoid receptors, i.e., the CB1 and CB2 receptor, their primary endogenous ligands AEA and 2-AG, and the AEA- and 2-AG-degrading enzymes FAAH and monoacylglycerol lipase (MAGL) (20, 42). At the molecular level, endocannabinoids play a crucial role in the regulation of various neurotransmitter systems, including the dopaminergic mesolimbic reward pathways (43). It is therefore suggested that the modulation of the endocannabinoid system might have the potential to normalize the dopamine signaling which is typically disrupted by heavy cannabis use, and, thus, appear to be a promising target in the treatment of CUD.

In this context, CBD, the second most abundant constituent of cannabis, is a cannabinoid of substantial interest with regard to the treatment of CUD (41). CBD has only minimal direct action at cannabinoid receptors but primarily acts as an inhibitor of the hydrolysis and reuptake of endocannabinoids (44) as well as a negative allosteric modulator of the CB1 receptor (45), thereby counteracting the acute psychoactive effects of THC (46). As a modulator of the endocannabinoid system, CBD reduced cannabis use, improved abstinence and attenuated withdrawal symptoms (33). The reductions in cannabis use at week-16 follow-up was only evident in the 400 mg-group and the reductions in cannabis withdrawal was only evident in the 800 mg-group. In this case, it can be assumed that the enhancement of the endocannabinergic activity might have contributed to the beneficial effects of CBD.

Another approach for the treatment of CUD also referred to the modulation of the endocannabinoid system by the inhibition of FAAH. In this context, the selective FAAH inhibitor URB597 has been reported to attenuate withdrawal symptoms in an animal model of CUD by increasing AEA concentrations (47). Moreover, mice with reduced FAAH expression due to a genetic variation have been shown to be less likely to develop CUD than the wild-type carriers (48). PF-04457845 is a novel and highly selective and potent human FAAH inhibitor which, so far, was mainly tested in patients with diverse pain syndromes (49). The only study in the present review which evaluated the efficacy of PF-04457845 in CUD showed reduced cannabis use and cannabis withdrawal symptoms compared to placebo (34). The authors suggested that the increase of endocannabinoid concentrations by selective FAAH inhibition might have been the key mechanism contributing to this outcome.

The results of the studies summarized in this review need to be interpreted with caution. First of all, the studies showed a large variety, particularly with regard to population characteristics (low motivation vs. high motivation to quit cannabis use), study setting (inpatient vs. outpatient), study design (forced abstinence vs. harm reduction), concomitant interventions (psychosocial therapy vs. no psychosocial therapy) and operationalization of outcome measures (e.g., cannabis use as assessed by the frequency or the amount of use). These differences impede a meaningful comparison of the efficacy of different cannabinoids. Second, the drop-out rate of about one third among all studies was relatively high which might have affected the significance of the respective study results. Third, specific populations were underrepresented, such as women, older people and individuals from different ethnicities or with comorbid mental disorders or other substance use disorders. In this respect, the study results are less generalizable to other populations and do not picture the real world. Finally, the effects of the different cannabinoids on the various outcome variables were rather modest, most probably due to the relatively small sample sizes.

The agonist substitution (replacement) approach with the CB1 receptor agonist dronabinol showed efficacy in the reduction of cannabis withdrawal symptoms but it was not able to demonstrate an influence on cannabis use or abstinence. In contrast, the modulation of the endocannabinoid system by CBD or the selective FAAH inhibitor PF-04457845 seems to be efficacious for both reducing withdrawal symptoms and improving cannabis use and abstinence. As endocannabinoid modulators, compared to CB1 receptor agonists, also produce lower abuse liability and less intoxication, they appear to be a promising group of drugs for the treatment of CUD (22). However, the evidence is at this time too weak to support any specific medication. Future studies should include greater sample sizes, more diverse populations, longer treatment periods and head-to-head comparisons.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

PR designed the study. CV and PR performed the literature searches, screened the studies, and wrote the first draft of the manuscript. CV, NS, UB, and PR critically contributed to the discussion and approved the final version of the article. All authors contributed to the article and approved the submitted version.

The publication of this study was supported by the Open Access Publication Fund of the University of Duisburg-Essen.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.