Quality & Science

A Review of Potential Pharmacological Treatments for Cannabis Abuse

by Diana Martinez, MD, and Pierre Trifilieff, PhD | April 13, 2015

Currently, only psychosocial therapies, such as cognitive-behavioral therapy, motivational enhancement therapy, contingency management and family-based therapies, are indicated for Cannabis Use Disorders (CUD) (see research from 2007 and 2013). Although the efficiency of these treatments has been shown in scientific studies, not every individual is responsive to such treatment, and the rate of relapse and drop out rates remain high (about 70%). Therefore, medications that can be combined with these therapies are being studied to improve the efficacy of CUD treatment, as reviewed recently by Balter et al

Agonist Pharmacotherapies
Agonist therapies aim at mimicking some effects of a drug of abuse, mainly to attenuate the effects of withdrawal and craving.  They show efficacy in treating some addictions, such as nicotine replacement therapy for smoking or methadone for opiate dependence. Cannabis contains many cannabinoids (>80), though delta 9-tetrahydrocannabinol (THC) is the primary psychoactive component. Oral THC produces the positive subjective effects and intoxication, and the THC concentration of the smoked plant correlates with ratings of “good drug effect” (see Balter et al.). The main receptor through which THC induces its effects is the cannabinoid type 1 (CB1) receptor, which is highly enriched in the brain. The brain produces its own ligands that bind to the CB1, anandamide (N-arachidonoylethanolamide, AEA) and 2-arachidonoylglycerol (2-AG). The strategy behind agonist therapies for cannabis is to stimulate the CB1 receptor - either with synthetic molecules or endogenous ligands – in order to partially mimic some effects of THC to diminish cannabis consumption and/or alleviate symptoms associated with craving and withdrawal.

Dronabinol (Marinol): Dronabinol is an oral form of synthetic THC that is FDA approved to treat chemotherapy-induced nausea and AIDS wasting syndrome. Human laboratory studies have been performed to investigate the effect of dronabinol on cannabis self-administration, cognition, and withdrawal symptoms in non-treatment seeking participants. In the studies, the volunteers were given the opportunity to self-administer cannabis in a laboratory setting, before and after dronabinol administration. While these self-administration studies were performed in volunteers who are not seeking treatment, they have the advantage of providing data on the potential efficacy of a medication in a small number of volunteers and evaluating the efficacy of a potential medication prior to embarking on clinical trials, which require large numbers of subjects and are long and expensive. 

The self-administration laboratory showed that that 50 to 120 mg/day dronabinol significantly reduced a number of withdrawal symptoms, including anxiety, misery, chills, sleep disturbance and appetite loss (see Haney et al. and Vandrey et al.). Given that withdrawal in CUD is a main factor of relapse, these results indicated that dronabinol might serve as a potential pharmacotherapy. 

However, some studies have shown that even though it reduces withdrawal, Dronabinol is not effective at reducing cannabis self-administration (Haney et al. and Hart et al.). Similarly, a clinical trial that combined dronabinol with behavioral therapies (motivational enhancement and relapse prevention therapy) also showed no benefit over placebo. However, the dronabinol group did have better treatment retention compared to the placebo group (77% vs 61%) and reported fewer withdrawal symptoms.

Nabilone: Nabilone is a synthetic analogue of THC, which is also FDA approved to treat chemotherapy-induced nausea. Recent human laboratory studies, using the methods described above, show that nabilone is promising as a potential treatment for CUD. Like dronabinol, nabilone reduces withdrawal symptoms. However, nabilone was shown to also decrease subjects’ choice to self-administer cannabis in the laboratory. 

The significant improvement of nabilone over dronabinol may result from: 1) the better bioavailability of nabilone (>60%) compared to dronabinol (<20%), since it is not metabolized as quickly; and 2) its longer duration of action (>6 hours) compared to dronabinol (4 hours), resulting in fewer daily doses needed, which improves compliance. Another advantage of dronabinol is that it does not interfere with urinary toxicology tests. Because nabilone produces urinary metabolites that are different from those produced by either dronabinol or cannabis use, clinicians can determine whether patients are taking the medication or continuing their cannabis use.

Fatty acid amide hydrolase (FAAH) and monoacylglycerol lipase (MGAL) inhibition: In the brain, the levels of endogenous cannabinoids are modulated by enzymes able to metabolize AEA and 2-AG, providing a “brake” on their signaling. These enzymes, fatty acid amide hydrolase (FAAH) and monoacylglycerol lipase (MGAL), can be pharmacologically inhibited. Inhibition of these enzymes results in stimulating the CB1 receptor by increasing the levels of its endogenous ligands. Experiments in mice chronically exposed to THC show that inhibiting these enzymes reduces withdrawal. Human research to investigate the safety and efficacy with a FAAH inhibitor (currently called PF-04457845) is now being performed at the Yale School of Medicine (NCT01618656).

Antagonist Pharmacotherapies
Whereas the goal of agonist therapies to treat CUD is to mimic some effects of THC, antagonist pharmacotherapies aim at “blocking” the CB1 receptor to prevent the action of THC. This antagonist strategy is often used to treat opiate addiction. Naltrexone (and naloxone) for example bind to and block the mu opioid receptor, preventing opiates from having their effect on the brain. for which treatment by naltrexone or naloxone prevents the ability of opiates to affect the brain by binding the mu opioid receptor.  Rimonabant is an antagonist at the CB1 receptor, and an early human laboratory study showed that rimonabant reduced the subjective effects and “high” of cannabis administration (see Huestis et al from 2001 and from 2009). However, clinical research investigating rimonabant as a potential treatment for obesity were stopped when it was discovered that it produced increased anxiety, depression and suicidality. As we described in ASAM Magazine previously, research is now being conducted to develop a medication that would partially blocks the CB1 receptor, in such a way that it would decrease the subjective effects of cannabis, but avoid these side effects. 

Non-Cannabinoid Pharmacologies
The opioid and cannabinoid receptor systems interact to regulate each other, and animal studies showed that opioid antagonists reduce cannabinoid self-administration (Navarro et al. and Justinova et al.).  However, human laboratory studies have mixed effects: acute dosing of naltrexone increases the “high” produced by cannabis, while repeated naltrexone administration reduces the positive subjective effects and may therefore serve as an adjunctive pharmacotherapy for CUD (Haney M et al. “Naltrexone maintenance decreases cannabis self-administration and subjective effects in daily cannabis smokers”. Neuropsychopharmacology, in press).

Preclinical data show that cannabinoid withdrawal results in hyperactivity of the noradrenergic neurotransmitter system. In animal models, agonists of the alpha-2 adrenergic receptor decrease activity of the noradrenergic transmission and reduce some symptoms of THC withdrawal. Lofexadine is an alpha-2 adrenergic receptor agonist that is used to treat opiate withdrawal in Europe. In human laboratory studies, lofexadine reduced cannabis withdrawal, and lofexadine combined with dronabinol showed even more beneficial effects by improving mood, attenuating withdrawal symptoms and craving, and reducing cannabis self-administration in the laboratory. A clinical trial is now being conducted at the New York State Psychiatric Institute to investigate the efficacy of lofexadine combined with dronabinol to reduce cannabis use in treatment seeking cannabis-dependent patients (NCT01020019).

Pharmacotherapies Targeting GABA and Glutamate
Numerous studies in animal models have demonstrated that THC – through its action on the CB1 receptor - alters both GABA and Glutamate transmission, which are two main regulation systems in the brain. Therefore, pharmacotherapies aiming at restoring the activity of these two neurotransmitter systems have been tested in human studies.

Baclofen, a derivative of GABA, did not improve withdrawal or relapse in a human laboratory study, although craving was reduced. Additionally, baclofen had deleterious effects on cognition.

Gabapentin is an antiepileptic and analgesic for neuropathic pain. Its mechanism of action is unclear since it binds poorly to GABA receptors, even though its structure is similar to GABA. However, it was found to increase GABA biosynthesis.  Gabapentin was tested in a clinical trial for CUD (1,200 mg/day) and although the dropout rate was high, it was shown to attenuate the severity of withdrawal and to reduce cannabis use compared to placebo. As a result of this initial finding, a larger clinical trial is currently being performed at Scripps Research Institute (NCT00974376).

N-acetylcysteine (NAC) is an over-the-counter supplement that up-regulates cystine-glutamate exchange and is thought to restore the normal glutamate activity that is disrupted by chronic drug use.  A recent clinical trial investigating NAC administration (1,200 mg, twice daily), in combination with contingency management as a behavioral treatment, showed that subjects receiving NAC were more likely to reduce their cannabis consumption compared to contingency management alone. A multi-site trial of NAC for the treatment of CUD is currently underway (NCT01675661).

Antidepressant and Antipsychotics
Both antidepressants and antipsychotics have been studied for CUD given their potential to address some of the symptoms of withdrawal, such as insomnia, anxiety and irritability. However, these studies show that, overall, these medications are not effective in the treatment of CUD.

A human laboratory study of mirtazapine demonstrated that it improved some symptoms of withdrawal, but did not reduce cannabis self-administration. Escitalopram, a selective serotonin reuptake inhibitor, was studied in a clinical trial, and failed to address the symptoms of depression or anxiety in cannabis-induced withdrawal and did not improve abstinence over placebo. A clinical trial with buspirone, a partial serotonin receptor antagonist, showed no significant effect on anxiety, withdrawal, or craving compared to placebo, and patients dropped out at a higher rate. 

Two clinical trials have investigated antidepressants as a treatment in co-morbid CUD and depression and these also demonstrated a lack of effect. One study used the selective serotonin reuptake inhibitor fluoxetine (10 to 20 mg/day) in adolescents, while another study used venlafaxine (375 mg/day), and showed that cannabis use actually increased in the treatment group compared to the control group.

The antipsychotic quetiapine was tested in a human laboratory study, and while it decreased some of the withdrawal symptoms in CUD, in actually increased both craving for cannabis and relapse. 

There is currently no pharmacological treatment approved for the CUD, but recent studies have shown that some medications may be effective in alleviating withdrawal symptoms and reducing cannabis use. These include nabilone, lofexadine combined with dronabinol, gabapentin and N-acetylcysteine. Larger trials are needed to confirm these findings and are currently underway. Overall, these studies indicate that treatment strategies that combine pharmacotherapy with established behavioral treatments may be the most effective approach.

Medications that have shown no benefit over placebo include the antidepressants, antipsychotics, baclofen, or rimonabant. There may be a role for naltrexone, but further studies are needed first. Clinical experience with the medications that block the CB1 receptor, such as rimonabant, showed that inhibiting transmission at this receptor produces serious side effects (depression and suicidality). Alternatively, animal experiments show that inhibiting the break down of brain’s own endogenous ligands that activate cannabinoid receptors – which increases CB1 receptor signaling - may be another treatment approach for CUD. These studies highlight the complexity of the endocannabinoid receptor system in the brain and its role in finely modulating emotions and cognition. A more thorough understanding of this system at a mechanistic level is needed to better understand CUD and to develop better pharmacotherapies.

Diana Martinez, MD

Dr. Martinez is an Associate Professor at Columbia University/New York State Psychiatric Institute. She is a psychiatrist and imaging researcher whose work has focused on using Positron Emission Tomography (PET) imaging in drug addiction. PET imaging allows the measurement of dopamine receptors and dopamine release in the human brain, and her work focuses on using this imaging technique, based on animal models of addiction, to better understand the neurochemistry of substance use disorders. Through these types of studies, her work is geared toward developing innovative treatments for addiction.

Pierre Trifilieff, PhD

Dr. Trifilieff is an Assistant Professor at INRA in the University of Bordeaux. His research focuses on the role of the mesolimbic dopaminergic transmission in physiologic and pathological conditions. Since the activity of the dopaminergic D2 receptor is altered in various psychiatric disorders that involve a dysregulation of the reward system, his work aims at unraveling the role of D2 receptor-dependent signaling in the modulation of reward processing and motivation. This includes studying the impact of D2 receptor manipulations on goal-directed behaviors as well as identifying environmental factors that impact D2-dependent signaling and related behaviors.