Effectiveness Study of Dronabinol and BRENDA for the Treatment of Cannabis Withdrawal

1. Background: Cannabis is the most widely used illicit drug in the U.S. According to the 2004 U.S. The 2004 NSDUH estimates there are 3.2 million daily or almost daily marijuana smokers in the United States. Recent reports have estimated that 1.5% of the US population meets DSM-IV diagnostic criteria for cannabis abuse or dependence, with the sharpest increases among young black men and women and young Hispanic men (2). Cannabis users commonly endorse adverse psychosocial and medical effects from their cannabis use, which include dysphoria, loss of control over use, cognitive impairment, and strained social relationships.

In humans, the cannabis withdrawal syndrome is a constellation of affective and behavioral symptoms that occur within 24 to 48 hours after abrupt cessation of marijuana use with a gradual return to baseline after 1-2 weeks. A recent review of the literature on cannabis withdrawal identified consistent reports of anxiety, irritability, physical discomfort, insomnia, and appetite suppression to be associated with abrupt cessation of cannabis use in heavy users across controlled inpatient and outpatient trials.

The agonist dronabinol (Marinol) has shown promise in the prevention of cannabis withdrawal in human lab studies and its reported amelioration of anxiety, misery, insomnia, drug craving and appetite suppression from abrupt cannabis cessation provides the rationale for its use as an agent to prevent cannabis withdrawal in an outpatient, treatment seeking population(9). Dronabinol has also been used as a brief "maintenance" treatment in a cannabis self-administration study in humans. Currently, dronabinol is indicated for treatment of AIDS-related anorexia and nausea associated with cancer chemotherapy. Two principle aims of our proposal are therefore to gain familiarity with cannabis withdrawal in a naturalistic outpatient treatment setting and compare an agonist therapy to placebo to attenuate cannabis withdrawal and enhance treatment outcome.

The general pattern of low rates of total abstinence and lack of significant differences between high intensity psychotherapy treatment vs. low intensity and voucher-based treatment holds across all psychosocial treatment studies. In keeping with prior evidence supporting the relative efficacy of brief interventions for cannabis dependence, the current proposal will apply a well-established brief medicalized psychotherapy known as B.R.E.N.D.A. for the treatment of cannabis dependence.

To date, studies evaluating cannabis patients have primarily focused on identifying neurotoxic drug effects such as impaired learning, memory, attention and executive function. We propose to more fully examine the behaviors with relevance to addiction vulnerability and relapse. Such behaviors include but are not limited to poor inhibition, high risk-taking, positive affective bias toward drug stimuli, poor affect regulation, and poor decision-making. We plan to utilize standardized tasks and questionnaires to objectively characterize dimensions of these addiction relevant behaviors in cannabis patients and to examine the effect of cannabis withdrawal and dronabinol treatment on these behavioral measures. We will also correlate brain differences (structural and functional) with behavioral measures in cannabis dependent patients before and after treatment with dronabinol as compared to placebo.

The search for effective treatment of addictive disorders is limited by our incomplete knowledge about the brain substrates critical for addiction and for addiction recovery. The recent use of PET and fMRI has allowed researchers to examine connections between brain activity and vulnerability to addiction and relapse. In marijuana users, PET and fMRI techniques have been used to study regional brain volumes, blood flow, metabolism. However, understanding cue induced craving is more immediately relevant to treatment and relapse prevention in cannabis addiction since cannabis users have a high rate of relapse compared to those found for other drugs of abuse. In addiction, cues that have been consistently associated with drug reward can trigger craving for the drug. Clinically, the craving for the drug is sufficiently compelling in many cases to precipitate a relapse despite significant effort on the patient's part to maintain abstinence and despite significant negative consequences. Although PET and fMRI techniques have been successfully used to study cue induced craving in other drugs of abuse (e.g. cocaine) leading to greater understanding of the mechanism of relapse (limbic activation in response to drug cues), these techniques have not been utilized in cannabis patients. Several preliminary studies used patient self-report to demonstrate cue induced craving in cannabis patients. We propose to utilize physiological measures and functional neuroimaging to objectively demonstrate cue-induced craving in cannabis dependent subjects and to identify the brain substrates that mediate this craving. Similarly, dronabinol reduced self-reported drug craving in one recent human laboratory study. We propose to examine whether dronabinol will change physiological measures and brain activation in response to drug cues.

1. Study objectives:
2. To evaluate the cannabis withdrawal syndrome in a naturalistic outpatient setting as one possible cause for the low rates of total abstinence seen in our trial and all other clinical trials of adult cannabis dependence.
3. Primary Hypothesis: Subjective ratings of cannabis withdrawal can be assessed reliably in an outpatient treatment setting.
4. Secondary Hypothesis: Subjects with high subjective ratings of cannabis withdrawal will have lower rates of abstinence regardless of treatment group (dronabinol vs. placebo).
5. To determine the feasibility and efficacy of a three-week trial of a cannabinoid agonist medication under double blind placebo controlled conditions to treat cannabis withdrawal.
6. Primary Hypothesis: Dronabinol can be used safely at a dosage of 10mg qid (four times per day) as an outpatient treatment for cannabis withdrawal
7. Secondary Hypothesis: Dronabinol will perform better than placebo in attenuating the symptoms of cannabis withdrawal.
8. To determine the feasibility and efficacy of providing brief psychotherapy (BRENDA) to patients seeking treatment for cannabis dependence.
9. Primary Hypothesis: Six sessions of BRENDA therapy can be applied to the treatment of cannabis dependence.
10. Secondary Hypothesis: Six-sessions of BRENDA therapy for cannabis dependence is an adequate "dose" of individual therapy for subjects motivated to stop smoking marijuana.
11. To characterize the neurocognitive aspects (e.g. attention, working memory, impulsivity, decision making, risk-taking, affective bias/drug preference, reward and punishment sensitivity, affect regulation) of cannabis dependence, cannabis withdrawal and dronabinol effects by utilizing a battery of tasks, questionnaires and interviews.
12. Primary Hypothesis: Poor performance on neurocognitive tasks and questionnaires will be correlated with duration and amount of cannabis use.
13. Secondary Hypothesis: Cannabis withdrawal will be correlated with impaired performance on neurocognitive tasks and questionnaires. Dronabinol treatment will reduce this impairment on tasks of affect regulation and affective bias/drug preference compared to placebo.
14. To objectively demonstrate cue-induced craving in cannabis dependent patients and to determine whether baseline measures of our cannabis patients' brain vulnerabilities (of structure, and of function; e.g., resting perfusion) can predict both brain and behavioral outcomes.
15. Primary Hypothesis: Cue-induced craving will be positively correlated with changes in physiological measures and increased amygdalar activation and negatively correlated with prefrontal activation in cannabis dependent patients during cannabis discontinuation, and dronabinol treatment will reduce the intensity of cue- induced physiological changes and brain activation compared to placebo.
16. Secondary Hypotheses: Measures of prefrontal cortex gray matter and resting perfusion in prefrontal cortex, cingulate cortex and amygdala will be positively correlated with performance on neurocognitive tasks and questionnaires as well as treatment outcome.
17. Location: This is a single site study. All patient recruitment and data collection will occur at the Treatment Research Center at the University of Pennsylvania, 3900 Chestnut Street, Philadelphia, PA 19104, except for the Neuroimaging which will take place at the HUP6 fMRI in the basement of the Hospital of the University of Pennsylvania, 34th and Spruce Streets, Philadelphia, PA 19104.

STUDY DESIGN

1. Research design \& methodology: This is a Phase 2 double blind placebo controlled trial. We will recruit 60 cannabis dependent subjects and treat them with the combination of dronabinol 10mg QID (four times per day) and BRENDA or placebo qid (four times per day) and BRENDA to reduce their consumption of cannabis. Subjects will be 60 men and women with current DSM-IV diagnosis of cannabis dependence. All patients will receive six sessions of BRENDA at visits 1, 3, 6, 9, 11, and 12. The study length for each patient will be one week for screening and baseline self-report and neurocognitive battery measures. Patients who qualify will be offered an opportunity to participate in the fMRI portion of the study with baseline fMRI obtained in the first week prior to starting medications and second fMRI obtained during week two of medications. This is followed by 3 weeks of medication and, after completing medications, three additional weekly visits (two for BRENDA and one for final assessment).
2. Duration: Each subject will participate in a 7 week treatment trial with the following schedule of visits: STUDY VISITS

Week 1 (Baseline Measures):

Week 2 - 4 (Medication Treatment):

Weeks 5-6 (Medication Washout)

Week 7 (Final Visit):

We anticipate that we will require two years to complete data collection with 60 subjects and will require one year for data analysis. Thus, the proposed study will be completed in approximately three years.

SUBJECT SELECTION \& WITHDRAWAL