After more than a decade of intensive efforts, researchers have yet to find a medication effective in human cocaine craving and dependence. Though brain dopamine (DA) systems have regularly been implicated both in cocaine reward and cocaine craving, direct manipulation of these systems with available agonists or antagonists has provided underwhelming clinical results: dopamimetics often seem to worsen craving (perhaps by acting as an internal cue reminiscent of cocaine), and neuroleptics have side-effects which make them unappealing as treatment agents. Recently, Roberts et al. discovered that GABA B agonists have, by far, the most selective effect on cocaine reward and incentive motivation of any compound (among dozens) systematically tested by their laboratory. Other evidence suggests GABA B agonists can inhibit DA release, offering a possible mechanism for their potential to blunt both the direct, and conditioned, effects of cocaine. Inspired by Dr. Roberts' preclinical findings, our clinical site initiated pilot work with baclofen, a prototypic GABA B agonist used safely in humans for decades as an anti-spastic. This research determined a dose-range of baclofen well-tolerated by cocaine patients, and additionally suggested that it may block cue-induced cocaine craving and """"""""high"""""""". The long-term objective of the application is to provide coordinated and cost-effective development and testing of the GABA B agonist baclofen in cocaine dependence, making use of the considerable clinical expertise of Dr. Childress' team in cocaine treatment, cocaine cue reactivity, medication trials, and safety trials employing cocaine administration. As any candidate medication for cocaine treatment will eventually be used in conjunction with cocaine, examining the interaction of the agent with cocaine under protected, controlled conditions is an important safety step in medication development. In drug development, medications are usually first tested for safety, then for efficacy (effect under optimal, controlled conditions), and finally for """"""""real-world"""""""" effectiveness. This sequence often entails several independent experiments over a very extended period of time. In contrast, Study 1 in the current application offers an innovative inpatient safety trial which will obtain several distinct efficacy measures, all within the same experiment. Specifically, Study 1 will determine 1) the safety of baclofen in combination with cocaine, 2) the efficacy of baclofen in blocking cocaine's subjective (particularly rewarding) effects, 3) the efficacy of baclofen in reducing the craving and arousal induced by cocaine-related cues, and 4) the efficacy of baclofen in blocking cocaine-triggered craving. Though this information could be collected in several independent experiments, the proposed design results in considerable savings in the total number of patients to be studied, and greatly increases the speed with which findings in these critical domains can be acquired. Three groups of cocaine inpatients (n=20 per group) will be studied in a double-blind, between-groups 10mg b.i.d., 10mg b.i.d., 20mg b.i.d.), random-assignment design. An additional placebo group will examine the effect of expecting cocaine on cue reactivity, an issue of direct relevance for use of cue methodologies in medications development. Study 2 will subsequently examine the effectiveness of baclofen vs. placebo (n=30 per group) in the treatment of cocaine outpatients, using a double-blind, random-assignment, controlled 12-week trial with subsequent medication and treatment taper. Dependent variables in Study 2 will include drug use as measured by frequent urine toxicologies, ambient and cue-elicited craving. treatment retention, and standard measures of clinical function.
