Dopamine antagonists have been used for their euphoria-blocking effects via receptor blockade although their ability to influence cocaine self-administration is problematic. Many dopamine receptor antagonists have had initial success in pre-clinical work and acute dosing studies but have failed in clinical trials. Ecopipam is a recent example of a dopamine antagonist where early success in attenuating cocaine effects was not sustained in clinical trials and indeed in one study increased cocaine induced euphoria and cocaine use (Haney, Ward, Foltin & Fischman, 2001; see also editorial by McCance-Katz, Kosten & Kosten, 2001; Nann-Vernotica, Donny, Bigelow & Walsh, 2001). Flupenthixol is a dopamine blocker that acts as an antidepressant at low oral doses (1-3 mg/day) and as a neuroleptic at higher intramuscular doses (30-100 mg/two week depot). It has shown promise in cocaine users with schizophrenia (Levin, Evans & Kleber, 1998) and also potential as an aversive agent (Gawin, Khalsa-Denison & Jatlow, 1996). However, a double blind placebo controlled study of the safety and tolerability of both oral and intramuscular flupenthixol in 23 cocaine users found low doses were well tolerated but ineffective in attenuating subjective and cardiovascular responses to intravenous cocaine, while high neuroleptic doses produced unacceptable dystonic reactions in subjects (Evans, Walsh, Levin, Foltin et al., 2001). The neuroleptic, haloperidol, produces receptor supersensitivity making it unsuitable for long-term treatment (Sevarino et al., 2000).

Antipsychotic agents have been examined due to their potential for those with comorbid psychostimulant dependence and psychotic disorders, for the treatment of psychostimulant-induced psychosis and for their dopamine antagonist properties. A randomised, double-blind study of risperidone, an atypical antipsychotic with 5HT and D2 antagonist properties, was conducted in 193 cocaine users without other psychiatric diagnoses (Grabowski, Rhoades, Schmitz, Silverman et al., 2000).

The study was terminated early due to poor retention and numerous side-effects in the risperidone groups with no reduction in cocaine use. The investigators speculated that highly selective antagonists would not be successful because they may not block the full range of neurobiological actions of psychostimulants or that the doses needed to effectively block drug effects could not be tolerated by patients. Antipsychotic agents may be useful in cocaine users with concurrent psychotic illness. However, as Sevarino et al. (2000) have commented, the high prevalence of cocaine use among neuroleptic-maintained individuals with schizophrenia does not auger well.

The anticonvulsant carbamazepine has been trialled in humans because it has been shown in animal studies to selectively inhibit the "kindling" effects (of increased limbic seizures) resulting from chronic cocaine use. Several open trial studies carried out in the late 1980s and early 1990s led to some optimism that this medication may effectively reduce cocaine use through reduced craving and blocking of euphoria. However, subsequent randomised controlled trials found no evidence for the superior efficacy of carbamazepine compared with placebo. Carbamazepine was the subject of a systematic meta-analysis including five placebo controlled trials and 455 cocaine-dependent subjects (de Lima et al., 2002). De Lima and colleagues concluded that there was no evidence supporting the clinical use of carbamazepine in the treatment of cocaine dependence. Furthermore, Withers, Pulvirenti, Koob and Gillin (1995) have cautioned that at least one study has noted increased cardiovascular effects of cocaine used in combination with carbamazepine. No benefits for carbamazepine either in retention or cocaine free urine results were found in a comparison with desipramine and placebo in 146 crack cocaine dependent users (Campbell et al., 2003).
Top of page
Crosby et al. (1996) assessed the effects of another anticonvulsant, phenytoin, and found that this drug was significantly associated with reduction in cocaine use compared with placebo. The study commenced with 44 subjects and ran for 12 weeks, at which time only 12 subjects (6 in each group) remained. The high dropout rate and the fact that 85% of the phenytoin group believed they were taking phenytoin, means that these results are inconclusive.