Antidepressants have been investigated in the treatment of comorbid depression, depressive symptomatology associated with psychostimulant withdrawal, or for their dopamine agonist properties. Agents have included tricyclic antidepressants, SSRIs and MAOIs. Generally, results of clinical trials of antidepressants have been equivocal, with a recent systematic review concluding that there was no evidence to support the use of antidepressants in the treatment of cocaine dependence (de Lima, Soares, Reisser & Farrell, 2002). Imipramine, a tricyclic antidepressant associated with more side-effects than desipramine, demonstrated no differential efficacy compared to placebo in reducing drug use in 113 cocaine users (Nunes, McGrath, Quitkin, Opeceek-Welkikson et al., 1995) or in 32 methamphetamine users (Galloway, Newmeyer, Knapp, Stalcup & Smith, 1996).

Among the SSRIs, fluoxetine (ProzacTM) has attracted most research attention, but with equivocal results. The review by de Lima et al. (2002) found too few fluoxetine trials to analyse. However, they noted a similarity in urinalysis results to those reported in early desipramine trials. Two further large-scale trials, both conducted at the same centre, arrived at negative conclusions. Grabowski et al. (1998) found fluoxetine ineffective in reducing cocaine use in 228 cocaine users with superior retention in the placebo group. Schmitz et al. (2001) found no improvement in cocaine use or depressive symptoms in 68 dually diagnosed patients in a double blind trial of placebo and 40mg/day fluoxetine. Indeed, the latter study concluded that fluoxetine may potentiate cocaine effects. Other smaller controlled trials of fluoxetine for cocaine users have found fluoxetine less effective than placebo in reducing cocaine use or craving (Batki, Washburn, Delucchi & Jones, 1996; Petrakis, Carroll, Nich, Gordon et al., 1998), less effective than desipramine or amantadine (Oliveto, Kosten, Schottenfeld, Falcioni & Ziedonis, 1995) and less effective than interpersonal psychotherapy (Covi, Hess, Kreiter & Haertzen, 1995). A placebo controlled trial of 40 mg/day fluoxetine in 60 methamphetamine users found no difference in retention, amphetamine positive urines or reported amphetamine free days although amphetamine craving was significantly reduced (Batki, Moon, Delucchi, Bradley et al., 2000).

Significant proportions of problematic psychostimulant users may have pre-existing untreated affective disorders or other concomitant psychiatric disorders that may predispose them to psychostimulant dependence (Tutton & Crayton, 1993). One possible reason for the variability in findings is that antidepressants may have differential effectiveness dependent upon the underlying pre-existing psychiatric conditions. For example, there is a tendency for desipramine and other tricyclic antidepressants to be more effective where there is pre-existing depression than in non-depressed cocaine users (Donovan & Nunes, 1998). Sevarino, Oliveto and Kosten (2000) also point to the heterogeneity among cocaine users and the need to develop specialised treatments for distinct subgroups of users.

The way in which pharmacotherapies are used may also influence outcomes. For example, it has been suggested that fixed dosages of desipramine may be counter-productive and that results might be improved by individually managing serum drug levels (Campbell, Nickel, Penick,Wallace et al., 2003; Platt, 1997). Another approach to the use of antidepressants that has been suggested but not yet trialed, is as an adjunct to a behavioural intervention aimed at teaching alternative reinforcers to cocaine, with the antidepressants used to reduce craving (Fischman & Foltin, 1998).
Top of page
There are disadvantages to the use of antidepressants. As identified above, fluoxetine may potentiate cocaine effects (Schmitz, Averill et al., 2001) and desipramine may be associated with negative cardiovascular side-effects (Platt, 1997). Furthermore, there is a two to three week delay before antidepressants such as desipramine become effective. As dropout rates in this early period tend to be very high, these medications may not have an opportunity to demonstrate efficacy.

MAOIs have also been used for their potentially aversive interaction with cocaine and amphetamines. However, no controlled studies exist and the risk of hypertensive reactions make their use questionable (Tutton & Crayton, 1993).

Early interest in buproprion (ZybanTM) was not sustained after a placebo controlled trial of 149 cocaine-dependent methadone patients found no significant differences between placebo and buproprion (Margolin, Kosten & Avants, 1995). Buproprion had no effect on the subjective or cardiovascular effects of intranasal cocaine in an open label drug interaction study involving ten primary cocaine users (Oliveto, McCance-Katz, Singha, Petrakis et al., 2001).

Srisurapanont et al. (2002), from a systematic review of treatment for amphetamine dependence and abuse, found that fluoxetine, amlodipine, imipramine and desipramine have very limited benefits. These reviewers found that fluoxetine decreased craving in short-term treatment and imipramine may increase duration of adherence to treatment in medium-term treatment.