Currently, no specific medication has been found to be of substantial value in the treatment of cocaine dependence. Recent reports indicate that cocaine intoxication or "high," dependence, and withdrawal symptomatology are probably mediated through the dopamine (DA) reward systems in the brain (1, 2). Animal studies have established that cocaine reward is mediated through the activation of dopamine circuits (3-7). The dopamine brain circuits include the nucleus accumbens and its mesolimbic and mesocortical dopaminergic pathways. The cocaine high is thought to be produced by a sudden increase of available dopamine in the synaptic cleft. With repeated administration of cocaine, a depletion of available dopamine occurs by degradation. Since dopamine is the neurotransmitter that mediates the reward pathways in the brain, this depletion leads to an anhedonic state which leads the patient back to using cocaine. This theory is called the "dopamine depletion hypothesis" (1). Additional support for this theory was gained by animal studies showing increased dopaminergic receptor binding after chronic cocaine exposure (6). One group of researchers (1) believes that decreased synaptic availability of dopamine may trigger cocaine craving.
Bromocriptine is a dopamine agonist which acts through stimulation of postsynaptic dopamine receptors. Preliminary research in humans (8) suggests that bromocriptine decreases the cocaine-induced anhedonic state by stimulating dopamine receptors and is, therefore, effective in decreasing cocaine craving. The same group found that thioridazine, which is a dopamine receptor blocker, had the effect of increasing cocaine craving. In another recent clinical study (9), both bromocriptine and amantadine hydrochloride were found to be effective in ameliorating cocaine withdrawal symptoms. Amantadine hydrochloride was chosen because it releases dopamine and norepinephrine from neuronal storage vesicles and delays their reuptake. Preliminary findings that bromocriptine decreases cocaine craving may be important in seeking an effective pharmacologic adjunct for treating the cocaine-dependent patient.
Hyperprolactinemia, regardless of etiology, is responsive to bromocriptine treatment, except when it is due to pituitary tumor. Dopamine is thought to be prolactin inhibitory factor (PIF). Several studies (10-12) have reported markedly elevated prolactin (PL) levels in cocaine addicts. Gawin and Kleber (13) reported a decrease in serum PL and an increase in serum growth hormone (GH) among their cohort. One study reported completely normal PL levels in abstinent cocaine users (14). PL secretion is tonically inhibited by dopamine neurons (15). Dackis and Gold concluded that their finding of hyperprolactinemia in chronic cocaine-dependent patients is consistent with the dopamine depletion hypothesis (11). Bromocriptine is also known to reduce the plasma level of GH in one-third of acromegalic patients. L-Dopa and catecholamines stimulate GH release (16). In both laboratory animals and in humans, PL secretion is known to be independent of GH secretion (16).
The present study was conducted to evaluate the safety and efficacy of bromocriptine in the treatment of patients experiencing withdrawal from cocaine. Serum PL and GH levels were evaluated. By monitoring the withdrawal symptomatology and hormonal status, as well as treatment outcome, we assessed the clinical value of prescribing bromocriptine to treat cocaine dependence.
Sixty-three male military veterans in inpatient treatment for substance abuse were entered into a study to determine the efficacy of bromocriptine in the treatment of cocaine withdrawal. Patients had to fulfill the following criteria: meet the DSM-III-R (17) criteria for cocaine dependence; the last cocaine use was within 6 days; no other drug dependence was present; alcohol abuse may be present, but not alcohol dependence; and no psychotropic medications other than Halcion for sleep for the first 3 days were prescribed.
Written consent to participate in the study was obtained after procedures and potential side effects were explained. The patients were randomly assigned to a double-blind trial of bromocriptine or an inactive placebo which was scored and matched in color and size to the active drug. The patients, ward staff, research assistant, and physician prescribing the medication were blind to whether the patient was receiving the active drug or the inactive placebo. The patients were seen daily by the research assistant and given several self-reports to complete on Day 1 (baseline) and on each subsequent day of the study up to 21 days: the Symptom Checklist 90-Revised (SCL-90-R) (18), the Beck Depression Inventory (BDI) (19), and a scale for rating degree of cocaine craving developed by the authors. The patients were also asked to rate changes related to appetite, sleeping patterns, level of anxiety, psychomotor activity, mood, energy level, level of enjoyment/ anhedonia, interest in sex, and suicidal ideation. On the second day of hospitalization, after baseline assessment measures were obtained, both groups were started on 1/2 tablet (1.25 mg bromocriptine for those in the bromocriptine group) orally two times a day. This dose was then titrated up to 10 mg bromocriptine per day within the first 7 days of the study. If side effects occurred during titration, the dosage was decreased and stabilized at a lower dose. Random urine toxicology screens were obtained twice weekly throughout the study. Serum GH and PL levels were obtained prior to administration of bromocriptine or placebo and on the last day of the study. Normal GH levels range up to 10 ng/mL. Normal prolactin levels range up to 15 ng/mL. Three samples were obtained 30 minutes apart between 9:00 A.M. and 12:00 P.M. after fasting overnight. The means of three samples were obtained to minimize normal fluctuations of hormone levels during the day.
1. Symptom Checklist 90-Revised (SCL-90-R) (18). This is a 90-item self-report clinical rating scale used to assess the patient's level of psychological distress in nine symptom categories: somatization, obsession-compulsion, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, and psychosis. The SCL-90 also provides a Global Severity Index (GSI), the best single indicator of current severity of the disorder; a Positive Symptom Distress Index (PSDI) which represents intensity of the distress; and a Positive Symptom Total (PST) which is an absolute count of the number of symptoms reported.
2. Beck Depression Inventory (BDI) (19). This is a 21-item inventory which focuses on the patient's subjective feelings of depression and is sensitive to changes in feeling status.
3. Symptoms related to cocaine withdrawal. The patients were also asked to rate changes related to appetite, sleeping patterns, level of anxiety, psychomotor activity, mood, energy level, level of enjoyment/anhedonia, interest in sex, and suicidal ideation. These nine items were chosen because each represents a symptom category associated with cocaine withdrawal (20, 21). Each patient was asked to identify which of five statements best describes the way the patient has been feeling in the last 24 hours compared to how they usually feel. The five choices and number of points assigned include: a significant increase in symptoms (5 points), a slight increase (4 points), no change in symptoms (3 points), a slight decrease (2 points), and a significant decrease (1 point). The only exception is the item related to suicidal ideation, which is one directional in choices, ranging from never thinking of suicide to a significant increase in thoughts of suicide and having plans for carrying it out.
4. Cocaine-Craving Scale. The authors developed a Cocaine-Craving Scale on which the patients were asked to rate their degree of craving for cocaine based on their own current subjective experience. The scale is presented as a single horizontal line which represents craving for cocaine on a continuum, starting with 0, representing no cocaine craving, up to 10, which represents the highest degree of cocaine craving they have ever experienced. This scale is similar to drug analogue scales used in other studies measuring degree of cocaine craving (8, 22).
All of the dependent variables that showed a significant difference between the groups on Day 1 were analyzed using the multivariate analysis of co variance (MANCOVA) approach for groups by days of treatment, using the first day's score as the covariate.
This, in theory, controls for the difference in baseline levels between groups so that when subsequent data points are examined, significant differences are based on differences between groups and not on original differences at baseline. The purpose and procedures of analysis of covariance have been fully described elsewhere (23). Multivariate analysis of covariance is used instead of analysis of covariance when there are multiple dependent variables.
Howell (1987) more recently described the utility of analysis of covariance in randomized studies similar to ours. "The ideal application for an ANCOVA is an experiment in which subjects are randomly assigned to treatments. In that situation, the expected value of each group or cell mean on the covariate is the same and any difference can be attributed only to chance assuming that the covariate was measured before the treatments were applied. In this situation, the analysis of covariance will primarily reduce the error term, but it will also, properly, remove any bias in the dependent variable means caused by chance, group differences on the covariate" (24).
All other dependent variables were analyzed using a multivariate analysis of variance (MANOVA) approach. Also, inspection of the data indicated that there was no significant improvement in the dependent variables after the first week of treatment, so the MANOVA or MANCOVA, as was appropriate, was also done for the first week of treatment. By looking at both the 18 days and the first week of treatment, it could be determined if any significant interactions between time in treatment and placebo or bromocriptine use were due to factors such as "floor" effects.
Demographic Characteristics of the Sample
Approximately 50% (N = 29) of the 63 subjects completed the 18-21 day treatment program. 86% (N = 54) were Black, and 15% (N = 9) were Caucasian. Approximately 60% (N = 38) of the subjects were unemployed, and 23.8% (N = 15) were married or in a monogamous relationship. The average age of the sample was 35.6, ranging from 25 to 61 years. 90.3% (N = 57) freebased cocaine, 54.8% (N = 35) reported intranasal use, while 4.8% (N = 3) used cocaine intravenously. The average age of onset of cocaine use was 27.9 years, (range = 15-57) with the duration of use averaging 7.9 years (range = 1-23). The average amount of cocaine used in a typical week was 9.2 grams (SD = 9.5; range = 0.50-50). The average number of days since last cocaine use was 3.2 (SD = 1.2).
Between Group Comparisons
There were 29 patients who completed most of 18 days of treatment; of these, 15 were receiving bromocriptine and 14 the placebo. The results from these patients were first analyzed to determine if there were differences between the groups at the first measurement. T-tests indicated that the groups differed significantly on the BDI (P = .006), all nine subscales (P = .000 to .033) and three global scales (P = .000, .002, .002) of the SCL-90-R, and on one question regarding cocaine withdrawal symptoms that was concerned with level of anxiety (P = .009). The difference between the groups approached significance on the Cocaine-Craving Scale scores (P = .058). In all of these instances the placebo group scored higher or more aberrant than did the bromocriptine group.
Cocaine Craving Scale
The MANCOVAs for both the 18 days and the first week of treatment indicated significant effect for treatment days but not for bromocriptine vs placebo (18 days: time effect F = 3.46, p < .001; first week: time effect F = 3.84, p < .002). Bromocriptine did not seem to reduce cocaine craving more expeditiously or quantitatively than placebo (Fig. 1).
Beck Depression Inventory
The BDI MANCOVAS showed a significant effect for days of treatment but not for drug group both for the 18 days (time effect F = 12.08, p < 001) and for the first week of treatment (time effect F = 13.58, p < .001). Again, the bromocriptine had no apparent effect on the BDI scores when the initial difference between the groups has been taken into account.
The nine clinical SCL-90-R subscales and the three global indices were all evaluated using the MANCOVA for 18 days and for the first week of treatment. The Hostility subscale and the PSDI showed no significant changes in scores over the first week or over the 18 days. The Somatization and Paranoid Ideation subscales showed significant improvements for both 18 days and the first week in treatment for all patients regardless of group assignment. The remaining six subscales and two global indices showed a significant drug by days interaction for the placebo group over the 18 days on the MANCOVA but only a significant days effect for the first week of treatment.
Figure 2 illustrates the mean GSI scores in relationship to days in treatment. The graphs of the other global index and the subscales illustrate the same trend. Inspection of the data suggest that the significant drug by days interaction was caused by a "floor" effect for the scores for the bromocriptine group which prevented them from further decreasing their score during the latter part of the treatment while the placebo group could continue to improve. These findings indicate that bromocriptine did not significantly affect the treatment course compared to placebo for the first week of treatment.
Cocaine Withdrawal Symptom Questions
Only one of the nine questions showed a significant difference between the bromocriptine and placebo groups on Day 1, therefore, the MANCOVA was used. MANOVAs for the 18 days and for the first week of treatment were used on the remaining eight questions. Most of the questions showed a significant effect for days of treatment that was not associated with group assignment. However, two questions showed significant effects for drug group.
The question concerning activity level showed a significant effect for bromocriptine versus placebo during the first week of treatment (F = 4.88, p = .038). The placebo group had an increase in their reported activity level while the level of the bromocriptine group remained essentially the same. The question concerning appetite level also showed significant differences between the bromocriptine and placebo group for the 18 days (F = 4.52, p = .048) and for the first week of treatment (F = 8.16, p = .009). The placebo group reported a significantly higher appetite level.
Prolactin (PL) and Growth Hormone (GH) Levels
PL levels and GH levels were obtained within 6 days of the last cocaine use prior to the patient receiving bromocriptine or placebo, and on the last day of the study (Days 18-21). There were no significant differences between the pre- and posttests for GH in either group. The mean pretest and posttest GH levels were within normal limits. The mean pretest and posttest GH levels for the placebo groups were 0.8840 and 0.5333 ng/mL, respectively. The mean pre- and posttest GH levels for the drug groups were 1.11146 and 0.7167 ng/mL, respectively. The mean pre- and posttest serum PL levels for both groups were within normal range. The mean PL level in the placebo group had increased at the posttest (p = .283) while the mean PL level in the drug group decreased (p = .001). This was expected as bromocriptine is known to decrease PL levels in humans. The mean pre- and posttest PL levels for the placebo group were 6.4040 and 8.8480 ng/ mL, respectively. The mean pre- and posttest PL levels for the drug group were 6.9986 and 5.2270 ng/mL, respectively. The individual mean GH and PL levels are listed in Table 1.
[TABULAR DATA 1 OMITTED]
Overall, there did not appear to be any advantage to receiving bromocriptine versus placebo during the first 3 weeks following cessation of cocaine use in most of the outcomes examined. The only statistical differences involved an increase in appetite and activity in the placebo group. A common side effect of bromocriptine is gastric distress which could account for the increase in appetite in the placebo group as compared to the drug group. An increase in appetite has been reported in the period following cocaine cessation. It is also possible that bromocriptine exerts a central mechanism of action which would explain the lower appetite and activity level found in the drug group. Restlessness or psychomotor agitation has been reported following cocaine cessation (17, 25). Restlessness or psychomotor agitation often results in as an increase in activity level.
All other assessments showed gradual improvement for number of days in treatment, but did not show any statistically significant changes between the bromocriptine and placebo groups. Patients in both groups improved significantly as measured by the Cocaine-Craving Scale, the BDI, the clinical subscales and two of the global indices of the SCL90-R, and the nine cocaine withdrawal symptom questions. This improvement appeared to progress according to length of treatment.
There has been conflicting literature published regarding the efficacy of bromocriptine in the treatment of cocaine withdrawal and craving. Preliminary reports indicated that low doses of bromocriptine ameliorated cocaine craving which is thought to play an important part in relapse (8, 22). One double-blind study of 14 subjects comparing amantadine and bromocriptine indicated that both were effective in treating cocaine withdrawal though amantadine was better tolerated and appears to decrease withdrawal symptoms more effectively (9). Another double-blind study with 24 patients comparing bromocriptine and placebo found that the bromocriptine group improved significantly over the placebo group as measured by the Brief Psychiatric Rating Scale (26). In a recent open trial of 25 patients given bromocriptine, little if any improvement occurred (27). In our double-blind trial involving 29 patients comparing bromocriptine and placebo, little, if any, difference in improvement was seen. However, our study did not evaluate the acute effects of a single dose of bromocriptine in craving (8, 28). This study also was not designed to study any delayed onset of action of bromocriptine following cocaine cessation (29).
There have been conflicting reports in the literature regarding serum PL levels in cocaine-dependent patients (10-13, 26). This study did not find any abnormality in mean serum PL or GH levels prior to treatment. The methodology differed in the current study. It was hoped that by calculating the mean level from three separate blood samples, the results would be more representative of the true serum levels. There is much variation in the individual GH and PL levels (Table 1).
These hormone levels vary throughout the day and are affected by a wide range of factors, including stress, diet, weight loss or gain, time of day, and sleep cycle. Stress secondary to impending discharge is one possible explanation for the high posttest PL levels in placebo patient 1 and drug patient 10. Further evaluation and research is needed in this area.
An interesting finding in this study involves the group of patients who had to be terminated from the study secondary to side effects (N = 19). Again, the group was distributed almost evenly (placebo = 9, drug = 10).
Approximately 50% of the subjects who were terminated secondary to side effects were receiving placebo (N = 9). This suggests that patients who are withdrawing from cocaine tend to report somatic symptoms. This is supported by the clinical experience of the authors treating this population and from the increase in the somatization subscale of the SCL-90-R. This finding should be taken into consideration in future studies as this added to the high dropout rate in this study. Patients were asked whether they experienced specific side effect symptoms related to bromocriptine. This approach may result in symptom endorsement.
Limitations to this study include the small sample size (N = 29), high dropout rate, and differences in degree of symptomatology in the two groups prior to initiation of treatment. The patients were randomly assigned to a double-blind trial; therefore, this initial difference can only be attributed to chance. This last concern was dealt with statistically by controlling for the initial difference.
Approximately 50%, or 34 of the 63 patients recruited for the study, did not complete the study. This high dropout rate was due in part to the pronounced number of patients reporting side effects as discussed earlier in this paper (N = 19). The 34 patients who dropped out were equally divided between groups (N = 17 for each group). This suggests that there was no significant difference in patient retention in treatment attributable to drug group assignment.
In conclusion, the results of this study indicated that patients received little benefit from the use of chronic bromocriptine therapy. The effects of acute or single doses of bromocriptine or delayed effects of bromocriptine therapy were not examined. Given the small sample size, the high dropout rate, and the fact that only male veterans were studied, additional research needs to be conducted to evaluate the efficacy of bromocriptine in the treatment of cocaine dependence.
The authors would like to acknowledge the support of the following at the Department of Veterans Affairs, Edward Hines, Jr. Hospital: the late Walter Dorus, M.D., Chief, ATEC/DDTC; Linda Knoll, Research Assistant; Kathleen Garvin, Statistical Assistant; and Angela Doup, Pat Flynn, and Geraldine Wlodarski, Editorial Assistants. We also acknowledge the Research and Development Service for their administrative support.
Figure 1 & 2 [GRAPH OMITTED]
[1.] Dackis, C. A., Gold, M. S., Davies, R. K., and Sweeney, D. R., Bromocriptine treatment for cocaine abuse: The dopamine depletion hypothesis, Int. J. Psychiatry Med. 15(2):125-135 (1985). [2.] Gold, M. S., and Washton, A. M., Cocaine abuse: Neurochemistry, phenomenology, and treatment, in Cocaine Use in America: Epidemiologic and Clinical Perspectives (N. J. Koze and E. H. Adams, Eds.), (NIDA Research Monograph Series, No. 61), National Institute on Drug Abuse, Rockville, Maryland, 1985. [3.] Olds, J., Pleasure centers in the brain, Sci. Am. 195:105-116 (1956). [4.] Goeders, N. E., and Smith, J. E., Critical dopaminergic involvement in cocaine reinforcement, Science 221:773-775 (1983). [5.] DeWitt, H., and Wise, R. A., A blockade of cocaine reinforcement in rats with the dopamine receptor blocker pimozide but not with the noradrenergic blockers phentolamine or phenoxybenzamine, Can. J. Psychol. 31:195-203 (1977). [6.] Taylor, D., Ho, B. T., and Fagen, J. D., Increased dopamine receptor binding in rat brain by repeated cocaine injections, Commun. Psychopharmacol. 3:137-142 (1979). [7.] Taylor, D., and Ho, B. T., Neurochemical effects of cocaine following acute and repeated injection, J Neurosci. Res. 3:95-101 (1977). [8.] Dackis, C. A., and Gold, M. S., Bromocriptine as treatment for cocaine abuse, Lancet pp. 1151-1152 (May 18, 1985). [9.] Tennant, F. S., and Sagherian, A. A., Double-blind comparison of amantadine and bromocriptine for ambulatory withdrawal from cocaine dependence, Arch. Intern. Med. 147:109-112 (1987). [10.] Dackis, C. A., Gold, M. S., Estroff, T. W., and Sweeney, D. R., Hyperprolactinemia in cocaine abuse, Soc. Neurosci. Abst. 10:1099 (1984). [11.] Dackis, C. A., and Gold, M. S., Neurotransmitter and neuroendocrine abnormalities associated with cocaine use, Psychiatr. Med. 3(4):472-473 (1987). [12.] Mendelson, J. H., Teoh, S. K., Lange, U., Mello, N. K., Weiss, R., Skuphy, A., and Ellingboe, J., Anterior pituitary, adrenal, and gonadal hormones during cocaine withdrawal, Am. J. Psychiatry 145(9):1094-1098 (1988). [13.] Gawin, F. H., and Kleber, H. D., Neuroendocrine findings in chronic cocaine abusers: A preliminary report, Br. J. Psychiatry 147:569-572 (1985). [14.] Swartz, C. M., Breen, K., and Leone, F., Serum prolactin levels during extended cocaine abstinence, Am. J. Psychiatry 147(6):777-779 (1990). [15.] MacLeod, R. M., Regulation of prolactin secretion, in Frontiers in Neuroendocrinology (L. Martin and W. F. Ganong, Eds.), Raven Press, New York, 1976, pp. 169-194. [16.] Morgan, H. E., Endocrine control system, in Best and Taylor's Physiologic Basis of Medical Practice, 9th ed. (J. R. Brobeck Ed.), Williams and Wilkins Press, Baltimore, 1973, pp. 8-28, 143-145. [17.] American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, 3rd ed., revised, Washington, D.C., 1987. [18.] Deragotis, L. R., SCL-90-R Administration, Scoring and Procedures Manual, John Hopkins School of Medicine, Clinical Psychometrics Research Unit, 1977. [19.] Beck, A. T., Ward, C. H., Mendelson, M., et al., An inventory for measuring depression, Arch. Gen. Psychiatry 4:561-571 (1961). [20.] Gawin, F. H., and Kleber, H. D., Abstinence symptomatology and psychiatric diagnosis in cocaine abusers, Arch. Gen. Psychiatry 43:107-113 (1986). [21.] Gawin, F. H., Chronic neuropharmacology of cocaine: Progress in pharmacotherapy, J. Clin. Psychiatry 49(2)(Suppl.):11-16 (1988). [22.] Extein, I., Gross, D. A., and Gold, M. S., Bromocriptine treatment of cocaine withdrawal symptoms, Am. J. Psychiatry 146(3):403 (1989). [23.] Guenther, W. C., Analysis of Variance, Prentice-Hall, Englewood Cliffs, New Jersey, 1964, pp. 143-164. [24.] Howell, D. C., Statistical Methods for Psychology, 2nd ed., Duxbury Press, Boston, 1987, p. 539. [25.] Brower, K. I., Maddahian, E., Blow, F. C., and Beresford, T. P., A comparison of self-reported symptoms and DSM-III-R criteria for cocaine withdrawal, Am. J. Drug Alcohol Abuse 14(3):347-356 (1988). [26.] Giannini, A. J., Baumgarter, P., and DiMarzio, L., Bromocriptine therapy in cocaine withdrawal, J. Clin. Pharmacol. 27:267-270 (1987). [27.] Teller, D. W., and Devenyi, P., Bromocriptine in cocaine withdrawal--Does it work?, Int. J. Addict. 23(11):1197-1205 (1988). [28.] Dackis, C. A., Gold, M. S., Sweeney, D. R., Byron, Jr., J. P., and Climko, R., Single-dose bromocriptine reverses cocaine craving, Psychiatr. Res. 20:261-264 (1987). [29.] Giannini, A. J., Folts, D. J., Feather, J. N., and Sullivan, B. S., Bromocriptine and amantadine in cocaine detoxification, Psychiatr. Res. 29:11-16 (1989).
COPYRIGHT 1995 Taylor & Francis Ltd.
COPYRIGHT 2004 Gale Group