- Email: [email protected]
A pilot trial of topiramate for the treatment of cocaine dependence
Drug and Alcohol Dependence 75 (2004) 233–240
A pilot trial of topiramate for the treatment of cocaine dependence Kyle M. Kampman a,b,∗ , Helen Pettinati a , Kevin G. Lynch a , Charles Dackis a , Thorne Sparkman c , Catherine Weigley a , Charles P. O’Brien a a
Department of Psychiatry, University of Pennsylvania School of Medicine, 3900 Chestnut Street, Philadelphia, PA 19104, USA b Department of Veterans Affairs Medical Center, University and Woodland Avenues, Philadelphia, PA 19104, USA c Department of Medicine, University of Pennsylvania School of Medicine, 3400 Spruce Street, Philadelphia, PA 19104, USA Received 14 July 2003; received in revised form 19 January 2004; accepted 10 March 2004
Abstract Background: Both GABAergic and glutamatergic neurons appear to be important modulators of the brain reward system and medications that affect GABA and glutamatergic neurotransmission may reduce the rewarding properties of cocaine and reduce cocaine craving. Topiramate, an anticonvulsant, raises cerebral GABA levels, facilitates GABAergic neurotransmission and inhibits glutametergic activity at AMPA/kainite receptors. Thus, it may be useful for treating cocaine dependence. Methods: The efficacy of topiramate for cocaine dependence was tested in a 13-week, double-blind, placebo-controlled pilot trial (n = 40). Topiramate was titrated gradually over 8 weeks to a dose of 200 mg daily. The primary outcome measure was cocaine abstinence verified by twice weekly urine benzoylecgonine tests (UBT). Results: Eighty-two percent of subjects completed the trial. Analysis of the UBT using a GEE model showed that after week 8, when the dose titration was completed, topiramate-treated subjects were more likely to be abstinent from cocaine compared to placebo-treated subjects (Z = 2.67, P = 0.01). Topiramate-treated subjects were also more likely to attain 3 weeks of continuous abstinence from cocaine (χ2 = 3.9, d.f. = 1, P = 0.05). Conclusion: Topiramate may be effective for the treatment of cocaine dependence. © 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Cocaine; Topiramate; Clinical trial; GABA; Glutamate
1. Introduction Multiple studies have suggested that increases in extracellular dopamine in the mesocorticolimbic dopamine system mediate the acute reinforcing effects of cocaine (Dworkin and Smith, 1988; Goeders et al., 1986; Koob et al., 1987; Roberts et al., 1980). In addition, the mesocorticolimbic dopamine system has been shown to be activated in response to conditioned stimuli associated with prior cocaine administration (Bowman et al., 1996; Schultz, 1998; Weiss et al., 2000). This activation may be associated with the subjective experience of cocaine craving in humans which often promotes relapse to cocaine use (Berger et al., 1996; Childress et al., 1993, 1999). Initial studies of pharmacological treatments for cocaine dependence focused largely on drugs that would directly influence dopaminergic func-
∗ Corresponding author. Tel.: +1-215-222-3200x109; fax: +1-215-386-6770. E-mail address: kampman [email protected] (K.M. Kampman).
tion, but these agents have been consistently found to be ineffective (for a review see de Lima et al., 2000). Another way of affecting dopaminergic function is indirectly through other modulatory neurotransmitter systems, such as the gamma amino butyric acid (GABA) system and the glutamate system. Several lines of evidence suggest that GABAergic neurons function as modulators of the mesocorticolimbic dopamine system and play an important role in mediating the reinforcing effects of cocaine (Dewey et al., 1998; Koob, 1992). Preclinical trials of medications that foster GABAergic neurotransmission have suggested that these compounds reduce the dopamine response to both cocaine administration and to conditioned cues of prior cocaine use (Dewey et al., 1992, 1997; Gerasimov et al., 1999). GABAergic medications also reduce the self-administration of cocaine in various animal models (Kushner et al., 1999; Roberts et al., 1996). Similarly, glutamatergic neurons appear to be important modulator of the mesocorticolimbic reward system. Glutamatergic neurons located in the prefrontal cortex innervate the nucleus accumbens and play a critical role in cocaine-reinforced behavior (McFarland and
0376-8716/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.drugalcdep.2004.03.008
234
K.M. Kampman et al. / Drug and Alcohol Dependence 75 (2004) 233–240
Kalivas, 2001). Inhibition of prefrontal glutamatergic neurons blocks cocaine induced reinstatement of drug seeking behavior in rats (Cornish and Kalivas, 2000; McFarland et al., 2003). Thus, medications that promote GABAergic activity or inhibit glutamatergic activity may be promising medications for cocaine dependence treatment. Topiramate is approved by the US Food and Drug Administration for the treatment of epileptic seizures. It has effects in both the GABA and glutamate systems (Privetera, 1997). Topiramate enhances GABA activity at GABA-A receptors and causes a general increase in GABA levels in the brain (Kuzniecky et al., 1998; Petroff et al., 1999; White et al., 1997). Topiramate also antagonizes glutamate transmission through effects at AMPA/kainite receptors (Severt et al., 1995). Thus, topiramate may reduce cocaine craving and cocaine use in cocaine dependent patients. Topiramate has been studied for the treatment of various addictions and published reports suggest it may be useful for both alcohol and opiate dependence. Topiramate reduced alcohol craving and alcohol use in a group of 24 alcohol dependent subjects in an open trial (Rubio et al., 2002). More recently, topiramate was found to be more effective than placebo in reducing alcohol use in a double-blind trial involving 150 alcoholics (Johnson et al., 2003). Finally, there is a case series published suggesting topiramate may be useful for the treatment of opiate withdrawal (Zullino et al., 2002). To our knowledge there are no published studies of topiramate for cocaine dependence. The current pilot study was intended to obtain more data regarding the safety, tolerability and potential efficacy of topiramate for the treatment of cocaine dependence.
2. Methods 2.1. Subjects The subjects were 40 DSM-IV treatment-seeking, cocaine-dependent men and women between the ages of 18 and 60 years. They were recruited at the University of Pennsylvania Treatment Research Center (TRC). The TRC recruits through advertisement in the local media as well as through professional referrals. Patients entering the trial were not involved in any other type of addiction treatment elsewhere. All subjects signed informed consent prior to participation in the trial, after trial procedures were explained to them by an investigator. The study was reviewed and approved by the Institutional Review Board (IRB) of the University of Pennsylvania. Subjects were reimbursed $5.00 at each visit for completing all research procedures; at the last visit they received $30.00 because of the greater number of research procedures done at the end of the study. Subjects received an additional $5.00 each week for returning the previous week’s medication package in order to facilitate the pill count compliance check. If needed, two transit tokens were provided at each visit.
Psychiatric diagnoses were obtained by a board-certified addiction psychiatrist (KK) through a clinical interview. Subjects were required to have self-reported at least $100 worth of cocaine use in the month prior to entry. Because topiramate was viewed as a medication that might worsen cocaine withdrawal symptoms by reducing dopaminergic tone (Westerink et al., 1998), subjects were selected who had low cocaine withdrawal symptom severity on the day of intake demonstrated by low scores (below 22) on the Cocaine Selective Severity Assessment (CSSA) (Kampman et al., 1998). Subjects also were required to maintain at least 3 days of self-reported abstinence from cocaine immediately prior to starting study medication. At our center, we have found that cocaine withdrawal symptom severity affects outcome in outpatient cocaine dependence treatment (Mulvaney et al., 1999; Kampman et al., 2000, 2001a,b, 2002) and we frequently use withdrawal symptom severity as a factor in assigning subjects to particular clinical trials. Medical screening included a complete medical history and physical examination conducted by a certified nurse practitioner. Baseline laboratory testing included a chemistry screen, complete blood count, urinalysis, eye exam with visual acuity and a 12 lead EKG. Women of childbearing potential were required to use barrier contraception during the trial. Women received urinary pregnancy testing prior to starting medications, and at monthly intervals throughout the study. Liver function tests were monitored monthly throughout the trial. Subjects with current dependence (DSM-IV criteria) on any additional drug except nicotine were excluded. Psychiatric exclusion criteria included psychosis, dementia, and the use of other psychotropic medications. Medical exclusion criteria included unstable medical illnesses, use of antiepileptic medications, a history of nephrolithiasis, a history of glaucoma, and a history of hypersensitivity to topiramate. 2.2. Measures The primary outcome measure for this trial was qualitative urine benzoylecgonine tests (UBT) obtained twice weekly. Urine collection was monitored by temperature checks. Samples less than 90 ◦ F, or greater than 100 ◦ F were not accepted. Samples were analyzed for benzoylecgonine by fluorescent polarization assay. Samples containing equal to or greater than 300 ng/ml of benzoylecgonine were considered to be positive. Secondary outcome measures for the trial included treatment retention, measured by the number of evaluation sessions attended. Severity of addiction-related problems was measured by the Addiction Severity Index (ASI) (McLellan et al., 1992) administered four times, at baseline, 4 weeks after starting medications, 8 weeks after starting medications, and at the end of the trial. The study physician rated illness severity and improvement weekly using the Clinical Global Impression Scale (CGI) (Guy, 1976). Cocaine crav-
K.M. Kampman et al. / Drug and Alcohol Dependence 75 (2004) 233–240
ing was measured weekly using the Brief Substance Craving Scale (BSCS) (Somoza et al., 1995). Safety measures included adverse events, which were monitored at each visit, and liver function tests monitored every 4 weeks during the trial. 2.3. Procedures Eligible subjects entered a 13-week trial that included a 1-week baseline phase during which all pretreatment measures were obtained and subjects began psychosocial treatment. Subjects were required to attend two visits during this week to be eligible to receive study medications. Eligible subjects were randomized to receive either topiramate or placebo. Because rapid dose titration of topiramate can result in uncomfortable central nervous system side effects such as sedation, slurred speech and trouble selecting words, the dose of topiramate was started very low (25 mg daily, given at bedtime) and the dose was increased 25 mg each week for 8 weeks until patients were receiving 200 mg of topiramate each day, or identical placebo capsules, given at bedtime. During the last week of the trial topiramate was rapidly reduced by 25 mg each night. Study medication was prepared by the Research Pharmacist at the Hospital of the University of Pennsylvania by over-encapsulating topiramate 25 and 100 mg tablets, purchased from a commercial pharmacy, and producing identical-appearing lactose-containing placebo capsules. Study medication was placed in blister packs with each day’s dose clearly marked. Medications were dispensed by the study physician each week and the previous week’s blister pack was collected. Compliance was measured by pill count. Patients were instructed to take their daily dose of medication at night before going to sleep. In addition to medication or placebo, subjects received twice weekly individual cognitive-behavioral relapse prevention therapy utilizing a Cognitive-Behavioral Coping Skills Therapy (CBT) manual. The CBT therapy manual and supporting materials were developed for the National Institute on Alcohol Abuse and Alcoholism Project MATCH (Kadden et al., 1995). The basic format was accepted, although specific procedures were adapted for treatment of cocaine dependence by our group. Therapy was provided by Master’s level therapists with additional training in CBT. 2.4. Statistical analysis Baseline measures between the topiramate and placebo groups were compared using t-tests for continuous variables and χ2 -tests for dichotomous variables. Measures were transformed when necessary to reduce skew. The number of sessions attended for each group during the trial was compared using a t-test. UBT results were compared using generalized estimating equations (GEE) models (Diggle et al., 1994). There were two urines collected per subject each
235
week, yielding a possible 24 urine drug screens (UDS) per subject. GEE approaches work best when the data are in the form of a small number of repeated measurements for each of many subjects (Sharples and Breslow, 1992). As our sample size is rather small, we chose to reduce the number of time points per subject by coding each of the 12 weeks as a use or non-use week rather than including each visit in the model. We initially labeled each study week as a cocaine non-use week or use week according to whether the subject provided two negative UBT (cocaine non-use week), or at least one positive UBT or one missing UBT (cocaine use week). In short, missing urine samples were labeled as positive. Using weeks as the time units of analysis, rather than visits, also reduces the risk of overestimating drug use due to benzoylecgonine carryover. Benzoylecgonine remains detectable in the urine at significant levels for several days, so one episode of use may lead to several BE positive samples. There was a gap of at least 3 days between the final UDS of 1 week and the first UDS of the next week, so the possibility of carryover from 1 week to the next should be negligible. The screening urine and two urines submitted during week 1 were used to determine use/non-use status prior to randomization, and this status was included as a covariate in the GEE model. Because topiramate required an 8-week dose run up to achieve full dose, we included a binary time-varying covariate to indicate the time before full dose was achieved (value 0 for weeks 2–8), or after full dose was achieved (value 1 for weeks 9–13). A set of continuously distributed secondary outcome measures gathered over varying time frames was analyzed using linear mixed effects models. The models used to compare the groups on the secondary outcome measures generally included the pre-treatment version of the response as a covariate, together with treatment group, and time effects. For responses with only a few time points, time was included as a categorical variable; for responses with a greater number of time points, linear time trends were found to be sufficient. The two-way interactions among pre-treatment response, medication group, and time were also considered for inclusion in these models by examining P-values of regression coefficients for the GEE models, and Akaike’s Information Criterion (AIC) (Littell et al., 1996) comparisons of overall model fit for the mixed effects models. For the GEE models for the UBT results, an exchangeable (compound symmetry) structure was used for the working correlation matrix. For the mixed effects models, covariance models were chosen by the methods described by Littell et al. (2000): models having exchangeable residuals fit best, sometimes with the inclusion of a random intercept. In the analyses of the secondary measures, the scales from the ASI, and the BSCS there were few significant effects. In the descriptions below, we report only the medication effects, and other significant effects, from the final model. In those descriptions of the analyses on the UBT and secondary responses, note that all Z and χ2 -statistics quoted are obtained from the models.
236
K.M. Kampman et al. / Drug and Alcohol Dependence 75 (2004) 233–240
ite Drug Scores and higher Hamilton Depression Rating Scale scores compared to placebo-treated subjects.
3. Results, intent-to-treat sample 3.1. Baseline demographic and drug use
3.2. Treatment retention On the whole, the two study groups, topiramate and placebo, were very similar in demographics and baseline drug use characteristics (see Table 1). The average age of the subjects was about 40. Most were African American men and most smoked crack cocaine. On average, subjects had used cocaine between 6 and 8 days in the month prior to treatment (range 1–25 days) and spent between $300 and $500 on drugs. Although all subjects met criteria for current DSM-IV cocaine dependence and had used at least $100 worth of cocaine in the 30 days prior to the trial, we selected patients with low cocaine withdrawal symptom severity for this trial. Therefore, the subjects were on the milder end of the spectrum of addiction severity measured by days of cocaine use and money spent for cocaine. Although the subgroup selected for this trial may have had less severe symptoms of addiction, the group represented a significant proportion of cocaine dependent subjects who presented for treatment at our center. During the 10 months we recruited for this trial, 186 cocaine dependent subjects were screened at intake and 72 had CSSA scores less than 22 and thus qualified for this trial. The topiramate and placebo groups differed in two ways, topiramate-treated subjects had, on average, significantly higher ASI Compos-
Only 6 of the 40 randomized subjects were lost to follow-up, 4/20 in the placebo group and 2/20 in the topiramate group (χ2 = 0.78, d.f. = 1, P = 0.38). Treatment retention, measured by the number of visits attended, did not differ significantly. On average, topiramate-treated subjects attended 17 visits and placebo-treated subjects attended 15 visits (t = −0.052, d.f. = 38, P = 0.96). 3.3. Urine benzoylecgonine test results We found a significant group by time by dose-period interaction, with topiramate-treated subjects becoming more likely to be cocaine abstinent after week 8, and placebo-treated subjects becoming less likely (see Fig. 1). There was no significant difference between the two groups during the dose run-up period of the trial: the placebo group had odds of use 1.31 times higher than those in the topiramate group (Z = 1.42, P = 0.10), and neither time effects (Z = 1.00, P = 0.32) nor group by time effects (Z = −1.10, P = 0.27) were significantly different from zero. There was a significant group by time interaction during the full-dose period: the odds of use in the placebo group
Table 1 Subject characteristics, expressed as percent or mean (standard deviation) Variable
Topiramate (N = 20)
Age Male (%)
40.10 (7.10) 95
Race African American (%) Caucasian (%)
95 5
Placebo (N = 20) 39.90 (6.38) 100 85 15
Years of education Days of alcohol use in past 30 daysa Days of cocaine use in past 30 daysa $ spent for drugsa Years of cocaine use, lifetime Number of prior treatments ASI Composite Drug Scorea,b ASI Composite Alcohol Score ASI Composite Employment Score ASI Composite Legal Scorea ASI Composite Family/Social Scorea ASI Composite Psychiatric Scorea ASI Composite Medical Scorea Hamilton Depression Rating Scalea,c Hamilton Anxiety Rating Scale
12.55 (1.47) 3.85 (5.15) 7.90 (7.07) 513.25 (342.62) 10.10 (6.35) 1.55 (1.54) 0.206 (0.049) 0.072 (0.086) 0.577 (0.245) 0.132 (0.248) 0.180 (0.211) 0.155 (0.187) 0.126 (0.257) 6.30 (4.210 3.15 (2.18)
12.35 (1.63) 6.25 (7.99) 6.40 (7.63) 364.75 (343.09) 12.75 (4.79) 1.60 (1.70) 0.162 (0.082) 0.124 (0.130) 0.522 (0.294) 0.090 (0.152) 0.078 (0.146) 0.062 (0.126) 0.143 (0.287) 3.45 (3.38) 2.55 (2.82)
Route of cocaine use Intranasal (%) Smoked (%)
10 90
15 85
a b c
Transformed prior to analysis to reduce skew. Significantly different t = −2.03, d.f. = 38, P = 0.05. Significantly different t = −2.57, d.f. = 38, P = 0.014.
K.M. Kampman et al. / Drug and Alcohol Dependence 75 (2004) 233–240
% Cocaine Abstinent
60%
237
cant group effect in the drug composite scores with lower scores in the topiramate group (F(1/34) = 4.12, P = 0.05). There was also a significant group by time interaction for the employment composite score (F(1/60) = 4.56, P = 0.04). Topiramate-treated subjects showed a significantly greater reduction in the employment composite score over the course of the trial than did the placebo-treated subjects.
50% 40% 30% 20% 10% 0% 1 2 3 4 5 6 7 8 9 10 11 12 13 Week
placebo, n = 20
topiramate, n = 20
Fig. 1. Percentage of subjects who were cocaine abstinent at each study week in the placebo and topiramate-treated groups.
increased over the period (odds ratio per week = 1.29, P = 0.14), while the odds of use in the topiramate group decreased over the period (odds ratio per week = 0.73, P = 0.04), with a significant difference between the two rates (Z = 2.67, P = 0.01). Finally, subjects who provided a clean UDS at baseline were less likely to use during the study than those who did not (odds ratio = 0.20, P = 0.01). 3.3.1. Missing data Although 33/40 subjects completed the study there were a number of missed visits resulting in missing data. The data most affected by missed visits were the UBT. Three hundred thirty-two urine samples out of 960 or 34.6% of the urine samples were missing. Because of the small sample size we addressed missing data using a clinical approach rather than through formal statistical modeling of the drop out process. In our initial model, missing urine samples were considered positive. However, this implies that patients who drop out of treatment always do so because of treatment failure. This may not always be the case. In order to ensure that we were not overestimating cocaine use in either group due to coding missed urine samples from dropouts as positive, we reran our original model excluding samples missing due to study dropout. Because the number of missed visits was not significantly different between groups ((t-test) t = −0.87, d.f. = 38, P = 0.39) and the time to last visit was not significantly different between groups (log-rankχ2 = 0.004, d.f. = 1, P = 0.98), we feel that this is a valid approach to account for the missing data. Using this model did not significantly change the conclusions: in particular, there was still a significant group by time by dose-period interaction (Z = 2.42, P = 0.03). 3.4. Results from the Addiction Severity Index Days of cocaine use, dollars spent on drugs and the legal composite scores tended to decline in both groups during the trial but there were no medication effects detected. The ASI Composite Drug Score declined significantly during the trial (F(1/60) = 11.97, P = 0.001). There was a signifi-
3.5. Results from Brief Substance Craving Scale The BSCS divides cocaine craving into three domains: intensity, duration and frequency. Each domain is measured on a 0–4 likert scale. By adding the three scores a craving composite measure can be derived. Although this measure is not completely continuous, its distribution was reasonably symmetric, and it was appropriate for the mixed effects models. Cocaine craving showed a significant decline over the trial in both groups: (F(1/285) = 17.08, P < 0.0001), and there was a significant interaction between baseline scores and time; subjects with higher craving at baseline showed a greater reduction in craving during the trial (F(1, 284) = 11.84, P = 0.0007). Although average craving scores declined slightly more in the topiramate group, this effect did not achieve statistical significance (F(1/34) = 2.70, P = 0.11). 3.6. Medication adherence Medication adherence was measured by pill count. The percentage of pills taken was calculated by subtracting the number of pills returned each week from the number of pills dispensed. Both groups showed good adherence. The mean percentage of prescribed pills taken by the topiramate-treated subjects was significantly higher compared to the placebo-treated subjects (94% versus 83%, t = −2.7, d.f. = 34, P = 0.01). 3.7. Safety analyses Adverse events were assessed at each visit. Topiramate was well tolerated. Adverse events were mainly mild and evenly distributed between the topiramate and placebo groups. The most commonly reported adverse events included headache (n = 6), fatigue (n = 6), diarrhea (n = 6), constipation (n = 4), and nausea (n = 4). There were no significant differences noted in the occurrence of any adverse event between the two medication groups. We did have one topiramate-treated subject who developed symptoms of kidney stones at week 9 and was referred to a urologist for treatment.
4. Results, evaluable cases In this trial there were four subjects who received their first dose of study medication and did not attend any
238
K.M. Kampman et al. / Drug and Alcohol Dependence 75 (2004) 233–240
subsequent evaluation visits, three in the topiramate group and one in the placebo group. It is impossible to evaluate what effect the study medication may have had on these subjects. Therefore, in a set of separate exploratory analyses, we looked at some outcome measures excluding these four subjects. We specifically examined rates of cocaine abstinence, and global ratings of improvement. We found that in this group of 36 subjects, topiramatetreated subjects were significantly more likely to achieve 3 or more weeks of continuous abstinence from cocaine measured by negative urine toxicology screens. We found that 10 out of 17 or 59% of topiramate subjects attained 3 or more continuous weeks of abstinence from cocaine compared to only 5 of 19 or 26% in the placebo group (χ2 = 3.9, d.f. = 1, P = 0.05). The higher rates of abstinence achieved by the topiramatetreated subjects were reflected in rates of improvement noted by the treating clinician at the last visit. Seventy one percent (12/17) of topiramate-treated subjects were rated as very much improved on the CGI compared to 32% (6/19) of placebo-treated subjects (χ2 = 5.46, d.f. = 1, P = 0.02).
5. Discussion Data from this pilot trial suggest that topiramate may be a promising medication for the treatment of cocaine dependence. In the intent-to-treat analysis, which included all 40 randomized subjects, topiramate was shown to be safe and well tolerated. In the primary outcome measure, cocaine abstinence measured by UBT results, there was a significant effect of topiramate, which was observable after week 8, when the dose titration was complete. After week 8, subjects received 200 mg of topiramate each night. This result is similar to the result noted by Johnson et al. in their study of topiramate for alcohol dependence. In that study, Johnson et al. (2003) found the greatest benefit from topiramate to occur after week 6, when subjects were receiving 200 mg daily of topiramate. Like Johnson, we cannot say whether what we observed was a dose–response effect, or simply an effect related to time of exposure to topiramate. ASI results also showed a beneficial effect of topiramate. Topiramate-treated subjects showed improvement in drug use problems and employment problems measured by ASI composite scores. In a set of exploratory analyses that included only subjects who returned for at least one evaluation visit postrandomization (n = 36), we found that topiramate-treated subjects were significantly more likely to attain a clinically significant period of continuous abstinence from cocaine compared to placebo-treated subjects. In previous cocaine pharmacotherapy trials, a stable period of continuous abstinence from cocaine (3–4 weeks) was been found to be predictive of long-term cocaine abstinence (Carroll et al., 1994; Kosten et al., 1992). Of the topiramate-treated subjects, 59% (10/17) attained 3 weeks of continuous abstinence from cocaine during the trial compared to 26% (5/19)
of the placebo-treated subjects. This difference in cocaine abstinence was reflected in CGI scores. Significantly more topiramate-treated subjects were rated as very much improved at the end of treatment compared to placebo-treated subjects (71% versus 32%, respectively). Adverse events were mainly mild and evenly distributed between the topiramate and placebo groups. We found that a slow dose titration successfully prevented the troublesome CNS side effects noted in other trials of topiramate. Thus, with a slow dose titration, topiramate appears to be safe and well tolerated by cocaine dependent patients. The current study has several weaknesses. First, the number of subjects included was small and there was only one woman entered into the trial. As a result, the sample may not be representative of all cocaine users in all treatment centers. For this trial, we selected a sample of cocaine dependent subjects who had relatively low cocaine withdrawal symptom severity at the start of the baseline week and all achieved a period of self-reported abstinence immediately prior to receiving medications. Subjects in this trial also had relatively less severe cocaine dependence measured by days of cocaine use prior in the 30 days prior to treatment and in dollars spent on drugs. Therefore, the results may not be generalizable to other populations of cocaine users with more severe cocaine dependence. Finally, the dose of topiramate selected was relatively low and perhaps a higher dose of topiramate would have yielded even better results. This was a small pilot trial and therefore the results will need to be confirmed in a larger trial. Nevertheless, the data suggest that topiramate may be able to promote abstinence in cocaine dependent patients undergoing outpatient treatment. In the intent-to-treat sample, topiramate was more efficacious than placebo in promoting abstinence from cocaine in the final 5 weeks of the trial. In a smaller sample of subjects who returned for at least one post-randomization visit, topiramate was better than placebo in promoting clinically-significant periods of sustained abstinence from cocaine, and topiramate-treated subjects had higher ratings of clinical improvement at the end of the trial. Topiramate may be a promising medication for the treatment of cocaine dependence and deserves further study in a larger well-controlled trial.
Acknowledgements This study was funded by NIDA grant DA12756.
References Berger, S.P., Hall, S., Mickalian, J.D., Reid, M., Crawford, C.A., Delucchi, K., Carr, K., Hall, S., 1996. Haloperidol antagonism of cue-elicited cocaine craving. Lancet 347, 504–508. Bowman, E.M., Aigner, T.G., Richmond, B.J., 1996. Neural signals in the monkey ventral striatum related to motivation for juice and cocaine rewards. J. Neurophysiol. 75, 1061–1073.
K.M. Kampman et al. / Drug and Alcohol Dependence 75 (2004) 233–240 Carroll, K., Rounsaville, B.J., Gordon, L.T., Nich, C., Jatlow, P., Gawin, F., 1994. One-year follow up of psychotherapy and pharmacotherapy for cocaine dependence: delayed emergence of psychotherapy effects. Arch. Gen. Psychiatry 51, 989–997. Childress, A., Ehrman, R., Damaris, J., Rohsenow, D., Robbins, S., O’Brien, C.P., 1993. Classically conditioned factors in drug dependence. In: Lowinson, J., Ruiz, P., Millman, R. (Eds.), Comprehensive Textbook of Substance Abuse. Williams and Wilkins, Baltimore, pp. 56–69. Childress, A.R., Mozley, P.D., McElgin, W., Fitzgerald, J., Reivich, M., O’Brien, C.P., 1999. Limbic activation during cue-induced craving. Am. J. Psychiatry 156, 11–18. Cornish, J.L., Kalivas, P.W., 2000. Glutamate transmission in the nucleus accumbens mediates relapse in cocaine addiction. J. Neurosci. 20, 1–5. de Lima, M.S., de Oliveira Soares, B.G., Reisser, A.A.P., Farrell, M., 2000. Pharmacological treatment of cocaine dependence: a systematic review. Addiction 97, 931–949. Dewey, S., Morgan, A., Ashby, C., Horan, B., Kushner, S., Logan, J., Volkow, N., Fowler, J., Gardner, E., Brodie, J., 1998. A novel strategy for the treatment of cocaine addiction. Synapse 30, 119–129. Dewey, S., Smith, G., Logan, J., Brodie, J., Yu, D., Ferrieri, R., King, P., MacGregor, R., Martin, T., Wolf, A.P., 1992. GABAergic inhibition of endogenous dopamine release measured in vivo with 11c-raclopride and positron emission tomography. J. Neurosci. 12, 3773–3780. Dewey, S.L., Chaurasia, C.S., Chen, C., Volkow, N.D., Clarkson, F.A., Porter, S., Straughter-Moore, R.M., Alexoff, D.L., Tedeschi, D., Russo, N.B., Fowler, J.S., Brodie, J.D., 1997. GABAergic attenuation of cocaine-induced dopamine release and locomoter activity. Synapse 25, 393–398. Diggle, P.J., Liang, K.Y., Zeger, S.L., 1994. Analysis of Longitudinal Data. Oxford University Press, New York. Dworkin, S.I., Smith, J.E., 1988. Neurobehavioral pharmacology of cocaine. In: Clouet, D., Ashgar, K., Brown, R. (Eds.), Mechanisms of Cocaine Abuse and Toxicity. National Institute on Drug Abuse Research Monograph #88. US Government Printing Office, Washington, DC, pp. 185–198. Gerasimov, M.R., Ashby, C.R., Gardner, E.L., Mills, M.J., Brodie, J.D., Dewey, S.L., 1999. Gamma-vinyl-GABA inhibits methamphetamine, heroin, or ethanol-induced increases in nucleus accumbens dopamine. Synapse 34, 11–19. Goeders, N.E., Dworkin, S.I., Smith, J.E., 1986. Neuropharmacological assessment of cocaine self-administration into the medial prefrontal cortex. Pharmacol. Biochem. Behav. 24, 1429–1440. Guy, W. (Ed.), 1976. Assessment Manual for Psychopharmacology: Publication ADM 76-338. US Department of Health Education and Welfare, Washington, DC. Johnson, B.A., Ait-Daoud, N., Bowden, C.L., DiClemente, C.C., Roache, J.D., Lawson, K., Javors, M.A., Ma, J.Z., 2003. Oral topiramate for the treatment of alcohol dependence: a randomized controlled trial. Lancet 361, 1677–1685. Kadden, R., Carroll, K., Donovan, D., Cooney, N., Monti, P., Abrams, D., Litt, M., Hester, R., 1995. Cognitive-Behavioral Coping Skills Therapy Manual: A Clinical Research Guide for Therapists Treating Individuals With Alcohol abuse and Dependence. National Institutes of Health, Rockville, MD. Kampman, K.M., Alterman, A.I., Volpicelli, J.R., Maany, I., Muller, E.S., Luce, D.D., Mulholland, E.M., Jawad, A., Parikh, G., Mulvaney, F.D., Weinieb, R.M., O’Brien, C.P., 2001a. Cocaine withdrawal symptoms and initial urine toxicology results predict treatment attrition in outpatient cocaine dependence treatment. Psychol. Addict. Behav. 15, 52–59. Kampman, K.M., Volpicelli, J.R., Alterman, A.I., Cornish, J.W., O’Brien, C.P., 2000. Amantadine in the treatment of cocaine dependent patients with severe cocaine withdrawal symptoms. Am. J. Psychiatry 157, 2052–2054. Kampman, K.M., Volpicelli, J.R., McGinnis, D.E., Alterman, A.I., Weinrieb, R.M., D’Angelo, L., Epperson, L., 1998. Reliability and valid-
239
ity of the Cocaine Selective Severity Assessment. Addict. Behav. 23, 449–461. Kampman, K.M., Volpicelli, J.R., Mulvaney, F.D., Alterman, A.I., Cornish, J., Gariti, P., Cnaan, A., Poole, S., Muller, E., Acosta, T., Luce, D., O’Brien, C.P., 2001b. Effectiveness of propranolol for cocaine dependence may depend on cocaine withdrawal symptom severity. Drug Alcohol Depend. 63, 69–78. Kampman, K.M., Volpicelli, J.R., Mulvaney, F.D., Rukstalis, M., Alterman, A.I., Pettinati, H., Weinieb, R.M., O’Brien, C.P., 2002. Cocaine withdrawal severity and urine toxicology results from treatment entry predict outcome in medication trials for cocaine dependence. Addict. Behav. 27, 251–260. Koob, G., Vaccarino, F., Amalric, M., Bloom, F., 1987. Positive reinforcement properties of drugs: search for neural substrates. In: Engel, J., Oreland, L. (Eds.), Brain Reward Systems and Abuse. Raven Press, New York, pp. 35–50. Koob, G.F., 1992. Neural mechanisms of drug reinforcement. Ann. N.Y. Acad. Sci. 654, 171–191. Kosten, T., Gawin, F.H., Kosten, T.A., Morgan, C., Rounsaville, B.J., Schottenfeld, R., Kleber, H.D., 1992. Six-month follow-up of short-term pharmacotherapy for cocaine dependence. Am. J. Addict. 1, 40–49. Kushner, S.A., Dewey, S.L., Kornetsky, C., 1999. The irreversible ␥-aminobutyric acid (GABA) transaminase inhibitor ␥-vinyl-GABA blocks cocaine self-administration in rats. J. Pharmacol. Exp. Ther. 290, 797–802. Kuzniecky, R., Hetherington, H., Ho, S., Pan, J., Martin, R., Gilliam, F., Hugg, J., Faught, E., 1998. Topiramate increases cerebral GABA in healthy humans. Neurology 51, 627–629. Littell, R.C., Milliken, G.A., Stroup, W.W., Wolfinger, R.D., 1996. SAS System for Mixed Models. SAS Institute Inc, Cary, NC. Littell, R.C., Pendergast, J., Natarajan, R., 2000. Modeling covariance structure in the analysis of repeated measures data. Stat. Med. 19, 1793–1819. McFarland, K., Kalivas, P.W., 2001. The circuitry mediating cocaine-induced reinstatement of drug-seeking behavior. J. Neurosci. 21, 8655–8663. McFarland, K., Lapish, C.C., Kalivas, P.W., 2003. Prefrontal glutamate release into the core of the nucleus accumbens mediates cocaine-induced reinstatement of drug seeking behavior. J. Neurosci. 23, 3531–3537. McLellan, A.T., Kushner, H., Metzger, D., Peters, R., Smith, I., Grissom, G., Pettinati, H., Argeriou, M., 1992. The fifth edition of the addiction severity index. J. Subst. Abuse Treat. 9, 199–213. Mulvaney, F.D., Alterman, A.I., Boardman, C.R., Kampman, K., 1999. Cocaine abstinence symptomatology and treatment attrition. J. Subst. Abuse Treat. 16, 129–135. Petroff, O.A., Hyder, F., Mattson, R.H., Rothman, D.L., 1999. Topiramate increases brain GABA, homocarnasine, and pyrrolidinone in patients with epilepsy. Neurology 52, 473–478. Privetera, M.D., 1997. Topiramate: a new antiepileptic drug. Ann. Pharmacother. 31, 1164–1173. Roberts, D.C.S., Koob, G.F., Klonoff, P., Figiber, H.C., 1980. Extinction and recovery of cocaine self-administration following 6-hydroxydopamine lesions of the nucleus accumbens. Pharmacol. Biochem. Behav. 12, 781–787. Roberts, D.C., Andrews, M.M., Vickers, G.J., 1996. Baclofen attenuates the reinforcing effects of cocaine in rats. Neuropsychopharmacology 15, 417–423. Rubio, G., Ponce, G., Jimenez-Arriero, M.A., Palomo, T., Ferre, F., 2002. Effectiveness of topiramate in control of alcohol craving. Eur. Neuropsychopharmacol. 12, S63. Schultz, W., 1998. Predictive reward signals of dopamine neurons. J. Neurophysiol. 80, 1–27. Severt, L., Coulter, D.S., Sombati, S., 1995. Topiramate selectively blocks kainate currents in cultured hippocampal neurons. Epilepsia 36, 38.
240
K.M. Kampman et al. / Drug and Alcohol Dependence 75 (2004) 233–240
Sharples, K., Breslow, N., 1992. Regression analysis of correlated binary data: some small sample results for the estimating equation approach. J. Stat. Comput. Simul. 42, 1–20. Somoza, E., Dyrenforth, S., Goldsmith, J., Mezinskis, J., Cohen, M., 1995. In search of a universal drug craving scale. Paper Presented at the Annual Meeting of the American Psychiatric Association, Miami, FL. Weiss, F., Maldanado-Vlaar, C.S., Parsons, L.H., Kerr, T.M., Smith, D.L., Ben-Shahar, O., 2000. Control of cocaine-seeking behavior by drug-associated stimuli in rats: effects on recovery of extinguished operant responding and extracellular dopamine levels in amygdala and nucleus accumbens. Proc. Natl. Acad. Sci. U.S.A. 97, 4321–4326.
Westerink, B.H.C., Enrico, P., Feiman, J., De Vries, J.B., 1998. The pharmacology of mesocortical dopamine neurons: a dual probe microdialysis study of the ventral tegmental area and prefrontal cortex of the rat brain. J. Pharmacol. Exp. Ther. 285, 143–154. White, H.S., Brown, S.D., Woodhead, J.H., Skeen, G.A., Wolof, H.H., 1997. Topiramate enhances GABA-mediated chloride flux and GABA-evoked chloride currents in murine brain neurons and increases seizure threshold. Epilepsy Res. 28, 167–179. Zullino, D.F., Cottier, A.C., Besson, J., 2002. Topiramate in opiate withdrawal. Prog. Neuropsychopharmacol. Biol. Psychiatry 26, 1221– 1223.
A pilot trial of topiramate for the treatment of cocaine dependence Kyle M. Kampman a,b,∗ , Helen Pettinati a , Kevin G. Lynch a , Charles Dackis a , Thorne Sparkman c , Catherine Weigley a , Charles P. O’Brien a a
Department of Psychiatry, University of Pennsylvania School of Medicine, 3900 Chestnut Street, Philadelphia, PA 19104, USA b Department of Veterans Affairs Medical Center, University and Woodland Avenues, Philadelphia, PA 19104, USA c Department of Medicine, University of Pennsylvania School of Medicine, 3400 Spruce Street, Philadelphia, PA 19104, USA Received 14 July 2003; received in revised form 19 January 2004; accepted 10 March 2004
Abstract Background: Both GABAergic and glutamatergic neurons appear to be important modulators of the brain reward system and medications that affect GABA and glutamatergic neurotransmission may reduce the rewarding properties of cocaine and reduce cocaine craving. Topiramate, an anticonvulsant, raises cerebral GABA levels, facilitates GABAergic neurotransmission and inhibits glutametergic activity at AMPA/kainite receptors. Thus, it may be useful for treating cocaine dependence. Methods: The efficacy of topiramate for cocaine dependence was tested in a 13-week, double-blind, placebo-controlled pilot trial (n = 40). Topiramate was titrated gradually over 8 weeks to a dose of 200 mg daily. The primary outcome measure was cocaine abstinence verified by twice weekly urine benzoylecgonine tests (UBT). Results: Eighty-two percent of subjects completed the trial. Analysis of the UBT using a GEE model showed that after week 8, when the dose titration was completed, topiramate-treated subjects were more likely to be abstinent from cocaine compared to placebo-treated subjects (Z = 2.67, P = 0.01). Topiramate-treated subjects were also more likely to attain 3 weeks of continuous abstinence from cocaine (χ2 = 3.9, d.f. = 1, P = 0.05). Conclusion: Topiramate may be effective for the treatment of cocaine dependence. © 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Cocaine; Topiramate; Clinical trial; GABA; Glutamate
1. Introduction Multiple studies have suggested that increases in extracellular dopamine in the mesocorticolimbic dopamine system mediate the acute reinforcing effects of cocaine (Dworkin and Smith, 1988; Goeders et al., 1986; Koob et al., 1987; Roberts et al., 1980). In addition, the mesocorticolimbic dopamine system has been shown to be activated in response to conditioned stimuli associated with prior cocaine administration (Bowman et al., 1996; Schultz, 1998; Weiss et al., 2000). This activation may be associated with the subjective experience of cocaine craving in humans which often promotes relapse to cocaine use (Berger et al., 1996; Childress et al., 1993, 1999). Initial studies of pharmacological treatments for cocaine dependence focused largely on drugs that would directly influence dopaminergic func-
∗ Corresponding author. Tel.: +1-215-222-3200x109; fax: +1-215-386-6770. E-mail address: kampman [email protected] (K.M. Kampman).
tion, but these agents have been consistently found to be ineffective (for a review see de Lima et al., 2000). Another way of affecting dopaminergic function is indirectly through other modulatory neurotransmitter systems, such as the gamma amino butyric acid (GABA) system and the glutamate system. Several lines of evidence suggest that GABAergic neurons function as modulators of the mesocorticolimbic dopamine system and play an important role in mediating the reinforcing effects of cocaine (Dewey et al., 1998; Koob, 1992). Preclinical trials of medications that foster GABAergic neurotransmission have suggested that these compounds reduce the dopamine response to both cocaine administration and to conditioned cues of prior cocaine use (Dewey et al., 1992, 1997; Gerasimov et al., 1999). GABAergic medications also reduce the self-administration of cocaine in various animal models (Kushner et al., 1999; Roberts et al., 1996). Similarly, glutamatergic neurons appear to be important modulator of the mesocorticolimbic reward system. Glutamatergic neurons located in the prefrontal cortex innervate the nucleus accumbens and play a critical role in cocaine-reinforced behavior (McFarland and
0376-8716/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.drugalcdep.2004.03.008
234
K.M. Kampman et al. / Drug and Alcohol Dependence 75 (2004) 233–240
Kalivas, 2001). Inhibition of prefrontal glutamatergic neurons blocks cocaine induced reinstatement of drug seeking behavior in rats (Cornish and Kalivas, 2000; McFarland et al., 2003). Thus, medications that promote GABAergic activity or inhibit glutamatergic activity may be promising medications for cocaine dependence treatment. Topiramate is approved by the US Food and Drug Administration for the treatment of epileptic seizures. It has effects in both the GABA and glutamate systems (Privetera, 1997). Topiramate enhances GABA activity at GABA-A receptors and causes a general increase in GABA levels in the brain (Kuzniecky et al., 1998; Petroff et al., 1999; White et al., 1997). Topiramate also antagonizes glutamate transmission through effects at AMPA/kainite receptors (Severt et al., 1995). Thus, topiramate may reduce cocaine craving and cocaine use in cocaine dependent patients. Topiramate has been studied for the treatment of various addictions and published reports suggest it may be useful for both alcohol and opiate dependence. Topiramate reduced alcohol craving and alcohol use in a group of 24 alcohol dependent subjects in an open trial (Rubio et al., 2002). More recently, topiramate was found to be more effective than placebo in reducing alcohol use in a double-blind trial involving 150 alcoholics (Johnson et al., 2003). Finally, there is a case series published suggesting topiramate may be useful for the treatment of opiate withdrawal (Zullino et al., 2002). To our knowledge there are no published studies of topiramate for cocaine dependence. The current pilot study was intended to obtain more data regarding the safety, tolerability and potential efficacy of topiramate for the treatment of cocaine dependence.
2. Methods 2.1. Subjects The subjects were 40 DSM-IV treatment-seeking, cocaine-dependent men and women between the ages of 18 and 60 years. They were recruited at the University of Pennsylvania Treatment Research Center (TRC). The TRC recruits through advertisement in the local media as well as through professional referrals. Patients entering the trial were not involved in any other type of addiction treatment elsewhere. All subjects signed informed consent prior to participation in the trial, after trial procedures were explained to them by an investigator. The study was reviewed and approved by the Institutional Review Board (IRB) of the University of Pennsylvania. Subjects were reimbursed $5.00 at each visit for completing all research procedures; at the last visit they received $30.00 because of the greater number of research procedures done at the end of the study. Subjects received an additional $5.00 each week for returning the previous week’s medication package in order to facilitate the pill count compliance check. If needed, two transit tokens were provided at each visit.
Psychiatric diagnoses were obtained by a board-certified addiction psychiatrist (KK) through a clinical interview. Subjects were required to have self-reported at least $100 worth of cocaine use in the month prior to entry. Because topiramate was viewed as a medication that might worsen cocaine withdrawal symptoms by reducing dopaminergic tone (Westerink et al., 1998), subjects were selected who had low cocaine withdrawal symptom severity on the day of intake demonstrated by low scores (below 22) on the Cocaine Selective Severity Assessment (CSSA) (Kampman et al., 1998). Subjects also were required to maintain at least 3 days of self-reported abstinence from cocaine immediately prior to starting study medication. At our center, we have found that cocaine withdrawal symptom severity affects outcome in outpatient cocaine dependence treatment (Mulvaney et al., 1999; Kampman et al., 2000, 2001a,b, 2002) and we frequently use withdrawal symptom severity as a factor in assigning subjects to particular clinical trials. Medical screening included a complete medical history and physical examination conducted by a certified nurse practitioner. Baseline laboratory testing included a chemistry screen, complete blood count, urinalysis, eye exam with visual acuity and a 12 lead EKG. Women of childbearing potential were required to use barrier contraception during the trial. Women received urinary pregnancy testing prior to starting medications, and at monthly intervals throughout the study. Liver function tests were monitored monthly throughout the trial. Subjects with current dependence (DSM-IV criteria) on any additional drug except nicotine were excluded. Psychiatric exclusion criteria included psychosis, dementia, and the use of other psychotropic medications. Medical exclusion criteria included unstable medical illnesses, use of antiepileptic medications, a history of nephrolithiasis, a history of glaucoma, and a history of hypersensitivity to topiramate. 2.2. Measures The primary outcome measure for this trial was qualitative urine benzoylecgonine tests (UBT) obtained twice weekly. Urine collection was monitored by temperature checks. Samples less than 90 ◦ F, or greater than 100 ◦ F were not accepted. Samples were analyzed for benzoylecgonine by fluorescent polarization assay. Samples containing equal to or greater than 300 ng/ml of benzoylecgonine were considered to be positive. Secondary outcome measures for the trial included treatment retention, measured by the number of evaluation sessions attended. Severity of addiction-related problems was measured by the Addiction Severity Index (ASI) (McLellan et al., 1992) administered four times, at baseline, 4 weeks after starting medications, 8 weeks after starting medications, and at the end of the trial. The study physician rated illness severity and improvement weekly using the Clinical Global Impression Scale (CGI) (Guy, 1976). Cocaine crav-
K.M. Kampman et al. / Drug and Alcohol Dependence 75 (2004) 233–240
ing was measured weekly using the Brief Substance Craving Scale (BSCS) (Somoza et al., 1995). Safety measures included adverse events, which were monitored at each visit, and liver function tests monitored every 4 weeks during the trial. 2.3. Procedures Eligible subjects entered a 13-week trial that included a 1-week baseline phase during which all pretreatment measures were obtained and subjects began psychosocial treatment. Subjects were required to attend two visits during this week to be eligible to receive study medications. Eligible subjects were randomized to receive either topiramate or placebo. Because rapid dose titration of topiramate can result in uncomfortable central nervous system side effects such as sedation, slurred speech and trouble selecting words, the dose of topiramate was started very low (25 mg daily, given at bedtime) and the dose was increased 25 mg each week for 8 weeks until patients were receiving 200 mg of topiramate each day, or identical placebo capsules, given at bedtime. During the last week of the trial topiramate was rapidly reduced by 25 mg each night. Study medication was prepared by the Research Pharmacist at the Hospital of the University of Pennsylvania by over-encapsulating topiramate 25 and 100 mg tablets, purchased from a commercial pharmacy, and producing identical-appearing lactose-containing placebo capsules. Study medication was placed in blister packs with each day’s dose clearly marked. Medications were dispensed by the study physician each week and the previous week’s blister pack was collected. Compliance was measured by pill count. Patients were instructed to take their daily dose of medication at night before going to sleep. In addition to medication or placebo, subjects received twice weekly individual cognitive-behavioral relapse prevention therapy utilizing a Cognitive-Behavioral Coping Skills Therapy (CBT) manual. The CBT therapy manual and supporting materials were developed for the National Institute on Alcohol Abuse and Alcoholism Project MATCH (Kadden et al., 1995). The basic format was accepted, although specific procedures were adapted for treatment of cocaine dependence by our group. Therapy was provided by Master’s level therapists with additional training in CBT. 2.4. Statistical analysis Baseline measures between the topiramate and placebo groups were compared using t-tests for continuous variables and χ2 -tests for dichotomous variables. Measures were transformed when necessary to reduce skew. The number of sessions attended for each group during the trial was compared using a t-test. UBT results were compared using generalized estimating equations (GEE) models (Diggle et al., 1994). There were two urines collected per subject each
235
week, yielding a possible 24 urine drug screens (UDS) per subject. GEE approaches work best when the data are in the form of a small number of repeated measurements for each of many subjects (Sharples and Breslow, 1992). As our sample size is rather small, we chose to reduce the number of time points per subject by coding each of the 12 weeks as a use or non-use week rather than including each visit in the model. We initially labeled each study week as a cocaine non-use week or use week according to whether the subject provided two negative UBT (cocaine non-use week), or at least one positive UBT or one missing UBT (cocaine use week). In short, missing urine samples were labeled as positive. Using weeks as the time units of analysis, rather than visits, also reduces the risk of overestimating drug use due to benzoylecgonine carryover. Benzoylecgonine remains detectable in the urine at significant levels for several days, so one episode of use may lead to several BE positive samples. There was a gap of at least 3 days between the final UDS of 1 week and the first UDS of the next week, so the possibility of carryover from 1 week to the next should be negligible. The screening urine and two urines submitted during week 1 were used to determine use/non-use status prior to randomization, and this status was included as a covariate in the GEE model. Because topiramate required an 8-week dose run up to achieve full dose, we included a binary time-varying covariate to indicate the time before full dose was achieved (value 0 for weeks 2–8), or after full dose was achieved (value 1 for weeks 9–13). A set of continuously distributed secondary outcome measures gathered over varying time frames was analyzed using linear mixed effects models. The models used to compare the groups on the secondary outcome measures generally included the pre-treatment version of the response as a covariate, together with treatment group, and time effects. For responses with only a few time points, time was included as a categorical variable; for responses with a greater number of time points, linear time trends were found to be sufficient. The two-way interactions among pre-treatment response, medication group, and time were also considered for inclusion in these models by examining P-values of regression coefficients for the GEE models, and Akaike’s Information Criterion (AIC) (Littell et al., 1996) comparisons of overall model fit for the mixed effects models. For the GEE models for the UBT results, an exchangeable (compound symmetry) structure was used for the working correlation matrix. For the mixed effects models, covariance models were chosen by the methods described by Littell et al. (2000): models having exchangeable residuals fit best, sometimes with the inclusion of a random intercept. In the analyses of the secondary measures, the scales from the ASI, and the BSCS there were few significant effects. In the descriptions below, we report only the medication effects, and other significant effects, from the final model. In those descriptions of the analyses on the UBT and secondary responses, note that all Z and χ2 -statistics quoted are obtained from the models.
236
K.M. Kampman et al. / Drug and Alcohol Dependence 75 (2004) 233–240
ite Drug Scores and higher Hamilton Depression Rating Scale scores compared to placebo-treated subjects.
3. Results, intent-to-treat sample 3.1. Baseline demographic and drug use
3.2. Treatment retention On the whole, the two study groups, topiramate and placebo, were very similar in demographics and baseline drug use characteristics (see Table 1). The average age of the subjects was about 40. Most were African American men and most smoked crack cocaine. On average, subjects had used cocaine between 6 and 8 days in the month prior to treatment (range 1–25 days) and spent between $300 and $500 on drugs. Although all subjects met criteria for current DSM-IV cocaine dependence and had used at least $100 worth of cocaine in the 30 days prior to the trial, we selected patients with low cocaine withdrawal symptom severity for this trial. Therefore, the subjects were on the milder end of the spectrum of addiction severity measured by days of cocaine use and money spent for cocaine. Although the subgroup selected for this trial may have had less severe symptoms of addiction, the group represented a significant proportion of cocaine dependent subjects who presented for treatment at our center. During the 10 months we recruited for this trial, 186 cocaine dependent subjects were screened at intake and 72 had CSSA scores less than 22 and thus qualified for this trial. The topiramate and placebo groups differed in two ways, topiramate-treated subjects had, on average, significantly higher ASI Compos-
Only 6 of the 40 randomized subjects were lost to follow-up, 4/20 in the placebo group and 2/20 in the topiramate group (χ2 = 0.78, d.f. = 1, P = 0.38). Treatment retention, measured by the number of visits attended, did not differ significantly. On average, topiramate-treated subjects attended 17 visits and placebo-treated subjects attended 15 visits (t = −0.052, d.f. = 38, P = 0.96). 3.3. Urine benzoylecgonine test results We found a significant group by time by dose-period interaction, with topiramate-treated subjects becoming more likely to be cocaine abstinent after week 8, and placebo-treated subjects becoming less likely (see Fig. 1). There was no significant difference between the two groups during the dose run-up period of the trial: the placebo group had odds of use 1.31 times higher than those in the topiramate group (Z = 1.42, P = 0.10), and neither time effects (Z = 1.00, P = 0.32) nor group by time effects (Z = −1.10, P = 0.27) were significantly different from zero. There was a significant group by time interaction during the full-dose period: the odds of use in the placebo group
Table 1 Subject characteristics, expressed as percent or mean (standard deviation) Variable
Topiramate (N = 20)
Age Male (%)
40.10 (7.10) 95
Race African American (%) Caucasian (%)
95 5
Placebo (N = 20) 39.90 (6.38) 100 85 15
Years of education Days of alcohol use in past 30 daysa Days of cocaine use in past 30 daysa $ spent for drugsa Years of cocaine use, lifetime Number of prior treatments ASI Composite Drug Scorea,b ASI Composite Alcohol Score ASI Composite Employment Score ASI Composite Legal Scorea ASI Composite Family/Social Scorea ASI Composite Psychiatric Scorea ASI Composite Medical Scorea Hamilton Depression Rating Scalea,c Hamilton Anxiety Rating Scale
12.55 (1.47) 3.85 (5.15) 7.90 (7.07) 513.25 (342.62) 10.10 (6.35) 1.55 (1.54) 0.206 (0.049) 0.072 (0.086) 0.577 (0.245) 0.132 (0.248) 0.180 (0.211) 0.155 (0.187) 0.126 (0.257) 6.30 (4.210 3.15 (2.18)
12.35 (1.63) 6.25 (7.99) 6.40 (7.63) 364.75 (343.09) 12.75 (4.79) 1.60 (1.70) 0.162 (0.082) 0.124 (0.130) 0.522 (0.294) 0.090 (0.152) 0.078 (0.146) 0.062 (0.126) 0.143 (0.287) 3.45 (3.38) 2.55 (2.82)
Route of cocaine use Intranasal (%) Smoked (%)
10 90
15 85
a b c
Transformed prior to analysis to reduce skew. Significantly different t = −2.03, d.f. = 38, P = 0.05. Significantly different t = −2.57, d.f. = 38, P = 0.014.
K.M. Kampman et al. / Drug and Alcohol Dependence 75 (2004) 233–240
% Cocaine Abstinent
60%
237
cant group effect in the drug composite scores with lower scores in the topiramate group (F(1/34) = 4.12, P = 0.05). There was also a significant group by time interaction for the employment composite score (F(1/60) = 4.56, P = 0.04). Topiramate-treated subjects showed a significantly greater reduction in the employment composite score over the course of the trial than did the placebo-treated subjects.
50% 40% 30% 20% 10% 0% 1 2 3 4 5 6 7 8 9 10 11 12 13 Week
placebo, n = 20
topiramate, n = 20
Fig. 1. Percentage of subjects who were cocaine abstinent at each study week in the placebo and topiramate-treated groups.
increased over the period (odds ratio per week = 1.29, P = 0.14), while the odds of use in the topiramate group decreased over the period (odds ratio per week = 0.73, P = 0.04), with a significant difference between the two rates (Z = 2.67, P = 0.01). Finally, subjects who provided a clean UDS at baseline were less likely to use during the study than those who did not (odds ratio = 0.20, P = 0.01). 3.3.1. Missing data Although 33/40 subjects completed the study there were a number of missed visits resulting in missing data. The data most affected by missed visits were the UBT. Three hundred thirty-two urine samples out of 960 or 34.6% of the urine samples were missing. Because of the small sample size we addressed missing data using a clinical approach rather than through formal statistical modeling of the drop out process. In our initial model, missing urine samples were considered positive. However, this implies that patients who drop out of treatment always do so because of treatment failure. This may not always be the case. In order to ensure that we were not overestimating cocaine use in either group due to coding missed urine samples from dropouts as positive, we reran our original model excluding samples missing due to study dropout. Because the number of missed visits was not significantly different between groups ((t-test) t = −0.87, d.f. = 38, P = 0.39) and the time to last visit was not significantly different between groups (log-rankχ2 = 0.004, d.f. = 1, P = 0.98), we feel that this is a valid approach to account for the missing data. Using this model did not significantly change the conclusions: in particular, there was still a significant group by time by dose-period interaction (Z = 2.42, P = 0.03). 3.4. Results from the Addiction Severity Index Days of cocaine use, dollars spent on drugs and the legal composite scores tended to decline in both groups during the trial but there were no medication effects detected. The ASI Composite Drug Score declined significantly during the trial (F(1/60) = 11.97, P = 0.001). There was a signifi-
3.5. Results from Brief Substance Craving Scale The BSCS divides cocaine craving into three domains: intensity, duration and frequency. Each domain is measured on a 0–4 likert scale. By adding the three scores a craving composite measure can be derived. Although this measure is not completely continuous, its distribution was reasonably symmetric, and it was appropriate for the mixed effects models. Cocaine craving showed a significant decline over the trial in both groups: (F(1/285) = 17.08, P < 0.0001), and there was a significant interaction between baseline scores and time; subjects with higher craving at baseline showed a greater reduction in craving during the trial (F(1, 284) = 11.84, P = 0.0007). Although average craving scores declined slightly more in the topiramate group, this effect did not achieve statistical significance (F(1/34) = 2.70, P = 0.11). 3.6. Medication adherence Medication adherence was measured by pill count. The percentage of pills taken was calculated by subtracting the number of pills returned each week from the number of pills dispensed. Both groups showed good adherence. The mean percentage of prescribed pills taken by the topiramate-treated subjects was significantly higher compared to the placebo-treated subjects (94% versus 83%, t = −2.7, d.f. = 34, P = 0.01). 3.7. Safety analyses Adverse events were assessed at each visit. Topiramate was well tolerated. Adverse events were mainly mild and evenly distributed between the topiramate and placebo groups. The most commonly reported adverse events included headache (n = 6), fatigue (n = 6), diarrhea (n = 6), constipation (n = 4), and nausea (n = 4). There were no significant differences noted in the occurrence of any adverse event between the two medication groups. We did have one topiramate-treated subject who developed symptoms of kidney stones at week 9 and was referred to a urologist for treatment.
4. Results, evaluable cases In this trial there were four subjects who received their first dose of study medication and did not attend any
238
K.M. Kampman et al. / Drug and Alcohol Dependence 75 (2004) 233–240
subsequent evaluation visits, three in the topiramate group and one in the placebo group. It is impossible to evaluate what effect the study medication may have had on these subjects. Therefore, in a set of separate exploratory analyses, we looked at some outcome measures excluding these four subjects. We specifically examined rates of cocaine abstinence, and global ratings of improvement. We found that in this group of 36 subjects, topiramatetreated subjects were significantly more likely to achieve 3 or more weeks of continuous abstinence from cocaine measured by negative urine toxicology screens. We found that 10 out of 17 or 59% of topiramate subjects attained 3 or more continuous weeks of abstinence from cocaine compared to only 5 of 19 or 26% in the placebo group (χ2 = 3.9, d.f. = 1, P = 0.05). The higher rates of abstinence achieved by the topiramatetreated subjects were reflected in rates of improvement noted by the treating clinician at the last visit. Seventy one percent (12/17) of topiramate-treated subjects were rated as very much improved on the CGI compared to 32% (6/19) of placebo-treated subjects (χ2 = 5.46, d.f. = 1, P = 0.02).
5. Discussion Data from this pilot trial suggest that topiramate may be a promising medication for the treatment of cocaine dependence. In the intent-to-treat analysis, which included all 40 randomized subjects, topiramate was shown to be safe and well tolerated. In the primary outcome measure, cocaine abstinence measured by UBT results, there was a significant effect of topiramate, which was observable after week 8, when the dose titration was complete. After week 8, subjects received 200 mg of topiramate each night. This result is similar to the result noted by Johnson et al. in their study of topiramate for alcohol dependence. In that study, Johnson et al. (2003) found the greatest benefit from topiramate to occur after week 6, when subjects were receiving 200 mg daily of topiramate. Like Johnson, we cannot say whether what we observed was a dose–response effect, or simply an effect related to time of exposure to topiramate. ASI results also showed a beneficial effect of topiramate. Topiramate-treated subjects showed improvement in drug use problems and employment problems measured by ASI composite scores. In a set of exploratory analyses that included only subjects who returned for at least one evaluation visit postrandomization (n = 36), we found that topiramate-treated subjects were significantly more likely to attain a clinically significant period of continuous abstinence from cocaine compared to placebo-treated subjects. In previous cocaine pharmacotherapy trials, a stable period of continuous abstinence from cocaine (3–4 weeks) was been found to be predictive of long-term cocaine abstinence (Carroll et al., 1994; Kosten et al., 1992). Of the topiramate-treated subjects, 59% (10/17) attained 3 weeks of continuous abstinence from cocaine during the trial compared to 26% (5/19)
of the placebo-treated subjects. This difference in cocaine abstinence was reflected in CGI scores. Significantly more topiramate-treated subjects were rated as very much improved at the end of treatment compared to placebo-treated subjects (71% versus 32%, respectively). Adverse events were mainly mild and evenly distributed between the topiramate and placebo groups. We found that a slow dose titration successfully prevented the troublesome CNS side effects noted in other trials of topiramate. Thus, with a slow dose titration, topiramate appears to be safe and well tolerated by cocaine dependent patients. The current study has several weaknesses. First, the number of subjects included was small and there was only one woman entered into the trial. As a result, the sample may not be representative of all cocaine users in all treatment centers. For this trial, we selected a sample of cocaine dependent subjects who had relatively low cocaine withdrawal symptom severity at the start of the baseline week and all achieved a period of self-reported abstinence immediately prior to receiving medications. Subjects in this trial also had relatively less severe cocaine dependence measured by days of cocaine use prior in the 30 days prior to treatment and in dollars spent on drugs. Therefore, the results may not be generalizable to other populations of cocaine users with more severe cocaine dependence. Finally, the dose of topiramate selected was relatively low and perhaps a higher dose of topiramate would have yielded even better results. This was a small pilot trial and therefore the results will need to be confirmed in a larger trial. Nevertheless, the data suggest that topiramate may be able to promote abstinence in cocaine dependent patients undergoing outpatient treatment. In the intent-to-treat sample, topiramate was more efficacious than placebo in promoting abstinence from cocaine in the final 5 weeks of the trial. In a smaller sample of subjects who returned for at least one post-randomization visit, topiramate was better than placebo in promoting clinically-significant periods of sustained abstinence from cocaine, and topiramate-treated subjects had higher ratings of clinical improvement at the end of the trial. Topiramate may be a promising medication for the treatment of cocaine dependence and deserves further study in a larger well-controlled trial.
Acknowledgements This study was funded by NIDA grant DA12756.
References Berger, S.P., Hall, S., Mickalian, J.D., Reid, M., Crawford, C.A., Delucchi, K., Carr, K., Hall, S., 1996. Haloperidol antagonism of cue-elicited cocaine craving. Lancet 347, 504–508. Bowman, E.M., Aigner, T.G., Richmond, B.J., 1996. Neural signals in the monkey ventral striatum related to motivation for juice and cocaine rewards. J. Neurophysiol. 75, 1061–1073.
K.M. Kampman et al. / Drug and Alcohol Dependence 75 (2004) 233–240 Carroll, K., Rounsaville, B.J., Gordon, L.T., Nich, C., Jatlow, P., Gawin, F., 1994. One-year follow up of psychotherapy and pharmacotherapy for cocaine dependence: delayed emergence of psychotherapy effects. Arch. Gen. Psychiatry 51, 989–997. Childress, A., Ehrman, R., Damaris, J., Rohsenow, D., Robbins, S., O’Brien, C.P., 1993. Classically conditioned factors in drug dependence. In: Lowinson, J., Ruiz, P., Millman, R. (Eds.), Comprehensive Textbook of Substance Abuse. Williams and Wilkins, Baltimore, pp. 56–69. Childress, A.R., Mozley, P.D., McElgin, W., Fitzgerald, J., Reivich, M., O’Brien, C.P., 1999. Limbic activation during cue-induced craving. Am. J. Psychiatry 156, 11–18. Cornish, J.L., Kalivas, P.W., 2000. Glutamate transmission in the nucleus accumbens mediates relapse in cocaine addiction. J. Neurosci. 20, 1–5. de Lima, M.S., de Oliveira Soares, B.G., Reisser, A.A.P., Farrell, M., 2000. Pharmacological treatment of cocaine dependence: a systematic review. Addiction 97, 931–949. Dewey, S., Morgan, A., Ashby, C., Horan, B., Kushner, S., Logan, J., Volkow, N., Fowler, J., Gardner, E., Brodie, J., 1998. A novel strategy for the treatment of cocaine addiction. Synapse 30, 119–129. Dewey, S., Smith, G., Logan, J., Brodie, J., Yu, D., Ferrieri, R., King, P., MacGregor, R., Martin, T., Wolf, A.P., 1992. GABAergic inhibition of endogenous dopamine release measured in vivo with 11c-raclopride and positron emission tomography. J. Neurosci. 12, 3773–3780. Dewey, S.L., Chaurasia, C.S., Chen, C., Volkow, N.D., Clarkson, F.A., Porter, S., Straughter-Moore, R.M., Alexoff, D.L., Tedeschi, D., Russo, N.B., Fowler, J.S., Brodie, J.D., 1997. GABAergic attenuation of cocaine-induced dopamine release and locomoter activity. Synapse 25, 393–398. Diggle, P.J., Liang, K.Y., Zeger, S.L., 1994. Analysis of Longitudinal Data. Oxford University Press, New York. Dworkin, S.I., Smith, J.E., 1988. Neurobehavioral pharmacology of cocaine. In: Clouet, D., Ashgar, K., Brown, R. (Eds.), Mechanisms of Cocaine Abuse and Toxicity. National Institute on Drug Abuse Research Monograph #88. US Government Printing Office, Washington, DC, pp. 185–198. Gerasimov, M.R., Ashby, C.R., Gardner, E.L., Mills, M.J., Brodie, J.D., Dewey, S.L., 1999. Gamma-vinyl-GABA inhibits methamphetamine, heroin, or ethanol-induced increases in nucleus accumbens dopamine. Synapse 34, 11–19. Goeders, N.E., Dworkin, S.I., Smith, J.E., 1986. Neuropharmacological assessment of cocaine self-administration into the medial prefrontal cortex. Pharmacol. Biochem. Behav. 24, 1429–1440. Guy, W. (Ed.), 1976. Assessment Manual for Psychopharmacology: Publication ADM 76-338. US Department of Health Education and Welfare, Washington, DC. Johnson, B.A., Ait-Daoud, N., Bowden, C.L., DiClemente, C.C., Roache, J.D., Lawson, K., Javors, M.A., Ma, J.Z., 2003. Oral topiramate for the treatment of alcohol dependence: a randomized controlled trial. Lancet 361, 1677–1685. Kadden, R., Carroll, K., Donovan, D., Cooney, N., Monti, P., Abrams, D., Litt, M., Hester, R., 1995. Cognitive-Behavioral Coping Skills Therapy Manual: A Clinical Research Guide for Therapists Treating Individuals With Alcohol abuse and Dependence. National Institutes of Health, Rockville, MD. Kampman, K.M., Alterman, A.I., Volpicelli, J.R., Maany, I., Muller, E.S., Luce, D.D., Mulholland, E.M., Jawad, A., Parikh, G., Mulvaney, F.D., Weinieb, R.M., O’Brien, C.P., 2001a. Cocaine withdrawal symptoms and initial urine toxicology results predict treatment attrition in outpatient cocaine dependence treatment. Psychol. Addict. Behav. 15, 52–59. Kampman, K.M., Volpicelli, J.R., Alterman, A.I., Cornish, J.W., O’Brien, C.P., 2000. Amantadine in the treatment of cocaine dependent patients with severe cocaine withdrawal symptoms. Am. J. Psychiatry 157, 2052–2054. Kampman, K.M., Volpicelli, J.R., McGinnis, D.E., Alterman, A.I., Weinrieb, R.M., D’Angelo, L., Epperson, L., 1998. Reliability and valid-
239
ity of the Cocaine Selective Severity Assessment. Addict. Behav. 23, 449–461. Kampman, K.M., Volpicelli, J.R., Mulvaney, F.D., Alterman, A.I., Cornish, J., Gariti, P., Cnaan, A., Poole, S., Muller, E., Acosta, T., Luce, D., O’Brien, C.P., 2001b. Effectiveness of propranolol for cocaine dependence may depend on cocaine withdrawal symptom severity. Drug Alcohol Depend. 63, 69–78. Kampman, K.M., Volpicelli, J.R., Mulvaney, F.D., Rukstalis, M., Alterman, A.I., Pettinati, H., Weinieb, R.M., O’Brien, C.P., 2002. Cocaine withdrawal severity and urine toxicology results from treatment entry predict outcome in medication trials for cocaine dependence. Addict. Behav. 27, 251–260. Koob, G., Vaccarino, F., Amalric, M., Bloom, F., 1987. Positive reinforcement properties of drugs: search for neural substrates. In: Engel, J., Oreland, L. (Eds.), Brain Reward Systems and Abuse. Raven Press, New York, pp. 35–50. Koob, G.F., 1992. Neural mechanisms of drug reinforcement. Ann. N.Y. Acad. Sci. 654, 171–191. Kosten, T., Gawin, F.H., Kosten, T.A., Morgan, C., Rounsaville, B.J., Schottenfeld, R., Kleber, H.D., 1992. Six-month follow-up of short-term pharmacotherapy for cocaine dependence. Am. J. Addict. 1, 40–49. Kushner, S.A., Dewey, S.L., Kornetsky, C., 1999. The irreversible ␥-aminobutyric acid (GABA) transaminase inhibitor ␥-vinyl-GABA blocks cocaine self-administration in rats. J. Pharmacol. Exp. Ther. 290, 797–802. Kuzniecky, R., Hetherington, H., Ho, S., Pan, J., Martin, R., Gilliam, F., Hugg, J., Faught, E., 1998. Topiramate increases cerebral GABA in healthy humans. Neurology 51, 627–629. Littell, R.C., Milliken, G.A., Stroup, W.W., Wolfinger, R.D., 1996. SAS System for Mixed Models. SAS Institute Inc, Cary, NC. Littell, R.C., Pendergast, J., Natarajan, R., 2000. Modeling covariance structure in the analysis of repeated measures data. Stat. Med. 19, 1793–1819. McFarland, K., Kalivas, P.W., 2001. The circuitry mediating cocaine-induced reinstatement of drug-seeking behavior. J. Neurosci. 21, 8655–8663. McFarland, K., Lapish, C.C., Kalivas, P.W., 2003. Prefrontal glutamate release into the core of the nucleus accumbens mediates cocaine-induced reinstatement of drug seeking behavior. J. Neurosci. 23, 3531–3537. McLellan, A.T., Kushner, H., Metzger, D., Peters, R., Smith, I., Grissom, G., Pettinati, H., Argeriou, M., 1992. The fifth edition of the addiction severity index. J. Subst. Abuse Treat. 9, 199–213. Mulvaney, F.D., Alterman, A.I., Boardman, C.R., Kampman, K., 1999. Cocaine abstinence symptomatology and treatment attrition. J. Subst. Abuse Treat. 16, 129–135. Petroff, O.A., Hyder, F., Mattson, R.H., Rothman, D.L., 1999. Topiramate increases brain GABA, homocarnasine, and pyrrolidinone in patients with epilepsy. Neurology 52, 473–478. Privetera, M.D., 1997. Topiramate: a new antiepileptic drug. Ann. Pharmacother. 31, 1164–1173. Roberts, D.C.S., Koob, G.F., Klonoff, P., Figiber, H.C., 1980. Extinction and recovery of cocaine self-administration following 6-hydroxydopamine lesions of the nucleus accumbens. Pharmacol. Biochem. Behav. 12, 781–787. Roberts, D.C., Andrews, M.M., Vickers, G.J., 1996. Baclofen attenuates the reinforcing effects of cocaine in rats. Neuropsychopharmacology 15, 417–423. Rubio, G., Ponce, G., Jimenez-Arriero, M.A., Palomo, T., Ferre, F., 2002. Effectiveness of topiramate in control of alcohol craving. Eur. Neuropsychopharmacol. 12, S63. Schultz, W., 1998. Predictive reward signals of dopamine neurons. J. Neurophysiol. 80, 1–27. Severt, L., Coulter, D.S., Sombati, S., 1995. Topiramate selectively blocks kainate currents in cultured hippocampal neurons. Epilepsia 36, 38.
240
K.M. Kampman et al. / Drug and Alcohol Dependence 75 (2004) 233–240
Sharples, K., Breslow, N., 1992. Regression analysis of correlated binary data: some small sample results for the estimating equation approach. J. Stat. Comput. Simul. 42, 1–20. Somoza, E., Dyrenforth, S., Goldsmith, J., Mezinskis, J., Cohen, M., 1995. In search of a universal drug craving scale. Paper Presented at the Annual Meeting of the American Psychiatric Association, Miami, FL. Weiss, F., Maldanado-Vlaar, C.S., Parsons, L.H., Kerr, T.M., Smith, D.L., Ben-Shahar, O., 2000. Control of cocaine-seeking behavior by drug-associated stimuli in rats: effects on recovery of extinguished operant responding and extracellular dopamine levels in amygdala and nucleus accumbens. Proc. Natl. Acad. Sci. U.S.A. 97, 4321–4326.
Westerink, B.H.C., Enrico, P., Feiman, J., De Vries, J.B., 1998. The pharmacology of mesocortical dopamine neurons: a dual probe microdialysis study of the ventral tegmental area and prefrontal cortex of the rat brain. J. Pharmacol. Exp. Ther. 285, 143–154. White, H.S., Brown, S.D., Woodhead, J.H., Skeen, G.A., Wolof, H.H., 1997. Topiramate enhances GABA-mediated chloride flux and GABA-evoked chloride currents in murine brain neurons and increases seizure threshold. Epilepsy Res. 28, 167–179. Zullino, D.F., Cottier, A.C., Besson, J., 2002. Topiramate in opiate withdrawal. Prog. Neuropsychopharmacol. Biol. Psychiatry 26, 1221– 1223.