Results of a baseline urine test predict levels of cocaine use during treatment.

Sixty-one cocaine dependent outpatients submitted a single urine sample at least 1 week prior to entry into a 4-week treatment study. Participants were then expected to provide three urine samples per week during the month of treatment. The 61 patients studied here all completed treatment and provided an average of more than 11 of 12 scheduled urine samples. Participants who submitted a cocaine-positive sample prior to treatment provided more positive urine samples during the 4-week trial, were less likely to be completely abstinent during the month, and took longer to reach an initial abstinence criterion of three consecutive cocaine-free urines. Thus, a single pretreatment urine test represents a powerful predictor of subsequent cocaine use. The results suggest that future randomized trials stratify group assignment based on the results of a baseline urine test.

A randomized, double-blind, placebo-controlled study of ritanserin pharmacotherapy for cocaine dependence.

Eighty cocaine-dependent individuals enrolled in outpatient treatment took part in a randomized, double-blind, placebo-controlled trial of ritanserin, a 5-HT(2) antagonist, as an adjunct therapy. Participants attended an outpatient day hospital therapy program each day and received tablets containing placebo or 10 mg ritanserin for a 4-week period. Primary outcome measures included retention in treatment, urine drug tests, and self-reports of craving. Secondary outcome measures were depression scores on the Beck and Hamilton inventories, negative mood as measured by the Profile of Mood States, and life functioning as measured by the Addiction Severity Index. Although participants showed improvement over the 4 weeks, there were no group differences on any of the measures. These results fail to support the use of ritanserin as a complement to outpatient psychosocial therapy for cocaine dependence.

Three decision-making tasks in cocaine-dependent patients: do they measure the same construct?

AIMS: Substance-abusing populations perform poorly on decision-making tasks related to delay and risk. These tasks include: (1) the Delay Discounting Procedure (DDP), in which choices are made between smaller-sooner and later-larger rewards, (2) the Gambling Task (GT), in which choices are made between alternatives varying in pay-off and punishment, and (3) the Rogers Decision-Making Task (RDMT) in which subjects choose between higher or lower probability gambles. We examine the interrelationship among these tasks. DESIGN: A test battery was created which included the DDP, GT and RDMT, as well as measures of impulsivity, intellectual functioning and drug use. SETTING: Subjects completed the test battery at an outpatient center, prior to beginning 12 weeks of treatment. PARTICIPANTS: Thirty-two treatment-seeking cocaine dependent individuals (primarily African-American males) participated. FINDINGS: Performance on the GT was significantly correlated with performance on the DDP (r = 0.37; p = 0.04). Reaction times on the RDMT correlated with performance on the GT (r = 0.36, p = 0.04) and DDP (r = 0.33, p = 0.07), but actual choices on the RDMT did not (p > 0.9 for both). While no significant relationships were observed between task performance and impulsivity, IQ estimate was positively correlated with both the GT (r = 0.44, p = 0.01) and RDMT (r = 0.41, p = 0.021). Split half reliability data indicated higher reliability when using only data from the latter half of the GT (r = 0.92 vs. r = 0.80). CONCLUSIONS: These data offer preliminary evidence of overlap in the decision-making functioning tapped by these tasks. Possible implications for drug-taking behavior are discussed.

Mood state and recent cocaine use are not associated with levels of cocaine cue reactivity.

Eighty-one cocaine-dependent outpatients were assessed for their reactions to cocaine-related cues in a laboratory setting. All subjects contributed a urine sample prior to the session. Compared with non-drug control cues, the cocaine stimuli produced increases in physiological arousal, self-reports of high, craving, and withdrawal, and self-reports of negative mood. Subjects who tested cocaine-positive on the day of testing differed only in skin resistance responding from those who tested cocaine-negative. Changes in cue-induced physiological and self-report measures were also not associated with between-subject variations in mood as measured by the Profile of Mood States (POMS) questionnaire administered prior to cue assessment. Thus, variations in baseline mood and recent cocaine use history do not introduce an additional source of variability in cue reactivity measurements. However, negative mood states at the start of a session were associated with higher levels of self-reported craving, high, and withdrawal both before and after cue exposure.

Open trials as a method of prioritizing medications for inclusion in controlled trials for cocaine dependence.

This paper describes a rapid and systematic method of using open trials to identify medications that may be useful for the treatment of cocaine dependence. Results of these open trials can be used to prioritize medications for inclusion in subsequent double-blind, placebo-controlled trials. Preliminary results are presented from the evaluation of propranolol, nefazodone, and the combination of phentermine and fenfluramine (phen/fen). Each medication was evaluated in an open trial, and results were compared to results obtained from a group that received a multivitamin. Outcome measures included treatment retention, urine toxicology screens, self-reported cocaine use, and changes on the Addiction Severity Index (ASI). Treatment retention was significantly better in the propranolol group than in the multivitamin group. Concurrent alcohol abuse was associated with increased rates of attrition in the multivitamin group, and the phen/fen group, but not in the propranolol group. Neither the nefazodone nor the phen/fen groups showed any outcome advantages over the multivitamin group. We conclude that propranolol may enhance retention among cocaine-dependent patients, especially among those who also abuse alcohol. These results encourage a double-blind, placebo-controlled trial of propranolol.

Comparing levels of cocaine cue reactivity in male and female outpatients.

Thirty-eight female and 26 male cocaine-dependent outpatients were exposed to cocaine cues in a laboratory setting. Stimuli consisted of an audiotape of patients discussing cocaine use, a videotape of simulated cocaine preparation and use, and the handling of cocaine paraphernalia. Overall, the stimuli produced significant decreases in skin temperature and skin resistance, and significant increases in heart rate, self-reported drug states (high, craving, and withdrawal), and self-reported negative moods. Females were more likely to report increased craving in response to the cues than males, but there were no other gender differences in any of the responses. Levels of reactivity in females were comparable to the results of previous studies with all male samples. These results support the use of a constant set of cues in future treatment studies employing gender-balanced patient samples.

Reliability and validity of the aggregate method of determining number of cigarettes smoked per day.

The authors evaluated the reliability of two pretreatment assessments (screening and intake) of cigarettes smoked per day (CPD) by the commonly used aggregate method. The validity of the aggregate method was also determined by comparison with results of the timeline followback (TLFB) method for the identical periods. The study participants were 49 outpatients undergoing nicotine patch treatment. The reliability of the two aggregate method evaluations of CPD was quite high by Pearson product-moment correlation (r) and good when based on the intraclass correlation. Correspondence between the CPD assessments based on the aggregate and TLFB methods for the two time-points ranged from fair (screening) to good (intake). Overall, the study findings indicate that the aggregate method provides reasonably consistent data.

Laboratory exposure to cocaine cues does not increase cocaine use by outpatient subjects.

Sixty-nine cocaine-dependent outpatients were exposed to cocaine-related stimuli and to non-drug events on separate days. Cocaine cue sessions were always followed by a meeting with a trained clinician designed to eliminate any craving that remained following cue presentations. Urine samples were collected before each laboratory session and 1 to 3 days later. Neither rates of cocaine use nor average urine metabolite values differed following the two sessions. Nearly 90% of subjects had the same urine test result both before and after the cocaine cue session. Thus, laboratory presentation of cocaine cues to outpatient subjects did not increase their risk of subsequent drug-taking. These results suggest that with proper clinical protections, cue exposure can be used as a treatment outcome measure and a behavioral intervention in outpatient settings without increasing the risk of drug use.

Cocaine use is associated with increased craving in outpatient cocaine abusers.

Sixty-one cocaine abuse patients provided self-reports of craving and urine samples 3 times a week. Within-subject analyses revealed several relationships between the measures. First, peak craving levels were higher for 2-3-day intervals during which cocaine use had occurred than for preceding or following abstinent intervals. Second, average craving ratings during cocaine use intervals were double the ratings given during abstinent intervals. Third, cocaine use was 4 times more likely to occur during a period of elevated craving than during comparison intervals. Finally, participants who provided at least 1 positive urine sample reported more craving increases over 4 weeks than did abstinent individuals. These results demonstrate a strong association between craving increases and naturally occurring cocaine use but do not allow a determination of whether craving caused cocaine use or cocaine use caused craving.

Conditioning factors in drug abuse: can they explain compulsion?

There is a good deal of clinical evidence suggesting that compulsion to resume drug taking is an important part of the addiction syndrome. The symptoms comprising motivation to resume drug use, namely craving and compulsion, have been studied experimentally in human subjects. While much work remains to be done, there is evidence showing that these symptoms are influenced by learning. The research has been guided by animal studies demonstrating that drug effects can be conditioned. Much attention has been directed toward demonstrating the existence of drug conditioning in human addicts and exploring the neurological structures that may underlie such learned responses. We do not yet know the relative importance of learning in the overall phenomenon of relapse, and treatments based on conditioning principles are still under investigation.

Comparing self-reported cocaine use with repeated urine tests in outpatient cocaine abusers.

Sixty-one participants in outpatient therapy for cocaine dependence provided urine samples and self-reports of cocaine use 3 times a week for 4 weeks. Participants later gave a retrospective self-report of cocaine use for the month on the addiction Severity Index (ASI). Comparisons with urine test values revealed substantial underreporting of cocaine use on both measures, and results from the 2 forms of self-report were only imperfectly correlated. More participants admitted to at least 1 cocaine use episode on the ASI than on the repeated self-reports, but the repeated reports provided a more accurate index of the relative frequency of cocaine use during the month. Self-reports can enhance cocaine use detection when urines are infrequently collected and can help determine whether consecutive positive urine samples represent elevated metabolite levels from a single drug use episode. However, self-reports cannot substitute for regular urine sampling.

Relationships among physiological and self-report responses produced by cocaine-related cues.

In response to cocaine cues, 150 subjects with a history of cocaine abuse showed decreases in skin temperature and skin resistance and increases in heart rate and reported craving, high, and withdrawal responses. These responses were consistent across four years of data collection. Craving reports were not consistently associated with either high or withdrawal responses, and many subjects endorsed increases in both high and withdrawal states. Correlations revealed no pattern of association among physiological variables and responding did not differ between subjects who did and those who did not report increases in each of the drug states. Finally, physiological variables did not predict reported drug states in discriminant analyses. Cocaine cue reactivity cannot be easily related to a unitary state of high, withdrawal, or craving. It is suggested that future studies focus more on the prediction and measurement of treatment outcome than on the form of cue responses.

Failure of ritanserin to block cocaine cue reactivity in humans.

As part of a double-blind placebo-controlled study of the effects of ritanserin on cocaine use and craving, reactivity to cocaine-related events was assessed both before and during medication. Twenty-two patients receiving ritanserin and 23 receiving placebo were exposed to cocaine cues while continuous measures of heart rate, skin temperature, and skin resistance were taken. Self-reports of high, withdrawal, and craving were also collected. The cues produced significant physiological responding as well as significant increases in high and craving during both sessions. Ritanserin reduced cue-elicited decreases in skin temperature, but had no effect on heart rate and skin resistance or on cue-induced high and craving. The results demonstrate that cue reactivity is a robust phenomenon across two assessment sessions but fail to support the use of ritanserin as a means of reducing cue-elicited drug states.

Effect of smoking cues and cigarette availability on craving and smoking behavior.

This study examined whether smokers respond differently to smoking cues than to affectively neutral or unpleasant cues without smoking content, and whether reactivity is affected by expectations regarding the opportunity to smoke. Expectancy was manipulated by telling subjects in group SMOKE that they could smoke, and subjects in group NO SMOKE that they could not smoke following each cue-reactivity session. The dependent variables were subjective ratings of "desire to smoke", "high", "withdrawal", and mood, as well as latency to initiate smoking measured in group SMOKE. Statistical analyses demonstrated that only group SMOKE (a) reported greater "desire to smoke" and "withdrawal" to the smoking cues compared to the baseline, (b) reported greater "desire to smoke" to the smoking cues than to the unpleasant or to the neutral cues, and (c) smoked faster after the smoking cues than after the neutral cues. Both groups rated the unpleasant cues as affectively more negative than the neutral cues. The data suggest that the impact of drug-cues on craving and subsequent drug-use is due to their drug-related content, and not to negative affect. Furthermore, the impact of drug-related cues appears to be influenced by perceived drug-availability.

Reliability and validity of 6-month timeline reports of cocaine and heroin use in a methadone population.

Fifty-nine persons addicted to heroin and maintained by methadone reported on daily heroin and cocaine use during 2 timeline calendar interviews administered 6 weeks apart. Retrospective reports covering 6 months were compared with urine samples taken weekly during the interval. Test-retest correlations were high and timeline estimates of drug use frequency were significantly correlated with the frequency of drug-positive urine results. Thus, timeline reports of drug-use frequency appeared both reliable and valid. Individual participants either over- or under-reported by an average of about 15%, and they did not identify instances of drug use with greater than chance accuracy when particular episodes of drug use occurred. These results support the use of timeline reports to make group comparisons of long-term drug use, but suggest that timeline data should not be used to identify specific drug-use episodes. Work with other drug-use population is necessary to extend these conclusions.

Can induced moods trigger drug-related responses in opiate abuse patients?

This study investigated the ability of four hypnotically induced mood states (euphoria, depression, anxiety, and anger) to trigger craving and other drug-related conditioned responses in detoxified opiate abuse patients. Hypnotically induced depression produced significant increases in drug craving for opiates. Depression also tended to increase global self-ratings of opiate withdrawal. Other trends included increases in self-rated craving by induced anxiety and increases in withdrawal symptoms by induced anger. These results suggest that negative mood states, perhaps in the context of repeated attempts at self-medication, may become conditioned stimuli capable of triggering craving and other drug-related conditioned responses. The ability of depression to produce reliable effects in this particular patient group may reflect the high lifetime prevalence of depression diagnoses for this sample. The implications of these findings for therapeutic strategies are discussed.

Developing treatments that address classical conditioning.

Repetitive use of psychoactive drugs produces a variety of learned behaviors. These can be classified in the laboratory according to an operant/classical paradigm, but in vivo the two types of learning overlap. The classical CRs produced by drugs are complex and bidirectional. There has been progress in classifying and predicting the types of CRs, but little is known of mechanisms. New techniques for understanding brain function, such as microdialysis probes in animals and advanced imaging techniques (positron emission tomography and single photon emission computerized tomography) in human subjects, may be utilized in conditioning paradigms to "open the black box." Because the existence of CRs in drug users is now well established, clinical studies have been instituted to determine whether modification of CRs can influence clinical outcome. A recently completed study in cocaine addicts has produced evidence that outcome can be improved by a passive extinction technique over an 8-week outpatient treatment program.

Cue reactivity and cue reactivity interventions in drug dependence.

Despite a venerable history dating back to Pavlov and countless testimonials from patients such as those in the opening paragraphs of this chapter, there is much that remains to be learned about drug signals and, particularly, about ways of reducing their adverse effects on human drug users. There is a substantial amount of data showing increased craving and signs of physiological arousal to drug-related versus neutral cues in drug users for both drug classes reviewed here. Additional controlled studies will be useful in refining which responses among those studied are, in fact, conditioned in origin and therefore can be subjected reasonably to learning-based interventions. Most attempts to modify cue responsivity for clinical benefit have met with only modest success, and there is ample room for creative, but controlled, treatment-outcome studies. In recent years, several other groups have joined in the effort to understand drug-related cue reactivity, extending the research area to alcohol and nicotine (Monti et al. 1987; Niaura et al. 1988, 1989; Cooney et al. 1984; Hodgson and Rankin 1982; Drummond 1990; Laberg 1990). The interested reader is referred to several additional reviews of cue reactivity and cue exposure research related to alcohol and nicotine (Niaura 1988; Drummond 1990; Laberg 1990), opiates (Powell 1990), opiates and cocaine (Childress et al. 1988b; O'Brien et al. 1990), and all the preceding areas (Rohsenow et al. 1991).

Using cue reactivity to screen medications for cocaine abuse: a test of amantadine hydrochloride.

The use of responding to drug-related stimuli as a dependent measure for studies of anticraving medications was assessed. Cocaine-dependent subjects receiving either amantadine hydrochloride, a putative anticraving agent, or placebo were exposed to drug-related cues prior to and 7 days after the initiation of the medication. Measurements of heart rate, skin resistance, skin temperature, and self-reported craving were taken during each stimulus session. Amantadine increased physiological reactivity to the drug-related cues compared to the placebo while having no effect on craving. Although the results discourage the use of amantadine as an anticraving medication, they do suggest that responses elicited by drug-related stimuli provide a valuable set of dependent measures for use in future medication trials of anticraving agents.