Recovery from addiction requires motivation to change, understanding of the process of addiction, and the ability to form new, stable patterns of behavior. Cognitive behavior therapy is currently a mainstay of treatment for cocaine dependence. Its aim is to change behavior and affective responses by teaching coping skills and by addressing and modifying dysfunctional thought patterns. Through learning and practice, a patient can gain increasing control over urges and behaviors that were previously considered to be unmanageable and immutable. In most substance abuse treatment programs, counseling also includes disease-model education, anger management, and motivational enhancement in both individual and group settings to provide alternative responses when an addict is faced with unmanageable feelings, urges, or circumstances.
The capacity to respond to such "psychosocial" intervention is largely dependent on a patient’s cognitive flexibility. Intellect, in the conventional sense, varies as a function of education and socioeconomic background. Educational history is relevant in cognitive therapy only as regards intellect in the intrinsic sense: the ability to learn and reason and abstract. The case reports presented here were chosen to illustrate the premise that intellect, in this less colloquial sense, changes over the period of early recovery from heavy drug exposure. We propose that this change in complex cognitive function is multidetermined, a product of effects of drug abuse at many levels of regulation of cerebral function, but focus on cerebral perfusion in these two cases.
Cocaine induces constriction of coronary and cerebral vessels in both humans and in animal models (1, 2). Over the last 10 years, several groups of investigators using positron emission tomography and single photon emission computed tomography to measure brain perfusion have found that chronic cocaine use (usually in the context of polysubstance abuse) is associated with multiple focal decreases in cerebral blood flow (CBF) (3, 4). Magnetic resonance angiography has provided the first direct demonstration of acute, cocaine-induced cerebral vasospasm in humans (5). Previous investigations have demonstrated both deficits in neuropsychological performance and abnormalities in brain perfusion or metabolism in chronic cocaine abusers and that both of these deficits can improve during abstinence (4, 6, 7). Although prior studies have indicated that the severity of these flow deficits is related to the degree and duration of drug abuse, their pathophysiology is unknown, and their consequences on brain function are unclear.
One corollary of these findings is that the degree of change in CBF during abstinence correlates with the degree of abnormality at the beginning of treatment. F1 presents the baseline image of the patient in case 2, Ms. B, in relation to that of a typical comparison subject. These images have been scaled and color-coded to show the same relative activity. Note that in addition to several cortical regions with markedly decreased perfusion, e.g., orbital and dorsal prefrontal, parieto-occipital and frontoparietal, CBF in the patient’s cortex is generally lower than that of a comparison subject. We have hypothesized that a measure of the change from baseline perfusion will correlate with a measure of the capacity to learn new behavior.
The cases were chosen to highlight the role of cognitive function and its relation to cerebral perfusion in recovery from cocaine abuse and to demonstrate an association between the response to behavioral therapy and improvement in cerebral perfusion. Both patients were in drug-free residential care during the two neuroimaging and neuropsychological testing sessions. We suggest that this correlation provides a hypothesis for further clinical study of the relationship between cerebral perfusion and the response to cognitive behavior therapy in recovery from cocaine dependence.
Ms. A, a 34-year-old woman, entered treatment 4 days after her last use of cocaine. She had a 19-year history of cocaine abuse: 10 years of intranasal use and 9 years of crack cocaine use. During the 3 months before her admission, she reported using varying amounts (0.5–1.5 g) of cocaine about 3 days per week. Her recent use of alcohol had been limited to about 1 pint of liquor on each day of cocaine use. She reported two previous substance abuse treatments: the first led to no sustained abstinence, and the second she did not complete.
Cognitive behavior therapy
Ms. A placed herself in the early-late stage of recovery on the basis of the Jellinek chart, a self-assessment instrument for staging the progress of addiction (8). From the outset, Ms. A was forthcoming with self-disclosure and receptive of feedback in group therapy, making excellent use of the group in learning how to handle stress. She had no difficulty identifying the consequences of her use, the most obvious external example being her inability to parent her two children.
During her second week in treatment, she reported a sporadic craving for cocaine and recognized that these cravings were playing a significant role in her perception of recovery and treatment. This was evidence of a fairly sophisticated ability to draw inferences. However, although she acknowledged the fallacy of her previous perceptions of drug use—"It will make things better"—she was less aware of the degree of denial that occurred in response to her craving for cocaine, manifested in statements such as, "I can control it."
During her third week in treatment, she remained active in group therapy and was often viewed by peers as a leader within the group. This contrasted significantly with her perception of herself as one consumed with insecurity; the feedback she received highlighted one of the greatest barriers to maintaining abstinence: her low self-esteem. She was able to discuss how her chronic feelings of shame and guilt were exacerbated by, but not entirely caused by, her addiction. After reflecting on her relationships, she agreed to a family therapy session and then chose to end the self-described "enabling" relationship she had with her boyfriend.
As noted, her cognitive state was impressive at admission and remained so throughout her treatment. She demonstrated the ability to draw inferences, to challenge unproductive thoughts, to recognize cravings and strategize responses to them, and to retain information. There was clearly a commitment on Ms. A’s part to assume responsibility for her recovery by changing thoughts, behaviors, and relationship dynamics. Although some of this profile may reflect familiarity with treatment, she clearly benefited from a high level of motivation combined with the capacity to integrate new concepts and information and readiness to change. At discharge, Ms. A had been clean and sober for 36 days. She was referred for ongoing treatment and entered a halfway house at discharge.
This right-handed woman had 13 years of formal education. She performed in the average range on most tests but did show mild impairment in tests that measured simple auditory attention, complex auditory processing, and simple motor repetition. On retesting, her performance fell in the average range on all measures. On her last examination day, Ms. A performed in the average range on most measures but scored above average on one motor test and a complex auditory processing task (the Paced Auditory Serial Addition Task) and below average on a verbal learning and memory task (the Auditory Verbal Learning Test). On this last task, mild impairment was noted both in learning and recall after both long and short delays, indicating lack of concentration or effort or both.
Overall, Ms. A demonstrated mild gains in performance over a 28-day study period; these were most likely due to the effects of practice. The results of the last day of testing show that although she showed significant improvement on some tasks, her effort was inconsistent during that session.
Statistically significant increases in CBF as determined by the use of statistical parametric mapping from baseline to 2 weeks for Ms. A are shown in F2(9, 10). The total number of regions that improved was not greater than that obtained by chance (i.e., the set-level inference was not significant); however, some individual brain regions showed significantly increased perfusion in both degree and spatial extent. The largest such cluster included the majority of the right temporal lobe. The other significant areas of improvement included the corresponding contralateral region, the left temporal pole, and the frontal opercular cortex bilaterally.
It is interesting to note that Ms. A’s performance on cognitive testing would predict changes in cerebral function in areas related to auditory and language processing. The psychometric test results showed some impairment in measured simple auditory attention and complex auditory processing at baseline, which had normalized by the time of the second test session administration and which was also the time of the second scan, analysis of which is presented here. This convergence of outcome data provides support for the correlation of biological and psychological processes.
Ms. B, a 32-year-old woman, entered treatment 6 days after her last use of cocaine. She had a 12-year history of cocaine use: 6 years of intranasal cocaine use and 6 years of crack cocaine use. During the 3 months before her admission, she reported using 4–6 g of cocaine per week. Despite a past history of alcohol abuse at ages 18–20, her use of alcohol in subsequent years was described as infrequent. Although she reported an extensive treatment history, her longest period of abstinence was estimated at "a few months."
Cognitive behavior therapy
Ms. B also placed herself in the early-late stage on a Jellinek chart of stages of addiction. For Ms. B, participation during group therapy was limited, despite peer and staff encouragement to involve herself. Initially, she recognized only the obvious consequences of her abuse: legal trouble resulting from a possession charge and the subsequent involvement of the Department of Children and Families. Her denial of the scope and severity of her addiction was at least partly related to an apparent inability to draw inferences, a relatively sophisticated cognitive process. Her first treatment plan was designed to address this deficiency.
During her second week in treatment, her participation during group therapy improved significantly. She identified with more clarity the progression of her drug use and the parallel progression of its consequences. She described what she would gain—control of her life, resumption of parental duties, employment, better health, all of which would improve her overall self-esteem and lose—nothing—if she eliminated drugs from her life. In response to this new recognition of cause and effect relationships, the next period of treatment focused on her underlying perceptions about drug use and the relation of those perceptions to her continued motivation to use. At this point she developed her second treatment plan, focusing on relapse-prevention planning.
During her third week in treatment she remained active in group therapy, although she frequently needed prompting to reveal more about herself and her progress in recovery. She identified her tendency to keep problematic issues to herself so as not to risk relapse. Her hesitancy to share information was not due to an inability to understand the nature of therapy for substance dependence. Indeed, she was able to identify relapse warning signs and effective coping strategies for them, but her motivation to address them proactively was inconsistent and, at times, nonexistent. She reported considerable past difficulty in sustaining motivation for recovery, due to, as she saw it, a nonsupportive family. However, she had never utilized self-help techniques or another support system to her advantage. In the hope of reinforcing motivation, therapy again focused on reasons to remain clean. Although she appeared genuinely "tired" of her life as it has been, her passive and at times apathetic attitude toward abstinence were significant barriers to her recovery. She declined an opportunity for family therapy, stating she had "heard it all before."
In order to overcome her passivity, further treatment addressed her patterns of thinking. This revealed that most of her thoughts reflected low self-esteem and low self-expectations. She was able to acknowledge that as long as she remained an active substance abuser, she would be unable to regain control of her life and realize her goals, such as custody of her child. She was encouraged to challenge her automatic, self-defeating thoughts and to communicate honestly with peers in treatment and self-help. Utilizing peer relationships in this way may have been complicated by the fact that she was the only woman in her treatment group. However, she was encouraged to make use of her clinician and other clinicians in the program who were female when this appeared to present an obstacle to her.
At discharge, Ms. B had been clean and sober for 30 days. As noted, her cognitive state had improved significantly, as evidenced by her ability to draw inferences, challenge unproductive thoughts, identify and strategize for high-risk situations and relapse warning signs and by her ability to retain information. Despite all that she gained cognitively, her ability to assume responsibility for her recovery by proactively changing thoughts, behaviors, and relationship dynamics was questionable. She was referred to another partial day program at discharge.
This 32-year-old, right-handed woman had 12 years of formal education. In the first test session, she performed in the average range on measures of simple auditory attention and concentration, visuomotor integration, verbal fluency, and learning. On tests that demanded a strong motor effort, her performance was slow (moderate or moderately severe impairment). In addition, her performance in retrieving verbal information after a long delay and on a complex working memory task were significantly below average (mild to moderate impairment). On her second testing session, she maintained her performance on most tests, yet demonstrated improvement on most motor coordination measures and on one task that requires visual scanning (digit vigilance). Her scores on indices of learning and memory were generally lower on repeat testing at 2 weeks than in her first testing session, consistent with the results of other research in cocaine-dependent individuals, suggesting persistence of declarative memory problems over other cognitive deficits (1). At her last examination, Ms. B performed in the average range on most measures. Exceptions to this were similar to the ones found in session 2: relative impairments were noted on both delayed memory and on complex mental control tasks. Further gains from the second session were noted on a simple motor repetition (finger oscillation) task.
Overall, Ms. B demonstrated mild to moderate gains in performance over the 28-day period, but especially on movement-related tasks. Some of the improvements demonstrated by Ms. B were undoubtedly due to the effects of practice. However, she did not exhibit consistent gains across trials (as one would expect with similar effort); therefore motivation must be considered a mitigating factor. This possibility was more apparent for tasks that demand greater effort.
Increases in CBF from baseline to 2 weeks for Ms. B are shown in F3. The number of regions of change, the set-level inference, is significant at the 5% error level. Several individual regions are also significant in degree of change in perfusion and in spatial extent: left and right occipital poles (including some visual association cortex), and a region of the right dorsolateral prefrontal cortex.
When we compare the results of previous cognitive and motor tests with areas of change in perfusion, we note a correspondence of these regions with the improvement noted in visual scanning tasks and hand-eye coordination. Although the perfusion imaging shows an increase in perfusion in the dorsolateral prefrontal cortex, the performance on memory tasks was worse on repeat testing; this divergence may reflect a greater demand on a brain region involved in performing working memory tasks. In addition to these specific regional findings, it is the overall (set-level) significance that we emphasize here. This is an index of the proportion of all cortical tissue with greater perfusion over this test period and provides evidence in this second case that the effect of abstinence on cerebral perfusion was distributed across the entire brain.
These two patients were matched on many factors that might affect their outcome in treatment (loss of child custody, low self-esteem, and unemployment), yet their overall performance in treatment was notably different. Ms. A was better prepared, psychologically, for taking control of her behavior, and Ms. A benefited to a significant degree from the education offered in treatment. The second patient, Ms. B, began treatment far less prepared to make any meaningful progress in her recovery, despite assessing herself as at the same overall stage in that process as Ms. A. However, Ms. B made significant advances in her level of insight and ability to make inferences about her behavior in relation to her thoughts and reactions. We hypothesize that this difference in cognitive flexibility is, in part, a reflection of the state of cerebral function during the period of treatment and that cerebral perfusion can provide a useful measure of this state.
Neuropsychological assessment in both cases was notable for variability among tests, consistent with previous studies in this population (11). Both patients showed below-normal performance on some tests, and most of these test scores improved with time and practice. Although the extent of overall change in perfusion was only significant in Ms. B, in both cases specific brain regions showing greater perfusion were correlated with expected measures of improvement on neuropsychological assessment.
Ms. A was able to utilize the tools presented in cognitive behavior therapy to make constructive changes in her self-image and behavior. At discharge, she had reshaped her behavior and thinking patterns in important ways by applying the principles presented in treatment. Direct cognitive testing generally revealed performance in the normal range, with changes detected in a fairly circumscribed area of function. Similarly, testing for changes in perfusion revealed only limited areas of improvement, whereas the overall pattern and degree of change was not significant. From these findings we conclude that only mild deficits in brain function were present in Ms. A at the outset of treatment.
In contrast, the second patient, Ms. B, clearly had much greater difficulties with a program of honest self-assessment and open communication, requiring far more effort to recognize the consequences of her addiction. However, in the course of treatment she made significant progress in adopting a new model of the relationship between her mood, thoughts, and behavior. Although direct cognitive testing primarily detected abnormalities in motor function, examination of cerebral perfusion detected significant overall change as well as specific brain regions of change. We propose that this finding of overall significance is an indication of more widespread abnormality in brain function at the outset of treatment. In keeping with this hypothesis, the corresponding progress in Ms. B’s recovery from addiction, as measured by increased readiness for behavioral change, for example, was much greater than that of Ms. A. Future research in this area will determine whether an index of cerebral function, such as the perfusion imaging presented here, can be used to optimize the choice of treatment strategy.
Cognitive behavior therapy is currently a mainstay of treatment for cocaine dependence; a complex learning process, its goal is a new relationship among mood, thought, and behavior. The capacity to respond to such "psychosocial" intervention is largely dependent on a patient’s cognitive flexibility. We propose that changes in cognitive function that occur during the period of early recovery from heavy drug abuse are multidetermined, reflecting alterations at many levels of regulation of cerebral function. Previous investigations have demonstrated both deficits in neuropsychological performance and abnormalities in brain perfusion or metabolism in chronic cocaine abusers and that both these deficits can improve during abstinence. We have chosen cerebral perfusion as a marker for recovery from cocaine’s complex pharmacological effects, which clearly outlast its serum half-life. We hypothesized that a measure of the change in cerebral perfusion during early abstinence from cocaine dependence will correlate with a measure of the capacity to learn new behavior. The cases presented here demonstrate an association between the response to cognitive behavior therapy and improvement in cerebral perfusion; this correlation encourages systematic clinical studies of the relationship between cerebral perfusion and the response to cognitive therapy in recovery from cocaine dependence.
Received Dec. 19, 1999; revision received Dec. 26, 2000; accepted Dec. 28, 2000. From the Departments of Psychiatry and Neurology, Yale School of Medicine, New Haven, Conn.; and the Substance Abuse Service, VA Connecticut Healthcare System, West Haven. Address reprint requests to Dr. Gottschalk, VACHS—West Haven, Mail Code Psychiatry 116-A, 950 Campbell Ave., West Haven, CT 06516; firstname.lastname@example.org (e-mail). Supported by grants from the National Institute on Drug Abuse (DA-04060, DA-09250, and DA-00454 to Dr. Kosten and DA-00167 to Dr. Gottschalk). The authors thank the staff of the Nuclear Medicine and Substance Abuse Services at the VA Connecticut Healthcare System—West Haven, Conn.; and Cheryl McCullough, Karen Hunkele, Maryjane Caporale, and Suemarie Sanabria for data collection and analysis.
Baseline Perfusion Scans of Ms. B, a Cocaine-Dependent Patient, and One Normal Comparison Subjecta
aRed represents the areas of highest perfusion, and yellow represents areas that are 60%–70% of maximum perfusion. Scans were acquired within 48 hours of last cocaine use for the patient. Both subjects were at rest, with eyes closed, in a dark, quiet room, without active task instructions.
Summary Statistical Parametric Map of Increases in Perfusion Over 2 Weeks in Ms. Aa
aThe colored areas represent areas of significant change 2 weeks after the baseline study. On the left are orthogonal views of the areas of greatest change in the right hemisphere. On the right are three-dimensional rendered views of the areas of maximum change on the external surface of the brain. The regional changes are less than that expected by chance, relative to the changes found in a comparison group.
Summary Statistical Parametric Map of Increases in Perfusion Over 2 Weeks in Ms. Ba
aThe colored areas represent areas of significant change 2 weeks after the baseline study. On the left are orthogonal views of the areas of greatest change in the left hemisphere. On the right are three-dimensional rendered views of the areas of maximum change on the external surface of the brain. The regional changes are more than that expected by chance, relative to the changes found in a comparison group.