This is Archived content. To view the latest NIDA Notes go here.
Cite this article

NIDA. (2007, February 1). Cocaine Abusers' Pretreatment Cue Responses Predict Recovery Success. Retrieved from

press ctrl+c to copy

In the future, patients' brain scans may help clinicians tailor addiction treatment to improve therapeutic outcomes.

February 01, 2007
Lori Whitten, NIDA Notes Staff Writer

A recent NIDA study strengthens prospects that brain imaging may one day help clinicians assign individual patients to treatment models that maximize their personal chances of a successful outcome. The study, conducted by Dr. Thomas Kosten and colleagues at Yale University School of Medicine, the University of Arkansas for Medical Sciences, and the Massachusetts Institute of Technology, correlated cocaine-addicted patients' regional brain responses to drug cues with their outcomes in subsequent treatment. The patients whose brain scans revealed rapid and strong activation in sensory, motor, and cognition- and emotion-processing brain areas were more likely to drop out of treatment and fail to achieve stable abstinence.

fMRI image showing activation of the posterior cingulate cortex - see captionfMRI Tracks Cue-Induced Brain Activity: When cocaine-addicted patients watched a drug-related videotape, activation of the posterior cingulate cortex (highlighted in the brain image) occurred more quickly in those who subsequently relapsed.

"A test that predicts treatment outcomes, especially vulnerability to relapse, could help guide individualized treatment. For example, a clinician might recommend an extended stay in residential treatment or more intense behavioral intervention for patients with a propensity for relapse," says Dr. Kosten, now at Baylor College of Medicine.

Dr. Kosten and colleagues pursued the implications of an intriguing finding made in a prior study of cocaine cue responses: In some patients, strong, rapid activation of brain areas associated with emotion and sensing preceded the onset of craving. Although craving itself does not generally predict relapse, Dr. Kosten's team speculated that cue-induced brain activation that occurs quickly and reflexively, below awareness, might do so. They hypothesized that patients who showed such responses during the first 30 seconds of cue exposure would also demonstrate poorer treatment outcomes.

To test their hypothesis, the investigators recruited 17 men and women who were participating in a trial of an antidepressant—sertraline—that is being evaluated as a possible treatment for cocaine addiction. The participants reported abusing cocaine 20 days, on average, during the month before the study. All met standard clinical criteria for cocaine addiction and had abused the drug for 6 years, on average. Most were new to treatment.

After being cocaine-free for 5 days, on average, each participant underwent functional magnetic resonance imaging (fMRI) while watching two 4-minute videotapes. The first minute of each tape reported on vegetable prices, and the participants' brain activity while hearing this emotionally neutral information served as a baseline for comparison. During the last 3 minutes, an actor pretended to smoke cocaine and experience a "rush." Immediately after viewing the tapes, each participant rated peak cocaine craving intensity on a scale from 0 to 10. After the imaging session, participants began taking either sertraline or a placebo daily and completed 2 weeks of residential treatment. During the 10-week outpatient phase of the trial, they were to continue their medication regimen, receive weekly individual cognitive-behavioral therapy, and submit urine samples three times a week.

Interplay Within Cingulate Cortex?

Nine of the 17 participants relapsed, defined by the investigators as submitting fewer than 15 of a possible 30 cocaine-free samples during the study and not completing outpatient treatment. Participants taking sertraline were just as likely as those taking the placebo to relapse. Relapsers and nonrelapsers reported cue-induced cravings of comparable intensity. The two groups differed, however, on brain activation during the first 30 seconds of the cocaine-cue videotapes. Relapsers showed greater cue-induced activation than nonrelapsers in several areas of the cortex: the left precentral (movement control), right superior temporal (auditory processing), right lingual and right inferior occipital (visual processing), and the left posterior cingulate cortices. The cingulate cortex is integral to attention, response inhibition, emotional regulation, and decisionmaking.

What Researchers Know About the Cingulate Cortex and Behavior

The cingulate cortex connects to both the limbic systems (emotion and motivation) and the prefrontal cortex (planning and control of behavior) and seems to integrate emotion and cognition. The anterior and posterior regions of the cingulate are connected to different brain areas and differ functionally.

Anterior Cingulate Cortex (ACC)

  • Paying attention—The ACC monitors inputs from the senses (competing options) and selects what we attend to.
  • Making decisions—Influenced by past experience, the ACC assesses risk, reward, and conflict. It works with areas of the frontal cortex to select a response.
  • Inhibiting responses—The ACC integrates input from the prefrontal cortex and detects and corrects errors in behavior.
  • Detecting and controlling emotions—The ACC monitors what is going on inside (feelings, pain, and bodily arousal) and controls voluntary suppression of these sensations.

Posterior Cingulate Cortex (PCC)

  • Responding reflexively—The PCC integrates sensory and movement information with established behavior patterns and acts "below awareness." The PCC responds to reward and positive feedback.
  • Reacting emotionally—The PCC processes emotion-related autobiographical memories and the emotional perspective of self and others. Its activity correlates with internal physiological responses (heart rate, anxiety, and arousal level).

The relapsers' greater activation of the posterior cingulate cortex (PCC) was the most notable of the findings. Also significant was the lack of any difference between the outcome groups in activation of the neighboring anterior cingulate cortex (ACC). This contrasts with findings from previous studies, in which ACC activation and craving were associated in patients who had longer abstinence (average 14-28 days) and were imaged for periods longer than 30 seconds after being shown cues.

Taken together Dr. Kosten says, these results suggest that an interplay occurs between the PCC and ACC following exposure to cocaine cues and changes with increasing stability of abstinence. In patients highly vulnerable to cues, intense PCC activation occurs within 30 seconds of cue exposure and is positively associated with risk for relapse. In less vulnerable patients, early PCC activation is less intense, and these patients are able to activate the ACC to counter the association with relapse risk.

Dr. Kosten's findings highlight the promise of imaging linked to behavioral assessments as a tool for guiding the treatment of addictions and other psychiatric disorders. They parallel a previous NIDA-funded study in which brain activity patterns during a decisionmaking task predicted treatment outcomes among patients addicted to methamphetamine (see "Brain Activity Patterns Signal Risk of Relapse to Methamphetamine").

"If researchers can determine changes in brain activity that predict responses to particular treatments, then clinicians could match therapy with individuals' scan results or even monitor progress in therapy," says Dr. Kosten. More generally, studies that examine biological and behavioral predictors of treatment response elucidate the physiology underlying addiction—particularly the neural circuitry integrating stress, craving, and the propensity to relapse. New tools—for example, scanners that highlight brain areas that are working together—are expected to reveal more about these physiological processes. "With such functional connectivity imaging, one could examine how the anterior and posterior cingulate 'talk' to each other during a drug cue or other experience," says Dr. Rajita Sinha, an investigator in the Kosten study.

"Eventually, researchers will integrate the findings of such studies into a complete picture that will specify therapeutic pathways or help in the development of targeted medications to reduce relapse probability," adds Dr. Harold Gordon of NIDA's Division of Clinical Neuroscience and Behavioral Research.


Kosten, T.R., et al. Cue-induced brain activity changes and relapse in cocaine-dependent patients. Neuropsychopharmacology 31(3):644-650, 2006. [Abstract]

Stress Cues Also Signal Relapse Risk

Exposing patients to stress cues at the beginning of cocaine addiction treatment triggers craving and measurable biological responses that may predict drug abuse outcomes during early recovery. NIDA-funded researchers found that stress-induced craving was associated with a shortened interval to relapse following inpatient treatment, while hormonal responses to stress predicted the amount of cocaine the patients consumed during relapse.

The findings were reported in a followup to prior research conducted by Dr. Rajita Sinha and colleagues at Yale University School of Medicine. In the previous study, patients who listened to tapes reminding them of a stressful experience and a drug-related experience demonstrated an elevated biological stress response and increased cocaine craving compared with their response to tapes of relaxing experiences (see "Cocaine-Related Environmental Cues Elicit Physiological Stress Responses").

Dr. Sinha and colleagues followed up with 49 of the 54 patients 3 months after completion of inpatient behavioral treatment. They found that patients who had experienced more intense cocaine craving while revisiting their stressful experiences via audiotape tended to relapse sooner. The probability of relapse 3 months after treatment was 56 percent among patients who reported no craving. Each unit increase on a craving intensity scale of 0 to 10 was associated with a 31 percent rise in the likelihood of relapse during the followup period.

Participants who released high levels of the stress hormones adrenocorticotropic hormone (ACTH) and cortisol in response to the stressful tapes consumed more cocaine than low-level responders during the followup. Three months after treatment, high-level responders had consumed about 8 g of cocaine cumulatively over their cocaine abuse periods, while low-level responders consumed about 3 g.

The findings of the study suggest that different components of the stress response are associated with various aspects of relapse: craving with reinitiating abuse and hormonal responses with the ability to control intake after reinitiating abuse. "Greater hormonal release during stress may 'prime' higher cocaine consumption or bingeing after return to abuse, perhaps by altering the rewarding effects of the drug," Dr. Sinha says.

Dr. Sinha and colleagues did not find a link between drug cue-induced craving and relapse outcomes, a result that is consistent with previous studies. However, because the drug cue imagery produced physiological reactions similar to those triggered by the stress cues, the researchers speculated that studies using a larger sample or exposure to actual drug cues, rather than just images of them, may show such an association.

Prior studies that did not find a link between cue-induced craving and relapse generally assessed only one or two dimensions of craving, Dr. Sinha points out. Studies that address multiple components—wanting the drug, feelings about the drug and about wanting it, drug-seeking behaviors, coping reactions, physiological arousal, and stress hormone levels—may better indicate vulnerability to relapse, she says.

"For people who are not addicted, knowing that you want a particular thing probably defines craving. Our findings suggest that for addicted people, craving is a 'state'—a multidimensional experience—comprised, in part, of stress-like arousal. In this state, desire becomes pathological, and people cannot delay gratification or divert their attention," says Dr. Sinha.

The results of Dr. Sinha's study suggest that stress-induced drug craving and physiological responses may be used as a diagnostic indicator of relapse propensity and might one day help clinicians tailor their interventions toward regulating stress and coping with stress-induced craving."Research on each component and the role that it plays in continued drug abuse is just beginning, but such studies ultimately may improve our ability to help people attain long-term recovery," she says.


Sinha, R., et al. Stress-induced cocaine craving and hypothalamic-pituitaryadrenal responses are predictive of cocaine relapse outcomes. Archives of General Psychiatry 63(3):324-331, 2006. [Abstract]