by Edythe D. London, Ph.D., Steven Grant , Ph.D., and David Newlin, Ph.D., NIDA

Schematic representation of the lateral (above) and medial (below) aspects of the brain. Blue dots indicate brain regions in which cocaine-related cues produced increases in glucose metabolism compared with the effects of neutral cues. Affected regions were located in prefrontal (superior, medial, and inferior frontal gyri; pregenual cingulate; and medial and posterior orbitofrontal gyri), temporal (superior and inferior temporal gyri), central (pre- and postcentral gyri), parietal (angular gyrus), limbic (parahippocampal gyrus), and occipital cortices (pre-cuneus, inferior fusiform gyrus, and middle occipital gyrus).

The epidemic of cocaine abuse in the United States has underscored the need for knowledge about the mechanisms by which cocaine alters behavior. Such information is critical in the development of medications to curb cocaine addiction. As part of this effort, we have applied the [18F]fluorodeoxyglucose method to assess changes in regional cerebral glucose metabolism, an index of local brain activity, by positron emission tomography (PET) in cocaine abusers exposed to cocaine-related cues. Our preliminary findings indicate that the presentation of a videotape with cocaine-specific cues, exposure to drug paraphernalia, and anticipation of cocaine self-administration produce self-reports of cocaine craving, electroencephalographic (EEG) arousal, and a characteristic metabolic pattern in the brain. Exposure to cocaine cues is associated with stimulation of areas of the occipital cortex, presumably because the cues are visual. Also activated are areas of the prefrontal, temporal, parietal, and limbic cortices. This work begins to define anatomical circuits that are important in the response to environmental stimuli. These circuits may link episodic memories of cocaine use to emotions and to planning future drug taking.

Although the role of the mesolimbic dopaminergic pathways in supporting reward and reinforcement from psychoactive drugs has dominated contemporary drug-abuse research (1), it has become increasingly evident that the pharmacological response to the drug per se is only one factor that maintains compulsive drug use. Drug-induced reward is a critical factor in the initiation or acquisition of self-administration behavior, but hedonic responses to the drug may not be critical to subsequent stages in addiction (2,3). In fact, Tiffany (4) has suggested that continued drug self-administration results in automatic cognitive and motoric habit patterns and that relapse results from re-exposure to cues that elicit these automatic patterns.

It is well-established that individuals who are drug-free can relapse to abuse long after detoxification, and it is, therefore, probable that chronic drug abuse produces persistent changes in the nervous system that outlast the immediate effects on brain reinforcement pathways. The presence of long-term changes suggests that learning and memory are critical to the addictive process. In this regard, almost 50 years ago, anecdotal reports of addicts who left the Addiction Research Center of NIDA in Lexington, Ky., and then relapsed after returning to their old neighborhoods first suggested that learning or conditioning factors may contribute to recidivism (5).

There are two broad and conflicting theories regarding the responses to drug-related environmental stimuli. Both propose that environmental cues promote drug-seeking behavior via classical conditioning mechanisms, but they differ with respect to the proposed relationship between acute drug effects and the responses elicited by the cues. The early proposals held that drug-related cues elicit an aversive motivational state, which the addict attempts to escape by further drug intake. The two major theories focusing on the aversive effects of cues posited 1) that conditioned withdrawal phenomena produced by drug-associated stimuli contribute to relapse (6), and 2) that exposure to drug cues activates an opposing process that contributes to tolerance (7). Either way, responses elicited by drug-related cues would be in the opposite direction of responses produced by the drug itself. For example, because cocaine increases arousal and heart rate and produces euphoria, cocaine-related cues would be predicted to decrease arousal and heart rate and to produce dysphoria.

More recently, researchers have proposed that responses to cues mimic aspects of the hedonic effects of drugs and, therefore, that drug-seeking behavior is motivated by approach rather than by avoidance (2,8,9). According to this view, cues activate reinforcement systems in the brain and the addict seeks to maintain this activation by engaging in drug-seeking behavior. These "approach" theories propose that responses to drug-related cues contribute to sensitization to drug effects because the conditioned response is in the same direction as the acute effect of the drug (2,10,11). In contrast to the "avoidance" predictions, the recent hypotheses propose that cues related to cocaine use would increase arousal and heart rate and would generate euphoria. Because there is empirical evidence of both drug-like and drug-opposite responses to drug-related cues (2,7,12,13), attempts have been made to reconcile these conflicting theories. One suggestion is that the nature of cue-elicited responses depends on the class of the drug taken. For example, cues associated with opioid and sedative drugs have been proposed to elicit drug-opposite and withdrawal-like responses, whereas cues associated with stimulants, such as cocaine, elicit drug-like responses (14). A second suggestion is that the direction of the cue-elicited effect depends on the specific response measured. Drug effects on afferent, or sensory, processes lead to drug-like responses to cues, whereas drug effects on efferent, or motor, processes produce drug-opposite cue-elicited responses (15).

Although conditioned responses to drug-related environmental stimuli were first described by Pavlov (16), the brain systems involved in drug conditioning have only recently been studied. Because direct measurement of many brain systems requires invasive techniques, these studies have been performed primarily in animals. Using the induction of the immediate early gene c-fos as a marker for neuronal activation, Fibiger and colleagues mapped the brain regions in rats that respond to environmental cues associated with cocaine administration (17). They found that drug-related cues increased c-fos expression throughout the limbic system, including cingulate cortex, amygdala, septum, and habenula. This distribution was similar to that of the direct effect of cocaine on c-fos expression, except that cocaine also increased c-fos expression in the caudate and accumbens nuclei. The absence of increased c-fos expression in any portion of the striatum is problematic for both "approach" and "avoidance" models of drug conditioning because they have emphasized the contribution of striatal areas to the responses elicited by drug-related cues (9,18). Furthermore, because dopaminergic input to these regions contributes to the reinforcing and hedonic effects of drugs of abuse, the incentive motivational theories propose that drug-related cues are reinforcing because they also increase dopaminergic tone. It is known that cues associated with delivery of biological reinforcers - such as food - increase the firing of dopaminergic neurons in the brainstem and increase dopamine release in the basal ganglia (2,19), but the evidence for dopamine release in response to drug-related cues is contradictory (2). Robinson and Berridge (2) recently suggested that although the reinforcing effects of drugs are initially dependent on alterations in dopaminergic tone, over time, other brain systems become dominant, especially those associated with drug-seeking behavior elicited by environmental stimuli.

Researchers have tested theories of drug conditioning in human volunteers by presenting them with drug cues in the absence of drug administration. Typically, the subjects view videotapes of drug-taking behavior or conduct a self-administration ritual and inject themselves with a placebo (14,20); this way, the conditioned response to drug-associated stimuli is not obscured by the drug. In studies of cocaine addicts, such experiments have shown that drug-related cues lead subjects to report cocaine craving, accompanied by reliable decreases in skin temperature and skin resistance, increases in heart rate, and EEG arousal (21,22).

Although drug abusers may attribute their addiction to craving, the extent to which craving drives drug-taking behavior may be limited (13). For example, research volunteers with histories of cocaine abuse reported less craving when treated with desipramine, a proposed therapy for cocaine abuse; however, the volunteers did not reduce self-administration of cocaine in the laboratory (23). In another study of the efficacy of desipramine in the treatment of cocaine abuse, decreases in self-reports of craving occurred weeks after a decrement in cocaine use was observed (24). Nonetheless, it is generally agreed that craving is "a subjective state in humans that is associated with drug dependence" (25). A current goal of our research on mechanisms of drug dependence is to elucidate the biological determinants of craving for abused drugs and the relationship, if any, between drug craving and drug dependence.

To study the effects of cocaine-related cues on cerebral metabolism, we paired PET measurements with psychophysiological assays and self-reported subjective assessments in cocaine abusers during two sessions. In one session, during the period when the radiotracer ([18F]fluorodeoxyglucose) was taken up after its intravenous injection, subjects viewed a neutral videotape on arts and crafts. In the second session, they were exposed to a cocaine-related stimulus complex - a videotape of cocaine-related activity and paraphernalia, the presence of actual paraphernalia, and a small amount of cocaine. Analysis of data from the first nine subjects revealed increases in self-reports of craving and overall arousal (decrease in EEG power) during presentation of the cocaine-related stimuli compared with the neutral session. Cerebral metabolic activation in response to cocaine-related stimuli differed from activation during the neutral session. The response also differed from the decrease in brain activity observed previously in response to acute administration of cocaine (26). The changes in response to cocaine-related stimuli were "drug-opposite": the stimuli caused selective increases in regional cerebral glucose metabolism, whereas acute cocaine administration reduced glucose metabolism globally. Increases due to cocaine-related cues, as distinct from neutral cues, were observed throughout the prefrontal cortex, occipital cortex and parahippocampal gyrus, with more restricted activations in the parietal and temporal lobes and the pre- and postcentral gyri (see figure). The metabolic changes produced by the cocaine-related cues are consistent with the view that the response to drug-related environmental stimuli may reflect activation of a distributed neuroanatomical network, involving areas that mediate retrieval of memories with affective components as well as those that may participate in the planning of future drug self-administration.

These and other recent findings concerning the activation of brain systems by exposure to cocaine cues raise interesting questions for further study. The study of neural systems involved in highly motivated behavior, such as addiction to drugs, may have important implications for other forms of both adaptive and maladaptive behaviors. In this regard, several medical disorders and risk factors may entail acquired motivation to engage in other self-destructive behaviors. Understanding cocaine craving may help us understand craving for tobacco, alcohol, and opiates and the brain substrates of normal and abnormal cravings for food in eating disorders. Related questions concern paraphilic sexual motivation and preoccupation with gambling. Moreover, it may be possible to determine whether brain systems underlying acquired motivation for stimuli that are not biologically relevant are similar to those of "normal" motivation for biologically relevant stimuli such as food, water, and sex.

Although our findings suggest that a specific anatomical network is activated by cocaine-related cues and that this activation is associated with cocaine craving, many questions remain unanswered. To date, our studies in this area have focused on identifying the brain regions important in responding to drug-related cues, but the specific neurotransmitters mediating cocaine craving are not known. The availability of radioligands for specific neurotransmitter receptors and the development of tracers and mathematical models for assessing neurotransmitter synthesis will facilitate further work with PET scanning.

In clinical applications, it will be important to determine whether the conditioned response to cocaine cues can be blocked pharmacologically, either by agonists such as bromocriptine (27) or by antagonists such as naltrexone (recently approved by the FDA for the treatment of alcoholism). Key questions will be whether blocking craving interferes with biologically necessary motivational systems and whether suppressing craving will also prevent the actual taking of abused drugs on the street.


We are indebted to our collaborators: Robert Phillips, Victor Villemagne, Xiang Liu, Alane Kimes, and Carlo Contoreggi at NIDA and Arthur Margolin at Yale University in New Haven, Conn. The NIDA Brain Imaging Facility is supported by the Operating Budget of the Intramural Research Program, NIDA, and the Office of National Drug Control Policy.


1. E.D. London, S.J. Grant, M.J. Morgan, and S.R. Zukin. "Neurobiology of drug abuse." In: Neuropsychiatry: A Comprehensive Textbook. B.S. Fogel, ed. (in press).

2. T.E. Robinson and K.C. Berridge. "The neural basis of drug craving: an incentive-sensitization theory of addiction." Brain Res. Rev. 18, 247 - 91 (1993).

3. J.S. Altman, B.J. Everitt, M. Fischman, S. Glautier, A. Markou, D. Nutt, et. al. "The biological, social and clinical bases of drug addiction." Psychopharmacology (Berl) (in press).

4. S.T. Tiffany. "A cognitive model of drug urges and drug-use behavior: role of automatic and nonautomatic processes." Psychol. Rev. 97, 147 - 68 (1990).

5. A. Wikler. "Conditioning factors in opiate addiction and relapse." In: Narcotics. D.M. Wilner and G.G. Kassebaum, eds. New York: McGraw Hill, pp. 85 - 100 (1948).

6. A. Wikler. "Dynamics of drug dependence: implications of a conditioning theory for research and treatment." Arch. Gen. Psychiatry 28, 611 - 16 (1973).

7. S. Siegel. "Classical conditioning, drug tolerance, and drug dependence." In: Research Advances in Alcohol and Drug Problems, Volume 7. R.G. Smart, F.B. Glaser, Y. Israel, H. Kalant, R.E. Popham, and W. Schmidt, eds. New York: Plenum, pp. 207 - 44 (1983).

8. J. Stewart, H. de Wit, and R. Eikelboom. "Role of unconditioned and conditioned drug effects in the self-administration of opiates and stimulants." Psychol. Rev. 91, 251 - 68 (1984).

9. J.P. Vincent, B. Kartalovski, P. Geneste, J.M. Kamenka, and M. Lazdunski. "Interaction of phencyclidine ("angel dust") with a specific receptor in rat brain membranes." Proc. Natl. Acad. Sci. USA 76, 4678 - 82 (1979).

10. P.W. Kalivas, B.A. Sorg, and M.S. Hooks. "The pharmacology and neural circuitry of sensitization to psychostimulants." Behav. Pharmacol. 4, 315 - 34 (1993).

11. R.M. Post and S.R.B. Weiss. "Sensitization and kindling: implications for the evolution of psychiatric symptomatology." In: Sensitization in the Nervous System. P.W. Kalivas and C.D. Barnes, eds. Caldwell, NJ.: Telford, pp. 257 - 93 (1993).

12. R.S. Niaura, D.J. Rohsenow, J.A. Binkoff, P.M. Monti, M. Pedraza, and D.B. Abrams. "Relevance of cue reactivity to understanding alcohol and smoking relapse." J. Abnorm. Psychol. 97, 133 - 52 (1988).

13. L.O. Bauer. "Psychobiology of craving." In: Substance Abuse. A Comprehensive Textbook. J.H. Lowinson, P. Ruiz, R.B. Millman, and J.G. Langrod, eds. Baltimore: Williams & Wilkins, pp. 51 - 5 (1992).

14. A.R., Childress, R. Ehrman, D.J. Rohsenow, S.J. Robbins, and C.P. O'Brien. "Classically conditioned factors in drug dependence." In: Substance Abuse. A Comprehensive Textbook.. J.H. Lowinson, P. Ruiz, R.B. Millman, and J.G. Langrod, eds. Baltimore: Williams & Wilkins, pp. 56 - 69 (1992).

15. R. Eikelboom and J. Stewart. "Conditioning of drug-induced physiological responses." Psychol. Rev. 89, 507 - 28 (1982).

16. I.P. Pavlov. In: Conditioned Reflexes; An Investigation of the Physiological Activity of the Cerebral Cortex. London: Oxford University Press, (1927)

17. E.E. Brown, G.S. Robertson, and H.C. Fibiger. "Evidence for conditioned neuronal activation following exposure to a cocaine-paired environment: role of forebrain limbic structures." J. Neurosci. 12, 4112 - 21 (1992).

18. G.F. Koob, L. Stinus, M. Le Moal, and F.E. Bloom. "Opponent process theory of motivation: neurobiological evidence from studies of opiate dependence." Neurosci. Biobehav. Res. 13, 135 - 40 (1989).

19. W. Schultz. "Activity of dopamine neurons in the behaving primate." Semin. Neurosci. 4, 129 - 38 (1992).

20. C.P. O'Brien, R.N. Ehrman, and J.W. Ternes. "Classical conditioning in human opioid dependence." In: Behavioral Analysis of Drug Dependence. S. Goldberg and I.P. Stolerman, eds. Orlando, FL.: Academic Press, pp. 329 - 56 (1986).

21. R.N. Ehrman, S.J. Robbins, A.R. Childress, and C.P. O'Brien. "Conditioned responses to cocaine-related stimuli in cocaine abuse patients." Psychopharmacology (Berl) 107, 523 - 9 (1992).

22. L.O. Bauer and H.R. Kranzler. "Electroencephalographic activity and mood in cocaine-dependent outpatients: effects of cocaine cue exposure." Biol. Psychiatry 36, 189 - 97 (1994).

23. M.W. Fischman, R.W. Foltin, G. Nestadt, and G.D. Pearlson. "Effects of desipramine maintenance on cocaine self-administration by humans." J. Pharmacol. Exp. Ther. 253, 760 - 70 (1990).

24. F.H. Gawin, H.D. Kleber, R. Byck, B.J. Rounsaville, T.R. Kosten, P.I. Jatlow, et al. "Desipramine facilitation of initial cocaine abstinence." Arch. Gen. Psychiatry 46, 117 - 21 (1989).

25. R.W. Pickens and C.E. Johanson. "Craving: consensus of status and agenda for future research." Drug Alcohol Depend. 30, 127 - 31 (1992).

26. E.D. London, N.G. Cascella, D.F. Wong, R.L. Phillips, R.F. Dannals, J.M. Links, et. al. "Cocaine-induced reduction of glucose utilization in human brain. A study using positron emission tomography and [fluorine 18]fluorodeoxyglucose." Arch. Gen. Psychiatry 4, 567 - 74 (1990).

27. H.R. Kranzler and L.O. Bauer. "Bromocriptine and cocaine cue reactivity in cocaine-dependent patients." Br. J. Addict. 87, 1537 - 48 (1992).

Table of Contents