* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Orbitofrontal Cortex and Human Drug Abuse: Functional Imaging
Environmental enrichment wikipedia , lookup
Optogenetics wikipedia , lookup
Feature detection (nervous system) wikipedia , lookup
Neurolinguistics wikipedia , lookup
Neurophilosophy wikipedia , lookup
Eyeblink conditioning wikipedia , lookup
Synaptic gating wikipedia , lookup
Embodied language processing wikipedia , lookup
Cortical cooling wikipedia , lookup
Executive functions wikipedia , lookup
Metastability in the brain wikipedia , lookup
History of neuroimaging wikipedia , lookup
Human brain wikipedia , lookup
Neuroplasticity wikipedia , lookup
Limbic system wikipedia , lookup
Biology of depression wikipedia , lookup
Neuropsychopharmacology wikipedia , lookup
Neuroanatomy of memory wikipedia , lookup
Neural correlates of consciousness wikipedia , lookup
Clinical neurochemistry wikipedia , lookup
Aging brain wikipedia , lookup
Cognitive neuroscience of music wikipedia , lookup
Neuroesthetics wikipedia , lookup
Time perception wikipedia , lookup
Affective neuroscience wikipedia , lookup
Neuroeconomics wikipedia , lookup
Cerebral cortex wikipedia , lookup
Insular cortex wikipedia , lookup
Prefrontal cortex wikipedia , lookup
Inferior temporal gyrus wikipedia , lookup
Orbitofrontal Cortex and Human Drug Abuse: Functional Imaging Edythe D. London, Monique Ernst, Steven Grant, Katherine Bonson and Aviv Weinstein Brain Imaging Center, National Institute on Drug Abuse, Baltimore, MD 21224, USA The orbitofrontal cortex (OFC), a paralimbic region, participates in association functions, integrating emotion with behavior and various sensory processes (Hof et al., 1995). Its dysfunction has been implicated in psychiatric disorders that involve inappropriate emotional and behavioral responses to stimuli. The best example is that of obsessive compulsive disorder (Rauch et al., 1994; Breiter and Rauch, 1996; Zald and Kim, 1996), but post-traumatic stress disorder (Semple et al., 1993), disorders of mood (Pardo et al., 1993; Baker et al., 1997), antisocial personality disorder (Meyers et al., 1992) and aggression (Fornazzari et al., 1992; Siever et al., 1999) are also included. A lthough symptoms may overlap among these and other disorders (e.g. compulsive drug-seeking by addicts, reminiscent of behavior that characterizes obsessive compulsive disorder), the range of pathologies linked to the OFC supports its central role in human behavior. More specifically, the OFC contributes to a variety of behavioral states and functions, including the processing of reward, emotion and decision making (Bechara et al., 2000; Rolls, 2000; Schultz et al., 2000), which are essential components of motivational-directed behavior. The motivational values can be innate (such as the drive for food); or they may be learned with repeated reinforcement. Substance abuse, which can be conceptualized as a dysregulation of motivationaldirected behavior, is an example of the latter case. The OFC is placed in a position to code the motivational attributes of responses to stimuli. It is a heterogeneous region that has connections with other prefrontal, limbic, sensory and premotor areas (Cavada et al., 2000; Öngür and Price, 2000). Linked to the mesolimbic dopamine system that is critical for drug reward (Di Chiara and Imperato, 1988; Koob and Bloom, 1988; London et al., 1996; Wise, 1996), it receives inputs from association areas of each sensory modality (olfactory, gustatory, visual, auditory and somatosensory) (Pandya and Yeterian, 1990; Morecraft et al., 1992; Cavada et al., 2000; Öngür and Price, 2000). Therefore, the OFC has the information needed to integrate the sensory characteristics of a stimulus-object (Carmichael and Price, 1996). The OFC can attach an emotional valence to the stimulus-object through its relationship with the amygdala (Barbas and De Olmos, 1990; Morecraft et al., 1992). Furthermore, it can evaluate these characteristics against previous experience through its connections with regions known to subserve memory [dorsolateral prefrontal cortex and mediodorsal nucleus of the thalamus, pars magnocellularis (Morecraft et al., 1992; Ray and Price, 1993), hippocampus and hippocampal gyrus (Cavada et al., 2000; Öngür and Price, 2000)]. Finally, the OFC–striatum–globus pallidus–thalamus–OFC loop (Alexander and Crutcher, 1990) can mediate reinforcement of the motivational attribute of a stimulus-object. A modular approach to the role of the OFC, although heuristic, may help tease out the various mechanisms that mediate substance abuse. Using such a strategy, individual components of aberrant behavior in substance abusers can be studied separately. One of these components is expectancy that is based on predictions of reward and attribution of probabilistic rewarding properties to the stimulus-object. Another is compulsive drive (motivational state) to use drugs, which is linked to craving. Lastly, decision making, based on the motivational attributes of the stimulus and the balance between expectation of immediate reward and long-term losses, is an important aspect of substance abuse behavior. While considering how the OFC is involved in these three behaviors, it is important to appreciate the heterogeneous nature of this region, histologically and anatomically (Morecraft et al., 1992; Carmichael et al., 1994). Anatomically, the OFC can be divided into anterior regions, which connect with higher association cortices (such as the dorsolateral prefrontal cortex), and posterior regions, which have selective connections with the amygdala, and entorhinal and perirhinal cortices. Therefore, anterior regions schematically would be important in decisionmaking processes, and posterior regions in the emotional and autonomic aspects of craving. Both regions could contribute to expectancy, which depends on the evaluation of the emotional valence of the stimulus. Along a transaxial plane, the medial OFC is connected with the ventral striatum and the lateral OFC is connected with the caudate nucleus, which suggests that the medial OFC would be more important to reinforcement than the lateral OFC. This review covers findings from brain imaging studies that support important contributions of the OFC to the persistent behavioral states characteristic of addiction. To date, most functional imaging studies have been unable to distinguish accurately between the different regions of the OFC that might take part in the respective behaviors. The enhanced spatial resolution of new generation positron emission tomography (PET) scanners and improved technology of functional mag- © Oxford University Press 2000 Cerebral Cortex Mar 2000;10:334–342; 1047–3211/00/$4.00 The orbitofrontal cortex (OFC) plays a central role in human behavior. Anatomically connected with association areas of all sensory modalities, limbic structures, prefrontal cortical regions that mediate executive function and subcortical nuclei, this brain region can serve to integrate the physical and emotional attributes of a stimulusobject and establish a motivational value based on estimation of potential reward. To the extent that addictive disorders reflect a dysregulation of the ability to evaluate potential reward against harm from drug self-administration, it would be anticipated that substance abuse disorder might reflect dysfunction of the OFC. With the application of brain imaging techniques to the study of human substance abuse, evidence has been obtained that activity in the OFC and its connections plays a role in several components of the maladaptive behavior of substance abuse, including expectancy, craving and impaired decision making. Figure 1. Images of rCMRglc in a selected subject from a group of polydrug abusers (right, n = 10) and a group of control subjects who were not substance abusers but were drawn form the same population (left, n = 10) (Stapleton et al., 1995). The arrow points to the orbitofrontal gyrus, which exhibited significantly higher rates of rCMRglc in the substance abuse group when metabolic rates were adjusted for global metabolism. Drug abuse subjects also showed higher adjusted rCMRglc in the superior and middle temporal gyri. netic resonance imaging (fMRI), however, offer the promise of isolating functional activity in subregions of the OFC. OFC Metabolism in Polysubstance Abusers: Personality Traits, Expectancy and Motivation PET studies, performed to elucidate abnormalities in brain function that might contribute to the perpetuation of addiction, have provided evidence for dysfunction of the OFC in drug abusers. In one of these studies, regional cerebral metabolic rate for glucose (rCMRglc), an index of local brain function (Sokoloff et al., 1977; Phelps et al., 1979; Reivich et al., 1979), was measured in polydrug abusers and in control subjects who were drawn from the same community (Stapleton et al., 1995). Twenty polydrug abusers took part in a study of the acute effects of cocaine on rCMRglc, measured using PET with the [18F]f luorodeoxyglucose (FDG) method (London et al., 1990). Polydrug abusers completed two test sessions, one in which cocaine was administered and another in which saline (placebo) was given. Ten control subjects, recruited for comparison, also participated in two PET sessions, but never received cocaine. Measures of rCMRglc during only the placebo administration were used for comparison with data from control subjects. To avoid order effects on rCMRglc (Stapleton et al., 1997), data from only those drug abusers (n = 10) who received placebo in their first PET session and data from the first PET session of the control subjects (n = 10) were used. Although the substance abusers did not know whether they would receive placebo or cocaine for their first session, control participants knew that they would receive placebo for all sessions. Prior to the PET scan, each substance abuser participated in four preliminary test sessions, two with placebo and two with cocaine (20 and 40 mg i.v.), in order to acquaint them with the subsequent test procedures. Mean global metabolic rate for glucose (mg/100 g tissue/min) in the substance abusers (6.91 ± 0.17) did not differ significantly from the rate in the control group (7.27 ± 0.21), but the pattern of rCMRglc differed between the two groups. Visual inspection of the brain images revealed that control participants showed relatively uniform cortical glucose metabolism compared with the substance abusers, whose rCMRglc was reduced in posterior regions (Fig. 1). When metabolic rate data for each region were submitted to analysis of covariance with the global metabolic rate as a covariate, the substance abusers showed higher glucose utilization than the control group in the OFC, superior frontal gyrus, middle temporal gyrus and insula (Table 1, Fig. 1). Notably, the greatest group difference was in the OFC (Stapleton et al., 1995). Of the ten substance abusers, six met criteria for antisocial personality disorder and one met criteria for pathological gambling. Therefore, differences in rCMRglc of the OFC between the groups could ref lect psychological characteristics associated with substance abuse. This view is consistent with previous findings of abnormal rCMRglc of OFC in individuals diagnosed with disorders of behavioral control, Cerebral Cortex Mar 2000, V 10 N 3 335 Table 1 Regional cerebral metabolic rates for glucose, adjusted for global metabolisma Control Left Frontal lobe Superior frontal gyrusA,B Middle frontal gyrusB,AB Inferior frontal gyrus Precentral gyrusB,AB Gyrus rectus Orbitofrontal cortexA Anterior cingulate gyrus Temporal lobe Temporal pole Superior temporal gyrusA,B Middle temporal gyrusA Inferior temporal gyrus InsulaA,AB Substance abuse Midline 8.05 9.47 10.10 10.68 Right Left 8.26 10.40 10.07 10.98 8.43 9.94 10.33 10.46 8.08 8.97 8.47 7.92 Right 8.86 9.95 10.04 11.37 9.00 10.06 6.07 6.59 7.84 8.61 10.09 Midline 9.12 10.63 6.00 6.19 8.09 8.23 9.81 6.17 7.26 8.54 8.66 10.53 6.18 7.01 8.71 8.70 11.09 a Each value is the mean regional cerebral metabolic rate for glucose (rCMRglc, mg/100 g/min), adjusted for global glucose metabolism as indicated in Materials and Methods. A Group factor: Substance Abuse different from Control, P < 0.05. B Laterality factor: Left different from Right, P < 0.05. AB Interaction of Group by Laterality factors, P < 0.05. such as obsessive compulsive disorder (Baxter 1990; Benkelfat et al. 1990). In addition, impulsive–aggressive patients showed a blunted metabolic response in OFC after fenf luramine challenge, consistent with a deficit in serotonergic modulation of this brain region in patients with personality disorders characterized by a high level of impulsivity and aggressiveness (Siever et al., 1999). Emotional state, related to expectancy of drug reward, may have also been ref lected in the higher rCMRglc in the OFC of the substance abusers. As they took part in a double-blind procedure in which the injection could have been placebo or cocaine, they likely experienced a negative emotional reaction (e.g. disappointment) when they realized that they had received placebo. In contrast, control participants would not have been as likely to have this emotional response. This interpretation is supported by our findings from a recent study (Grant et al., 1996) in which normalized rCMRglc was negatively correlated with self-reported mood in a number of OFC regions when neutral (not drug-related) visual cues were presented. The elevated rCMRglc in other cortical areas (insula, temporal lobe and superior frontal gyrus) in addition to OFC in the substance abusers suggests that the response to the placebo injection included activation of a circuit related to a component of motivation. Classic work of Mesulam and Mufson in the old world monkey has demonstrated connections between the OFC and insula, which also receives projections from the temporal pole and other paralimbic areas (Mesulam and Mufson, 1982a–c). On the basis of this and other work, these authors suggest that the paralimbic brain, particularly its insulo-orbitotemporopolar component, provides integration between extrapersonal stimuli and the internal milieu (Mesulam and Mufson, 1982a,c). In this light, elevated rCMRglc in the OFC and insula (and also in temporal and superior frontal areas) is consistent with an interpretive response to placebo administration in the context of expectancy of drug reward. Such integration can be critical to the ‘motivational state’ or drive to use drugs. Although the PET study situation was markedly different from the typical conditions of drug self-administration, the substance abusers may have shown conditioned drug effects, as they had 336 Orbitofrontal Cortex and Drug Abuse • London et al. previously received cocaine under test conditions that were in some ways similar to the PET session. Thus, the higher rCMRglc as compared with control values may ref lect activation in response to conditioned, drug-related cues (i.e. an injection given by an investigator who administered cocaine previously in a test situation). Nonetheless, substance abusers also exhibit higher normalized rCMRglc under conditions where there are no drug-related cues or expectancy of drug administration. A supplementary analysis of the data collected in our study of cue-elicited cocaine craving (Grant et al., 1996) compared rCMRglc in seven OFC brain regions of substance abusers and controls only in the session where a non-drug (arts and crafts) stimulus complex was presented. Drug abusers exhibited a significantly higher normalized rCMRglc than controls in the medial prefrontal cortex. Role of the OFC and its Connections in Drug Craving In addition to the aforementioned evidence that the OFC shows an abnormality in rCMRglc, which may be related to some degree to conditioned responses and motivation, findings obtained with PET and fMRI have suggested that activation of the OFC and its connections accompanies cocaine craving (Table 2). Although it is not definitively proven that the activation reported ref lects the mechanism by which craving is induced, the findings are consistent with an integrative function of the OFC, as supported by its anatomical connections. It is tempting to speculate that activation of cortical regions with sensory and limbic functions, along with activation of the OFC, ref lects an interplay of related networks. Thus, sensory information about environmental (or internal) stimuli might be interpreted at the level of the OFC, which connects with the amygdala in evaluating the motivational value of the stimuli, and labeling the emotional response as craving. One approach to the study of the neurobiological substrates of craving has involved the induction of craving by presentation of stimuli that have previously been associated with drug use. There is substantial evidence that exposure to cues, such as the paraphernalia used during drug self-administration, which are strongly associated with previous drug taking, can trigger craving for drugs of abuse (Childress et al., 1992). Our research group studied cue-elicited cocaine craving in cocaine abusers (n = 13) and control participants (n = 5), who were presented with a stimulus complex, including a cocainerelated videotape and paraphernalia in one session and a videotape and objects related to arts and crafts in another session (Grant et al., 1996). We also measured rCMRglc by the FDG method and PET. Exposure to cocaine-related cues produced self-reports of craving and activation in the medial OFC and five other composite cortical brain regions in cocaine abusers: dorsolateral prefrontal, peristriate, temporal/parietal, retrosplenial and temporal (see Fig. 2). Furthermore, self-reports of craving were positively correlated with metabolic increases in the dorsolateral prefrontal cortex, medial temporal lobe (amygdala) and cerebellum. These effects were not seen in non-drug-abusing control participants. The activation pattern seen in our study, involving the OFC and several areas to which it is linked anatomically, is consistent with an integrative function of OFC, involving brain areas that participate in sensory processing (peristriate cortex) and episodic memory (dorsolateral prefrontal cortex and temporal areas), as well as emotional coding (amygdala). Childress et al., by assay of regional cerebral blood f low (rCBF) with 15O-labeled water, also investigated the response to Table 2 Brain regions showing altered function during craving for drugs of abuse Type of study Study Treatment effects Correlations with craving Drug administration Cocaine-induced craving fMRI (Breiter et al., 1997) nucleus accumbens, subcallosal cortex, striatum, basal forebrain, thalamus, cingulate, parahippocampal gyrus, lateral prefrontal cortex, striate/extrastriate (increases) temporal pole, medial prefrontal cortex, amygdala (decreases) anterior cingulate, right thalamus, cerebellum nucleus accumbens, subcallosal cortex, parahippocampal gyrus, lateral prefrontal cortex (positive correlation) right amygdala (negative correlation) Methylphenidate-induced craving in cocaine abusers Cue-induced craving Cocaine cues PET rCMRglc (Volkow et al., 1999) Cocaine cues fMRI (Maas et al., 1998) Cocaine cues PET rCBF O15 (Childress et al., 1999) Cocaine cues PET rCMRglc (Wang et al., 1999) dorsolateral prefrontal cortex, medial OFC, temporal/parietal, dorsolateral prefrontal cortex, medial temporal lobe, peristriate, temporal, retrosplenial cerebellum (positive correlation) anterior cingulate, left dorsolateral prefrontal anterior cingulate, left dorsolateral prefrontal cortex (positive correlation) amygdala and anterior cingulate (increased) not reported basal ganglia (decreased) medial OFC, left insula, cerebellum right insula (positive correlation) PET rCMRglc (Volkow et al., 1991) PET rCMRglc (Grant et al., 1996) medial OFC medial prefrontal cortex (relative to control subjects) Spontaneous craving For cocaine For cocaine PET rCMRglc (Grant et al., 1996) right OFC, right striatum (positive correlation) striatum (relative to control subjects) anterior OFC (negative correlation) Figure 2. Images of rCMRglc, determined in a selected cocaine abuser (from a group with n = 13), during two test sessions in which either neutral (left) or cocaine-related (right) stimuli were presented (see Grant et al., 1996). The subjects abstained from cocaine use for at least 36 h before each assay of rCMRglc. The figure shows pseudo-colored metabolic (PET) images superimposed on the respective structural (MRI) images. The arrow points to the medial orbitofrontal cortex (MO), which exhibited a significant increase in rCMRglc during the active cues session. Subjects also showed increases of rCMRglc in other cortical regions (temporal lobe, TL; parahippocampal gyrus, PH; peristriate cortex, PS) and reported cocaine craving. In contrast to the metabolic increases in the cocaine abuser group, control subjects tended to show decreases in rCMRglc. exposure to cocaine-related cues (Childress et al., 1999). While watching a cocaine-related videotape, cocaine users experienced craving and showed a pattern of increase in limbic (amygdala and anterior cingulate) normalized rCBF and decreases in basal ganglia normalized rCBF relative to their responses to a neutral videotape (nature film). These findings generally supported our previous observations (1996) in demonstrating that limbic activation accompanies cue-induced cocaine craving, but did not show activation of the OFC or the dorsolateral prefrontal cortex. Differences in experimental conditions Cerebral Cortex Mar 2000, V 10 N 3 337 may account for the discrepancies. For example, the two studies differed in the time courses of the measurements (greater temporal resolution with rCBF than with rCMRglc), in the spatial resolution afforded by the respective measurements (greater spatial resolution with FDG than with 15O-labeled water), in the stimuli (shorter duration of the videotape, repeated more often in the rCMRglc than in the rCBF study) and in the characteristics of the research participants (cocaine abusers had previously demonstrated reactivity to the cues in the rCBF study but not the rCMRglc study). The issue of the anatomical substrates of cue-induced craving was also addressed with fMRI (Maas et al., 1998). Audiovisual stimuli containing alternating intervals of drug-related and neutral scenes were presented to six men with histories of crack cocaine use and six control participants. Activation was detected in the anterior cingulate and left dorsolateral prefrontal cortex in the cocaine users. In addition, self-reported levels of craving were correlated with activation in these regions. No data were presented for the OFC or the amygdala, which are difficult to evaluate with fMRI because of the susceptibility to artifacts in regions near the sinuses induced by the air-tissue interface. A pattern of activation that was generally consistent with the studies reviewed above was observed when drug craving was elicited by script-driven imagery. In an FDG–PET study, Wang et al. conducted interactive interviews in 13 active cocaine abusers about either neutral themes or cocaine-themes during the FDG uptake period (Wang et al., 1999). Cocaine-related paraphernalia were also presented to the subject during the cocaine-theme session. Increases in both absolute and normalized rCMRglc were seen in the medial OFC and left insular cortex, and increased normalized glucose metabolism was seen in the cerebellum. Since no control subjects were studied, it is possible that differences in the physical properties of the stimuli in the neutral versus cocaine-cues session could have contributed to some portion of the observed changes in cerebral metabolism. Self-reported cocaine craving was correlated with absolute and normalized rCMRglc in the right insular region. Finally, preliminary results have shown rCBF activation with PET in the basal ganglia, insula, cerebellum and left parahippocampal gyrus in heroin-dependent individuals who listened to personalized audiotapes (Weinstein et al., 1998). Subjects listened to six of their own audiotaped neutral scripts and six audiotaped craving scripts at random, each script lasting for 1.5 min, with a rest of 8 min between scripts (120 min total). The main regions of activation were the medial prefrontal gyrus and the anterior cingulate. Activation in the anterior cingulate was positively correlated with self-reports of craving. Craving induced by drug administration has been studied with both PET and fMRI (Breiter et al., 1997; Volkow et al., 1999). When cocaine was administered to cocaine-dependent subjects in an fMRI study, craving was correlated with early and sustained activation of regions that included the nucleus accumbens, left amygdala and right parahippocampal gyrus, and some regions of the lateral prefrontal cortex (Breiter et al., 1997). Sustained negative signal change in the amygdala was correlated with ratings of craving. Cocaine-induced perceptions of ‘rush’ were associated with short-term activation of the ventral tegmentum, pons, basal forebrain, caudate, cingulate and most regions of the lateral prefrontal cortex. In a FDG–PET study, craving for cocaine in cocaine abusers was elicited by injection of methylphenidate, which has a pharmacological mechanism of action similar to that of cocaine (Volkow, 1999). Cocaine craving was correlated with increased normalized rCMRglc in 338 Orbitofrontal Cortex and Drug Abuse • London et al. the posterior portion of the OFC and the striatum of the right hemisphere. Although the report on the fMRI study indicated no signal changes in the OFC, it is unclear whether this absence ref lected susceptibility artifacts that precluded imaging of this region, as in the study by Maas et al. (Maas et al., 1998). Even though these findings suggest that activation of the OFC is common to both cue- and drug-elicited craving, it is important to note that there are differences in the circuitry affected when craving is induced by cocaine itself or by cocaine-related cues. Drug-related cues activated the amygdala and medial temporal lobe, simulating the direct effects of cocaine. However, other regions, such as the nucleus accumbens, which presented no significant change in cue activation studies, showed direct pharmacological effects of cocaine. In addition to these findings, related to experimentally induced craving, findings from brain imaging studies where spontaneous craving was noted also showed involvement of the OFC. For example, in recently abstinent cocaine abusers, rCMRglc in the OFC and dorsoprefrontal area were correlated with retrospective self-reports of cocaine craving during the week that preceded the PET scan (Volkow et al., 1991). In contrast we observed a negative correlation of cocaine-craving with rCMRglc in the OFC in additional analysis of data from the study by Grant et al. (Grant et al., 1996). Here, increased levels of self-reported cocaine craving during presentation of neutral stimuli was related to lower levels rCMRglc in the right anterior OFC. One possible reason for the differences in direction of the correlations is that in the study by Grant et al., subjects had been abstinent from cocaine for only 48 h, whereas the subjects studied by Volkow were abstinent for up to 1 week. Another possibility may be related to the greater level of cocaine use in the subjects studied by Volkow. These data suggest that the OFC is involved in the induction of craving, at least for cocaine; however, methodological factors complicate the issue. Craving induced by cocaine itself, though perceived by users as the drive to acquire and use more drug, may be anatomically distinct from craving that is triggered by stimuli conditioned through repetitive drug use. Whereas humans may not be able to discriminate between these states, the observations that different neurobiological substrates are associated with these states may have profound implications for therapeutic interventions. The OFC and Decision Making in Drug Abusers In light of the aforementioned evidence for a functional contribution of the OFC to expectancy states and drug craving, it is of interest to determine how dysfunction of this brain region might also inf luence cognitive processes, particularly decision making in substance abusers. Despite the behavioral alterations associated with OFC lesions, patients with such brain damage failed to show impairments on classic neuropsychological tests of frontal lobe function (Stuss et al., 1998; Bechara et al., 2000). However, novel tasks that can serve as specific probes for the functional role of the OFC have been developed (Freedman et al., 1998; Bechara et al., 2000). As a first step toward determining whether drug abusers have a functional deficit involving the OFC, we have utilized a task that was specifically designed to evaluate human patients with lesions of the ventromedial portion of the OFC (VmOFC) Bechara et al., 2000). Patients with these lesions demonstrate generally irresponsible behaviors, including poor judgement in business and personal decisions. The Gambling Task was developed to capture these traits in a laboratory setting by requiring subjects to make choices based on the long-term consequences of complex reward/punishment contingencies. The task is performed as a card game in which the participant selects cards from four decks that differ with respect to the payoff for each card selection and the frequency and severity of penalties, indicated by certain cards. It has face validity for drug abuse since poor performance results from making choices that yield large short-term rewards even when such choices eventually lead to net losses. Indeed, one of the DSM IV criteria for substance abuse disorder states (on p. 181) that ‘substance abuse is continued despite knowledge of having a recurrent physical or psychological problem that is likely to have been caused or exacerbated by the substance’ (American Psychiatric Association, 1994). Similarly, the World Health Organization (1992) uses the criterion for disorders due to psychoactive substance use of ‘persisting . . . despite harmful consequences’ (p. 321). Although the Gambling Task models some aspects of drug abuse behavior, the balance between reward and negative consequences differs somewhat from the real life situation of the drug abuser. One difference is the fact that the certainty of reward is greater when considering self-administration of a drug than drawing a card in the model task. In our study, polydrug abusers were compared with a group of subjects who did not use illicit drugs of abuse (Grant et al., 2000). The scores on the GT were nearly 50% lower in drug abusers (n = 30) than in controls (n = 24), but not as low as those of patients with VmOFC lesions (Bechara et al., 1994). In contrast to GT results, there was no difference in performance between drug abusers and controls on the Wisconsin Card Sorting Task, which is sensitive to lesions of the dorsolateral prefrontal cortex but not the OFC (Stuss et al., 1998). These results suggest the working hypothesis that there is a functional abnormality, as opposed to a frank lesion, specifically in the ventromedial prefrontal cortex of drug abusers, but not necessarily in the frontal lobe as a whole. Further, they are consistent with the notion that the OFC is linked to the ability to balance short-term gains against longer-term consequences in the process of decision making. Other studies have also shown that drug abusers are likely to make maladaptive decisions when faced with short-term versus long-term outcomes, especially under conditions that involve risk and uncertainty. In one of these, opioid-dependent participants placed less value on delayed monetary rewards than control subjects (Madden et al., 1997), underscoring the concept that drug abusers are overly biased towards the immediate value of rewards and less able to evaluate the consequences of their actions in the future. In another study, amphetamine abusers, opiate abusers, non-drug-using comparison subjects and patients with VmOFC lesions were tested on a decision-making task similar to the Gambling Task (Rogers et al., 1999). Amphetamine users and lesion patients performed more poorly than subjects in the other groups, although the amphetamine users did not do as badly as the lesioned patients. Furthermore, there was a negative correlation between task performance and years of stimulant use. Interestingly, experimentally induced depletion of tryptophan, the precursor of the neurotransmitter serotonin, in non-drug-using comparison subjects led to impaired task performance. These results suggest that specific classes of drugs of abuse may have a differential impact on the processes underlying decision-making abilities, and the impact may be inf luenced by the drug’s effect on serotonergic, and perhaps other aminergic, neurotransmission. Our research group has begun a series of experiments to investigate whether impaired performance on the Gambling Task in drug abusers is directly due to dysfunction of the VmOFC. The aim of the first set of studies was to demonstrate that the VmOFC is involved in Gambling Task performance in neurologically intact subjects. We therefore conducted PET assays of rCMRglc during performance of the Gambling Task. Preliminary results from seven subjects indicate that the level of performance on the Gambling Task is positively correlated with activation of the VmOFC (Grant et al., 1999) (see Fig. 3). These results suggest that successful Gambling Task performance requires anatomical integrity as well as adequate metabolic activity of the VmOFC. Our results are convergent with those of previous neuroimaging studies showing activation of the ventromedial prefrontal cortex in subjects who had to guess about task-related reward contingencies (Elliott et al., 1997) or in subjects who experienced violations in expectancies (Nobre et al., 1999; Elliot et al., 2000). It is important to emphasize that these imaging studies do not preclude the participation of other brain regions in complex decision making, especially regions with anatomical connections with the VmOFC. For example, in our study, there was also a strong positive correlation (r = 0.93) between rCMRglc in the amygdala and Gambling Task performance. This is consistent with the recent demonstration of impaired performance on the Gambling Task in patients with amygdala lesions (Bechara et al., 1999). As reviewed in the previous section, work in our laboratory (Grant et al., 1996) and others (Table 2) has implicated the amygdala in behavioral processes related to addiction, such as craving. Having obtained a preliminary description of the brain circuitry associated with Gambling Task performance, we have initiated PET studies to test the hypotheses that during performance of the Gambling Task, drug abusers will exhibit little or no activation in the VmOFC or other relevant brain regions, such as the amygdala. Conclusion This review presents findings of neuroimaging studies that implicate the OFC in several behavioral aspects of substance abuse, including anticipation of drug reward, craving for the drug, and possible impairments in judgement that could inf luence the decision to abstain or take the drug. For example, cocaine-abusing research participants have manifested elevated normalized rCMRglc in the OFC when they were in a test situation where they anticipated possible cocaine administration but were drug-free. Furthermore, the OFC and some of its afferents (e.g. amygdala, insula) showed activation when drug abusers are presented with drug-related cues, such as videotapes or their own audiotapes relating their drug experiences, which induced drug craving. Craving induced by psychostimulant drug administration was also correlated with changes in signal measured from the amygdala in a fMRI study and with normalized rCMRglc in the OFC. Lastly, individuals with recent histories of drug abuse show impairment on cognitive tasks that require decision making, particularly the Gambling Task, their performance appearing to be related to activation of the OFC. The extent to which differences between substance abusers and naive control subjects, particularly related to the function of the OFC, ref lect a pre-existing condition, which confers vulnerability to addiction and promotes its perpetuation, is not known. Although unproven, it is reasonable to hypothesize that the negative consequences of drug abuse may include a negative impact on the function of the OFC. If this were the case, drug Cerebral Cortex Mar 2000, V 10 N 3 339 Figure 3. Saggital, transaxial and coronal sections through the VmOFC, and corresponding scores of two subjects who performed the Gambling Task. The subject, whose data are shown in the top panel, exhibited a marked increase in metabolic activity in the VmOFC and amygdala, and performed well as indicated by positive scores in the top right graph. The subject in the bottom panel exhibited a marked decrease in metabolic activity in the VmOFC and amygdala, and performed as poorly as subjects with VmOFC lesions, as indicated by negative scores in the bottom right graph. Images show color-coded differences in normalized glucose metabolism measured at two times: when the subject performed a control task and the Gambling Task. Metabolic data are superimposed on the subject’s MRI. The more red the color, the greater the decrease. Cross hairs show the plane of section. Overall performance on the Gambling Task during the PET session is broken down into 10 min blocks (100 cards/block). Negative numbers indicate poorer performance. abuse would damage the very substrates of brain function that are required for recovery. The initial impression from the work reviewed supports the view that the OFC, through its anatomical connections with sensory and limbic systems, serves as a critical ‘node’ in the processing of environmental and internal cues to generate feeling states that inf luence the motivation to seek and selfadminister a drug. Even if this inference ultimately were proven to be true, however, several important tasks would still lay before us. We would need to identify the neural networks that code for the different behavioral components of substance abuse, such as drug expectancy, craving, and impaired decision making. It then would be important to characterize the role of the OFC in these networks as primary or secondary; and to evaluate the contribution of possible dysfunction of the OFC to the deviant behaviors of substance abuse (e.g. necessary but insufficient, sufficient but not necessary, or insufficient and not necessar y). A final step would be identification of the 340 Orbitofrontal Cortex and Drug Abuse • London et al. neurochemical messengers that operate in the networks. Even before these tasks are tackled, however, it is important to consider the findings reviewed above in the design of therapeutic approaches for substance abuse. This instruction is particularly relevant to the implementation of a cognitive therapy that would require executive functioning that requires integrity of the OFC and its connections. Notes Address correspondence to Edythe D. London, PhD, Brain Imaging Center, NIDA, 5500 Nathan Shock Drive, Baltimore, MD 21224, USA. Email: [email protected]. References A lexander GE, Crutcher MD (1990) Functional architecture of basal ganglia circuits: neural substrates of parallel processing. Trends Neurosci 13:266–271. American Psychiatric Association (1994) Diagnostic and statistical manual of mental disorders, 4th edn. Washington, DC: American Psychiatric Association. Baker SC, Frith CD, Dolan RJ (1997) The interaction between mood and cognitive function studied with PET. Psychol Med 27:565–578. Barbas H, De Olmos J (1990) Projections from the amygdala to basoventral and mediodorsal prefrontal regions in the rhesus monkey. J Comp Neurol 300:549–571. Baxter LR (1990) Brain imaging as a tool in establishing a theory of brain pathology in obsessive compulsive disorder. J Clin Psychiat 51:22–25. Bechara A, Damasio AR, Damasio H, Anderson SW (1994) Insensitivity to future consequences following damage to human prefrontal cortex. Cognition 50:7–15. Bechara A, Damasio H, Damasio AR, Lee GP (1999) Different contributions of the human amygdala and ventromedial prefrontal cortex to decision-making. J Neurosci 19:5473–5481. Bechara A, Damasio H, Damasio AR (2000) Emotion, decision making, and the orbitofrontal cortex. Cereb Cortex 10:295–307. Benkelfat C, Nordahl TE, Semple WE, King C, Murphy DL, Cohen RM (1990) Local cerebral glucose metabolic rates in obsessive-compulsive disorder: patients treated with clomipramine. Arch Gen Psychiat 47: 840–848. Breiter HC, Rauch SL (1996) Functinal MRI and the study of OCD; from symptom provocation to cognitive-behavioral probes of corticostriatal systems and the amygdala. Neuroimage 3:S127–138. Breiter HC, Gollub RL, Weisskoff RM, Kennedy DN, Makris N, Berke JD, Goodman JM, Kantor HL, Gastfriend DR, Riorden JP, Mathew RT, Rosen BR, Hyman SE (1997) Acute effects of cocaine on human brain activity and emotion. Neuron 19:591–611. Carmichael ST, Price JT (1996) Connectional networks within the orbital and medial prefrontal cortex of macaque monkeys. J Comp Neurol 371:179–207. Carmichael ST, Clugneet MC, Price JL (1994) Central olfactory connections in the macaque monkey. J Comp Neurol 346:403–434. Cavada C, Compañy T, Tejedor J, Cruz-R izzolo RJ, Reinoso-Suárez F (2000) The anatomical connections of the macaque monkey orbitofrontal cortex. A review. Cereb Cortex 10:220–242. Childress AR, Ehrman R, Rohsenow DJ, Robbins SJ, O’Brien CP (1992) Classically conditioned factors in drug dependence. In: Substance abuse. A comprehensive textbook, 2nd edn (Lowinson JH, Ruiz P, Millman RB, Langrod JG, eds), pp. 56–69. Baltimore, MD: Williams & Wilkins. Childress AR, Mozley PD, McElgin W, Fitzgerald J, Reivich M, O’Brien CP (1999) Limbic activation during cue-induced cocaine craving. Am J Psychiat 156:11–18. Di Chiara G, Imperato A (1988) Drugs abused by humans preferentially increase synaptic dopamine concentrations in the mesolimbic system of freely moving rats. Proc Natl Acad Sci USA 85:5274–5278. Elliott R, Frith CD, Dolan RJ (1997) Differential neural response to positive and negative feedback in planning and guessing tasks. Neuropsychologia 35:1395–1404. Elliot R, Dolan RJ, Frith CD (2000) Neuroimaging and orbitofrontal cortex. Cereb Cortex 10:308–317. Fornazzari L, Farcnik K, Smith I, Heasman GA, Ichise M (1992) Violent visual hallucinations and aggression in frontal lobe dysfunction: clinical manifestations of deep orbitofrontal foci. J Neuropsychiat Clin Neurosci 4:42–44. Freedman M, Black S, Ebert P, Binns M (1998) Orbitofrontal function, object alternation and perseveration. Cereb Cortex 8:18–27. Grant S, London ED, Newlin DB, Villemagne VL, Liu X, Contoreggi C, Phillips RL, Kimes AS, Margolin A (1996) Activation of memory circuits during cue-elicited cocaine craving. Proc Natl Acad Sci USA 93:12040–12045. Grant SJ, Bonson KR, Contoreggi CC, London ED (1999) Activation of the ventromedial prefrontal cortex correlates with gambling task performance: a FDG–PET study. Abstr Soc Neurosci 25:1551. Grant SJ, Contoreggi CC, London ED (2000) Drug abusers show impaired performance in a laboratory test of decision making. Neuropsychologia (in press). Hof PR, Mufson EJ, Morrison JH (1995) Human orbitofrontal cortex; cytoarchitecture and quantitative immunohistochemical parcellation. J Comp Neurol 359:48–68. Koob GF, Bloom FE (1988) Cellular and molecular mechanisms of drug dependence. Science 242:715–723. London ED, Cascella NG, Wong DF, Phillips RL, Dannals RF, Links JM, Herning R, Grayson R, Jaffe JH, Wagner HN Jr (1990) Cocaine-induced reduction of glucose utilization in human brain. A study using positron emission tomography and [f luorine 18]-f luorodeoxyglucose. Arch Gen Psychiat 47:567–574. London ED, Grant SJ, Morgan MJ, Zukin SR (1996) Neurobiology of drug abuse. In: Neuropsychiatry: a comprehensive textbook (Fogel BS, Schiffer RB, Rao SM, eds), pp. 635–678. Baltimore, MD: William & Wilkins. Maas LC, Lukas SE, Kaufman MJ, Weiss RD, Daniels SL, Rogers V W, Kukes TJ, Renshaw PF (1998) Functional magnetic resonance imaging of human brain activation during cue-induced cocaine craving. Am J Psychiat 155:124–126. Madden GJ, Petry NM, Badger GJ, Bickel WK (1997) Impulsive and self-control choices in opioid-dependent patients and non-drug-using control participants: drug and monetary rewards. Exp Clin Psychopharmacol 5:256–262. Mesulam MM, Mufson EJ (1982a) Insula of the old world monkey. I. Architectonics in the insulo-orbito-temporal component of the paralimbic brain. J Comp Neurol 212:1–22. Mesulam MM, Mufson EJ (1982b) Insula of the old world monkey. II. A fferent cortical input and comments on the claustrum. J Comp Neurol 212:23–37. Mesulam MM, Mufson EJ (1982c) Insula of the old world monkey. III. Efferent cortical output and comments on function. J Comp Neurol 212:38–52. Meyers CA; Berman SA; Scheibel RS; Hayman A (1992) Case report: acquired antisocial personality disorder associated with unilateral left orbital frontal lobe damage. J Psychiat Neurosci 17:121–125. Morecraft RJ, Geula C, Mesulam MM (1992) Cytoarchitecture and neural afferents of orbitofrontal cortex in the brain of the monkey. J Comp Neurol 323:341–358. Nobre AC, Coull JT, Frith CD, Mesulam MM (1999) Orbitofrontal cortex is activated during breaches of expectation in tasks of visual attention. Nature Neurosci 2:11–12. Öngür D, Price JL (2000) Intrinsic and extrinsic connections of networks within the orbital and medial prefrontal cortex. Cereb Cortex 10:206–219. Pandya DN, Yeterian EH (1990) Prefrontal cortex in relation to other cortical areas in rhesus monkey: architecture and connections. Prog Brain Res 85:63–94. Pardo JV, Pardo PJ, Raichle ME (1993) Neural correlates of self-induced dysphoria. Am J Psychiat 150:713–719. Phelps ME, Huang SC, Hoffman EJ, Selin C, Sokoloff L, Kuhl DE (1979) Tomographic measurement of local cerebral glucose metabolic rate in humans with (F-18)2-f luoro-2-deoxy-D-glucose: validation of method. Ann Neurol 6:371–388. Rauch SL, Jenike MA, Alpert NM, Baer L, Breiter HC, Savage CR, Fischman AJ (1994) Regional cerebral blood f low measured during symptom provocation in obsessive-compulsive disorder using oxygen 15labeled carbon dioxide and positron emission tomography. Arch Gen Psychiat 51:62–70. Ray JP, Price JL (1993) The organization of projections from the mediodorsal nucleus of the thalamus to orbital and medial prefrontal cortex in macaque monkeys. J Comp Neurol 337:1–31. Reivich M, Kuhl D, Wolf A, Greenberg J, Phelps M, Ido T, Casella V, Fowler J, Hoffman E, Alavi A, Som P, Sokoloff L (1979) The [18F]f luorodeoxyglucose method for the measurement of local cerebral glucose utilization in man. Circ Res Jan 44:127–37. Rogers RD, Everitt BJ, Baldacchino A, Blackshaw AJ, Swainson R, Wynne K, Baker NB, Hunter J, Carthy T, Booker E, London M, Deakin JF, Sahakian BJ, Robbins TW (1999) Dissociable deficits in the decision-making cognition of chronic amphetamine abusers, opiate abusers, patients with focal damage to prefrontal cortex, and tryptophan-depleted normal volunteers: evidence for monoaminergic mechanisms. Neuropsychopharmacology 20:322–339. Rolls ET (2000) The orbitofrontal cortex and reward. Cereb Cortex 10:284–294. Schultz W, Tremblay L, Hollerman JR (2000) Reward processing in primate orbitofrontal cortex and basal ganglia. Cereb Cortex 10: 272–283. Semple WE, Goyer P, McCormick R, Morris E, Compton B, Muswick G, Nelson D, Donovan B, Leisure G, Berridge M, Miraldi F, Schulz SC (1993) Preliminary report: brain blood f low using PET in patients with posttraumatic stress disorder and substance-abuse histories. Biol Psychiat 34:115–118. Cerebral Cortex Mar 2000, V 10 N 3 341 Siever LJ, Buchsbaum MS, New AS, Speigel-Cohen J, Wei T, Hazlett EA, Sevin E, Nunn M, Mitropoulou V (1999) d,l-Fenf luramine response in impulsive personality disorder assessed with [18F]f luorodeoxyglucose positron emission tomography. Neuropsychopharmacology 20:413–423. Sokoloff L, Reivich M, Kennedy C, Des Rosiers MH, Patlak CS, Pettigrew KD, Sakurada O, Shinohara M (1977) The [14C]deoxyglucose method for the measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neurochem 28:897–916. Stapleton JM, Morgan MJ, Phillips RL, Wong DF, Yung BC, Shaya EK, Dannals RF, Liu X, Grayson RL, London ED (1995) Cerebral glucose utilization in polysubstance abuse. Neuropsychopharmacology 13: 21–31. Stapleton JM, Morgan MJ, Liu X, Yung BC-K, Phillips RL, Wong DF, Shaya EK, Dannals RF, London ED (1997) Cerebral glucose utilization is reduced in second test session. J Cereb Blood Flow Metab 17: 704–712. Stuss DT, Benson DF, Kaplan EF, Weir WS, Naeser MA, Lieberman I, Erril D (1998) The involvement of orbitofrontal cerebrum in cognitive tasks. Neuropsychologia 21:235–247. Volkow ND, Fowler JS, Wolf AP, Hitzemann R, Dewey S, Bendriem B, 342 Orbitofrontal Cortex and Drug Abuse • London et al. Alpert R, Hoff A (1991) Changes in brain glucose metabolism in cocaine dependence and withdrawal. Am J Psychiat 148:621–626. Volkow ND, Wang G-J, Fowler JS, Hitzemann R, Angrist B, Gatley SJ, Logan J, Ding Y-S, Pappas N (1999) Association of methylphenidateinduced craving with changes in right striato-orbitofrontal metabolism in cocaine abusers: implications in addiction. Am J Psychiat 156:19–26. Wang G-J, Volkow ND, Fowler JS, Cervany P, Hitzemann RJ, Pappas NR, Wong CT, Felder C (1999) Regional brain metabolic activation during craving elicited by recall of previous drug experiences. Life Sci 64: 775–784. Weinstein A, Feldtkeller B, Malizia A, Wilson S, Bailey J, Nutt D (1998) Integrating the cognitive and physiological aspects of craving. J Psychopharmacol 12:31–38. Wise R A (1996) Neurobiology of addiction. Curr Opin Neurobiol 6:243–251. World Health Organization (1992) International statistical classification of diseases and related health problems, 10th rev. Geneva: World Health Organization. Zald DH, Kim SW (1996) anatomy and function of the orbital frontal cortex, I. Anatomy, neurochemistry; and obsessive-compulsive disorder. J Neuropsychiat Clin Neurosci 8:125–138.