Addiction and the Prefrontal Cortex
An Interview with Rita Z. Goldstein, Ph.D.

by Brenda Patoine

April, 2007

Q: What is I-RISA and how does it apply to the study of addiction?

A: Together with Dr. Nora D. Volkow, we have recently emphasized that human drug addiction can be characterized by Impaired Response Inhibition and Salience Attribution (I-RISA), where the motivation to procure drugs overpowers the drive to attain most other non-drug-related goals. In this model, we mapped the core clinical symptoms in drug addiction, including craving, or “drug wanting” to the brain mechanisms that underlie the ability to control behavior, especially in an emotionally salient (e.g., drug-related) context. In this model we postulated that drug-addicted individuals attribute excessive salience (i.e., importance, relevance) to the drug and drug-related cues. At the same time, insufficient salience is attributed to non-drug-related reinforcers, stimuli such as food or social relationships that increase the probability of a subsequent behavior. We further hypothesized that this change in salience attribution, which is modulated by prefrontal cortical brain regions, would be predictive of impaired control of behavior (impulsivity).

This I-RISA model advances the notion that drug addiction cannot be fully understood without looking beyond the “pleasure principle,” the classical brain-reward circuit. The circuit encompasses subcortical regions such as the ventral tegmental area, where dopamine, a neurotransmitter critically implicated in drug self-reinforcement, is manufactured, and the nucleus accumbens, where dopamine is released. We also emphasized the importance of cortical brain regions within this reward circuit, particularly the prefrontal cortex (PFC), which includes the orbitofrontal cortex (OFC), and anterior cingulate cortex (ACC).

These PFC regions are involved in higher-order cognition (e.g., decision making) and emotion (regulation). In our I-RISA model we have specifically implicated these PFC regions in determining the salience of a given reinforcer. Knowing this, it may be possible to increase the salience of a non-drug reinforcer (e.g., the prospect of being employed) to buffer a strong emotional response to a drug-related reinforcer. These notions are consistent with a modern view of dopamine function that advances its role beyond reward to salience and novelty processing.

Q: What have you learned about how drug-addicted individuals respond to non-drug rewards?

A: In our translational research we use a combination of cognitive tests, self-report questionnaires, and brain-recording tools. For example, we record study participants’ behavioral responses on reaction-time tasks. We ask subjects how they think/feel about certain stimuli by using subjective rating scales. Finally, we image subjects’ brain structures and functions while they are performing these tasks, using functional magnetic resonance imaging (fMRI), positron emission tomography (PET), or event-related potential (ERP) recordings.

Preliminary results from our laboratory using this multimodal research approach(1) indicate differences in responses to non-drug rewards as a function of addiction. We found that when drug-addicted individuals think about a hypothetical situation during which they are “under the influence,” the importance of a drug reward exceeds

that of other primary reinforcers, such as food. Using a self-report instrument we’ve recently developed, we also asked subjects to rank their feelings of “wanting” vs. “liking” the drug while they were thinking about this hypothetical situation. Overall, cocaine-addicted individuals—but not healthy controls—ranked “wanting” a drug higher than “liking” a drug. Furthermore, the addicted subjects with the highest “wanting” over “liking” rankings had both a higher frequency of recent drug use and greater reactivity to drug cues. This supports the fact that they indeed may be craving the drug (or at least unable to ignore it) even when the drug is no longer pleasurable.

Our results further suggest a compromise in the ability to process the relative value of secondary non-drug-related rewards (e.g., money). In a group of 16 individuals with cocaine-use disorders, nine (56%) subjects demonstrated decreased sensitivity to differences between levels of abstract monetary rewards. When asked to rate seven monetary amounts ($10, $20, $50, $100, $200, $500, $1000) on a scale of 0 (not at all valuable) to 10 (most valuable), these subjects rated $10 to be equally valuable to $1000; all amounts received a rating of 10. Only 2 of 13 (15%) control subjects demonstrated this flattened sensitivity to monetary reward, a statistically significant group difference (Figure 1A). In the drug-addicted subjects, the PFC response (as measured with fMRI) accounted for 85% of the variability in this compromised sensitivity to monetary reward. In particular, this compromise in subjective sensitivity to relative monetary reward was paralleled by the OFC response to money: while OFC activity monotonically increased in the healthy control subjects, its activity was reduced and not linear in the cocaine subjects (Figure 1B). Importantly, we did not ask our study volunteers to choose between $10 and $1000; instead we asked them about the subjective value of these amounts.

This result makes sense if one thinks about the desire to use drugs in drug addiction: even very small amounts of money can bring an individual closer to this goal. It still remains to be determined whether this compromised sensitivity to gradients in reward predicts choice behavior—whether it will predict more severe drug use symptomatology, for example. This may indeed be the case because if the relative context of reward is compromised, the addicted individual may be more amenable to making disadvantageous decisions such as trading something of high personal value for the opportunity to get high.


Q: How does this relate to the propensity for drug relapse?

A: Our results also point to a disrupted perception of inner motivational drives (or the inability to translate perception into action), which could contribute to impairments in self-control in the drug-addicted individuals. Thus, while healthy control subjects were able to modify behavior based on the perceived relative value of a reward, drug-addicted individuals were not able to do so. This impairment may represent not only a compromise in perceiving the value of a reward, but also in utilizing this knowledge to modify behavior.

These results suggest an underlying cognitive-emotional mechanism in drug-related situations: when the value of a drug stimulus is higher than all other available rewards, which are perceived as equally less important than the drug, the ability to use non-drug reinforcement to control drug-taking behavior would necessarily be compromised. This would predispose an individual to relapse and drug use. Indeed, preliminary studies from other laboratories suggest that, in initially abstinent drug-addicted subjects, stronger drug cue- or stress-induced brain activations in the PFC during the early abstinence period are predictive of earlier or more severe relapse.

Q: What is your current thinking regarding the neural mechanisms underlying this flattened sensitivity to non-drug rewards?

A: Individuals with lesions to certain regions of their PFC, including the OFC, have difficulties in modifying behavior appropriately in response to altered reinforcement situations in their environment. Similarly, drug-addicted individuals also have PFC structural changes (e.g., reduced volumes), OFC and ACC functional changes (e.g., increased response when craving), and parallel behavioral changes (e.g., increased impulsivity). These findings led us to ask what role the OFC and ACC play in the drug-addicted individual’s ability to modify behavior based on the salience and value of a given reinforcer.

We think that drug addiction may be better understood as a disorder of neural regulation. Here’s why: even though the OFC and ACC are not sufficiently engaged in the processing of non-drug-related rewards, they are activated—in addicted individuals but not controls—in response to drug-related cues (e.g., words/pictures/videos of drug taking or pharmacologically similar drugs). Indeed, our preliminary fMRI results suggest that a possible communication breakdown between PFC sub-regions (OFC and dorsolateral PFC) may underlie the disrupted perception of motivational drive and the impaired control of behavior that characterized the drug-addicted individuals in our study.

Q: What does this work suggest in terms of clinical implications for treating drug addiction?

A: Consistent with the compulsive and chronically relapsing nature of drug addiction, our findings may help explain why efforts to control addiction through reinforcement can be compromised. It is possible that instead, efforts should be focused on devising new training and skill-development strategies and on supervised pharmacological interventions, all with the goal of decreasing the reinforcing effects of the drug, enhancing the relative value attributed to non-drug-related rewards, and increasing control of behavior. Together, these approaches may enhance the ability to control drug-taking behavior even in situations when the desire for the drug exceeds that for other rewards.

More research is required to delineate the optimal treatment approaches. As we acquire basic knowledge about addiction-related brain circuits and their interaction with environmental variables, dual approaches pairing behavioral interventions with medications will likely offer new and effective treatments for drug addiction and its associated neurobiological changes. For example, one could conceive of interventions designed to “exercise” brain circuits using specific cognitive and behavioral therapies to remediate and strengthen the circuits affected by chronic drug use, analogous to some interventions currently used for reading disabilities and traumatic brain injury. Such dual interventions that specifically activate and strengthen circuits involved in inhibitory control and salience attribution (e.g., the PFC) may increase successful abstinence from drug taking.

Q: What are the next steps in your research?

A: In our next project we target the I-RISA model using a newly developed fMRI task, the drug Stroop task, to directly test the effect of salient cues on inhibitory control in cocaine-addicted individuals. While participants perform this fMRI task, we plan to administer a pharmacological challenge (methylphenidate, or Ritalin) that increases extracellular dopamine and enhances the striatal-PFC activity that marks an event as salient. We will test whether response to this pharmacological salience enhancement predicts clinical outcome at follow-up. Results of this study may be helpful in devising intervention strategies to counteract the overwhelming salience that drugs of abuse have on addicted individuals, with the goal of minimizing relapse. This study will also allow us to directly probe the dopaminergic circuit in human subjects addicted to cocaine, to be accomplished for the first time with pharmacological fMRI.

In addition, we plan to test the predictive utility of our self-report instruments vis-à-vis choice behavior. We want to know if drug-addicted individuals who report wanting drugs more than liking drugs and who show flattened sensitivity to non-drug-related rewards would choose drugs over other salient reinforcers (e.g., money) more frequently and/or despite severe consequences.

Lastly, our interest in personality traits (e.g., a tendency to avoid harm vs. approach risk) has led us to examine genetic vulnerabilities in drug-addicted individuals. Can we associate heightened I-RISA risk with a modified genotype? If so, what are the gene candidates most related to the underlying neurocognitive I-RISA mechanisms? A better understanding of the interactions between genes, environment, and neurobiology may offer new targets for the development of pharmacological and non-pharmacological interventions.

Such future studies could help elucidate the following questions: did the neurocognitive impairments develop secondary to drug abuse and addiction, or were they a predisposing factor? The answer probably lies between these two possibilities and varies among individuals. Most importantly, can we identify susceptible individuals before addiction develops, thus preventing the onset of this vicious cycle? And, can we offer the intense treatment needed to individuals at highest risk for the most severe forms of addiction, reducing the high morbidity and mortality associated with this chronic disease?

1 I want to acknowledge the contribution to our studies of many talented and dedicated researchers, including Nelly Alia-Klein, Dardo Tomasi, Patricia Woicik, and Thomas Maloney from the Neuropsychoimaging group and also Frank Telang, Gene-Jack Wang, Joanna Fowler, Chris Wong, and numerous others from the PET group.