Skip Navigation
MUSC mobile menu
Illustration for Restoring Choice article

Restoring Choice

Two decades of MUSC research suggest that an over-the-counter antioxidant could help those recovering from addiction avoid relapse by controlling intrusive thoughts

By Tonisha Kearney-Ramos, Ph.D. and Nouhou Ibrahim, Ph.D.
Illustration by Emma Vought

Addiction robs its victims of free will, leaving compulsion in its place. Peter W. Kalivas, Ph.D., chair of the Department of Neuroscience (Research) at the Medical University of South Carolina (MUSC), has spent two decades elucidating the biological mechanisms underlying that compulsion and identifying a means to restore choice.

The answer could come in the form of an inexpensive, over-the-counter oral antioxidant called N-acetyl-cysteine (NAC), which he has shown in preclinical models to help prevent reinstatement of drug seeking by acting on a pathway whose importance in addiction he helped identify. Excitement over his findings has led to studies worldwide to test NAC’s efficacy in a wide variety of addictions and compulsive disorders.

Clinical investigators at MUSC have worked closely with Kalivas to show NAC’s potential for treating adolescent marijuana and alcohol abuse, smoking, cocaine addiction and for substance use that is comorbid with other compulsive disorders, such as post-traumatic stress disorder (PTSD). Collectively, their findings suggest that NAC, while not capable of causing an active user to cease taking drugs, can help those who are abstinent remain sober by reducing the intensity and frequency of intrusive, drug-seeking thoughts.

“NAC doesn’t cure addiction,” says Kalivas. “People still require therapy and have to relearn their life if they have been an addict for twenty years. But this drug will make it easier for them to relearn their life without drugs.”

Their work could one day lead to an inexpensive and easily administered adjunct therapy to low-intensity psychosocial treatments for addiction. With its low cost and good side effect profile at therapeutic doses, NAC should translate easily to the primary care clinic, where it can benefit the most people.

But none of this would have been possible without decades of patient and tenacious basic science research to elucidate the underlying mechanisms of drug addiction and identify a drug target.

Peter W Kalivas
Dr. Peter W. Kalivas has spent two decades elucidating the biological mechanisms underlying the intrusive thoughts characteristic of addiction.

Hijacking the brain’s reward system

The neurotransmitter dopamine is the primary currency of the brain’s reward system. An action that elicits an increase in dopamine is experienced as pleasurable. The brain “learns” to associate that action with reward. The pleasurable experience of natural rewards such as food and sex are signaled through dopamine release in the reward system as a way to ensure continued pursuit of actions that enable survival.

Drugs of abuse, however, put the reward system into overdrive by eliciting unnaturally high increases in dopamine release. As a result, the brain “overlearns” the association between taking the drug and reward. As that association is strengthened through continued drug use, it becomes increasingly difficult for the user to think of anything other than the drug.

Yet, while dopamine drives formation of the drug-reward associations, decades of research by Kalivas have shown that another neurotransmitter, glutamate, is also implicated in drug addiction. Typically, glutamate would be released in the synapse in response to cues that signal reward but would then be cleared by a glutamate transporter (GLT-1) — a protein that regulates glutamate signaling.1 This balanced release and clearance of glutamate facilitates the brain’s ability to fine-tune and adapt learned behaviors in response to environmental stimuli.

Kalivas has shown reduced numbers of GLT-1 transporters in animals using addictive drugs. As a result, the released glutamate is not cleared from the synapse and instead “spills over” into the extrasynaptic space, activating glutamate receptors in a broader area.

This spillover and broader activation of glutamate receptors cause the brain to attribute greater weight to actions related to drug use, thereby reducing the user’s willpower — the ability to choose other actions in the face of intrusive drug-related thoughts. Continued dominance of drug-related thoughts and actions precipitates the compulsive drug-seeking and use behaviors that characterize drug addiction.

Kalivas illustrates this hijacking of the brain by drug-related thoughts with an anecdote:

I’m at work and my wife calls me and says ‘Remember, I have to be out tonight, you’ve got to be home to cook dinner and help your daughter with homework.’ So, I’m driving home, then I get a call from a buddy who’s just arrived in town and everybody’s meeting up at my favorite bar downtown. Now I’ve got two things, my wife’s phone call and the phone call from my friend. So, what am I going to do? If I’m an addict, what happens is you’ve got these two competing plans, and pretty soon all I’m thinking about is that drinking plan, and the plan with my kids fades into the background. What we have been figuring out is how the drinking plan becomes the dominant, prepotent plan and causes the person to make the wrong choice — something that can be seen as a lack of willpower.

Target practice — Homing in on the NAC mechanism

The observation that the glutamate system was dysregulated in addiction sparked two decades of research by Kalivas. In a 1996 study published in the Journal of Neuroscience, Kalivas’ laboratory showed that repeated cocaine administration in rodents produced sustained effects on glutamate release in the reward system, which resulted in glutamate spillover at the synapse. These findings — the first clear indication of drug-mediated glutamate dysregulation of the reward system — identified the glutamate system as a potential therapeutic target.1

Kalivas and colleagues next pursued the source of the glutamate spillover. In 2002, they published a study revealing the involvement of the cystine/glutamate exchanger, a protein known to play a role in regulating extracellular glutamate levels. They showed that withdrawal from repeated cocaine administration resulted in down-regulated cystine/glutamate exchangers, significantly altering extracellular glutamate levels.2

This led them to hypothesize that targeting the cystine/glutamate exchanger using NAC, a cystine precursor, might be a potential way to rebalance glutamate homeostasis — the balance between glutamate release and clearance — that is disrupted
in addiction.

In 2003, Kalivas and colleagues published a study showing that restoration of cystine/glutamate exchange by NAC normalized the levels of glutamate in cocaine-treated rodents. On a behavioral level, reinstatement of drug seeking — the preclinical model of relapse — was prevented by stimulating cystine/glutamate exchange with NAC and restoring extracellular glutamate levels. This study demonstrated that it was possible to therapeutically alter the glutamate system and see a difference in addictive behavior.3

However, later work by Kalivas revealed that they may have been aiming at the wrong target. An article published in 2010 showed that repeated cocaine use reduces GLT-1, another regulator of extracellular glutamate, and that NAC inhibits cocaine seeking through restoring GLT-1 activity.4

The identification of an alternative pathway led Kalivas and colleagues to be less certain about their previously proposed mechanism of action for NAC in cocaine use. As a result, in 2015, Kalivas conducted a preclinical study directly aimed at disentangling whether NAC inhibits reinstatement of cocaine seeking through the restoration of GLT-1 and/or the cystine/glutamate exchanger. This was tested by determining whether blocking the ability of NAC to restore one or the other protein would prevent the ability of NAC to inhibit cocaine seeking. Kalivas discovered that restoring GLT-1, but not the cystine/glutamate exchanger, is the key mechanism by which daily NAC administration reduces cocaine reinstatement.5

Through many years of dedicated research and persistence in the face of misidentified scientific targets, Kalivas was able to uncover the mechanisms critical to an understanding of how NAC works in the glutamate system.

Taking NAC to rehab

Kalivas’s groundbreaking work — which could only have been accomplished in preclinical models — paved the way for human clinical studies by researchers at MUSC and elsewhere that have brought NAC to the forefront as a promising new treatment for those addicted to a wide variety of substances as well as those afflicted with other complex psychiatric disorders.


In 2007, Kalivas and Robert J. Malcolm, M.D., professor of Psychiatry, Family Medicine and Pediatrics at MUSC, published the first two human clinical studies investigating NAC for treatment in substance abuse. These small pilot studies suggested early promise for NAC use in cocaine users by demonstrating that short-term NAC treatment could stop or reduce cocaine use, as well as cocaine craving.6,7

However, it was not until years later in 2013 that they published the first large-scale, double-blind, placebo-controlled clinical trial that truly enabled them to investigate the efficacy of NAC treatment in substance users. Cocaine-dependent participants (n = 111) were randomized to receive daily doses of 1,200 mg of NAC, 2,400 mg of NAC or placebo. Participants were followed for eight weeks, with urine samples collected and tested for cocaine use at periodic visits.8

Overall, the primary results for the clinical trial were negative — they did not find reduced cocaine use for NAC-treated vs. placebo-treated participants. However, when only the subset of participants who had achieved abstinence at trial entry was considered, results favored the 2,400 mg NAC group relative to placebo, with the 2,400 mg group showing longer times to relapse and lower craving ratings.

While the trial failed to demonstrate that NAC could reduce cocaine use in actively using cocaine-dependent individuals, this was evidence that it could prevent relapse in individuals who had already achieved abstinence from cocaine by reducing intrusive thoughts (craving) about drug use.

“If you can get someone abstinent from cocaine and then put them on NAC, that is where it seems to help them over a placebo — as a relapse preventive agent but not a curative agent,” Malcolm explains.


Marijuana use has spiked among adolescents, and yet current cessation programs, which rely primarily on psychosocial treatments, have yielded low rates of abstinence. Currently, there is no FDA-approved pharmacotherapy for marijuana use disorder.

Evidence that NAC could play a role in treating substance abuse led Kevin M. Gray, M.D., a professor in the Department of Psychiatry and Behavioral Sciences at MUSC, and colleagues to conduct the first double-blind, randomized controlled trial to assess NAC as a potential pharmacotherapy for marijuana dependence in adolescents. The results of the study were published in the August 2012 issue of the American Journal of Psychiatry.9

Led by Gray, the eight-week trial randomized marijuana-dependent adolescents (aged 15-21) to either 1200 mg of NAC or placebo twice daily, along with twice-weekly contingency management (i.e., abstinence and adherence were rewarded with small cash payments) and a weekly cessation counseling session lasting less than ten minutes.

At the trial’s end, participants receiving NAC were more than twice as likely to have negative urine cannabinoid tests as those in the placebo group (adjusted odds ratio: 2.4). In the NAC group, 41 percent of participants had a negative urine screen on the last day of treatment vs. 27 percent in the placebo group.

“This is the first fully powered trial of a pharmacological agent conducted in any age group to have a positive finding as an adjunct to psychosocial treatment, and it happened to be in kids,” adds Gray.

Enthusiasm over these results in adolescents led the National Institute of Drug Addiction to fund Gray to lead a randomized controlled trial in adults, the findings of which he reported in April 2016 at the American Society of Addiction Medicine’s annual conference in Baltimore, MD. Unlike the adolescent trial, the six-site, 12-week trial, which enrolled 302 adults, showed that the NAC group was no more likely than the placebo group to have negative urine cannabinoid tests (unpublished data; manuscript under review).

“The effects of NAC did not seem to translate from adolescents to adults, and so I put my thinking cap on to see what the differences might be,” says Gray. “A key is that quantitative urine cannabinoid levels and years of regular marijuana use at baseline were much higher in the adult than in the adolescent trial. The threshold to have a treatment effect was much higher to reach. Neurochemically, we might have been doing the same thing, but maybe it wasn’t quite enough.” Medication adherence was also poorer in adult patients.

Subgroup analysis of participants aged 18-21 in the multisite trial, however, suggested that those taking NAC were twice as likely to have a negative test as those in the placebo group, similar to the findings reported in the 2012 article.

The next step for Gray is to replicate the adolescent trial using low-intensity cessation counseling but without contingency management. The low-intensity counseling mirrors what would be widely available in primary care clinics where many adolescents seek care. If the findings are positive, NAC should be poised to translate easily into front-line care as an adjunct therapy for marijuana use disorder.

In a secondary analysis of the data obtained from the first adolescent marijuana trial, Lindsay M. Squeglia, Ph.D., an assistant professor in the Department of Psychiatry and Behavioral Sciences, and Gray showed that, instead of compensating for reduced marijuana use by drinking more, the NAC-treated group actually decreased its alcohol use as well.10 This is particularly compelling given the participants were not actively attempting to reduce their alcohol use, nor were they engaged in a combined behavioral treatment for alcohol use. These findings suggest NAC effects may generalize from marijuana to other substances and could be useful in decreasing adolescent alcohol use.


MUSC investigators interested in tapping NAC’s potential to develop new therapies for nicotine addiction, including Gray, Brett E. Froeliger, Ph.D., of the Department of Neuroscience, and Erin A. McClure, Ph.D., of the Department of Psychiatry and Behavioral Sciences, noted that the design of clinical trials of NAC did not always map onto preclinical findings. Preclinical studies conducted by Kalivas and others had shown NAC to be effective at preventing reinstatement of drug-seeking behavior after a period of extinction, and yet many NAC trials were focused instead on cessation in active users. The MUSC team hypothesized that NAC would be more effective if given to smokers who had achieved at least a brief abstinence.

In 2015, they published the results of a small proof-of-concept trial that randomized 16 non-treatment-seeking smokers, who were paid to remain abstinent, to NAC or placebo for 3.5 days.11 On day four, participants underwent functional MRI imaging. The NAC group reported less craving than the placebo group, and neuroimaging suggested that NAC positively affected the dysregulated corticostriatal connectivity characteristic of addiction.

This early evidence of NAC’s potential efficacy in abstinent smokers led the group to wonder if it could prove a powerful adjunct therapy for varenicline (VAR; Chantix®, Pfizer, New York, NY), the front-line pharmacotherapy for smoking cessation.

“Compared to all smoking cessation therapies, VAR is far and away the best, but the majority of smokers that use VAR still relapse,” says Froeliger. “It helps people quit (smoking) but does not do as good of job at keeping them off cigarettes.”

Since VAR and NAC use different mechanisms of action, the nicotine addiction research team at MUSC hypothesized that combining VAR, which acts on the cholinergic pathway to promote cessation, and NAC, which acts on the glutamate pathway to prevent relapse, could synergistically promote long-term abstinence at higher rates in smokers than either medication alone.

In 2015, they published the results of a pilot trial in 19 adult cigarette smokers that suggested the combination regimen was safe — most side effects were mild and in line with what would be expected of the individual agents — and well tolerated and that it was feasible to take it forward into clinical trial.12

The group then launched a clinical trial of the combination therapy in 2015 that will eventually enroll 100 smokers. Now in year two, the trial has already enrolled a quarter of those patients.

To ensure that clinical trial design is aligned with preclinical findings of efficacy and that preclinical studies are informed by real-world results, the project’s principal investigators include Kalivas, who will focus on understanding the neurobiological basis of NAC/VAR in a rodent model; Froeliger, a cognitive neuroscientist who will focus on imaging of human brains to explore whether NAC/VAR works on similar circuits in humans as in rodents; and Gray, who will lead the clinical trial. The ultimate goal is to compile the preliminary data needed to take the combination therapy forward into a much larger randomized clinical trial that would be powered to show efficacy and potentially gain FDA approval for a new smoking cessation approach.

PTSD with a comorbid substance use disorder

NAC may also have a role to play in the treatment of psychiatric disorders characterized by intrusive thoughts, such as PTSD, in which the glutamate pathway is dysregulated, as in addiction. Because NAC acts to restore balance to the glutamate system that is disrupted in both addiction and PTSD, researchers at MUSC hypothesized that it could potentially benefit patients with addiction and comorbid PTSD. Currently, there are no well-explored pharmacological treatments for patients with co-occurring addiction and PTSD, a particularly difficult-to-treat population.

Sudie E. Back, Ph.D., a professor in the Department of Psychiatry and Behavioral Sciences at MUSC and a staff psychologist at the Ralph H. Johnson VA Medical Center, randomized 35 veterans with an alcohol or drug use disorder and comorbid PTSD, all of whom were receiving group cognitive behavioral therapy (CBT) for their addiction, to either 2400 mg/day of NAC or placebo for eight weeks.11 To be included, veterans had to have abstained from substance use for at least seven days. This was the first randomized controlled trial of NAC as a pharmacotherapy for PTSD and a broad range of comorbid addictions.

NAC plus CBT reduced symptoms of PTSD, cravings and depression significantly more than CBT alone in these veterans. Veterans in the NAC-treated group showed a 46 percent reduction in scores on the Clinical-Administered PTSD Scale (CAPS), a gold-standard measure of PTSD symptoms, vs. a 25 percent reduction in the placebo group. The threshold CAPS score for diagnosis of PTSD is 50 and “as a group, the NAC-treated veterans were below diagnostic level for PTSD at the end of treatment,” says Back. “These are some of the best outcomes we have seen in the literature for a medication.”

The NAC-treated group saw greater decreases than the placebo group in craving (81 vs. 32 percent) and depression (48 vs. 15 percent). “Craving is a key component of substance use and relapse,” says Back. “If you have a medication that can really reduce craving, that will go a long way to helping people stay clean and sober.”

It is important to note that both groups received CBT and that NAC appears to work best in those who are at least briefly abstinent. As such, NAC should not be used as a monotherapy or substitute for evidence-based behavioral treatment but instead be seen as a medication that can be added to therapy to help patients with addiction and PTSD attain positive outcomes.

The positive results seen in the pilot study of NAC among veterans with PTSD and addiction led the Department of Defense and the National Institute on Alcohol Abuse and Alcoholism to fund Back, Gray, Kalivas and colleagues at MUSC to conduct two larger-scale, randomized controlled trials in veterans and civilians with alcohol use disorders and PTSD. Both projects were initiated in 2016 and are currently enrolling participants.

Can we cure addiction?

Two decades of preclinical research with NAC is coming to clinical fruition. Collectively, these trials suggest that adding NAC to standard-of-care psychosocial therapies could help prevent relapse in those who have achieved at least short-term abstinence from a variety of addictive substances. And yet, NAC is far from a cure for addiction.

“To think that we will cure drug addiction tomorrow is to say we will get in a space shuttle and colonize the moons of Saturn. We can conceive it, we can almost taste it, we can see how we could get in orbit and maybe do it. But it’s really quite a long way away — probably decades away,” explains Kalivas. “And that’s where we are with the brain. We have to understand how the brain works to develop the really precise interventions that will cure psychiatric disorders. For the brain, that’s 20 billion cells and 100 trillion connections. That will take time, and we will just never get there without preclinical research.”


1 Pierce RC, et al. J Neurosci 1996;15:1550-1560.

2 Baker DA, et al. J Neurosci 2002;22:9134-9141.

3 Baker DA, et al. Nat Neurosci 2003;6(7):743-749.

4 Knackstedt LA, et al. Biol Psychiatry 2010;67(1):81-84.

5 Reissner KJ, et al. Addict Biol 2015;20(2):316-323.

6 Mardikian PN, et al. Prog Neuropsychopharmacol Biol Psychiatry 2007;31(2):389-394.

7 LaRowe SD, et al. Am J Psychiatry 2007;164(7):1115-1117.

8 LaRowe SD, et al. Am J Addict 2013;22(5):443-452.

9 Gray KM, et al. Am J Psychiatry 2012 Aug;169(8):805-812.

10 Squeglia LM, et al. Addict Behav 2016 Dec;63:172-177.

11 Froeliger B, et al. Drug Alcohol Depend 2015;156:234-242.

12 McClure EA, et al. Am J Drug Alcohol Abuse 2015 Jan;41(1):52-56.

13 Back SE, et al. J Clin Psychiatry 2016 Nov;77(11):e1439-e1446.