Cerebrum Article

Searching for a Drug to Extinguish Fear

Psychotherapists often work for years to reduce patients’ anxiety, whether that anxiety has taken the form of panic attacks, a phobia such as fear of heights, obsessive-compulsive disorder, or post-traumatic stress disorder. Things may be easier for future practioners. Building on new understanding of how the brain can be trained to “extinguish” learned fears, scientists are finding promising evidence that a medication now used to treat tuberculosis can dramatically accelerate cognitive-behavioral therapy for many crippling fears and phobias.

Published: July 1, 2005

People with severe obsessive-compulsive disorder (OCD) may spend an hour or more each morning trying to decide whether to step out of bed with their left foot first or their right foot first, and people with post-traumatic stress disorder (PTSD) can have the same nightmares night after night, year after year. Someone who has had a panic attack and has become terrified of having another one may be housebound for years. People with fear of heights often drive miles out of their way every day to avoid driving over a high bridge, and people with social phobia, characterized by extreme anxiety about being judged by others or behaving in a way that might cause embarrassment or ridicule, will not only avoid parties but will take only jobs in which they never have to speak in public.

Anxiety disorders such as these are the most common type of psychiatric illness, affecting an estimated 19 million children and adults. All are painful and disruptive, and many of the physical symptoms associated with them are similar, including trembling, heart palpitations, chest pain or discomfort, sweating, insomnia, abdominal upsets, dizziness, and irritability. But the specific characteristics and triggers can be quite different.

Generalized anxiety disorder is characterized by excessive, often unrealistic worry that can last for months or years. The worry may be about money, career, or some other aspect of life, but it is out of proportion to the actual situation. People with OCD have often irrational, recurring thoughts—obsessions—such as excessive worry about germs. These obsessional thoughts may lead them to compulsive behavior such as washing their hands 20 times a day. For people who have panic attacks, in most cases their first episode occurs “out of the blue”; in the grip of an attack, a person feels he may die, cannot catch his breath, and is afraid he will lose control, which leads to an overall sense of unreality. Although both genetic and environmental influences appear to be at work, the underlying causes of many of these disorders are still largely unknown.

On the other hand, PTSD, some specific phobias, and avoidance behavior in people with panic disorder seem to result from a process known as fear conditioning, in which a person learns to associate a particular stimulus or place with an aversive event. Thus, PTSD can develop after a traumatic event, such as a sexual or physical assault or witnessing a natural or manmade disaster (for example, 9/11). People with PTSD often “relive” the trauma through flashbacks and nightmares and avoid the place where the trauma took place. This pattern of avoidance can go from specific to general, so that, for example, a woman will first avoid the alley where she was assaulted, then avoid going out at night, and then entirely stop leaving home. A similar pattern happens with people who have panic attacks: they first avoid the place where the panic attack took place but eventually generalize to the point where they will not leave home.

Certain phobias also are clearly tied to a prior traumatic event, for example, being bitten by a dog or falling into deep water as a child. In these situations, the anxiety disorder depends on fear conditioning. The assault, dog bite, or near drowning is the aversive event, and the place where the assault took place or the sight of the dog or of the water is the stimulus (called the conditioned stimulus) that brings the aversive event to mind. Although some phobias are difficult to connect directly to a particular traumatic event, psychiatrists still generally believe children learn the fear reaction— often from their parents’ own fears or their parents’ reactions to situations children get themselves into. For example, a mother’s fear of spiders may be unwittingly transferred to her child, or a father’s reaction to seeing his child about to fall off the top of a table may lead the child to become afraid of heights.

All of these anxiety disorders are painful and disruptive, not only to the people experiencing them but also to their families. Although certain forms of therapy are often successful in treating anxiety disorders, strategies to make therapy even more effective would be enormously beneficial. As researchers learn more about the neurobiological processes involved in conditioned fear and in its extinction, they are beginning to see the real possibility of pharmaceutical assistance in retraining the fearful brain.

Facing and Eliminating Fear through Therapy

Even though anxiety disorders have different causes and triggering stimuli, they all respond to a form of psychotherapy called cognitive-behavioral therapy (CBT). A critical component of CBT is getting patients to face their fears, to talk about and imagine what makes them afraid. This technique, usually called “exposure therapy,” has several variants. In one type of exposure therapy, called “systematic desensitization,” the patient is asked to visualize the traumatic event (or stimuli that remind the patient of the traumatic event), while at the same time trying to remain calm and relaxed, using deep breathing and reassuring imagery, such as thinking about sitting before a cozy fire. The patient initially tries to visualize the least fearful aspects of the traumatic memory or its associated cues, then to face increasingly more and more fearful aspects. At each step, the therapist watches the patient and gets feedback to see whether he is coping with the cues or images, before the therapist allows the patient to take on more fearful ones. Thus, systematic desensitization uses a graduated approach to the traumatic cues in combination with a relaxation strategy.

Another variety of CBT is called “flooding,” in which the full traumatic event is imagined in the presence of a supportive therapist. Initially, this technique elicits lots of anxiety, which then eventually subsides. This process is repeated in the presence of the therapist, who often suggests alternative ways for the patient to think about what is happening. Eventually, the patient is able to imagine the entire traumatic event with little or no anxiety, although reaching this point can take a long time.

Central to all of these variants of exposure therapy is the repetitive recall of a traumatic event or cues that remind a patient of that event. For example, people with panic attacks are taken back to the place where they first had such an attack to show them it will not always happen in that place. Or people with OCD may be exposed to dirty towels but not allowed to wash their hands, again in concert with a supportive therapist.

The process of exposure therapy is similar to a procedure called extinction studied extensively by animal psychologists. In fact, a considerable problem in PTSD and certain other types of anxiety disorders is the inability to suppress or inhibit terrible memories, which is, in essence, a failure of extinction. For this reason, researchers are focusing in both animals and humans on the way in which unwanted memories are normally inhibited and the reasons for difficulty in doing so after traumatic fear conditioning. 

How the Brain Learns to Extinguish Fear

Inhibition of acquired fear, whether in humans or animals, is studied in the laboratory with a simple procedure called extinction training.1 First, the animal or person is conditioned to fear some normally neutral stimulus, such as a light or tone, by pairing it with an aversive stimulus, such as a mild shock. After the conditioning, the new fear stimulus is presented repeatedly in the absence of the shock. This process results in the gradual decline and ultimate disappearance of the fear response, because the subject learns that the stimulus no longer predicts that the aversive event will follow. The result is known as extinction.

It is essential to realize that extinction is a form of new learning in its own right.

It is not an “unlearning” or a forgetting of previous learning. After extinction training, fear memories can return over time (spontaneous recovery), when the fear stimulus is presented in a place different from the place where extinction training took place (renewal), or after an intervening stress (reinstatement). In each of these instances, the reemergence of the fear response indicates that fear was not actually lost through extinction, but, rather, that fear was actively suppressed through an additional learning process. Therefore, extinction is considered a form of acquired inhibition. When acquired, it counteracts or suppresses fear responses that are no longer adaptive because the stimulus that elicits the response is no longer associated with an aversive consequence. 

Converging evidence from many different laboratories indicates that the amygdala, located in the brain’s temporal lobe, is critically involved in both the formation and expression of aversive memories.2 The amygdala receives highly processed sensory information of all kinds and communicates it to the many parts of the brain involved in all aspects of fear and anxiety. Researchers studying fear conditioning in rats discovered that when the animal was exposed to a light or a tone paired with a shock, a protein in the amygdala called the NMDA receptor was activated. This allowed calcium to enter neurons in the amygdala, triggering a complex pattern of intracellular changes. These changes, in turn, led to long-term structural changes, allowing conditioned fear to become more or less permanent. Researchers have also learned that lesions or inactivation of the amygdala in animals blocks both the acquisition and expression of conditioned fear.

Normally, when people look at pictures of scary faces, remember traumatic events, or perceive a cue previously paired with shock, blood flow in the amygdala increases. This does not happen with people who have damage to the amygdala. They have difficulty detecting fear in other peoples’ faces and do not find even sinister-looking faces to be potentially dangerous but instead see everyone as trustworthy.3

Much less is known about the neural underpinnings of fear extinction. It has been established, however, that both extinguishing and acquiring fear depend on the NMDA receptors within the amygdala. This discovery was made by William Falls, Ph.D., Mindy Miserendino, Ph.D, and me. We reported that when we infused the amygdala with a compound that interferes with activity of this receptor shortly before extinction training, this blocked the process of extinction to a degree directly related to the size of the dose.4 This effect could not be attributed to the compound’s action on NMDA receptors outside the amygdala, to damage to or destruction of the amygdala, or to impairment of sensory transmission during extinction training. Additional studies using systemic administration of other compounds that block this receptor confirmed this effect. Furthermore, blocking NMDA receptors after extinction training also blocks extinction, suggesting these receptors are important not just for extinction per se but also for making the effects of the extinction long-lasting, a process called consolidation.5

Enhancing Extinction

Because the process of extinguishing the fear response depends on the NMDA receptor, it was logical to ask if enhancing the functioning of that receptor would also enhance extinction. My colleagues and I were aware that a compound called D-cycloserine binds to the NMDA receptor and makes it work better. We decided to test this compound by giving it to laboratory rats before extinction training.

In a series of experiments, we administered D-cycloserine either systemically or directly into the rats’ amygdala before extinction training and then tested retention of extinction the next day.6 As we predicted, when we gave the drug in combination with repeated exposure to the feared stimulus (a light) without a shock, the process of extinction was enhanced. This enhancement did not occur with control rats that received the drug alone, without extinction training. From this experiment, we concluded that the positive effects of the D-cycloserine were specifically connected with extinction and did not result from a general dampening of fear expression.

Further information about the effects of D-cycloserine has continued to accumulate. Other researchers found that the compound, given either systemically or directly into the amygdala, facilitates extinction of a fear response called conditioned freezing.

Most interestingly, the drug still facilitates extinction even when it is given up to about three hours after extinction training, 7 a result consistent with the idea that it helps with consolidation of extinction. More recently, that same research group (Lana Lederwood, Ph.D., Jacelyn Cranney, Ph.D., and Rick Richardson, Ph.D.) investigated the effects of D-cycloserine on processes that normally disrupt extinction. A new stress given after extinction training can reinstate the fear response, and control rats in their study that were given a foot shock exhibited the typical return of conditioned fear. Rats previously treated with D-cycloserine, however, showed much less reinstated fear even after they received the unexpected shock.8

Surprisingly, both our laboratory and the Richardson laboratory found that D-cycloserine only enhances extinction. It does not facilitate fear conditioning and, in fact, may actually interfere with it. Why might that be so?

D-cycloserine is an analogue of (that is, has a similar structure and function as) a naturally occurring chemical in the brain, D-serine. D-serine and glycine bind to the same site on the NMDA receptor as D-cycloserine does and make the receptor work better. So it is possible that NMDA receptors involved in fear conditioning are already saturated with D-serine or glycine, making these receptors work optimally. This optimal functioning would be adaptive, because it is important for any living creature to learn quickly what stimuli are dangerous so as to avoid them in future. If the receptors are already saturated, D-cycloserine would not be able to have any further effect. In fact, D-cycloserine is actually less effective than either D-serine or glycine. Thus, if the site on the NMDA receptors involved in fear conditioning is fully saturated, D-cycloserine might actually reduce the activity of the receptors by displacing the more effective endogenous chemicals already there. This could explain why D-cycloserine appears to actually inhibit fear conditioning.

But how does this theory explain the ability of D-cycloserine to facilitate extinction? Most likely the NMDA receptors involved in extinction are different from those involved in fear conditioning (for example, the two types of receptors might be on different neurons), or perhaps the receptors involved in extinction are not saturated with glycine or D-serine. This would suggest that these particular NMDA receptors do not work as efficiently, an explanation for why extinction takes much longer to develop than does fear conditioning. Because these receptors are not already saturated with natural chemicals, the effect of giving D-cycloserine would be to facilitate NMDA transmission and, therefore, extinction.

But Will it Work in People?

In tests, when people with anxiety disorders such as PTSD are presented with reminders of their trauma, their amygdalas show much greater activation than when patients who also experienced a traumatic event but did not develop PTSD are given reminders of their trauma.9 For this reason, researchers hypothesized that inappropriate and excessive fear in humans results from abnormal fear-learning processes that may reflect irregularities in the circuitry of the amygdala and related structures.

As mentioned, treatments for PTSD and other anxiety disorders typically involve a process similar to extinction. The patient who repeatedly recalls fearful events or cues related to them in the presence of a supportive therapist can be seen as having a roughly parallel experience to the experimental animal that has been trained to associate a light with a shock and is then exposed to the light without a shock following it. Although the patient may become anxious—and, in fact, must become at least somewhat anxious for therapy to work—this anxiety eventually declines with repeated sessions, just as the rat’s fear reaction to the light lessens. Because exposure-based psychotherapy for fear disorders is so similar to extinction, understanding the biological process of extinction in animals should help scientists develop better techniques and tools for human treatment. Given the positive effect of D-cycloserine on facilitating extinction in rats, my research group decided to investigate whether it might facilitate the loss of fear in human patients.

Kerry Ressler, M.D., Ph.D., Barbara Rothbaum, Ph.D., I, and others conducted a study to evaluate the usefulness of D-cycloserine as a pharmacologic adjunct to exposure therapy for people with an inordinate fear of heights (acrophobia).10 Although fear of heights is not a major psychiatric disorder like schizophrenia or severe depression, it is debilitating to people who have it. For example, they may be so terrified that they avoid working in an office or staying in a hotel if they have to use an elevator, or they may drive miles out of the way to avoid an elevated roadway. We could immediately test the possibility of using D-cycloserine because, like many other medications, it has different effects at low and high doses. At low doses, it acts at the glycine regulatory site on the NMDA receptor to make it work better. Because the NMDA receptor has been implicated in all types of learning, D-cycloserine has been used as a cognitive enhancer in autism, Alzheimer’s disease, and schizophrenia, but with little success because patients rapidly show a decreasing response to repeated doses (tolerance). At high doses, it has antibacterial effects and has been used to treat tuberculosis and has shown few side effects.

Scientists at Emory University used computer simulation to give patients with acrophobia the virtual experience of rising up in a glass elevator and, as shown here, looking down from the top of the elevator ride to the bottom of the building. Gradually extinguishing their fear by means of exposure therapy was much more effective and lasting when they were given D-cycloserine before the therapy sessions. Image courtesty of Michael Davis

Although the antibacterial effects have nothing do with the process of extinction, the use of D-cycloserine for many years as a well-tolerated drug gave us confidence in exploring its use in a potential new human application. Moreover, because we wanted to give D-cycloserine only infrequently— that is, just before therapy—we thought we could avoid the problem of tolerance.

We used an unusual form of exposure therapy for the subjects with acrophobia in our study: a virtual reality situation developed by Barbara Rothbaum and her colleagues in which patients ride in a computer-simulated glass elevator to progressively higher floors. This situation is frightening to patients just entering treatment but becomes considerably less so with increasing exposure to the virtual environment, typically about six to eight sessions.

For our study, we matched 30 participants for their initial fear of heights and other variables and gave them only two exposure training sessions in the virtual reality set-up, separated by one or two weeks. The participants took single doses of either a placebo or D-cycloserine (50 or 500 mg) two to four hours before each of the two sessions; neither the therapist nor the patients knew which medication had been taken.

During the sessions, the participants reported their level of discomfort at each floor of the virtual elevator, and, in addition, we measured what are called “spontaneous galvanic skin conductance fluctuations,” a physiological indication of overall anxiety, at each session. Both measures were also taken at a one-week follow-up session, and the patients’ level of discomfort was assessed as well at a three-month follow-up session. Participants also were asked to report the number of times they exposed themselves to real-life height encounters during the three-month period after the exposure training.

The results were gratifying. When tested in the virtual elevator set-up at both one week and three months after the initial training sessions, participants who took the D-cycloserine experienced significantly greater improvement in their acrophobia symptoms than participants who received the placebo. They also showed significantly greater decreases when we measured their skin conductance fluctuations. Perhaps most important, these study participants also reported greater improvement in their real-world acrophobia. This improvement was evident early on and was still present when measured at three months, using scales that rated behaviors and other aspects of the phobia, such as avoidance of situations that might cause acrophobia, anxiety about the acrophobia, attitudes toward heights, overall clinical improvement, and the number of times patients chose to experience real-world heights.

The Wider Possibilities

Ours was a small study, which needs to be replicated. Also, we tested people with a specific phobia, so it remains to be seen whether D-cycloserine will improve cognitive-behavioral therapy for more complex disorders, such as PTSD. Since we first reported the results of this clinical trial in 2003, however, other research groups have begun to combine D-cycloserine with exposure-based psychotherapy for the treatment of social phobia, OCD, panic disorder, PTSD, and even anorexia. Already, encouraging reports have appeared in connection with social phobia. We plan to try D-cycloserine in people with fear of public speaking, using virtual reality that involves exposure to a “virtual audience.” Thus, in a few years, we may know whether this new method will be useful not only for simple phobias but also for more complex anxiety disorders.

The discovery that D-cycloserine facilitated exposure therapy for people with acrophobia raises several exciting possibilities. First, combining this or similar medications with psychotherapy may offer patients with all kinds of phobias (and, perhaps, more complex anxiety disorders, such as PTSD) a greater likelihood of overcoming their fears with as little stress as possible during therapy and of maintaining that improvement over time. Second, seeing the benefits of D-cycloserine for the human participants in our study confirms what has been learned through basic research about the role of NMDA receptors in extinction in rodents and extends the principle identified in that research to humans. Finally, this is a classic example of translational research. Basic knowledge of the physiologic processes underlying fear and fear extinction in animals has been translated into improving a whole category of therapies for the many disorders in which abnormal fear runs roughshod over people’s lives.


  1. Myers, KM, and Davis, M. “Behavioral and neural analysis of extinction: A Review.” Neuron 2002; 36: 567-584.
  2. Aggleton, JP. (Ed.) The Amygdala (Vol. 2). Oxford, United Kingdom. Oxford University Press, 2000.
  3. Adolphs, R, Tranel, D, and Damasio, AR. “The human amygdala in social judgment.” Nature 1998; 393: 470-474.
  4. Falls, WA, Miserendino, MJ, and Davis, M. “Extinction of fear-potentiated startle: blockade by infusion of an NMDA antagonist into the amygdala.” Journal of Neuroscience 1992; 12(3): 854-863.
  5. Santini, E, Muller, RU, and Quirk, GJ. “Consolidation of extinction learning involves transfer from NMDA- independent to NMDA-dependent memory.” Journal of Neuroscience 2001; 21(22): 9009-9017.
  6. Walker, DL, Ressler, KJ, Lu, K-T, et al. “Facilitation of conditioned fear extinction by systemic administration or intra-amygdala infusions of D-cycloserine as assessed with fear-potentiated startle in rats.” Journal of Neuroscience 2002; 22: 2343-2351.
  7. Ledgerwood, L, Richardson, R, and Cranney, J. “D-cycloserine facilitates extinction of conditioned fear as assessed by freezing in rats.” Behavioral Neuroscience 2003; 117: 341-349.
  8. Ledgerwood, L, Richardson, R, and Cranney, J. “D-cycloserine and reinstatement.” Behavioral Neuroscience 2004; 118: 505-513.
  9. Rauch, SL, Shin, LM, and Wright, CI. “Neuroimaging studies of amygdala function in anxiety disorders.” Annals of the New York Academy of Science 2003; 985: 389-410.
  10. Ressler, KJ, Rothbaum, BO, Tannenbaum, L, et al. “Cognitive enhancers as adjuncts to psychotherapy:  Use of D-cycloserine in phobics to facilitate extinction of fear.” Archives of General Psychiatry 2004; 61: 1136-1144.