Vaughan & Asburys General Ophthalmology, 17th ed.

Chapter 9. Anxiety disorders and anxiolytics

   Symptom dimensions in anxiety disorders

    When is anxiety an anxiety disorder?

    Overlapping symptoms of major depression and anxiety disorders

    Overlapping symptoms of different anxiety disorders

   The amygdala and the neurobiology of fear

   Cortico-striato-thalamo-cortical (CSTC) loops and the neurobiology of worry

   GABA and benzodiazepines

    GABAA receptor subtypes

    Benzodiazepines as positive allosteric modulators or PAMs

    Benzodiazepines as anxiolytics

   Alpha-2-delta ligands as anxiolytics

   Serotonin and anxiety

   Noradrenergic hyperactivity in anxiety

   Fear conditioning versus fear extinction

    Fear conditioning

    Novel approaches to the treatment of anxiety disorders

   Treatments for anxiety disorder subtypes

    Generalized anxiety disorder

    Panic disorder

    Social anxiety disorder

    Posttraumatic stress disorder

   Summary

This chapter will provide a brief overview of anxiety disorders and their treatments. Included here are descriptions of how the anxiety disorder subtypes overlap with each other and with major depressive disorder. Clinical descriptions and formal criteria for how to diagnose anxiety disorder subtypes are mentioned only in passing. The reader should consult standard reference sources for this material. The discussion here will emphasize how discoveries about the functioning of various brain circuits and neurotransmitters – especially those centered on the amygdala – impact our understanding of fear and worry, symptoms that cut across the entire spectrum of anxiety disorders.

The goal of this chapter is to acquaint the reader with ideas about the clinical and biological aspects of anxiety disorders in order to understand the mechanisms of action of the various treatments for these disorders discussed along the way. Many of these treatments are extensively discussed in other chapters. For details of mechanisms of anxiolytic agents used also for the treatment of depression (i.e., certain antidepressants), the reader is referred to Chapter 7; for those anxiolytic agents used also for chronic pain (i.e., certain anticonvulsants), the reader is referred to Chapter 10. The discussion in this chapter is at the conceptual level, and not at the pragmatic level. The reader should consult standard drug handbooks (such as Stahl's Essential Psychopharmacology: the Prescriber's Guide) for details of doses, side effects, drug interactions, and other issues relevant to the prescribing of these drugs in clinical practice.

Symptom dimensions in anxiety disorders

When is anxiety an anxiety disorder?

Anxiety is a normal emotion under circumstances of threat and is thought to be part of the evolutionary “fight or flight” reaction of survival. Whereas it may be normal or even adaptive to be anxious when a saber-tooth tiger (or its modern-day equivalent) is attacking, there are many circumstances in which the presence of anxiety is maladaptive and constitutes a psychiatric disorder. The idea of anxiety as a psychiatric disorder is evolving rapidly, and is characterized by the concept of core symptoms of excessive fear and worry (symptoms at the center of anxiety disorders in Figure 9-1), compared to major depression, which is characterized by core symptoms of depressed mood or loss of interest (symptoms at the center of major depressive disorder in Figure 9-1).



Figure 9-1. Overlap of major depressive disorder and anxiety disorders. Although the core symptoms of anxiety disorders (anxiety and worry) differ from the core symptoms of major depression (loss of interest and depressed mood), there is considerable overlap among the rest of the symptoms associated with these disorders (compare the “anxiety disorders” puzzle on the right to the “MDD” puzzle on the left). For example, fatigue, sleep difficulties, and problems concentrating are common to both types of disorders.

Anxiety disorders have considerable symptom overlap with major depression (see those symptoms surrounding core features shown in Figure 9-1), particularly sleep disturbance, problems concentrating, fatigue, and psychomotor/arousal symptoms. Each anxiety disorder also has a great deal of symptom overlap with other anxiety disorders (Figures 9-2 through 9-5). Anxiety disorders are also extensively comorbid, not only with major depression, but also with each other, since many patients qualify over time for a second or even third concomitant anxiety disorder. Finally, anxiety disorders are frequently comorbid with many other conditions such as substance abuse, attention deficit hyperactivity disorder, bipolar disorder, pain disorders, sleep disorders, and more.



Figure 9-2. Generalized anxiety disorder (GAD). The symptoms typically associated with GAD are shown here. These include the core symptoms of generalized anxiety and worry as well as increased arousal, fatigue, difficulty concentrating, sleep problems, irritability, and muscle tension. Many of these symptoms, including the core symptoms, are present in other anxiety disorders as well.



Figure 9-3. Panic disorder. The characteristic symptoms of panic disorder are shown here, with core symptoms of anticipatory anxiety as well as worry about panic attacks. Associated symptoms are the unexpected panic attacks themselves and phobic avoidance or other behavioral changes associated with concern over panic attacks.



Figure 9-4. Social anxiety disorder. Symptoms of social anxiety disorder, shown here, include the core symptoms anxiety or fear over social performance plus worry about social exposure. Associated symptoms are panic attacks that are predictable and expected in certain social situations as well as phobic avoidance of those situations.



Figure 9-5. Posttraumatic stress disorder (PTSD). The characteristic symptoms of PTSD are shown here. These include the core symptoms of anxiety while the traumatic event is being re-experienced as well as worry about having the other symptoms of PTSD, such as increased arousal and startle responses, sleep difficulties including nightmares, and avoidance behaviors.

So, what is an anxiety disorder? These disorders all seem to maintain the core features of some form of anxiety or fear coupled with some form of worry, but their natural history over time shows them to morph from one into another, to evolve into full syndrome expression of anxiety disorder symptoms (Figure 9-1) and then to recede into subsyndromal levels of symptoms only to reappear again as the original anxiety disorder, a different anxiety disorder (Figures 9-2 through 9-5), or major depression (Figure 9-1). If anxiety disorders all share core symptoms of fear and worry (Figures 9-1 and 9-6) and, as we shall see later in this chapter, are all basically treated with the same drugs, including many of the same drugs that treat major depression, the question now arises, what is the difference between one anxiety disorder and another? Also, one could ask, what is the difference between major depression and anxiety disorders? Are all these entities really different disorders, or are they instead different aspects of the same illness?



Figure 9-6. Anxiety: the phenotype. Anxiety can be deconstructed, or broken down, into the two core symptoms of fear and worry. These symptoms are present in all anxiety disorders, although what triggers them may differ from one disorder to the next.

Overlapping symptoms of major depression and anxiety disorders

Although the core symptoms of major depression (depressed mood or loss of interest) differ from the core symptoms of anxiety disorders (fear and worry), there is a great deal of overlap with the other symptoms considered diagnostic both for a major depressive episode and for several different anxiety disorders (Figure 9-1). These overlapping symptoms include problems with sleep, concentration, and fatigue as well as psychomotor/arousal symptoms (Figure 9-1). It is thus easy to see how the gain or loss of just a few additional symptoms can morph a major depressive episode into an anxiety disorder (Figure 9-1) or one anxiety disorder into another (Figures 9-2 through 9-5).

From a therapeutic point of view, it may matter little what the specific diagnosis is across this spectrum of disorders (Figures 9-1 through 9-5). That is, first-line psychopharmacological treatments may not be much different for a patient who currently qualifies for a major depressive episode plus the symptom of anxiety (but not an anxiety disorder) versus a patient who currently qualifies for a major depressive episode plus a comorbid anxiety disorder with full criteria anxiety symptoms. Although it can be useful to make specific diagnoses for following patients over time and for documenting the evolution of symptoms, the emphasis from a psychopharmacological point of view is increasingly to take a symptom-based therapeutic strategy to patients with any of these disorders because the brain is not organized according to the DSM, but according to brain circuits with topographical localization of function. That is, specific treatments can be tailored to the individual patient by deconstructing whatever disorder the patient has into a list of the specific symptoms a given patient is experiencing (see Figures 9-2 through 9-5), and then matching these symptoms to hypothetically malfunctioning brain circuits regulated by specific neurotransmitters in order to rationally select and combine psychopharmacological treatments to eliminate all symptoms and get the patient to remission.

Overlapping symptoms of different anxiety disorders

Although there are different diagnostic criteria for different anxiety disorders (Figures 9-2 though 9-5), these are constantly changing, and many do not even consider obsessive–compulsive disorder to be an anxiety disorder any longer (OCD is discussed in Chapter 14 on impulsivity). All anxiety disorders have overlapping symptoms of anxiety/fear coupled with worry (Figure 9-6). Remarkable progress has been made in understanding the circuitry underlying the core symptom of anxiety/fear based upon an explosion of neurobiological research on the amygdala (Figures 9-7 through 9-14). The links between the amygdala, fear circuits, and treatments for the symptom of anxiety/fear across the spectrum of anxiety disorders are discussed throughout the rest of this chapter.



Figure 9-7. Linking anxiety symptoms to circuits. Anxiety and fear symptoms (e.g., panic, phobias) are regulated by an amygdala-centered circuit. Worry, on the other hand, is regulated by a cortico-striato-thalamo-cortical (CSTC) loop. These circuits may be involved in all anxiety disorders, with the different phenotypes reflecting not unique circuitry but rather divergent malfunctioning within those circuits.



Figure 9-8. Affect of fear. Feelings of fear are regulated by reciprocal connections between the amygdala and the anterior cingulate cortex (ACC) and the amygdala and the orbitofrontal cortex (OFC). Specifically, it may be that overactivation of these circuits produces feelings of fear.



Figure 9-9. Avoidance. Feelings of fear may be expressed through behaviors such as avoidance, which is partly regulated by reciprocal connections between the amygdala and the periaqueductal gray (PAG). Avoidance in this sense is a motor response and may be analogous to freezing under threat. Other motor responses are to fight or to run away (flight) in order to survive threats from the environment.



Figure 9-10. Endocrine output of fear. The fear response may be characterized in part by endocrine effects such as increases in cortisol, which occur because of amygdala activation of the hypothalamic–pituitary–adrenal (HPA) axis. Prolonged HPA activation and cortisol release can have significant health implications, such as increased risk of coronary artery disease, type 2 diabetes, and stroke.



Figure 9-11. Breathing output. Changes in respiration may occur during a fear response; these changes are regulated by activation of the parabrachial nucleus (PBN) via the amygdala. Inappropriate or excessive activation of the PBN can lead not only to increases in the rate of respiration but also to symptoms such as shortness of breath, exacerbation of asthma, or a sense of being smothered.



Figure 9-12. Autonomic output of fear. Autonomic responses are typically associated with feelings of fear. These include increases in heart rate (HR) and blood pressure (BP), which are regulated by reciprocal connections between the amygdala and the locus coeruleus (LC). Long-term activation of this circuit may lead to increased risk of atherosclerosis, cardiac ischemia, change in BP, decreased HR variability, myocardial infarction (MI), or even sudden death.



Figure 9-13. The hippocampus and re-experiencing. Anxiety can be triggered not only by an external stimulus but also by an individual's memories. Traumatic memories stored in the hippocampus can activate the amygdala, causing the amygdala, in turn, to activate other brain regions and generate a fear response. This is termed re-experiencing, and it is a particular feature of posttraumatic stress disorder.



Figure 9-14. Linking anxiety symptoms to circuits to neurotransmitters. Symptoms of anxiety/fear are associated with malfunctioning of amygdala-centered circuits; the neurotransmitters that regulate these circuits include serotonin (5HT), γ-aminobutyric acid (GABA), glutamate, corticotropin-releasing factor (CRF), and norepinephrine (NE), among others. In addition, voltage-gated ion channels are involved in neurotransmission within these circuits.

Worry is the second core symptom shared across the spectrum of anxiety disorders (Figure 9-7). This symptom is hypothetically linked to the functioning of cortico-striato-thalamo-cortical (CSTC) loops. The links between the CSTC circuits, “worry loops,” and treatments for the symptom of worry across the spectrum of anxiety disorders are discussed later in this chapter (see also Figures 9-15 through 9-179-26, and 9-29). We shall see that what differentiates one anxiety disorder from another may not be the anatomical localization, or the neurotransmitters regulating fear and worry in each of these disorders (Figures 9-6 and 9-7), but the specific nature of malfunctioning within these same circuits in various anxiety disorders. That is, in generalized anxiety disorder (GAD), malfunctioning in the amygdala and CSTC worry loops may be hypothetically persistent, and unremitting, yet not severe (Figure 9-2), whereas malfunctioning may be theoretically intermittent but catastrophic in an unexpected manner for panic disorder (Figure 9-3) or in an expected manner for social anxiety (Figure 9-4). Circuit malfunctioning may be traumatic in origin and conditioned in posttraumatic stress disorder (PTSD: Figure 9-5).



Figure 9-15. Linking worry symptoms to circuits to neurotransmitters. Symptoms of worry are associated with malfunctioning of cortico-striato-thalamo-cortical (CSTC) loops, which are regulated by serotonin (5HT), γ-aminobutyric acid (GABA), dopamine (DA), norepinephrine (NE), glutamate, and voltage-gated ion channels.



Figure 9-16. Worry/obsessions circuit. Shown here is a cortico-striato-thalamo-cortical (CSTC) loop originating and ending in the dorsolateral prefrontal cortex (DLPFC). Overactivation of this circuit may lead to worry or obsessions.



Figure 9-17. COMT genetics and life stressors. Activity in the cortico-striato-thalamo-cortical (CSTC) loop may vary during cognitive tasks depending on the variant of catechol-O-methyl-transferase (COMT) that an individual has (upper portion of figure). Thus, those with the Met genotype for COMT (i.e., those who have lower COMT activity and thus higher dopamine levels) may have “normal” activation and no problems with performance during a cognitive task, whereas those with the Val genotype may exhibit inefficiency of cognitive information processing, require overactivation of this circuit, and potentially make more errors during the same task. These latter individuals may also be at increased risk for schizophrenia. Similarly, the variant of COMT that an individual has may affect response to stress, since the CSTC loop also regulates worry. In this case, however, the beneficial genotype may be reversed. That is, because individuals with the Met genotype have lower COMT activity and thus higher dopamine levels, dopamine release in response to stress may be excessive and contribute to worry and risk for anxiety disorders. Those with the Val genotype, on the other hand, may be less reactive to stress because COMT can destroy the excess dopamine.

The amygdala and the neurobiology of fear

The amygdala, an almond-shaped brain center located near the hippocampus, has important anatomical connections that allow it to integrate sensory and cognitive information and then determine whether there will be a fear response. Specifically, the affect or feeling of fear may be regulated via reciprocal connections that the amygdala shares with key areas of prefrontal cortex that regulate emotions, namely the orbitofrontal cortex and the anterior cingulate cortex (Figure 9-8). However, fear is not just a feeling. The fear response can also include motor responses. Depending upon the circumstances and one's temperament, those motor responses could be fight, flight, or freezing in place. Motor responses of fear are regulated in part by connections between the amygdala and the periaqueductal gray area of the brainstem (Figure 9-9).

There are also endocrine reactions that accompany fear, in part due to connections between the amygdala and the hypothalamus, causing changes in the HPA (hypothalamic–pituitary–adrenal) axis, and thus of cortisol levels. A quick boost of cortisol may enhance survival when encountering a real but short-term threat. However, chronic and persistent activation of this aspect of the fear response can lead to increased medical comorbidity, including increased rates of coronary artery disease, type 2 diabetes, and stroke (Figure 9-10), and potentially to hippocampal atrophy as well (discussed in Chapter 6 and shown in Figure 6-39B). Breathing can also change during a fear response, regulated in part by the connections between amygdala and the parabrachial nucleus in the brainstem (Figure 9-11). An adaptive response to fear is to accelerate respiratory rate when having a fight/flight reaction to enhance survival, but in excess this can lead to unwanted symptoms of shortness of breath, exacerbation of asthma, or a false sense of being smothered (Figure 9-11), all symptoms common during anxiety, and especially during attacks of anxiety such as panic attacks.

The autonomic nervous system is attuned to fear, and is able to trigger responses from the cardiovascular system such as increased pulse and blood pressure for fight/flight reactions and survival during real threats. These autonomic and cardiovascular responses are mediated by connections between the amygdala and the locus coeruleus, home of the noradrenergic cell bodies (Figure 9-12; noradrenergic neurons are discussed in Chapter 6, and noradrenergic pathways and neurons are illustrated in Figures 6-25 through 6-30, and also Figure 6-32). When autonomic responses are repetitive and inappropriately or chronically triggered as part of an anxiety disorder, this can lead to increases in atherosclerosis, cardiac ischemia, hypertension, myocardial infarction, and even sudden death (Figure 9-12). “Scared to death” may not always be an exaggeration or a figure of speech! Finally, anxiety can be triggered internally from traumatic memories stored in the hippocampus and activated by connections with the amygdala (Figure 9-13), especially in conditions such as posttraumatic stress disorder.

The processing of the fear response is regulated by the numerous neuronal connections flowing into and out of the amygdala. Each connection utilizes specific neurotransmitters acting at specific receptors (Figure 9-14). What is known about these connections is that not only are several neurotransmitters involved in the production of symptoms of anxiety at the level of the amygdala, but numerous anxiolytic drugs have actions upon these specific neurotransmitter systems to relieve the symptoms of anxiety and fear (Figure 9-14). The neurobiological regulators of the amygdala, including the neurotransmitters GABA, 5HT, and NE, the voltage-gated calcium channels, and anxiolytics that act upon these neurotransmitters in order to mediate their therapeutic actions, are specifically discussed later in this chapter.

Cortico-striato-thalamo-cortical (CSTC) loops and the neurobiology of worry

Dopamine and born worried?

The second core symptom of anxiety disorders, worry, involves another unique circuit (Figure 9-15). Worry, which can include anxious misery, apprehensive expectations, catastrophic thinking, and obsessions, is linked to cortico-striato-thalamo-cortical (CSTC) feedback loops from the prefrontal cortex (Figures 9-15 and 9-16). Some experts theorize that similar CSTC feedback loops regulate the related symptoms of ruminations, obsessions, and delusions, all of these symptoms being types of recurrent thoughts. Several neurotransmitters and regulators modulate these circuits, including serotonin, GABA, dopamine, norepinephrine, glutamate, and voltage-gated ion channels (Figure 9-15). These overlap greatly with many of the same neurotransmitters and regulators that modulate the amygdala (Figure 9-14). Since various genotypes for the enzyme COMT (catechol-O-methyl-transferase) regulate the availability of one of these neurotransmitters, namely dopamine, in the prefrontal cortex, differences in dopamine availability may impact the risk for worry and anxiety disorder and help to determine whether you are “born worried” and vulnerable to developing an anxiety disorder, particularly under stress (Figure 9-17).

Warriors versus worriers

In Chapter 4 the impact of genetic variants of COMT on cognitive functioning are mentioned in relation to schizophrenia. Specifically, normal controls with the Met variant of COMT have more efficient information processing in the dorsolateral prefrontal cortex (DLPFC) during a cognitive task such as the n-back test. These subjects have lower COMT activity, higher dopamine levels, and presumably better information processing during tasks of executive functioning that recruit circuits in the DLPFC. Because of more efficient cognitive information processing, such subjects also have a lower risk for schizophrenia than subjects who are Val carriers of COMT (Figure 4-44).

At first glance, it would seem that all the biological advantages go to those with the Met variant of COMT. However, that is not necessarily true when it comes to processing stressors that cause dopamine release. With the Met genotype and its low COMT activity and high dopamine levels, stressors can hypothetically produce too much dopamine activity, which actually decreases the efficiency of information processing under stress and creates the symptoms of anxiety and worry (“worriers”). Under stress, therefore, it appears that Val carriers of COMT with their higher enzyme activity and lower dopamine levels are hypothetically able to handle the increased dopamine release that comes with stress by optimizing their information processing; thus they are “warriors” who are not afraid or worried when stressed. Dopamine is just one of the potential regulators of worry circuits and CSTC loops.

GABA and benzodiazepines

GABA (γ-aminobutyric acid) is one of the key neurotransmitters involved in anxiety and in the anxiolytic action of many drugs used to treat the spectrum of anxiety disorders. GABA is the principal inhibitory neurotransmitter in the brain and normally plays an important regulatory role in reducing the activity of many neurons, including those in the amygdala and those in CSTC loops. Benzodiazepines, perhaps the best-known and most widely used anxiolytics, act by enhancing GABA actions at the level of the amygdala and at the level of the prefrontal cortex within CSTC loops to relieve anxiety. To understand how GABA regulates brain circuits in anxiety, and to understand how benzodiazepines exert their anxiolytic actions, it is important to understand the GABA neurotransmitter system, including how GABA is synthesized, how GABA action is terminated at the synapse, and the properties of GABA receptors (Figures 9-18 through 9-24).



Figure 9-18. Gamma-aminobutyric acid (GABA) is produced. The amino acid glutamate, a precursor to GABA, is converted to GABA by the enzyme glutamic acid decarboxylase (GAD). After synthesis, GABA is transported into synaptic vesicles via vesicular inhibitory amino acid transporters (VIAATs) and stored until its release into the synapse during neurotransmission.



Figure 9-19. Gamma-aminobutyric acid (GABA) action is terminated. GABA's action can be terminated through multiple mechanisms. GABA can be transported out of the synaptic cleft and back into the presynaptic neuron via the GABA transporter (GAT), where it may be repackaged for future use. Alternatively, once GABA has been transported back into the cell, it may be converted into an inactive substance via the enzyme GABA transaminase (GABA-T).



Figure 9-20. Gamma-aminobutyric acid (GABA) receptors. Shown here are receptors for GABA that regulate its neurotransmission. These include the GABA transporter (GAT) as well as three major types of postsynaptic GABA receptors: GABAA, GABAB, and GABAC. GABAA and GABAC receptors are ligand-gated ion channels; they are part of a macromolecular complex that forms an inhibitory chloride channel. GABAB receptors are G-protein-linked receptors that may be coupled with calcium or potassium channels.



Figure 9-21. Gamma-aminobutyric acid-A (GABAA) receptors. (A) Shown here are the four transmembrane regions that make up one subunit of a GABAA receptor. (B) There are five copies of these subunits in a fully constituted GABAA receptor, at the center of which is a chloride channel. (C) Different types of subunits (also called isoforms or subtypes) can combine to form a GABAA receptor. These include six different alpha (α) isoforms, three different beta (β) isoforms, three different gamma (γ) isoforms, delta (δ), epsilon (ε), pi (π), theta (θ), and three different rho (ρ) isoforms. The ultimate type and function of each GABAA receptor subtype will depend on which subunits it contains. Benzodiazepine-sensitive GABAA receptors (middle two) contain γ and α1–3 subunits and mediate phasic inhibition triggered by peak concentrations of synaptically released GABA. Benzodiazepine-sensitive GABAA receptors containing α1 subunits are involved in sleep (second from left), while those that contain α2 and/or α3 subunits are involved in anxiety (second from right). GABAA receptors containing α4, α6, γ1, or δ subunits (far right) are benzodiazepine-insensitive, are located extrasynaptically, and regulate tonic inhibition.



Figure 9-22. GABAA mediation of tonic and phasic inhibition. Benzodiazepine-sensitive GABAA receptors (those that contain γ and α1–3 subunits) are postsynaptic receptors that mediate phasic inhibition, which occurs in bursts triggered by peak concentrations of synaptically released GABA. Benzodiazepine-insensitive GABAA receptors (those containing α4, α6, γ1, or δ subunits) are extrasynaptic and capture GABA that diffuses away from the synapse as well as neurosteroids that are synthesized and released by glia. These receptors mediate inhibition that is tonic (i.e., mediated by ambient levels of extracellular GABA that has escaped from the synapse).



Figure 9-23. Positive allosteric modulation of GABAA receptors. (A) Benzodiazepine-sensitive GABAA receptors, like the one shown here, consist of five subunits with a central chloride channel and have binding sites not only for GABA but also for positive allosteric modulators (e.g., benzodiazepines). (B) When GABA binds to its sites on the GABAA receptor, it increases the frequency of opening of the chloride channel and thus allows more chloride to pass through. (C) When a positive allosteric modulator such as a benzodiazepine binds to the GABAA receptor in the absence of GABA, it has no effect on the chloride channel. (D) When a positive allosteric modulator such as a benzodiazepine binds to the GABAA receptor in the presence of GABA, it causes the channel to open even more frequently than when GABA alone is present.



Figure 9-24. Flumazenil. The benzodiazepine receptor antagonist flumazenil is able to reverse a full agonist benzodiazepine acting at its site on the GABAA receptor. This may be helpful in reversing the sedative effects of full agonist benzodiazepines when administered for anesthetic purposes or when taken in overdose by a patient.

Specifically, GABA is produced, or synthesized, from the amino acid glutamate (glutamic acid) via the actions of the enzyme glutamic acid decarboxylase, or GAD (Figure 9-18). Once formed in presynaptic neurons, GABA is transported by vesicular inhibitory amino acid transporters (VIAATs) into synaptic vesicles, where GABA is stored until released into the synapse during inhibitory neurotransmission (Figure 9-18). GABA's synaptic actions are terminated by the presynaptic GABA transporter (GAT), also known as the GABA reuptake pump (Figure 9-19), analogous to similar transporters for other neurotransmitters discussed throughout this text. GABA action can also be terminated by the enzyme GABA transaminase (GABA-T), which converts GABA into an inactive substance (Figure 9-19).

There are three major types of GABA receptors and numerous subtypes. The major types are GABAA, GABAB, and GABAC receptors (Figure 9-20). GABAA and GABAC receptors are both ligand-gated ion channels and are part of a macromolecular complex that forms an inhibitory chloride channel (Figure 9-21). Various subtypes of GABAA receptors are targets of benzodiazepines, sedative hypnotics, barbiturates, and/or alcohol (Figure 9-21), and are involved with either tonic or phasic inhibitory neurotransmission at GABA synapses (Figure 9-22). The physiological role of GABAC receptors is not well clarified yet, but they do not appear to be targets of benzodiazepines. GABAB receptors, by contrast, are members of a different receptor class, namely, G-protein-linked receptors. GABAB receptors may be coupled to calcium and/or potassium channels, and may be involved in pain, memory, mood, and other CNS functions.

GABAA receptor subtypes

GABAA receptors play a critical role in mediating inhibitory neurotransmission and as targets of the anxiolytic benzodiazepines. The molecular structure of GABAA receptors is shown in Figure 9-21. Each subunit of a GABAAreceptor has four transmembrane regions (Figure 9-21A). When five subunits cluster together, they form an intact GABAA receptor with a chloride channel in the center (Figure 9-21B). There are many different subtypes of GABAAreceptors, depending upon which subunits are present (Figure 9-21C). Subunits of GABAA receptors are sometimes also called isoforms, and include α (with six isoforms, α1 to α6), β (with three isoforms, β1 to β3), γ (with three isoforms, γ1 to γ3), δ, ε, π, θ, and ρ (with three isoforms, ρ1 to ρ3) (Figure 9-21C). What is important for this discussion is that, depending upon which subunits are present, the functions of a GABAA receptor can vary significantly.

Benzodiazepine-insensitive GABAA receptors

Benzodiazepine-insensitive GABAA receptors are those with α4, α6, γ1, or δ subunits (Figure 9-21C). GABAA receptors with a δ subunit rather than a γ subunit, plus either α4 or α6 subunits, do not bind to benzodiazepines. Such GABAA receptors do bind to other modulators, namely the naturally occurring neurosteroids, as well as to alcohol and to some general anesthetics (Figure 9-21C). The binding site for these non-benzodiazepine modulators is located between the α and the δ subunits, one site per receptor complex (Figure 9-21C). Two molecules of GABA bind per receptor complex, at sites located between the α and the β subunits, sometimes referred to as the GABA agonist site (Figure 9-21C). Since the site for the modulators is in a different location from the agonist sites for GABA, the modulatory site is often called allosteric (literally, “other site”), and the agents that bind there are called allosteric modulators.

Benzodiazepine-insensitive GABAA receptor subtypes (with δ subunits and α4 or α6 subunits) are located extrasynaptically, where they capture not only GABA that diffuses away from the synapse, but also neurosteroids synthesized and released by glia (Figure 9-22). Extrasynaptic, benzodiazepine-insensitive GABAA receptors are thought to mediate a type of inhibition at the postsynaptic neuron that is tonic, in contrast to the phasic type of inhibition mediated by postsynaptic benzodiazepine-sensitive GABAA receptors (Figure 9-22). Thus, tonic inhibition may be regulated by the ambient levels of extracellular GABA molecules that have escaped presynaptic reuptake and enzymatic destruction. Tonic inhibition is thought to set the overall tone and excitability of the postsynaptic neuron, and to be important for certain regulatory events such as the frequency of neuronal discharge in response to excitatory inputs.

Since the GABAA receptors that modulate this action are not sensitive to benzodiazepines, they are not likely to be involved in the anxiolytic actions of benzodiazepines in various anxiety disorders. However, novel hypnotics as well as anesthetics have targeted these extrasynaptic benzodiazepine-insensitive GABAA receptors, and it is possible that novel synthetic neurosteroids that also target benzodiazepine-insensitive GABAA receptor subtypes could some day become novel anxiolytics. Indeed, anxiety itself may in part be dependent upon having the right amount of tonic inhibition in key anatomic areas such as the amygdala and cortical areas of CSTC loops. Furthermore, naturally occurring neurosteroids may be important in setting that inhibitory tone in critical brain areas. If this tone becomes dysregulated, it is possible that abnormal neuronal excitability could become a factor in the development of various anxiety disorders.

Benzodiazepine-sensitive GABAA receptors

Benzodiazepine-sensitive GABAA receptors have several structural and functional features that make them distinct from benzodiazepine-insensitive GABAA receptors. In contrast to benzodiazepine-insensitive GABAA receptors, for a GABAA receptor to be sensitive to benzodiazepines, and thus to be a target for benzodiazepine anxiolytics, there must be two β units plus a γ unit of either the γ2 or γ3subtype, plus two α units of the α1, α2, or α3 subtype (Figure 9-21C). Benzodiazepines appear to bind to the region of the receptor between the γ2/3 subunit and the α1/2/3 subunit, one benzodiazepine molecule per receptor complex (Figure 9-21C). GABA itself binds with two molecules of GABA per receptor complex to the GABA agonist sites in the regions of the receptor between the α and the β units (Figure 9-21C).

Benzodiazepine-sensitive GABAA receptor subtypes (with γ subunits and α1–3 subunits) are thought to be postsynaptic in location and to mediate a type of inhibition at the postsynaptic neuron that is phasic, occurring in bursts of inhibition that are triggered by peak concentrations of synaptically released GABA (Figure 9-22). Theoretically, benzodiazepines acting at these receptors, particularly the α2/3 subtypes clustered at postsynaptic GABA sites, should exert an anxiolytic effect due to enhancement of phasic postsynaptic inhibition. If this action occurs at overly active output neurons in the amygdala or in CSTC loops, it would theoretically cause anxiolytic actions with reduction of both fear and worry.

Not all benzodiazepine-sensitive GABAA receptors are the same. Notably, those benzodiazepine-sensitive GABAA receptors with α1 subunits may be most important for regulating sleep and are the presumed targets of numerous sedative-hypnotic agents, including both benzodiazepine and non-benzodiazepine positive allosteric modulators of the GABAA receptor (Figure 9-21C). The α1 subtype of GABAAreceptors and the drugs that bind to it are discussed further in Chapter 11 on sleep. Some of these agents are selective for only the α1 subtype of GABAA receptor.

On the other hand, benzodiazepine-sensitive GABAA receptors with α2 (and/or α3) subunits may be most important for regulating anxiety and are the presumed targets of the anxiolytic benzodiazepines (Figure 9-21C). However, currently available benzodiazepines are nonselective for GABAA receptors with different α subunits. Thus, there is an ongoing search for selective α2/3 agents that could be utilized to treat anxiety disorders in humans. Such agents would theoretically be anxiolytic without being sedating. Partial agonists selective for α2/3 subunits of benzodiazepine-sensitive GABAA receptors hypothetically would cause less euphoria, be less reinforcing and thus less abusable, cause less dependence, and cause fewer problems in withdrawal. Such agents are being investigated but have not yet been introduced into clinical practice. Abnormally expressed γ2, α2, or δ subunits have all been associated with different types of epilepsy. Receptor subtype expression can change in response to chronic benzodiazepine administration and withdrawal, and could theoretically be altered in patients with various anxiety-disorder subtypes.

Benzodiazepines as positive allosteric modulators or PAMs

Since the benzodiazepine-sensitive GABAA receptor complex is regulated not only by GABA itself, but also by benzodiazepines at a highly specific allosteric modulatory binding site (Figure 9-23), this has led to the notion that there may be an “endogenous” or naturally occurring benzodiazepine synthesized in the brain (the brain's own Xanax!). However, the identity of any such substance remains elusive. Furthermore, it is now known that synthetic drugs that do not have a benzodiazepine structure also bind to the “benzodiazepine receptor.” These developments have led to endless confusion with terminology, since non-benzodiazepines also bind to the “benzodiazepine receptor!” Thus, many experts now call the “benzodiazepine site” the GABAA allosteric modulatory site and anything that binds to this site, including benzodiazepines, allosteric modulators.

Acting alone, GABA can increase the frequency of opening of the chloride channel, but only to a limited extent (compare Figure 9-23A and B). The combination of GABA with benzodiazepines is thought to increase the frequency of opening of inhibitory chloride channels but not to increase the conductance of chloride across individual chloride channels, nor to increase the duration of channel opening. The end result is more inhibition. More inhibition supposedly yields more anxiolytic action. How does this happen? The answer is that benzodiazepines act as agonists at the allosteric modulatory site of GABA binding. They are positive allosteric modulators, or PAMs, but have no activity on their own. Thus, when benzodiazepines bind to the allosteric modulatory site, they have no activity when GABA is not simultaneously binding to its agonist sites (compare Figure 9-23A and C).

So, how do benzodiazepines act as PAMs? This can occur only when GABA is binding to its agonist sites. The combination of benzodiazepines at the allosteric site plus GABA at its agonist sites increases the frequency of opening of the chloride channel to an extent not possible with GABA alone (compare Figure 9-23B and D).

The actions of benzodiazepines essentially as agonists at their positive allosteric sites can be reversed by the neutral antagonist flumazenil (Figure 9-24). Flumazenil is a short-acting intravenously administered antagonist to benzodiazepines that can reverse overdoses or anesthesia from benzodiazepines but can also induce seizures or withdrawal in patients dependent upon benzodiazepines.

Benzodiazepines as anxiolytics

A simplified notion of how benzodiazepine anxiolytics might modulate excessive output from the amygdala during fear responses in anxiety disorders is shown in Figure 9-25. Excessive amygdala activity (shown in Figures 9-8through 9-12 and in Figure 9-25A) is theoretically reduced by enhancing the phasic inhibitory actions of benzodiazepine PAMs at postsynaptic GABAA receptors within the amygdala to blunt fear-associated outputs, hypothetically reducing the symptom of fear (Figure 9-25B). Benzodiazepines also theoretically modulate excessive output from worry loops (Figure 9-26A) by enhancing the actions of inhibitory interneurons in CSTC circuits (Figure 9-26B), hypothetically reducing the symptom of worry.

Figure 9-25. Potential therapeutic actions of anxiolytics on anxiety/fear. (A) Pathological anxiety/fear may be caused by overactivation of amygdala circuits. (B) GABAergic agents such as benzodiazepines may alleviate anxiety/fear by enhancing phasic inhibitory actions at postsynaptic GABAA receptors within the amygdala. (C) Agents that bind to the α2δ subunit of presynaptic N and P/Q voltage-sensitive calcium channels can block the excessive release of glutamate in the amygdala and thereby reduce the symptoms of anxiety. (D) The amygdala receives input from serotonergic neurons, which can have an inhibitory effect on some of its outputs. Thus, serotonergic agents may alleviate anxiety/fear by enhancing serotonin input to the amygdala.



Figure 9-26. Potential therapeutic actions of anxiolytics on worry. (A) Pathological worry may be caused by overactivation of cortico-striato-thalamo-cortical (CSTC) circuits. (B) GABAergic agents such as benzodiazepines may alleviate worry by enhancing the actions of inhibitory GABA interneurons within the prefrontal cortex. (C) Agents that bind to the α2δ subunit of presynaptic N and P/Q voltage-sensitive calcium channels can block the excessive release of glutamate in CSTC circuits and thereby reduce the symptoms of worry. (D) The prefrontal cortex, striatum, and thalamus receive input from serotonergic neurons, which can have an inhibitory effect on output. Thus, serotonergic agents may alleviate worry by enhancing serotonin input within CSTC circuits.

Alpha-2-delta ligands as anxiolytics

We have discussed voltage-sensitive calcium channels (VSCCs) in Chapter 3 and have illustrated presynaptic N and P/Q subtypes of VSCCs and their role in excitatory neurotransmitter release (see Figures 3-19 and 3-22 through 3-24). Gabapentin and pregabalin, also known as α2δ ligands, since they bind to the α2δ subunit of presynaptic N and P/Q VSCCs, block the release of excitatory neurotransmitters such as glutamate when neurotransmission is excessive, as postulated in the amygdala to cause fear (Figure 9-25A) and in CSTC circuits to cause worry (Figure 9-26A). Hypothetically, α2δ ligands bind to open, overly active VSCCs in the amygdala (Figure 9-25C) to reduce fear, and in CSTC circuits (Figure 9-26C) to reduce worry. The α2δ ligands pregabalin and gabapentin have demonstrated anxiolytic actions in social anxiety disorder and panic disorder, and are also proven to be effective for the treatment of epilepsy and certain pain conditions, including neuropathic pain and fibromyalgia. The actions of α2δ ligands on VSCCs are discussed in Chapter 10 on pain and illustrated in Figures 10-17 through 10-19. α2δ ligands clearly have different mechanisms of action compared to serotonin reuptake inhibitors or benzodiazepines, and thus can be useful for patients who do not do well on SSRIs/SNRIs or benzodiazepines. Also, α2δ ligands can be useful to combine with SSRIs/SNRIs or benzodiazepines in patients who are partial responders and are not in remission.

Serotonin and anxiety

Since the symptoms, circuits, and neurotransmitters linked to anxiety disorders overlap extensively with those for major depressive disorder (Figure 9-1), it is not surprising that drugs developed as antidepressants have proven to be effective treatments for anxiety disorders. Indeed, the leading treatments for anxiety disorders today are increasingly drugs originally developed as antidepressants. Serotonin is a key neurotransmitter that innervates the amygdala as well as all the elements of CSTC circuits, namely, prefrontal cortex, striatum, and thalamus, and thus is poised to regulate both fear and worry (serotonin pathways are discussed in Chapters 5 and 6 and illustrated in Figure 6-33). Antidepressants that can increase serotonin output by blocking the serotonin transporter (SERT) are also effective in reducing symptoms of anxiety and fear in every one of the anxiety disorders illustrated in Figures 9-2 through 9-5 – namely, GAD, panic disorder, social anxiety disorder, and PTSD. Such agents include the well-known SSRIs (selective serotonin reuptake inhibitors; discussed in Chapter 7 and their mechanism of action illustrated in Figures 7-12through 7-17), as well as the SNRIs (serotonin–norepinephrine reuptake inhibitors; also discussed in Chapter 7 and their mechanism of action illustrated in Figures 7-12 through 7-17 plus Figures 7-33 and 7-34).

A serotonin 1A (5HT1A) partial agonist, buspirone, is recognized as a generalized anxiolytic, but not as a treatment for anxiety disorder subtypes. 5HT1A partial agonists as augmenting agents to antidepressants are discussed in Chapter 7, as are antidepressants combining 5HT1A partial agonism with serotonin reuptake inhibition (i.e., SPARIs and vilazodone: see Figures 7-25 through 7-29), which should theoretically be anxiolytic as well as antidepressant in action. The 5HT1A partial agonist properties of numerous atypical antipsychotics are discussed in Chapter 5 and illustrated in Figures 5-155-165-25, and 5-26.

The potential anxiolytic actions of buspirone could theoretically be due to 5HT1A partial agonist actions at both presynaptic and postsynaptic 5HT1A receptors (Figure 9-27 and Figures 5-155-165-257-25through 7-29), with actions at both sites resulting in enhanced serotonergic activity in projections to the amygdala (Figure 9-25D), prefrontal cortex, striatum, and thalamus (Figure 9-26D). SSRIs and SNRIs theoretically do the same thing (Figures 9-25D and 9-26D). Since the onset of anxiolytic action for buspirone is delayed, just as it is for antidepressants, this has led to the belief that 5HT1A agonists exert their therapeutic effects by virtue of adaptive neuronal events and receptor events (Figures 7-12 through 7-17 and 7-25 through 7-29), rather than simply by the acute occupancy of 5HT1A receptors. In this way, the presumed mechanism of action of 5HT1A partial agonists is analogous to the antidepressants – which are also presumed to act by adaptations in neurotransmitter receptors – and different from the benzodiazepine anxiolytics – which act relatively acutely by occupying benzodiazepine receptors.



Figure 9-27. 5HT1A partial agonist actions in anxiety. 5HT1A partial agonists such as buspirone may reduce anxiety by actions both at presynaptic somatodendritic autoreceptors (left) and at postsynaptic receptors (right). The onset of action of buspirone, like that of antidepressants, is delayed, suggesting that the therapeutic effects are actually related to downstream adaptive changes rather than acute actions at these receptors.

Noradrenergic hyperactivity in anxiety

Norepinephrine is another neurotransmitter with important regulatory input to the amygdala (Figure 9-28) and to the prefrontal cortex and thalamus in CSTC circuits (Figure 9-29). Excessive noradrenergic output from the locus coeruleus can not only result in numerous peripheral manifestations of autonomic overdrive, as discussed above and illustrated in Figures 9-8 through 9-12, but can also trigger numerous central symptoms of anxiety and fear, such as nightmares, hyperarousal states, flashbacks, and panic attacks (Figure 9-28A). Excessive noradrenergic activity can also reduce the efficiency of information processing in the prefrontal cortex and thus in CSTC circuits and theoretically cause worry (Figure 9-29A). Hypothetically, these symptoms may be mediated in part by excessive noradrenergic input onto α1- and β1-adrenergic postsynaptic receptors in the amygdala (Figure 9-28A) or prefrontal cortex (Figure 9-29A). Symptoms of hyperarousal such as nightmares can be reduced in some patients with α1-adrenergic blockers such as prazocin (Figure 9-28B); symptoms of fear (Figure 9-28C) and worry (Figure 9-29B) can be reduced by norepinephrine reuptake inhibitors (also called NET or norepinephrine transporter inhibitors). The clinical effects of NET inhibitors can be confusing, because symptoms of anxiety can be made transiently worse immediately following initiation of an SNRI or selective NET inhibitor, when noradrenergic activity is initially increased but the postsynaptic receptors have not yet adapted. However, these same NET inhibitory actions, if sustained over time, will downregulate and desensitize postsynaptic NE receptors such as β1 receptors, and actually reduce symptoms of fear and worry long term (Figure 9-29B).



Figure 9-28. Noradrenergic hyperactivity in anxiety/fear. (A) Norepinephrine provides input not only to the amygdala but also to many regions to which the amygdala projects; thus it plays an important role in the fear response. Noradrenergic hyperactivation can lead to anxiety, panic attacks, tremors, sweating, tachycardia, hyperarousal, and nightmares. α1- and β1-adrenergic receptors may be specifically involved in these reactions. (B) Noradrenergic hyperactivity may be blocked by the administration of α1-adrenergic blockers, which can lead to the alleviation of anxiety and other stress-related symptoms. (C) Noradrenergic hyperactivity may also be blocked by the administration of a norepinephrine transporter (NET) inhibitor, which can have the downstream effect of downregulating β1-adrenergic receptors. Reduced stimulation via β1-adrenergic receptors could therefore lead to the alleviation of anxiety and stress-related symptoms.



Figure 9-29. Noradrenergic hyperactivity in worry. (A) Pathological worry may be caused by overactivation of cortico-striato-thalamo-cortical (CSTC) circuits. Specifically, excessive noradrenergic activity within these circuits can reduce the efficiency of information processing and theoretically cause worry. (B) Noradrenergic hyperactivity in CSTC circuits may be blocked by the administration of a norepinephrine transporter (NET) inhibitor, which can have the downstream effect of downregulating β1-adrenergic receptors. Reduced stimulation via β1-adrenergic receptors could therefore lead to the alleviation of worry.

Fear conditioning versus fear extinction

Fear conditioning

Fear conditioning is a concept as old as Pavlov's dogs. If an aversive stimulus such as footshock is coupled with a neutral stimulus such as a bell, the animal learns to associate the two and will develop fear when it hears a bell. In humans, fear is learned during stressful experiences associated with emotional trauma and is influenced by an individual's genetic predisposition as well as by an individual's prior exposure to environmental stressors that can cause stress sensitization of brain circuits (e.g., child abuse: see Chapter 6 and Figures 6-40 through 6-43). Often, fearful situations are managed successfully and then forgotten. Some fears are crucial for survival, such as appropriately fearing dangerous situations, and thus the mechanism of learned fear, called fear conditioning, has been extremely well conserved across species, including humans. However, other fears that are “learned” and not “forgotten” may hypothetically progress to anxiety disorders or a major depressive episode. This is a big problem, since almost 30% of the population will develop an anxiety disorder, due in large part to stressful environments, including exposure to fearful events during normal activities in twenty-first-century society, but in particular during war and natural disasters. Hearing an explosion, smelling burning rubber, seeing a picture of a wounded civilian, and seeing or hearing flood waters are all sensory experiences than can trigger traumatic re-experiencing and generalized hyperarousal and fear in PTSD. Panic associated with social situations will “teach” the patient to panic in social situations in social anxiety disorder. Panic randomly associated with an attack that happens to occur in a crowd, on a bridge, or in a shopping center will also trigger another panic attack when the same environment is encountered in panic disorder. These and other symptoms of anxiety disorders are all forms of learning known as fear conditioning (Figure 9-30).



Figure 9-30. Fear conditioning versus fear extinction. When an individual encounters a stressful or fearful experience, the sensory input is relayed to the amygdala, where it is integrated with input from the ventromedial prefrontal cortex (VMPFC) and hippocampus, so that a fear response can be either generated or suppressed. The amygdala may “remember” stimuli associated with that experience by increasing the efficiency of glutamate neurotransmission, so that on future exposure to stimuli, a fear response is more efficiently triggered. If this is not countered by input from the VMPFC to suppress the fear response, fear conditioning proceeds. Fear conditioning is not readily reversed, but it can be inhibited through new learning. This new learning is termed fear extinction and is the progressive reduction of the response to a feared stimulus that is repeatedly presented without adverse consequences. Thus the VMPFC and hippocampus learn a new context for the feared stimulus and send input to the amygdala to suppress the fear response. The “memory” of the conditioned fear is still present, however.

The amygdala is involved in “remembering” the various stimuli associated with a given fearful situation. It does this by increasing the efficiency of neurotransmission at glutamatergic synapses in the lateral amygdala as sensory input about those stimuli comes in from the thalamus or sensory cortex (Figure 9-30). This input is then relayed to the central amygdala, where fear conditioning also improves the efficiency of neurotransmission at another glutamate synapse there (Figure 9-30). Both synapses are restructured and permanent learning is embedded into this circuit by NMDA receptors triggering long-term potentiation and synaptic plasticity, so that subsequent input from the sensory cortex and thalamus is very efficiently processed to trigger the fear response as output from the central amygdala every time there is sensory input associated with the original fearful event (Figure 9-30; see also Figures 9-8 through 9-13).

Input to the lateral amygdala is modulated by the prefrontal cortex, especially the ventromedial prefrontal cortex (VMPFC), and by the hippocampus. If the VMPFC is unable to suppress the fear response at the level of the amygdala, fear conditioning proceeds. The hippocampus remembers the context of the fear conditioning and makes sure fear is triggered when the fearful stimulus and all its associated stimuli are encountered. Most contemporary psychopharmacological treatments for anxiety and fear act by suppressing the fear output from the amygdala (Figures 9-25 and 9-28) and therefore are not cures, since the fundamental neuronal learning underlying fear conditioning in these patients remains in place.

Novel approaches to the treatment of anxiety disorders

Once fear conditioning is in place, it can be very difficult to reverse. Nevertheless, there may be two ways to neutralize fear conditioning: either by facilitating a process called extinction or by blocking a process called reconsolidation.

Fear extinction

Fear extinction is the progressive reduction of the response to a feared stimulus, and occurs when the stimulus is repeatedly presented without any adverse consequence. When fear extinction occurs, it appears that the original fear conditioning is not really “forgotten” even though the fear response can be profoundly reduced over time by the active process of fear extinction. Rather than reversing the synaptic changes described above for fear conditioning, it appears that a new form of learning with additional synaptic changes in the amygdala occurs during fear extinction. These changes can suppress symptoms of anxiety and fear by inhibiting the original learning but not by removing it (Figure 9-30). Specifically, activation of the amygdala by the VMPFC occurs while the hippocampus “remembers” the context in which the feared stimulus did not have any adverse consequences and fear is no longer activated (Figure 9-30). Fear extinction hypothetically occurs when inputs from the VMPFC and hippocampus activate glutamatergic neurons in the lateral amygdala that synapse upon an inhibitory GABAergic interneuron located within the intercalated cell mass of the amygdala (Figure 9-30). This sets up a gate within the central amygdala, with fear output occurring if the fear conditioning circuit predominates, and no fear output occurring if the fear extinction circuit predominates.

Fear extinction theoretically predominates over fear conditioning when synaptic strengthening and long-term potentiation in the new circuit are able to produce inhibitory GABAergic drive that can overcome the excitatory glutamatergic drive produced by the pre-existing fear conditioning circuitry (Figure 9-30). When fear extinction exists simultaneously with fear conditioning, memory for both are present, but the output depends upon which system is “stronger,” better remembered,” and has the most robust synaptic efficiency. These factors will determine which gate will open, the one with the fear response or the one that keeps the fear response in check. Unfortunately, over time, fear conditioning may have the upper hand over fear extinction. Fear extinction appears to be more labile than fear conditioning, and tends to reverse over time. Also, fear conditioning can return if the old fear is presented in a context different than the one “learned” to suppress the fear during fear extinction, a process termed renewal.

Novel treatment approaches to anxiety disorders seek to facilitate fear extinction rather than just suppress the fear response triggered by fear conditioning, which is how current anxiolytic drugs work (Figures 9-25 through 9-30). Among currently effective treatments for anxiety, cognitive behavioral therapies that use exposure techniques and that require the patient to confront the fear-inducing stimuli in a safe environment may come closest to facilitating fear extinction, hypothetically because when these therapies are effective, they are able to trigger the learning of fear extinction in the amygdala (Figure 9-30). Unfortunately, because the hippocampus “remembers” the context of this extinction, such therapies are often context-specific and do not always generalize once the patient is outside the safe therapeutic environment; thus fear and worry may be “renewed” in the real world. Current psychotherapy research is investigating how contextual cues can be used to strengthen extinction learning so that the therapeutic learning generalizes to other environments. Current psychopharmacology research is investigating how specific drugs might also strengthen extinction learning by pharmacologically strengthening the synapses on the fear-extinction side of the amygdala gate disproportionately to the synapses on the fear-conditioned side of the amygdala gate. How could this be done?

Based on successful animal experiments of extinction learning, one idea, shown in Figure 9-31, is to boost NMDA receptor activation at the very time when a patient receives systematic exposure to feared stimuli during cognitive behavioral therapy sessions. This can be done either with direct-acting agonists such as D-cycloserine or with indirect glycine-enhancing agents such as selective glycine reuptake inhibitors (SGRIs). This approach to boosting activity at NMDA synapses is discussed in Chapter 5 in relation to schizophrenia and is illustrated in Figure 5-90. As applied to novel anxiolytic therapy, the idea is that as therapy progresses, learning occurs, because glutamate release is provoked in the lateral amygdala and in the intercalated cell mass at inhibitory GABA neurons by the psychotherapy. If NMDA receptors at these two glutamate synapses could be pharmacologically boosted to trigger disproportionately robust long-term potentiation and synaptic plasticity, timed to occur at the exact time this learning and therapy is taking place and thus exactly when these synapses are selectively activated, it could theoretically result in the predominance of the extinction pathway over the conditioned pathway. Animal studies support this possibility, and early clinical studies are encouraging but not always robust or consistent to date. In the meantime, prudent psychopharmacologists are increasingly leveraging their current anxiolytic drug portfolio with concomitant psychotherapy, since many patients have already received enhanced therapeutic benefit from this combination.



Figure 9-31. Facilitating fear extinction with NMDA receptor activation. Strengthening of synapses involved in fear extinction could help enhance the development of fear extinction learning in the amygdala and reduce symptoms of anxiety disorders. Administration of the N-methyl-D-aspartate (NMDA) coagonist D-cycloserine while an individual is receiving exposure therapy could increase the efficiency of glutamate neurotransmission at synapses involved in fear extinction. Likewise, administration of indirect glycine-enhancing agents such as selective glycine reuptake inhibitors (SGRIs) during exposure therapy could boost NMDA receptor activation. If this leads to long-term potentiation and synaptic plasticity while the synapses are activated by exposure therapy, it could result in structural changes in the amygdala associated with the fear extinction pathway and thus the predominance of the extinction pathway over the conditioned pathway.

Reconsolidation

Blocking reconsolidation of fear memories is a second mechanism that could theoretically be therapeutic for patients with anxiety disorders. Although classically, emotional memories that have been fear-conditioned were thought to last forever, recent animal experiments show that emotional memories can in fact be weakened or even erased at the time they are re-experienced. When fear is first conditioned, that memory is said to be “consolidated” via a molecular process that some have thought was essentially permanent. Hints at the mechanism of the initial consolidation of fear conditioning come from observations that both β blockers and opioids can potentially mitigate the conditioning of the original traumatic memory, even in humans, and some studies show that these agents can potentially reduce the chances of getting PTSD after a traumatic injury (Figure 9-32). Furthermore, once emotional memories have been consolidated as fear conditioning, animal experiments now show that they are not necessarily permanent, but can change when they are retrieved. Reconsolidation is the state in which reactivation of a consolidated fear memory makes it labile, and requires protein synthesis to keep the memory intact. Beta blockers disrupt reconsolidation of fear memories as well as formation of fear conditioning (Figure 9-32). Future research is trying to determine how to use psychotherapy to provoke emotional memories and reactivate them by producing a state where a pharmacologic agent could be administered to disrupt reconsolidation of these emotional memories and thereby relieve symptoms of anxiety. These are early days in terms of applying this concept in clinical settings, but this notion supports the growing idea that psychotherapy and psychopharmacology can be synergistic. Much more needs to be learned as to how to exploit this theoretical synergy.



Figure 9-32. Blocking fear conditioning and reconsolidation. When fear is first conditioned, the memory is said to be “consolidated” via a molecular process that once was thought to be permanent. However, there is some research to suggest that administration of either β-adrenergic blockers or opioids can potentially mitigate the conditioning of the original traumatic memory. Furthermore, research also now shows that even when emotional memories have been consolidated as fear conditioning, they can change when they are retrieved. Reconsolidation is the state in which reactivation of a consolidated fear memory makes it labile. This requires protein synthesis to keep the memory intact and, like fear conditioning, may also be disrupted by β blockers.

Treatments for anxiety disorder subtypes

Generalized anxiety disorder

Treatments for generalized anxiety disorder (GAD) overlap greatly with those for other anxiety disorders and depression (Figure 9-33). First-line treatments include SSRIs and SNRIs, benzodiazepines, buspirone, and α2δ ligands such as pregabalin and gabapentin. Some prescribers are reluctant to give benzodiazepines for anxiety disorders in general and for GAD in particular, because of the long-term nature of GAD and the possibility of dependence, abuse, and withdrawal reactions with benzodiazepines.



Figure 9-33. Generalized anxiety disorder (GAD) pharmacy. First-line treatments for GAD include α2δ ligands, selective serotonin reuptake inhibitors (SSRIs), benzodiazepines (BZs), serotonin–norepinephrine reuptake inhibitors (SNRIs), and buspirone. Second-line treatments include tricyclic antidepressants (TCAs), mirtazapine, trazodone, and serotonin partial agonist/reuptake inhibitors (SPARIs, e.g., vilazodone). Adjunctive medications that may be helpful include hypnotics or an atypical antipsychotic; cognitive behavioral therapy (CBT) is also an important component of anxiety treatment.

While it is not a good idea to give benzodiazepines to a GAD patient who is abusing other substances, particularly alcohol, benzodiazepines can be useful when initiating an SSRI or SNRI, since these serotonergic agents are often activating, difficult to tolerate early in dosing, and have a delayed onset of action. α2δ ligands are a good alternative to benzodiazepines in some patients. Both benzodiazepines and α2δ ligands can have a role in some patients as augmenting agents, especially when initiating treatment with another agent that may be slower-acting or even activating. In other patients, benzodiazepines can be useful to “top up” an SSRI or SNRI for patients who have experienced only partial relief of symptoms. Benzodiazepines can also be useful for occasional intermittent use when symptoms surge and sudden relief is needed.

It should be noted that remission from all symptoms in patients with GAD who are taking an SSRI or SNRI may be slower in onset than it is in depression, and be delayed for 6 months or longer. If a GAD patient is not doing well after several weeks to months of treatment, switching to another SSRI/SNRI or buspirone or augmenting with a benzodiazepine or an α2δ ligand can be considered. Failure to respond to first-line treatments can lead to trials of sedating antidepressants such as mirtazapine, trazodone, or tricyclic antidepressants, or even sedating antihistamines such as hydroxyzine. Although not well studied, the SPARI vilazodone should theoretically have efficacy for GAD and can be considered as a second-line agent as well. Adjunctive treatments that can be added to first- or second-line therapies for GAD include hypnotics for continuing insomnia, atypical antipsychotics for severe, refractory, and disabling symptoms unresponsive to aggressive treatment, and cognitive behavioral psychotherapy. Old-fashioned treatments for anxiety such as barbiturates and meprobamate are not considered appropriate today, given the other choices shown in Figure 9-33.

Panic disorder

Panic attacks occur in many conditions, not just panic disorder, and panic disorder is frequently comorbid with the other anxiety disorders and with major depression. It is thus not surprising that contemporary treatments for panic disorder overlap significantly with those for the other anxiety disorders and with those for major depression (Figure 9-34). First-line treatments include SSRIs and SNRIs, as well as benzodiazepines and α2δ ligands, although benzodiazepines are often used second line, during treatment initiation with an SSRI/SNRI, for emergent use during a panic attack, or for incomplete response to an SSRI/SNRI. α2δ ligands are approved for the treatment of anxiety in Europe and other countries but not in the US.



Figure 9-34. Panic pharmacy. First-line treatments for panic disorder include α2δ ligands, selective serotonin reuptake inhibitors (SSRIs), benzodiazepines (BZs), and serotonin–norepinephrine reuptake inhibitors (SNRIs). Second-line treatments include monoamine oxidase inhibitors (MAOIs), tricyclic antidepressants (TCAs), mirtazapine, and trazodone. Cognitive behavioral therapy (CBT) may be beneficial for many patients. In addition, adjunctive medications for residual symptoms may include hypnotics or an atypical antipsychotic.

Second-line treatments include older antidepressants such as tricyclic antidepressants. Mirtazapine and trazodone are sedating antidepressants that can be helpful in some cases, and are occasionally used as augmenting agents to SSRIs/SNRIs when these first-line agents have only a partial treatment response. The MAO inhibitors, discussed in Chapter 7, are much neglected in psychopharmacology in general and for the treatment of panic disorder in particular. However, these agents can have powerful efficacy in panic disorder and should be considered when first-line agents and various augmenting strategies fail.

Cognitive behavioral psychotherapy can be an alternative or an augmentation to psychopharmacologic approaches, and can help modify cognitive distortions and, through exposure, diminish phobic avoidance behaviors.

Social anxiety disorder

The treatment options for this anxiety disorder (Figure 9-35) are very similar to those for panic disorder, with a few noteworthy differences. The SSRIs and SNRIs and α2δ ligands are certainly first-line therapies, but the utility of benzodiazepine monotherapy for first-line treatment is not generally as widely accepted as it might be for GAD and panic disorder. There is also less evidence for the utility of older antidepressants for social anxiety disorder, particularly the tricyclic antidepressants, but also other sedating antidepressants such as mirtazapine and trazodone. Beta blockers, sometimes with benzodiazepines, can be useful for some patients with very discrete types of social anxiety, such as performance anxiety. Listed as adjunctive treatments are agents for alcohol dependence/abuse, such as naltrexone and acamprosate, since many patients may discover the utility of alcohol in relieving their social anxiety symptoms and develop alcohol dependence/abuse. Cognitive behavioral psychotherapy can be a powerful intervention, sometimes better than drugs for certain patients, and often helpful in combination with drugs.



Figure 9-35. Social anxiety pharmacy. First-line treatments for social anxiety disorder include α2δ ligands, selective serotonin reuptake inhibitors (SSRIs), and serotonin–norepinephrine reuptake inhibitors (SNRIs). Monoamine oxidase inhibitors (MAOIs) have been shown to be beneficial and may be a second-line option; other second-line options include benzodiazepines (BZs) and β blockers. Several medications may be used as adjuncts for residual symptoms; cognitive behavioral therapy may (CBT) be useful as well.

Posttraumatic stress disorder

Although many treatments are shown in Figure 9-36, psychopharmacologic treatments for PTSD in general may not be as effective as these same treatments are in other anxiety disorders. Also, PTSD is so highly comorbid that many of the psychopharmacologic treatments are more effectively aimed at comorbidities such as depression, insomnia, substance abuse, and pain than at core symptoms of PTSD. SSRIs and SNRIs are proven effective and are considered first-line treatments, but often leave the patient with residual symptoms, including sleep problems. Thus, most patients with PTSD do not take monotherapy. Benzodiazepines are to be used with caution, not only because of limited evidence from clinical trials for efficacy in PTSD, but also because many PTSD patients abuse alcohol and other substances. A unique treatment for PTSD is the administration of α1 antagonists at night to prevent nightmares. Pre-emptive treatment with β blockers or opioids is discussed above, but is not a proven or practical treatment option at this point. Much more effective treatments for PTSD are greatly needed.



Figure 9-36. Posttraumatic stress disorder (PTSD) pharmacy. First-line pharmacological options for PTSD are selective serotonin reuptake inhibitors (SSRIs) and serotonin–norepinephrine reuptake inhibitors (SNRIs). In PTSD, unlike other anxiety disorders, benzodiazepines (BZs) have not been shown to be as helpful, although they may be considered with caution as a second-line option. Other second-line treatments include α2δ ligands, tricyclic antidepressants (TCAs), and monoamine oxidase inhibitors (MAOIs). Several medications may be used as adjuncts for residual symptoms, and cognitive behavioral therapy (CBT) is typically recommended as well.

Much of the advance in treatment of PTSD has been in using drugs to treat comorbidities and psychotherapies to treat core symptoms. Exposure therapy is perhaps most effective among psychotherapies, but many forms ofCBT are being investigated and used in clinical practice, depending upon the training of the therapist and the specific needs of the individual.

Summary

Anxiety disorders have core features of fear and worry that cut across the entire spectrum of anxiety disorder subtypes, from generalized anxiety disorder to panic disorder, social anxiety disorder, and posttraumatic stress disorder. The amygdala plays a central role in the fear response, and cortico-striato-thalamo-cortical (CSTC) circuits are thought to play a key role in mediating the symptom of worry. Numerous neurotransmitters are involved in regulating the circuits that underlie the anxiety disorders. GABA (γ-aminobutyric acid) is a key neurotransmitter in anxiety and the benzodiazepine anxiolytics act upon this neurotransmitter system. Serotonin, norepinephrine, α2δ ligands for voltage-gated calcium channels, and other regulators of anxiety circuits are also discussed as approaches to the treatment of anxiety disorders. The concept of opposing actions of fear conditioning versus fear extinction within amygdala circuits hypothetically is linked to the production and maintenance of symptoms in anxiety disorder and provides a substrate for potential novel therapeutics combining psychotherapy and drugs. Numerous treatments are available for anxiety disorders, most of which are similar for the entire anxiety disorder spectrum and are also used for the treatment of depression.