Panic Disorder: Neuroanatomical Hypothesis

Neuroanatomical Hypothesis

Gorman and colleagues (2000) begin with the observation that there is a remarkable similarity between the physiological and behavioral consequences of panic attacks in humans and condi-tioned fear responses in animals. Similarities include autonomic arousal, fear evoked by specific cues (i.e., contextual fear) and avoidance of these cues. Animal research indicates that condi-tioned fear responses are mediated by a “fear network” in the brain, consisting of the amygdala and its afferent and efferent projections, particularly its connections with the hippocampus, medial prefrontal cortex, hypothalamus and brainstem. Animal studies also show that activation of this network produces biolog-ical and behavioral reactions that are similar to those associated with panic attacks. Thus, Gorman and colleagues (2000) posit that a similar network is involved in panic disorder.

The fear network consists of a complex matrix of intercon-nections, implicating a number of brain structures and neurotrans-mitter systems. Sensory input passes through the anterior thalamus to the lateral nucleus of the amygdala. Input is then transferred to the central nucleus of the amygdala, which coordinates autonomic and behavioral responses. Direct sensory input to the amygdala from brainstem structures and the sensory thalamus enables a rapid response to potentially threatening stimuli. The central nu-cleus of the amygdala projects to the following structures:

·        the parabrachial nucleus, producing an increase in respira-tory rate;

·        the lateral nucleus of the hypothalamus, causing autonomic arousal and sympathetic discharge;

·        the locus coeruleus, leading to an increase in norepinephrine and to increases in blood pressure, heart rate and behavioral fear responses (e.g., freezing);

·        the paraventricular nucleus of the hypothalamus, resulting in an increase in the release of adrenocorticoids; and

·        the periaqueductal gray region, leading to avoidance beha-viors.

In addition, there are reciprocal connections between the amy-gdala and the sensory thalamus, prefrontal cortex, insula, and primary somatosensory cortex.

Panic attacks arise from excessive activation of the fear network (Gorman et al., 2000). The fear network becomes sensi-tized (conditioned) to respond to noxious stimuli such as internal (bodily sensations) and external (contexts or situations) that the person associates with panic. Sensitization of the network may be manifested by the strengthening of various projections from the central nucleus of the amygdala to brainstem sites (such as the locus ceruleus, periaqueductal gray region and hypothalamus). The network could be over-activated if brainstem inputs to the amygdala are dysregulated. However, autonomic activation (e.g., increased respiration and heart rate) and neuroendocrine activa-tion (e.g., increased cortisol secretion) does not occur in all panic attacks. Moreover, a variety of biological agents with diverse physiological properties can trigger panic attacks in people with panic disorder (e.g., sodium lactate, yohimbine, CO₂, caffeine, cholecystokinin-4). It is, therefore, unlikely that a single brain-stem dysregulation is responsible for panic or, in turn, that brain-stem dysregulation is the only way of producing an over-active fear network.

There may be other ways of activating the fear network. For example, the amygdala receives input from cortical regions involved in the processing and evaluation of sensory information. Therefore, a neurocognitive deficit in these cortico-amygdala pathways could result in the catastrophic misinterpretation of sensory information (i.e., misinterpretation of bodily sensations), leading to an inappropriate activation of the fear network. No-tice that this pathway resembles the cognitive model of panicdescribed earlier. Thus, Gorman and colleagues (2000) model integrates the cognitive model and places it in a neuroanatomical context.

In addition to playing a role in panic disorder, the fear network is thought to play a role in other anxiety disorders and in mood disorders. This is consistent with the comorbidity be-tween panic disorder and these disorders. Abnormalities in the fear network may vary from disorder to disorders. For example, the strength of various connections between components of the network may distinguish various disorders.

Medications, particularly selective serotonin reuptake inhibitors (SSRIs), are thought to desensitize the fear network. This may happen in a number of ways. SSRIs increase sero-tonergic transmission in the brain. Serotonergic neurons origi-nate in the brainstem raphe and project throughout the central nervous system and some of these projections have inhibitory influences. For example, the greater the activity in the raphe, the greater the inhibition of noradrenergic neurons in the locus ceruleus, resulting in a reduction of cardiovascular symptoms associated with panic attacks, such as tachycardia. Similarly, the greater the activity in the raphe, the greater the inhibition in the periaqueductal gray region, resulting in a reduction in avoid-ance behavior. Increased serotonergic activity also may reduce hypothalamic release of corticotropin-releasing factor, thereby resulting in a reduction of cortisol and a reduction in activity of the locus ceruleus thereby leading to a reduction in fear. SSRIs may also directly inhibit activity of the lateral nucleus of the amygdala. Thus, there appear to be several ways in which SS-RIs could desensitize the fear network. Effective psychological therapies are thought to reduce contextual fear and catastrophic misinterpretations at the level of the medial prefrontal cortex and hippocampus.

Environmental and Genetic Factors

The fear network is thought to be influenced by genetic factors and stressful life events, particularly events in early childhood. The search for genetic markers and candidate genes for panic disorder has revealed several possible loci but, to date, none has been replicated across studies. Research with monozygotic and dizygotic twins show that panic disorder is moderately heritable, with 32 to 46% of variance in liability for panic being attributed to genetic factors (Kendler et al., 1993).

Vulnerability to panic disorder appears to result from a combination of disorder-specific and disorder-nonspecific fac-tors. The importance of nonspecific genetic factors is consistent with observation that panic disorder is often comorbid with other disorders. Twin studies suggest that nonspecific factors influence the vulnerability to several disorders, including panic disorder, bulimia nervosa, generalized anxiety disorder and alcohol de-pendence. Genetic factors specific to panic disorder may be those that influence the tendency catastrophically to misinterpret bod-ily sensations. This cognitive tendency is a distinguishing feature of panic disorder, as described above. Recent twin research indi-cates that it is moderately heritable in women but not men (Jang et al., 1999). Thus, some specific genetic factors in panic disorder appear to be sex-linked.

Environmental events occurring during particular devel-opmental phases such as separation from the primary caregiver during early childhood may activate the genes that modulate the fear network, thereby creating a vulnerability to panic disorder. Research suggests that later events, occurring during adolescence or early adulthood, then precipitate panic disorder in vulnerable individuals. These events may stress the individual at a psycho-logical or physiological level. Events commonly associated with the onset of panic disorder include:

·        separation, loss, or illness of a significant other;

·        being the victim of sexual assault or other forms of interper-sonal violence;

·        financial or occupational stressors; and

·        intoxication with, or withdrawal from, a psychoactive sub-stance such as marijuana, cocaine, or anesthetic

Dynamic Models

The most promising psychodynamic models for understanding panic disorder are those that focus specifically on this disorder. Rather than review all the models, we will summarize the model developed by the Cornell Panic-Anxiety Study Group (Milrod et al., 1997; Shear et al., 1993) because it has led to a promis-ing treatment. According to the Cornell group, people at risk for panic disorder have 1) a neurophysiological vulnerability to panic attacks, and/or 2) multiple experiences of developmental trauma. These factors lead the child to become frightened of unfamiliar situations and to become excessively dependent on the primary caregiver to provide a sense of safety. The caregiver is unable to provide support always, so the child develops a fearful depend-ency. This leads, in turn, to the development of unconscious con-flicts about dependency (independence versus reliance on others) and anger (expression versus inhibition). The dependency conflict is said to express itself in a number of ways. Some panic-vulnera-ble people are sensitive to separation and overly reliant on others, while others are sensitive to suffocation and overly reliant on a sense of independence. These conflicts can activate conscious or unconscious fantasies of catastrophic danger, which can trigger panic attacks. In addition, the conflicts evoke aversive emotions, such as anxiety, anger and guilt. The otherwise benign arousal sensations accompanying these emotions can become the focus of “conscious as well as unconscious cognitive catastrophizing”, thereby leading to panic attacks.