Mood Disorders: Neurobiology, Evaluation, and Treatment

Transcription

1 neurology Board Review Manual Statement of Editorial Purpose The Hospital Physician Neurology Board Review Manual is a peer-reviewed study guide for residents and practicing physicians preparing for board examinations in neurology. Each manual reviews a topic essential to the current practice of neurology. PUBLISHING STAFF PRESIDENT, Group PUBLISHER Bruce M. White editorial director Debra Dreger EDITOR Tricia Faggioli, ELS assistant EDITOR Farrawh Charles executive vice president Barbara T. White executive director of operations Jean M. Gaul PRODUCTION Director Suzanne S. Banish PRODUCTION assistant Nadja V. Frist ADVERTISING/PROJECT director Patricia Payne Castle sales & marketing manager Deborah D. Chavis NOTE FROM THE PUBLISHER: This publication has been developed without involvement of or review by the American Board of Psychiatry and Neurology. Mood Disorders: Neurobiology, Evaluation, and Treatment Editor: Alireza Atri, MD, PhD Instructor in Neurology, Harvard Medical School; Assistant in Neurology, Massachusetts General Hospital, Boston, MA; Associate Director, Center for Translational Cognitive Neuroscience, Geriatric Research Education and Clinical Center, VA Medical Center, Bedford, MA Associate Editor: Tracey A. Milligan, MD Instructor in Neurology, Harvard Medical School; Associate Neurologist, Brigham and Women s and Faulkner Hospitals, Boston, MA Contributor: William R. Marchand, MD Acting Associate Director, Department of Veterans Affairs VISN 19 MIRECC; Assistant Professor, Department of Psychiatry, University of Utah School of Medicine; Adjunct Assistant Professor, Department of Psychology, University of Utah, Salt Lake City, UT Table of Contents Introduction Neural Circuits Involved in Regulating Emotion Neurobiology of Mood Disorders Principles of Evaluation and Treatment of Mood Disorders... 9 Conclusion References Cover Illustration by Kathryn K. Johnson Copyright 2008, Turner White Communications, Inc., Strafford Avenue, Suite 220, Wayne, PA ,. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, mechanical, electronic, photocopying, recording, or otherwise, without the prior written permission of Turner White Communications. The preparation and distribution of this publication are supported by sponsorship subject to written agreements that stipulate and ensure the editorial independence of Turner White Communications. Turner White Communications retains full control over the design and production of all published materials, including selection of topics and preparation of editorial content. The authors are solely responsible for substantive content. Statements expressed reflect the views of the authors and not necessarily the opinions or policies of Turner White Communications. Turner White Communications accepts no responsibility for statements made by authors and will not be liable for any errors of omission or inaccuracies. Information contained within this publication should not be used as a substitute for clinical judgment. Neurology Volume 12, Part 6

2 Neurology Board Review Manual Mood Disorders: Neurobiology, Evaluation, and Treatment William R. Marchand, MD INTRODUCTION Two categories of mood disorders are currently recognized in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) unipolar (depressive) spectrum disorders and bipolar spectrum disorders. 1 Despite the high prevalence of depression and the significant burden of suffering caused by unipolar and bipolar syndromes, the neurobiology of mood disorders remains incompletely understood. However, the recent development of functional neuroimaging methods has allowed the characterization of neurobiologic correlates of mood disorders and has contributed to evolving theories of causation. In particular, findings from functional neuroimaging studies combined with evidence from postmortem lesion, molecular biology, and genetic studies increasingly implicate neural circuits that regulate emotional behavior in the pathophysiology of mood disorders. Neural networks allow the functional integration of multiple brain regions to facilitate neural processing. There is considerable evidence that some networks function abnormally in mood disorders. This may occur as a result of underlying molecular pathology. In some cases, however, neural network dysfunction associated with mood disorders also might represent the brain s response or adaptation to illness or be the result of changes related to illness progression. Nonetheless, functional abnormalities of neural networks may be the final common pathway of symptom development and eventually may provide clinically useful biomarkers of illness. This manual provides an overview of our current understanding of the neurobiology of mood disorders, with a particular focus on abnormalities of neural circuitry associated with these conditions. The aim of this manual is to provide a basic understanding of the current state of knowledge about the etiology of these conditions as well as review the fundamentals of evaluation and treatment. NEURAL CIRCUITS INVOLVED IN REGULATING EMOTION In 1937, James Papez, a neuroanatomist at Cornell University, described a neural circuit for regulation of emotion in what had previously been labeled the limbic lobe by the French neuroanatomist, Paul Broca. Over time, the concept of an emotional control circuit has been revised, and this circuit is generally referred to as the limbic system (Figure 1). Our current understanding is that limbic functions are best described by 2 neural networks that are critical for human affect generation and modulation the ventral and the dorsal emotional control networks. 2 VENTRAL AND DORSAL NETWORKS The ventral network, which includes the ventral prefrontal cortex, amygdala, insula, ventral striatum, thalamus, orbitofrontal cortex, ventral anterior cingulate cortex, and brainstem nuclei, is involved with the perception of emotional stimuli, the generation of affect, and the production of an autonomic response. 2 These functions occur primarily by way of the amygdala, which receives sensory inputs from multiple areas. External sensory inputs provide information about the environment and include olfactory input from the olfactory bulb, visual information from the inferior temporal cortex, and auditory information from the thalamus. Internal information about the current state of bodily functions comes from the hypothalamus. The outflow of information from the amygdala provides control over the body s response to the environment. For example, in response to potentially dangerous situations, amygdala output results in sympathetic nervous system activation (fight or flight response) by way of the hypothalamus. Also, output to the reticular pontine nuclei leads to the startle response, and output to the central gray nucleus leads to the freezing behavior seen in fear states. Connections from the amygdala to the hypothalamus result in adrenocorticotropin release and the stress hormone Hospital Physician Board Review Manual

3 Ventral striatum Hippocampus Cortex Cortex Amygdala Thalamus GPe D 2 D 1 Input nucleus (striatum) Ventral tegmental area Locus ceruleus Thalamus Hypothalamus Input nucleus (STN) SNc & VTA Figure 1. General organization of the limbic circuits and major connections. The amygdala is the structure most involved with formulating emotional response. The cognitive experience of emotion is mediated by connections from the amygdala to the cortex. Amygdala input to the locus ceruleus and ventral tegmental area control the release of norepinephrine and dopamine, respectively. response of corticosteroid release. The cognitive experience of emotion is mediated by connections from the amygdala to the cingulate gyrus and orbitofrontal cortex. Finally, amygdala input to the locus ceruleus and ventral tegmental area control norepinephrine release and dopamine release, respectively. The dorsal network includes the dorsolateral prefrontal cortex, medial prefrontal cortex, dorsal anterior cingulate cortex, and hippocampus. In contrast to the rapid unconscious generation of emotion by the ventral network, the dorsal circuit consists primarily of cognitive brain regions and is important for effortful conscious regulation of resulting affective states. 2 CORTICOSTRIATAL CIRCUITS Neurotransmitters: Glutamate GABA Dopamine Output nuclei (GPi/SNr) Figure 2. General organization of the corticostriatal circuits. The direct pathway is shown in red, and the indirect pathway is shown in gray. D = dopamine; GABA = g-aminobutyric acid; GPe = globus pallidus external segment; GPi = globus pallidus internal segment; SNc = substantia nigra pars compacta segment; SNr = substantia nigra reticular segment; STN = subthalamic nucleus; VTA = ventral tegmental area. (Adapted from Marchand WR, Lee JN, Thatcher JW, et al. Motor deactivation in the human cortex and basal ganglia. Neuroimage 2007;38: Copyright 2007, with permission from Elsevier.) The corticostriatal circuitry also plays a role in emotional processing. The ventral striatum is a component of the ventral network, and the corticostriatal emotional processing circuits provide excitatory or inhibitory feedback to cortical components of both the ventral and dorsal networks. The corticostriatal circuits are composed of the cortex, basal ganglia, and thalamus (Figure 2). The basal ganglia structures consist of large numbers of cell bodies organized into nuclei and include the caudate nucleus, putamen, nucleus accumbens, substantia nigra (pars compacta and pars reticulata), globus pallidus (pars interna and pars externa), and subthalamic nucleus. Five parallel corticostriatal subcircuits are generally recognized: the skeletomotor and oculomotor circuits (which modulate motor control), and the dorsolateral prefrontal, orbitofrontal, and anterior cingulate circuits (which play a role in executive function, limbic control, and motivated behavior, respectively). 3 Corticostriatal information processing occurs via signals that originate in the cortex, pass through basal ganglia to the thalamus, and then flow back to the cortex. Models of corticostriatal circuit function have been based on the 2-pathway hypothesis. 3,4 This hypothesis suggests that there are 2 parallel and opponent pathways direct and indirect which exist in a dynamic balance and provide a mechanism by which the 5 corticostriatal subcircuits can modulate the cortex. The direct pathway tends to have a net disinhibitory effect on the thalamus and increases excitatory outflow to the Neurology Volume 12, Part 6

4 cortex, while the indirect pathway tends to inhibit the thalamus and decreases excitatory drive to the cortex. The direct and indirect pathways are thought to have reciprocal functions, with the direct pathway being responsible for sustaining and the indirect pathway for halting various behavioral routines. 4 In addition to cortical input, the ventral striatal regions receive amygdala projections. 5 8 Amygdala input to the striatum modulates corticostriatal feedback to cortical regions and contributes to the mediation of goal-directed behaviors. 5 7 In addition to amygdala input, nucleus accumbens output neurons receive input from the hippocampus, cingulate gyrus, and prefrontal cortex as well as from the ventral tegmental area. 9,10 Therefore, the corticostriatal circuits integrate information from the amygdala, hippocampus, and cortex at the level of the striatum and then provide either excitatory or inhibitory modulation of cortical components of the ventral and dorsal emotional control networks. NEUROBIOLOGY OF MOOD DISORDERS Historically, theories of mood disorder pathology have focused on abnormalities of monoaminergic neurotransmitter systems. 11 Monoaminergic involvement was initially hypothesized when monoamine-depleting agents (eg, reserpine) were found to produce depressive symptoms. Additional evidence for the monoamine theory came from the finding that monoamine oxidase inhibitors and tricyclic antidepressants increased synaptic levels of norepinephrine and serotonin. Specifically, the theory postulated that depression was the result of insufficient monoaminergic transmission, and treatment was achieved by normalizing serotonin and norepinephrine levels in the synapse. Conversely, mania was explained as a relative increase in monoamine activity. It is now well accepted that this theory is an oversimplification and that mood disorder neuropathology must be considered at the level of the gene, the neuron, neurotransmitters, and neural circuits. 11 genetic and other RISK FACTORS Twin and family studies show that genetic susceptibility plays a significant role in the development of mood disorders. 12 However, the evidence suggests that complex interactions occur between multiple genes, with each exerting small effects on vulnerability. 12 Thus, it has been difficult to determine exactly how a specific genetic vulnerability results in expression of unipolar or bipolar symptoms. In addition, mood disorders likely represent heterogeneous neurobiologic conditions, with several neurobiologic abnormalities resulting in the symptom complexes recognized as either major depression or bipolar disorder. 12 This heterogeneity of genetic vulnerability and neurobiology likely account for the fact that pharmacologic treatments can be effective for some individuals with a mood disorder and ineffective for others. Finally, psychological and medical stressors constitute acquired risk factors that interact with genetic vulnerability in the development of mood disorders. 12 Eventually, the identification of biomarkers associated with subtypes of illness and/or endophenotypes may allow the identification of disorders based on neurobiology rather than symptom-based diagnoses. INTRACELLULAR SIGNALING PATHWAY ABNORMALITIES There is growing evidence to suggest that mood disorders may, at least in part, involve impaired neuroplasticity at the level of the neuron. Much research has focused on abnormalities of intracellular signaling pathways in mood disorders, 11,13,14 and these pathways may serve as a primary link between genetic vulnerability and the eventual development of illness. Complex intracellular pathways allow the neuron to process and respond to stimuli as well as modulate the signal generated by chemical neurotransmission. 11 These pathways also are involved in the regulation of neuroplasticity and cellular resilience by way of neurotrophic factors, 11 such as brain-derived neurotrophic factor (BDNF). Neurotrophic factors are necessary for survival and function of the neuron 15 and promote the expression of cytoprotective proteins, such as B cell lymphoma protein 2 (Bcl-2). Evidence for impaired neuroplasticity in mood disorders includes the association of these disorders with brain changes thought to result from neuronal damage and death. Structural imaging and postmortem brain studies have demonstrated reductions in gray matter volumes, glial cell counts, and neuron size in the prefrontal cortex, ventral striatum, hippocampus, and amygdala of mood-disordered individuals. 11,16,17 Disruption of intracellular signaling mechanisms and impaired neuroplasticity may occur as a result of underlying genetic vulnerabilities; however, stress-induced cellular injury and neuronal damage from multiple mood episodes could also play a role. 17 Stress (psychosocial, physiologic, medical) can lead to impairment of cellular resilience, 11 and it is well known that stress can precipitate episodes of mood symptoms. 18 In some cases, stress may lead to neuronal damage in individuals genetically predisposed to develop mood disorders. In rodents, stress can lead to atrophy and death of hippocampal neurons, 19 and in humans depression is associated with hippocampal atrophy. 20 Stress-induced neuronal atrophy is believed to be at least partially mediated by Hospital Physician Board Review Manual

5 activation of the hypothalamic-pituitary-adrenal axis and resultant high plasma concentrations of glucocorticoids. 19,20 Glucocorticoids may lead to cellular atrophy by facilitation of glutamatergic signaling, inhibition of glucose transport, and impairment of neurogenesis These mechanisms could account for the role of stress in the onset of mood episodes. Finally, multiple mood episodes may lead to further neuronal damage. Perhaps the most convincing evidence for the neuroplasticity hypothesis in mood disorders is the influence of antidepressants and mood stabilizers on neurotrophic pathways. 14 Antidepressant administration increases levels of serotonin and norepinephrine at the level of the synapse. It was historically thought that this increase in neurotransmitter levels was the primary mechanism of action of antidepressant agents. However, it is now known that increased synaptic concentrations of serotonin and norepinephrine also activate intracellular signal transduction cascades, 14 causing enhanced expression of BDNF and its receptor. 14,21,23 Thus, enhancement of neurotrophic mechanisms may constitute part or all of the mechanism of action of antidepressants. There is also evidence that these neurotrophic effects result in regeneration of catecholamine axon terminals in the cortex and enhanced synaptic plasticity in the hippocampus and may attenuate hippocampal atrophy Additional evidence that upregulation of BDNF may contribute to the therapeutic mechanism of antidepressants is the fact that infusion of BDNF into the midbrain has an antidepressant-like influence on animal models of depression. 24,25 In regard to mood-stabilizing medications, studies have shown that lithium increases the levels of Bcl-2 in the rodent brain and in cells of human neuronal origin. 26,27 In human clinical studies, lithium treatment has been shown to increase a marker of neuron function in gray matter 28 and to increase total gray matter content in the brain. 29 The mood stabilizer valproate also activates Bcl-2, and both valproate and lithium increase the expression of BDNF. 14 Taken together, these findings constitute compelling evidence that mood disorders are, at least in part, disorders of neuroplasticity and that many treatments for these disorders act by enhancing neuronal health and survival. Table 1 summarizes the evidence supporting the neuroplasticity hypothesis of mood disorders. NEUROTRANSMITTER ABNORMALITIES Historically, serotonin, norepinephrine, and dopamine were the primary neurotransmitters implicated in mood disorders. It is now known that other neurotransmitters, including acetylcholine, histamine, γ-aminobutyric acid (GABA), and glutamate, also contribute to the pathophysiology of these disorders. Table 1. Evidence for Impaired Neuroplasticity in Mood Disorders Psychosocial stress often precipitates mood episodes and is linked to impaired neuroplasticity Brain atrophy in mood disorders is likely secondary to impaired neuroplasticity Antidepressants increase the neurotrophic factor BDNF Mood stabilizers increase BDNF and the cytoprotective protein Bcl-2 Transcranial magnetic stimulation and electroconvulsive therapy increase BDNF BDNF = brain-derived neurotrophic factor. (Data from references 11, ) Serotonin The serotonin system originates in the raphe nuclei of the midbrain and projects throughout the brain. In regard to the role of this neurotransmitter in mood disorders, projections to the amygdala and hippocampus are of particular interest. Multiple studies implicate serotonin abnormalities in patients with depression. Studies have found decreased serotonin uptake in the platelets of depressed patients. 30 Further, depletion of serotonin precursors can reverse antidepressantinduced remission among those with depression, 31 and serotonin precursors may have antidepressant effects. 32 Suicidal behavior has also been associated with alterations in the serotonin system. For example, a postmortem study of suicide victims who suffered from major depression revealed a significant decrease in the number of serotonin binding sites in the hippocampus, 33 while another study revealed increased serotonin binding sites in the frontal cortex of suicide victims. 34 Another postmortem study showed that the total number of serotonin receptors was lower in the dorsal raphe nucleus of suicide victims compared with controls. 35 Thus, while our understanding of serotonergic function in mood disorders is incomplete, there is compelling evidence of dysfunction in major depression and evidence for a specific role in suicidal behavior. Norepinephrine The norepinephrine system primarily originates in the locus ceruleus and projects to brain regions involved in emotional control, such as the hippocampus, amygdala, and cortex. While evidence suggests that some antidepressants impact norepinephrine function, 36,37 there has been no consistent relationship between norepinephrine metabolites in the cerebrospinal fluid (CSF), serum, or urine of depressed patients. 37 Thus, while it is likely that norepinephrine dysfunction plays a role in major depression, the exact mechanism Neurology Volume 12, Part 6

6 is incompletely understood. In regard to bipolar disorder, there is some evidence that lithium may impact norepinephrine function. 38 Dopamine Three dopaminergic subsystems regulate motor activity and cognitive functioning. Of these, the mesolimbic system, which originates in the ventral tegmental area and projects to the ventral striatum and prefrontal cortex, is involved with regulation of emotion and response to reward. The best evidence supporting a role of dopamine in unipolar depression is from a recent review of the available literature by Dunlop and Nemeroff. 39 Studies reviewed include postmortem investigations demonstrating reduced concentrations of dopamine metabolites in both the CSF and the brain of depressed patients, as well as studies showing altered response of depressed subjects to dextroamphetamine, indicating abnormal dopamine signaling. The authors also cite studies showing the efficacy of medications that directly enhance dopaminergic neurotransmission (eg, monoamine oxidase inhibitors, pramipexole) for the treatment of depression. The authors conclude that the evidence clearly supports diminished dopaminergic neurotransmission in the pathophysiology of unipolar depression. The best evidence of a possible role of dopamine in bipolar disorder comes from a recent review by Anand and Charney 40 and a recent study by Anand and colleagues. 41 Anand and Charney 40 conclude that multiple studies suggest an increase in dopamine and other catecholamines in mania and a decrease in bipolar depression. Further, bipolar disorder may involve dysregulation of the corticostriatal circuit involved in mood regulation. 40 GABA and Glutamate GABA is the main inhibitory neurotransmitter, and glutamate is the main excitatory neurotransmitter. Both GABA and glutamate have been implicated in mood disorders. Some studies suggest that lower plasma concentrations of GABA are present in patients with unipolar depression. 42,43 Decreased synthesis and release of GABA also occur in response to acute and chronic stress. Recent studies have reported higher plasma concentrations of glutamate in patients with major depression 44 and increased levels of glutamate in the left dorsolateral prefrontal cortex of patients with acute mania. 45 Acetylcholine Acetylcholine is the neurotransmitter used by the cholinergic system. This system projects from the basal forebrain nucleus to almost all portions of the cortex. As a consequence, acetylcholine exerts multiple effects on neuronal and mental functioning. There is some evidence suggesting abnormalities of acetylcholine in mood disorders, 46 although further research is needed. The exact mechanism by which cholinergic dysfunction could lead to mood symptoms is poorly understood but could relate to the role of acetylcholine in modulating monoamine neurotransmission, particularly dopamine. 47 Histamine Histaminergic neurons originate in the posterior hypothalamus and project to the limbic system and neocortex. These neurons regulate learning and memory, endocrine homeostasis, the sleep-wake cycle, appetite control, and emotion. 48 High levels of histamine have been associated with depressive symptoms, and recent research suggests that decreased histamine H 1 receptor binding may correlate with the severity of depressive symptoms. 49 NEURAL CIRCUITRY DISRUPTION Evidence increasingly suggests that the neurocircuitry involved in the regulation of normal human emotional experience functions abnormally in mood disorders. 2,12,50 Structural imaging studies reveal abnormalities in regions within both the ventral and dorsal emotional control networks. Many morphologic abnormalities found in these regions are similar in both unipolar and bipolar illness, such as nonspecific signs of atrophy and similar volumetric abnormalities in prefrontal, cingulate, and temporal structures. 12 Also, reduced gray matter has been reported in the anterior cingulate cortex and medial frontal regions in bipolar disorder, in affective disorders in general, and in the hippocampus and left amygdala in major depressive disorder. 55 However, the striatum tends to be smaller in major depressive disorder and larger in bipolar illness, 56 suggesting one area where structural differences between disorders may exist. In recent years, functional neuroimaging studies have also revealed evidence of disruption of the dorsal and ventral emotional control networks as well as dysfunction of the corticostriatal circuitry in mooddisordered patients. Mood dysregulation could theoretically occur as a result of dysfunction in the dorsal or ventral network or both. Dysregulation of the ventral network could result in the pathologic mood states of mania and depression at the level of affect generation. Impairment in the dorsal network could result in mood episodes as a result of inadequate modulation of ventral Hospital Physician Board Review Manual

7 mechanisms. Table 2 summarizes key functional neuroimaging findings in mood disorders. Abnormalities of the Ventral Emotional Control Network There is considerable evidence that the amygdala is functionally abnormal in both bipolar and major depressive disorders. Amygdala overactivation appears to be a characteristic during depressive episodes in both conditions, although the mechanism underlying this phenomenon is incompletely understood. Functional imaging studies have provided compelling evidence of increased as well as sustained 61,62 amygdala activation in patients with major depressive disorder as compared with control subjects. Further, amygdala hyperactivity has been shown to be a neural substrate of negatively biased automatic emotion processing that is characteristic of depression, 59 and severity of depression has been shown to correlate with the activity of the left amygdala. 63 Left amygdala hyperarousal as demonstrated by functional magnetic resonance imaging (MRI) has also been shown to normalize with antidepressant treatment. 64 One neuroimaging study demonstrated that disrupted top-down control by the prefrontal cortex of the amygdala may underlie the abnormal response to negative feedback in unipolar depression. 65 Also, a functional MRI study found subjects with major depressive disorder showed greater functional connectivity between the amygdala and subgenual cingulate (ventral system) and less between the amygdala and supragenual cingulate (dorsal system) as compared with controls. 66 Further, the study revealed that greater depressive symptom severity correlated positively with decreased functional connectivity between bilateral amygdala and supragenual cingulate. 66 Finally, another functional MRI study revealed that patients with major depressive disorder had increased activation of cortical and limbic regions and decreased functional connectivity between the anterior cingulate cortex, amygdala, striatum, and thalamus. 67 Taken together, these findings suggest that loss of control of the amygdala by the dorsal system could be the underlying cause of exaggerated amygdala response in major depressive disorder. Functional neuroimaging studies have also revealed amygdala dysfunction in bipolar disorder Studies of homogeneous mood states (ie, all study subjects in the same phase of illness) have reported increased 71 and decreased 72 amygdala activation in response to emotional tasks during mania. One study reported increased activation in response to an emotional task during depression. 75 During euthymia, studies have Table 2. Summary of Functional Abnormalities of Emotional Control Circuitry in Mood Disorders Amygdala activation is increased in both unipolar and bipolar depression Abnormal amygdala output likely results in the development of symptoms of depression Amygdala dysregulation may be the result of deficient prefrontal modulation Striatal function is significantly different in unipolar versus bipolar disorder There is compelling evidence of significant disruption of both the dorsal and ventral emotional control networks in both major depressive disorder and bipolar disorder shown increased 69 and decreased 68 amygdala activation in response to emotional tasks and increased activation in response to a cognitive task. 77 Studies have also shown disruption of the orbitofrontal cortex, insula, and ventral prefrontal cortex in major depressive disorder and bipolar disorder. Increased orbitofrontal activation has been reported in patients with major depressive disorder, 78 and severity of depression has been shown to correlate with the activity of the bilateral inferior orbitofrontal areas. 63 A morphometry functional MRI study found an association between structural alterations in the orbitofrontal cortex and disturbed functional activation in the emotional compartment of the anterior cingulate cortex. 79 Studies of bipolar depression have shown some orbitofrontal regions with increased activation and other areas with decreased activation. 80 However, in studies of mania, only decreased activation has been reported during cognitive 80 and emotional 71,81 tasks. In insular regions, both increased and decreased activation have been reported in major depressive disorder. 57,63 Studies of bipolar disorder have reported increased insula activation in both mania 72 and depression 75 in response to emotional tasks and in euthymia in response to a cognitive task. 77 Finally, several studies have demonstrated abnormal functioning of the ventral prefrontal cortex in bipolar disorder. Functional MRI studies of homogeneous mood states have revealed increased 77,82 and decreased activation of the ventral prefrontal cortex in euthymia in response to cognitive tasks. However, studies using emotional paradigms have consistently demonstrated decreased activation in euthymia. 68,69,87 In mania, a study involving an emotional task revealed increased activation, 81 while a cognitive paradigm revealed decreased activation of the ventral prefrontal cortex. 88 Studies of depression have reported increased Neurology Volume 12, Part 6

8 activation of the ventral prefrontal cortex in response to emotional 75,89 and cognitive 88 tasks. Abnormalities of the Dorsal Emotional Control Network Studies of major depressive disorder have revealed decreased 61,90 and increased 91 dorsolateral prefrontal cortex activity. There is also evidence of left and right dorsolateral prefrontal imbalance in major depressive disorder, with left dorsolateral prefrontal hypoactivity and right dorsolateral prefrontal hyperactivity. 92 In bipolar disorder, increased activation of the dorsolateral prefrontal cortex has been reported in response to both motor 93 and emotional 75 paradigms. However, decreased activation in response to an emotional paradigm has also been reported. 80 In regard to dorsolateral prefrontal cortex function in the euthymic state, increased activation has been reported by studies using motor 94 and emotional 69 tasks, whereas decreased activation has been reported in response to cognitive 83 86,95,96 and emotional 87 paradigms. The only study of mania reported increased activation in response to an emotional task. 81 In regard to other dorsal system regions, a study of major depressive disorder revealed increased activation of the dorsal anterior cingulate cortex. 57 Further, increased 58 and decreased 63 hippocampal activation has been reported in major depressive disorder, while increased activation in depression 80 and in euthymia 69 has been reported in bipolar disorder. Finally, increased activation of the left medial prefrontal cortex has been reported in major depressive disorder, 97 and both increased 83,86 and decreased 96 activation have been reported in euthymic bipolar disorder. Abnormalities of the Corticostriatal Circuits Several authors have hypothesized that corticostriatal circuit abnormalities may play a direct role in the etiology of both unipolar depression 16,98,99 and bipolar disorder. 98, These theories derive in part from studies of individuals who develop secondary mood symptoms after suffering a focal brain lesion. Lesions of corticostriatal circuits produce circuit-specific behavioral syndromes, 98,103,104 and 3 syndromes produce symptoms similar to those seen in affective illness. The anterior cingulate syndrome, sometimes termed akinetic mutism in its most severe form, produces profound apathy, motor and verbal inactivity, and indifference to thirst or hunger. 98,103 Less severe forms result in loss of motivation, psychomotor slowing, and blunted affect. 104 This syndrome has similarities to depression. The orbitofrontal syndrome involves behavioral disinhibition and labile emotions 98,103 and has many similarities to mania. Finally, the dorsolateral prefrontal syndrome includes symptoms of executive dysfunction, such as difficulty focusing and sustaining attention as well as reduced verbal fluency and motor programming. 103 Cognitive symptoms are associated with both unipolar and bipolar illness. Thus, all 3 syndromes involve symptoms seen in mood disorders, suggesting that corticostriatal dysfunction could play a prominent role in the etiology of affective illness. Dopamine is the principal modulator of striatal circuit function. 40 This occurs by way of dopaminergic projections from the ventral tegmental area to the ventral striatum and the substantia nigra pars compacta to the dorsal striatum. 104 Evidence suggests diminished dopaminergic neurotransmission in major depressive disorder, 39 while a recent study by Anand and colleagues 41 provides evidence of enhanced dopaminergic modulation of the striatum in bipolar disorder. These findings would predict striatal hyperactivation in bipolar disorder and hypoactivation in major depressive disorder. Our previous work comparing bipolar depressed subjects with controls using a motor activation paradigm found increased striatal activation in bipolar depressed subjects, 93 which persisted in the euthymic state. 94 Caligiuri and colleagues 105 also reported increased activation in basal ganglia regions in both depressed and manic bipolar subjects in response to a motor task. Most functional neuroimaging studies of bipolar disorder have used cognitive or emotional tasks, and many but not all of these have reported increased striatal activation. 106 In regard to studies of unipolar subjects, many studies show decreased striatal activation These findings suggest that corticostriatal modulation of the dorsal and ventral emotional control systems may be fundamentally different in bipolar versus unipolar spectrum disorders. Summary In summary, functional neuroimaging studies have provided important insight into the neurobiology of mood disorders by revealing evidence of dysfunction of the dorsal and ventral emotional control as well as corticostriatal circuits. However, the interpretation of these studies requires consideration of important confounding variables. One potential confounder is that mood-disordered subjects sometimes do not perform as well as normal controls on cognitive tasks typically used as activation paradigms in functional neuroimaging studies. 95,110 If task performance differs between groups, differences in brain activation may reflect the diminished ability of patients to complete the task instead of actual differences in brain function. 77 Also, studies of patients who are experiencing a mood episode may reveal differences in brain activation that represent epiphenomena Hospital Physician Board Review Manual

9 of that mood state but are not a trait of the disorder. 77 Further, the impact of medication on functional studies is a significant concern. Most studies have examined medicated subjects, which introduces significant potential confounding. Finally, a variety of different functional imaging methodologies and study populations have been used, which makes comparisons of results across studies challenging. Thus, while these studies provide useful information about neurocircuit dysfunction in mood disorders, they do not yet reveal a mechanism by which circuit abnormalities result in the expression of specific clinical symptoms. PRINCIPLES OF EVALUATION AND TREATMENT OF MOOD DISORDERS CASE 1 PRESENTATION A 23-year-old man presents with the complaint that that he has been feeling depressed for 2 months. The history reveals symptoms of dysphoria, hypersomnia, weight gain, and anhedonia. The patient reports that he has experienced similar symptoms in the past, particularly during the changing of the seasons, but he tried to shrug them off and eventually they went away. However, he is now in a serious relationship, and his girlfriend has expressed concern about his depression. He admits that he also is worried, because he has stopped doing many of the things he enjoys and sometimes can t even get motivated to get out of bed. The patient reports no history of heart disease, hypertension, diabetes, or other major medical conditions. He also denies any history of headache, change in speech, paralysis, vertigo, weakness, or loss of consciousness. The general physical examination is unremarkable. The patient is alert and oriented to time, place, person, and situation. Language, recall, concentration, and judgment are intact, although the patient appears to exhibit mild psychomotor slowing. Neurologic evaluation reveals mild bradykinesia. Cranial nerves are intact. Strength is normal and symmetric. There is no evidence of atrophy or fasciculations. Stance and gait are normal. Reflexes are symmetric. Babinski sign is absent bilaterally. What further information is important to obtain in the assessment of this patient? EVALUATING THE PATIENT WITH MOOD SYMPTOMS Table 3 lists key elements in the evaluation of patients with mood symptoms, and Table 4 summarizes the major Table 3. Assessment of Patients with Mood Symptoms Obtain symptoms and duration of current manic or depressive episode Inquire about past manic, hypomanic, or depressive episodes Assess for substance abuse Determine if there is a family history of mood disorders Ask about past treatment with psychiatric medications and response Evaluate for risk of harm to self or others criteria required for diagnosis of mood disorders. This patient presents with symptoms that clearly warrant a diagnosis of a major depressive episode. He meets the DSM-IV-TR criteria of having at least 5 depressive symptoms that have persisted for at least 2 weeks and resulted in significant distress or impairment. 1 However, an episode of depression can occur with either major depressive disorder or bipolar disorder. Bipolar spectrum illness is characterized by episodes of mood elevation as well as episodes of depression in most cases; therefore, bipolar conditions must always be ruled out during an evaluation of a depressive episode. In this case, the patient relates several symptoms suggestive of bipolar depression (Table 5), making this line of inquiry even more important. In particular, the patient should be asked if he has ever experienced symptoms of mania or hypomania (Table 6). There is strong evidence that bipolar disorder is underdiagnosed. Ghaemi and colleagues 111 found that bipolar disorder was misdiagnosed as unipolar depression in 40% of hospitalized patients. At least 2 studies have found that 34% of patients with bipolar spectrum disorder wait at least 10 years after symptom onset before a correct diagnosis is given. 112,113 The delay in receiving an accurate diagnosis has a substantial adverse result for many bipolar patients, including unnecessary suffering caused by not receiving treatment that adequately controls their symptoms. Another even more concerning result of misdiagnosis is that many patients receive antidepressants for what is thought to be unipolar depression. This can lead to a worsening of mood cycling and exacerbation of manic symptoms. 111 Alhough the mechanisms underlying this phenomenon are poorly understood, it likely reflects distinct differences in the neurobiology of unipolar versus bipolar spectrum disorders. CASE 1 CONTINUED Further evaluation reveals that the patient has had episodes of mood symptoms for about 5 years. Most of the episodes have been depression lasting from a few days to several months. However, when asked specifically about manic episodes, he also reports Neurology Volume 12, Part 6

10 Table 4. Mood Disorder Episode Criteria Disorder Depressive Episodes Manic or Mixed Episodes Hypomanic Episodes Bipolar I disorder Common but not required 1 required Common but not required Bipolar II disorder 1 required None allowed 1 required Cyclothymic disorder Required, but not major depression None allowed Numerous periods over 2 years required Bipolar disorder NOS Common but not required None allowed Required, but do not meet criteria for a specific bipolar disorder Major depressive disorder 1 required None allowed None allowed Dysthymic disorder 2 years required but not major depression None allowed None allowed NOS = not otherwise specified. (Data from American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. text revision. DSM-IV-TR. Washington [DC]: The Association; 2000: ) Table 5. Presentation of Depressed Patients Who May Have Bipolar Disorder Hyperphagia/weight gain Hypersomnia Melancholic features Severe anhedonia Seasonal mood changes Psychomotor slowing Psychotic features History of poor response to antidepressants History of recurrent but brief depressive episodes History of antidepressant-induced mania or hypomania Family history of bipolar illness in a first-degree relative Early age of onset Postpartum onset a 2-week period during the past year when his mood was too high, but he felt great and had so much energy he needed only 2 or 3 hours of sleep a night. He recalls that his friends questioned why he was in such a good mood. He reports that, during this period, his thoughts were going very fast, he had trouble staying focused and paying attention, and his friends complained that he talked too fast. He also reports hearing voices on and off. Finally, he states that he got in a lot of trouble during that time, because he charged thousands of dollars to his credit card while on a shopping spree. As a result, he has been in debt and having trouble paying his rent. The patient reports no other episodes of this nature. When asked whether he knows if any family members have had similar experiences, he notes that his father has behaved similarly at times but has never been diagnosed with or treated for a mood disorder. Is this sufficient clinical evidence to make a diagnosis of bipolar disorder? What other symptoms should the physician inquire about during this visit? DIAGNOSIS OF BIPOLAR DISORDER The DSM-IV-TR describes 3 types of manic episodes manic, hypomanic, and mixed episodes. 1 Table 6 lists specific symptoms seen during the 3 states. All require an abnormally elevated, expansive, or irritable mood. Associated symptoms are grandiosity, decreased need for sleep, increased speech, flight of ideas or racing thoughts, distractibility, increased activity, and excessive involvement in pleasurable activities. Four subtypes of bipolar disorder are described in the DSM-IV-TR, which are based on the existence and severity of manic, mixed, and depressive episodes. 1 These are bipolar I disorder, bipolar II disorder, cyclothymic disorder, and bipolar disorder not otherwise specified (NOS). The case patient has bipolar I disorder. A diagnosis of bipolar I disorder requires at least 1 manic or mixed episode (Table 4). The previous episode the patient described meets criteria for a manic episode (Table 6). Although patients with bipolar disorder typically experience multiple manic episodes over the course of their illness, only 1 is required to establish the diagnosis. Like this patient, virtually all patients with bipolar disorder experience long periods of depression. 114 Depression almost always causes more subjective distress than mania, and this results in patients typically seeking help when they are depressed rather than when manic. Also, patient insight is more impaired in mania than in depression; thus, reliance on patients self-report about their own manic symptoms might contribute to underdiagnosis of mania. In this case, if the physician had not inquired about a history of mania or hypomania, the patient could easily have been misdiagnosed as having recurrent major depressive disorder. 10 Hospital Physician Board Review Manual

11 Table 6. Symptoms of Manic, Hypomanic, and Mixed States Diagnostic Criteria Manic Episode Hypomanic Episode Mixed Episode Mood state Elevated or irritable Same as for manic episode Same as for manic episode Coexisting depression Associated symptoms (3 required or 4 if the mood is irritable but not elevated) Duration of episode required for diagnosis Impairment Grandiosity Decreased sleep Increased speech Flight of ideas Distractibility Increased activity Excessive pleasure-seeking Same as for manic episode 1 week or hospitalization At least 4 days At least 1 week Marked impairment or hospitalization or psychosis Does not cause marked impairment Same as for manic episode Also meets criteria for a major depressive episode Same as for manic episode Data from American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. text revision. DSM-IV-TR. Washington (DC): The Association; 2000: Patients with mood disorders must always be evaluated for suicide risk. Like individuals suffering from unipolar depression, 115 patients with bipolar depression are at high risk of suicide. 116 One study of adult bipolar I patients found that more than 50% had attempted suicide. 117 Suicide attempts had occurred during depressive episodes in 93% of these cases. Therefore, clinicians must carefully evaluate depressed bipolar patients for suicide risk and, if necessary, consider hospitalization or an emergency consultation with a psychiatrist. Table 7 lists factors that may increase the suicide risk in patients with bipolar disorder. CASE 1 CONtinued The physician inquires about suicidal thoughts, and the patient states he is not having thoughts of death or suicide. He agrees to let the physician know immediately if he starts feeling suicidal. However, he reiterates that his depression has begun to cause him distress, and he requests treatment. The physician informs the patient that his depression is a component of bipolar disorder and explains the recommended treatment. What is the appropriate treatment for bipolar disorder? TREATMENT OF BIPOLAR DISORDER In bipolar disorder, treatment selection is based on the phase of illness the patient is currently experiencing. However, guidelines from the American Psychiatric Association (APA) do not recommend antidepressant monotherapy because of the risk of causing a switch into mania. 116 If antidepressants are used, they should be combined with a mood stabilizer. Thus, the approach Table 7. Factors That May Increase Suicide Risk in Bipolar Disorder Severe depression Loss of significant other Single motherhood Lack of medical insurance Less education Lower income Being single Comorbid eating, anxiety, or personality disorder History of physical or sexual abuse Family history of suicide Adapted from Post RM. New findings on suicide attempts, substance abuse, obesity and more. Curr Psychiatry 2002;1: to treating bipolar depression is to use a mood stabilizer that is effective for both mania and depression. The APA guidelines recommend lithium or lamotrigine as the first-line pharmacologic intervention, 116 and lithium would be a good choice for this patient. Ideally, this treatment would lead to remission of the current depressive episode and prevent future episodes of depression and mania. However, bipolar disorder is difficult to treat, and relapse of illness is common. Physicians frequently need to try several agents or combinations of agents before adequate symptom control is obtained. Table 8 outlines strategies for treating bipolar depression. Table 9 outlines the use of mood stabilizers for the treatment of mania. Psychotherapy, particularly interpersonal therapy and cognitive behavioral therapy, may be beneficial along with pharmacotherapy. In severe cases, electroconvulsive therapy may be considered. 118 Neurology Volume 12, Part 6 11

12 Table 8. Management of Bipolar Depression Presentation of Depressive Symptoms Breakthrough depressive episode and currently on mood stabilizers Acute depressive episode Severe refractory depression with suicidal ideation and/or psychosis Psychotic depression Interventions Optimize dose/blood level of maintenance medications Consider adjunctive psychotherapy If no or incomplete response, follow guidelines below Lithium or lamotrigine monotherapy is first-line treatment Antidepressant monotherapy not recommended Lithium in combination with an antidepressant is an alternative, especially for more seriously ill patients If no response, add lamotrigine, bupropion, or a selective serotonin reuptake inhibitor Consider adjunctive psychotherapy Electroconvulsive therapy Augment with an atypical antipsychotic agent Electroconvulsive therapy may be necessary if no response Adapted from Practice guideline for the treatment of patients with bipolar disorder. In: American Psychiatric Association. American Psychiatric Association practice guidelines for the treatment of psychiatric disorders compendium Washington (DC): The Association; 2002: Mechanisms and Use of Mood Stabilizers Mood stabilizers are the mainstay of pharmacologic treatment of bipolar disorder. The mechanism of action of these agents remains incompletely understood. 119 Most research has focused on how mood stabilizers may modulate neurotransmitters. Lithium historically has been considered the gold standard for the treatment of bipolar disorder, but the exact mechanism of action remains unclear. Recent research suggests that it is unlikely that lithium modulates neurotransmission at the level of the synapse. 120 Currently, attention has focused on the possibility that lithium exerts its therapeutic action via modulation of postreceptor, signal transduction mechanisms. 120 This research suggests that lithium modulates norepinephrine, dopamine, and serotonin function. 119 Lithium works preferentially for classic, euphoric mania and is typically less effective for patients with rapid cycling and for those with mixed depression and mania. There is a risk of birth defects if lithium is administered to pregnant women during the first trimester of pregnancy. Therefore, all females of child-bearing age must have a pregnancy test before the initiation of therapy. In addition, lithium has many potential side effects that can limit its usefulness, including thirst, polyuria, tremor, memory loss, weight gain, diarrhea, impairment of renal tubular function, hypothyroidism, and benign leukocytosis. Finally, lithium toxicity is a risk, and overdose can be fatal. Functional imaging studies also are beginning to add to our understanding of the mechanisms of mood stabilizer action. For example, Chang and colleagues 121 studied 8 adolescents with bipolar disorder using functional MRI at baseline and after 8 weeks of lamotrigine treatment. Clinical improvement was correlated with decreased right amygdala activation. This suggests that, at a circuit level, lamotrigine may contribute to normalization of the ventral emotional control circuit. Also using functional MRI, Haldane and colleagues 122 detected lamotrigine-induced changes in brain activity between drug-free and post-lamotrigine monotherapy conditions. They found lamotrigine monotherapy was associated with increased activation mostly within the prefrontal cortex and cingulate gyrus. Another functional MRI study revealed that following lamotrigine monotherapy, patients demonstrated reduced left temporal and right precentral activation and increased bilateral inferior frontal, right thalamus, and medial frontal and left precentral activation. 87 These latter 2 studies suggest lamotrigine may also modulate the dorsal emotional control circuit. As previously discussed, mood stabilizers influence neurotrophic pathways. 14 Lithium increases the levels of Bcl-2 in cells of human neuronal origin, 26,27 and lithium treatment has been shown to increase a marker of neuron function in gray matter 28 as well as to increase total gray matter content in the brain. 29 Valproate also activates Bcl-2, and both valproate and lithium increase the expression of BDNF. 14 Thus, it is likely that mood stabilizers normalize neurocircuit function by restoring neuron function at the cellular level. CASE 2 PRESENTATION A 39-year-old woman presents with a request for treatment of her depression. The patient says that she has been feeling sad for about 6 months. In addition, she reports poor concentration, feelings of worthlessness, insomnia, and loss of appetite. The patient says she started having these moods in her late teens and has had too many episodes to count. The episodes have generally lasted for 2 to 3 months but have ranged from a few days to 8 months. She denies any history consistent with a manic or hypomanic episode. She also denies any suicidal ideation. The patient reports no history of major medical conditions or neurologic 12 Hospital Physician Board Review Manual