18 - Approaches to the Treatment of Depression
Editors: Shader, Richard I.
Title: Manual of Psychiatric Therapeutics, 3rd Edition
Copyright 2003 Lippincott Williams & Wilkins
> Table of Contents > 18 - Approaches to the Treatment of Depression
function show_scrollbar() {}
18
Approaches to the Treatment of Depression
Ronald W. Pies
Richard I. Shader
The basic features of single or recurring episodes of clinically significant depression, which hereafter will be called major depressive disorder (MDD) have been recognized since the time of Hippocrates. Ironically, the humoral notions about the etiology of depression put forward by Hippocratic-era writers have striking resemblances to contemporary theories of neuroendocrine and neurotransmitter dysfunction in mood disorders. Although the lifetime expectancies vary somewhat because of the survey methodologies used, the lifetime expectancy for developing MDD is about 12% for women and 6% for men. With anxiety disorders, MDD is among the most common psychiatric disorders seen in adults. Some episodes of MDD may be triggered by traumatic life events like losses or disappointments; once triggered, however, the mood disturbance and other features of this disorder may become autonomous, just as they are in spontaneously occurring episodes (i.e., those that appear to arise out of the blue ). MDD is not simply a severe form of unhappiness or despair, nor is it synonymous with grief (see Chapter 16) or demoralization. Given the morbidity and mortality of MDD, which are nearly comparable with those of chronic cardiovascular disease, the recognition and vigorous treatment of this disorder are critical. Recent data suggest that depression is second only to ischemic heart disease in its total disease burden.
Despite the prevalence of MDD, better public and professional educational programs, and improvement in screening procedures for depression, the rate of outpatient treatment in the United States was only 2.33 per 100 persons in 1997. Although this rate shows a dramatic increase from 1987 (0.73 per 100 persons), most people in the United States who are depressed do not receive treatment.
Because of important similarities in clinical presentation and treatment, MDD and dysthymic disorder are considered together in this chapter.
I. Clinical Features of Major Depressive Disorder and Dysthymic Disorder
A. Major Depressive Disorder
A variety of research and clinical criteria has been proposed to define MDD. In the typical patient, however, the clinical presentation is quite characteristic. Central features are a loss of interest, satisfaction, or pleasure in almost all activities that lasts at least 2 weeks; some degree of appetite and sleep disturbance; decreased energy, concentration, or libido; low self-esteem or excessive guilt; and recurrent thoughts of death or suicide. Psychomotor agitation or retardation is usually observed. Psychotic features that are either mood congruent or incongruent may also be present. Some clinicians argue that MDD with psychotic features is a distinct clinical entity because it usually does not respond to classic antidepressant therapy when given alone. In a more severe form of MDD that is sometimes called melancholia, neurovegetative features such as weight loss or early morning awakening are especially prominent. The essential features of MDD are summarized in Table 18.1.
In addition to these typical features, MDD patients may reveal atypical signs and symptoms (see section VIII). Some patients, particularly adolescents and the elderly, may not report and may even deny subjective feelings of depression. Instead, depression may be manifested as behavioral disturbances, such as delinquency in adolescents, or as multiple vague somatic complaints. Crying spells or tearfulness are common. Some patients, however, may report being on the verge of tears or being unable to cry despite feelings of deep sadness or despair. Great variability in the nature of the appetite or sleep disturbance in MDD is observed. Many patients report loss of appetite and weight; others, often those with atypical features, may report
P.241
excessive eating and weight gain. Similarly, although early morning awakening is the classic complaint, some patients with MDD will report excessive sleep or daytime somnolence. In the Diagnostic and Statistical Manual of Mental Disorders (DSM), 4th edition, the term major depressive episode can be used to describe either an episode of MDD or the depressed phase of bipolar disorder (see Chapter 19); no symptomatic differences are outlined. Many clinicians, however, have noted the predominance of hypersomnia (excessive sleeping), psychomotor retardation, and weight gain in patients who are in the depressed phase of bipolar (versus unipolar) disorder. These features overlap with patterns seen in some patients with atypical depression (AD) or seasonal affective disorder (SAD) (see section VIII).
TABLE 18.1. DIAGNOSTIC FEATURES OF MAJOR DEPRESSIVE DISORDER | ||
|---|---|---|
|
Importance is often attached to the autonomy versus the reactivity of the patient's mood. Patients with the most severe forms of MDD often show nearly complete autonomy of mood (i.e., no mood elevation even in the presence of marked psychosocial stimulation, support, or reward). Other depressed patients (see section VIII.A) are markedly, albeit transiently, responsive to admiration, attention, or other ego boosts. The presence of psychotic features in MDD generally predicts a more disabling episode and a slower recovery than those seen in nonpsychotic MDD patients; however, eventual recovery appears equally likely in both groups. Delusional features are often of the paranoid or nihilistic variety (e.g., fixed and exaggerated ideas of worthlessness, guilt, or bodily decay). The prototypical patient may believe that he or she is the greatest sinner of all time or that his or her body is being eaten away by worms because I'm so evil. Such delusions are often deemed mood congruent because they resonate with the patient's depressive affect. The significance of so-called mood-incongruent delusions is not yet clear; some overlap with schizoaffective disorder (see Chapters 19 and 20) may be present in such patients. As has been noted, the presence of psychotic features in MDD usually suggests the need for different or additional treatment modalities (see also section VII.B.6).
P.242
B. Dysthymic Disorder
Dysthymic disorder (Table 18.2) may be to MDD what cyclothymia is to bipolar disorder an attenuated, although perhaps more chronic, form or presentation of the illness. For many years, terms such as neurotic depression or chronic characterological depression were applied to patients who are now diagnosed with dysthymic disorders. These earlier terms did not provide useful operational definitions, and their prognostic implications were never established. Dysthymic disorder is a chronic disturbance involving depressed mood (or irritability in children) most of the time for at least 2 years (or 1 year in children or adolescents). During these periods of depressed mood, some or all of the following features are present: poor appetite or overeating; insomnia or hypersomnia; low energy or fatigue; low self-esteem; poor concentration or difficulty making decisions; and feelings of hopelessness. Unlike MDD, dysthymic disorder may lead to only mild or moderate impairment in social and occupational functioning; however, its chronicity may predispose the individual to substance abuse, possibly as an attempt to self-treat the experiences of dysphoria and distress. Patients with dysthymic disorder do not show the profound neurovegetative signs seen in MDD: rarely is severe weight loss or psychomotor change seen. Suicidality and psychotic features are not common in uncomplicated dysthymic disorder; either, however, may be present in patients with dysthymic disorder who have superimposed or concomitant MDD ( double depression ).
Dysthymic disorder is subclassified as either primary or secondary and as either late onset or early onset. In primary dysthymic disorder, the mood disturbance must not be linked to a preexisting, chronic, non mood disorder (e.g., other axis I or axis III pathology). Early-onset dysthymic disorder begins before the age of 21. Some evidence seems to suggest that early-onset primary-type dysthymic disorder may be a distinct clinical entity.
II. Prevalence, Demographics, and Comorbidity of Major Depressive Disorder and Dysthymic Disorder
A. Major Depressive Disorder
Although the prevalence of bipolar disorder is roughly the same in men and women, most studies find that MDD is approximately twice as prevalent
P.243
in women as in men. Worldwide surveys report prevalence rates of 4.7% to 25.8% for women and of 2.1% to 12.3% for men. The large multicenter Epidemiological Catchment Area study in the United States found the lifetime risk for major depressive episodes (per the DSM, 3rd edition [-III]) to be from 4.9% to 8.7% in women and from 2.3% to 4.4% in men. Most cross-national data suggest that these differences are not an artifact of differential treatment-seeking behavior but rather that they are a consequence of biologic and developmental differences between men and women; cross-national studies find the lifetime prevalence rates to be from 4.4% to 18% for major depressive episodes. Evidence from the mid-1980s suggests that race and social class do not appreciably affect the risk for DSM-III defined depression. However, one reanalysis suggests that poverty (as defined by federal guidelines) does increase the risk for major depressive episodes. Although a major depressive episode or MDD can occur at any age, data from the Epidemiological Catchment Area study show that the highest lifetime prevalence is in the age cohort of 25 to 44 years. Data from the Cross-National Collaboration Group and other surveys suggest that more recent birth cohorts are at increased risk for major depression.
TABLE 18.2. DIAGNOSTIC FEATURES OF DYSTHYMIC DISORDER | ||
|---|---|---|
|
B. Dysthymic Disorder
The lifetime prevalence of dysthymic disorder is roughly 3% in the United States, making this a common psychiatric disorder. Cross-national data yield similar lifetime prevalence rates of 3.1% to 3.9%. Dysthymic disorder appears to be about twice as frequent in adult women as in men, but it is distributed equally in children. In one study, 65% of patients with dysthymia (the DSM-III, revised, term for dysthymic disorder) and 59% of patients with major depression received at least one additional axis I diagnosis, usually of social phobia or generalized anxiety disorder. Just over 10% of patients with dysthymia received additional diagnoses of alcohol abuse or dependence. This study points out the vexing problem of comorbidity in patients with mood disorders, as well as the risks of associated substance abuse. Dysthymic disorder and MDD also frequently occur together. The genetics, epidemiology, course, and treatment of patients with comorbid axis I disorders may differ, depending on which disorder developed first (e.g., did the major depression precede or follow the onset of panic disorder?). See the Additional Reading for more detailed discussions of comorbidity.
III. Etiology of Major Depressive Disorder and Dysthymic Disorder
A. Psychosocial Theories
No single psychodynamic or cognitive theory of depression has been verified by systematic research; however, each theory may contribute to the clinician's understanding and treatment of depressed people. Psychoanalytic formulations have evolved primarily from the work of Freud and Abraham. Freud believed that, in melancholia (as opposed to normal mourning or grief), hostile feelings previously directed against the lost person become directed at the self, which, through introjection, has incorporated the lost person. The commonly used formulation of anger turned inward to explain depression stems from Freud's view. Abraham believed that early and repeated disappointments in childhood predisposed the individual to later depression in the face of similar stressors. The ego, according to Abraham, retreats or regresses from its maturely functioning state.
Cognitive theories of depression have evolved, guided principally by the work of Beck and Ellis, and these are now generally accepted and widely taught. The cognitive model views depression as the consequence of irrational or negative thinking. The depressed person is seen as the victim of his or her own internal propaganda (i.e., irrational or self-defeating statements that distort reality). Thus, the depressed patient may espouse beliefs, such as If I am not perfect in everything I do, I am no good. Therapy with these patients tries to correct this faulty mode of thinking (see section VII.A.3).
P.244
A variety of psychosocial factors is correlated with the onset of depression, particularly in women. These include the absence of close interpersonal ties, marital discord and separation, the absence of employment outside the home, the loss of one's mother before the age of 11, more than three children under the age of 14 at home, and recent death or illness in the family. Ethnic and cultural factors may also shape the symptomatic presentation of depression; for example, African Americans with mood disorders appear to be more likely than whites to have hallucinations or delusions even after socioeconomic status has been considered.
Although psychodynamic theories are not based on reproducible data, these theories suggest that primary dysthymic disorder results from some as yet ill-defined problem in personality or ego development that culminates in difficulty adapting to adolescence and young adulthood. Chronic stress and acute loss may also predispose the individual to dysthymic disorder. As was noted above, secondary dysthymic disorder may develop as a consequence, feature, or complication of both psychiatric and medical disorders like anorexia nervosa, somatization disorder, or rheumatoid arthritis, or it may be present as an independent comorbid condition.
B. Biologic Theories
A large number of biochemical and neuroendocrine studies of depression have accrued over the past 50 years. Nevertheless, no single biologic theory accounts for all of the amassed data. Epidemiologic studies clearly suggest that both bipolar disorder and MDD have familial and genetic components as both generally breed true (i.e., if one monozygotic twin has unipolar illness, the other twin, if affected, is likely to have unipolar illness). However, some twin pairs follow a unipolar bipolar pattern, suggesting that both disorders may be associated with a common element in their genetic vulnerability. In general, association and linkage studies have not demonstrated a single-gene susceptibility for MDD. That MDD is a heterogeneous condition encompassing several subtypes seems likely. The prevalence of unipolar depression ranges from 11% to 18% in relatives of MDD patients, a rate that is threefold higher than that seen in control subjects.
A number of biologic hypotheses have been offered to explain MDD. About 40 years ago, a relatively simple deficiency hypothesis attributed depression to low levels of one or more monoamines, usually serotonin (5-HT) or norepinephrine (NE). This model is clearly inadequate; it cannot account even for the time course of response to antidepressant treatment. Much of the evidence today points to a complex interplay of neurobiologic factors, including compromised neurons affected by insufficient amounts of nerve growth factors (e.g., brain-derived nerve growth factor), altered concentrations of relevant precursors for neurotransmitter synthesis from diet or other reasons, relative amounts of neurotransmitter in the synaptic cleft, the sensitivity of numerous postsynaptic receptors, the feedback sent to the presynaptic neuron via autoreceptors, and altered second-messenger systems. Recently, interest has been focused on the role of cyclic adenosine monophosphate (AMP), cyclic AMP response element binding protein, and a cyclic AMP response element binding protein regulated gene that codes for brain-derived neurotrophic factor. Some have suggested that up-regulation of the cyclic AMP and brain-derived neurotrophic factor systems may provide a model of antidepressant action and perhaps of depression as well. For example, many effective antidepressant treatments increase the expression of brain-derived neurotrophic factor in the hippocampus. Despite this focus on gene activation, continued interest in the role of altered neuronal receptors in depression is seen.
Evidence derived from in vitro studies suggests that most, but not all, effective antidepressants down-regulate postsynaptic -adrenergic receptors over a period of 1 to 3 weeks, corresponding to the time course of drug effect. The situation is undoubtedly more complicated, however, because the 5-HT system is intimately linked to the down-regulation of -adrenergic receptors. Modest evidence also indicates that the 2-adrenergic autoreceptor that normally
P.245
mediates inhibition of NE release may be overactive in some depressed patients. This could lead to deficient release of NE. Although this might suggest a critical noradrenergic mechanism in depression, the situation is undoubtedly more complicated because, as noted, the down-regulation of -adrenergic receptors is intimately linked to the 5-HT system.
Of all the hypotheses put forth to date to explain MDD, the most robust involves 5-HT dysregulation, regardless of its cause. Numerous studies support the association of low 5-HT function with the presence of MDD. Both reduced reuptake of presynaptic 5-HT and antagonism at or down-regulation of postsynaptic 5-HT2A receptors appear to be involved in the mechanism of action of most effective antidepressants. Nefazodone, for example, appears to act primarily as a 5-HT2A receptor antagonist.
Cholinergic overactivity has also been linked to depression, a hypothesis consistent with the euphoric buzz that some patients experience from the overuse of anticholinergic agents.
Neuroendocrine factors have also been implicated in MDD. As with neurotransmitter findings, separating cause from effect is difficult. About 50% of MDD patients before treatment will show elevated 24-hour plasma concentrations of cortisol; a similar percentage (although not necessarily the same patients) will show early escape of cortisol secretion after suppression by the ingestion of dexamethasone (an abnormal dexamethasone suppression test [DST] ). Recently, the role of corticotropin-releasing factor has been scrutinized because corticotropin-releasing factor hypersecretion may contribute to depressive symptoms. A blunted thyroid-stimulating hormone response to the administration of thyroid-releasing hormone has also been observed in about 25% of patients with MDD. Some data suggest that reduced thyroid-stimulating hormone responses in women are associated with the presence of suicidality, agitation, or concomitant panic attacks. Whether these various abnormalities are markers for MDD or are instead integral reflections of its pathophysiology is not yet clear. Finally, a number of sleep abnormalities have been linked to major depression, including decreased rapid eye movement latency and increased rapid eye movement density in the first half of sleep. The traditional interpretation was that changes in sleep architecture or physiology were secondary to the underlying depression; however, the effectiveness of sleep deprivation as a temporarily helpful treatment suggests that abnormal sleep may itself induce or contribute to depression.
The pathophysiology of dysthymic disorder is even more murky. Some primary dysthymic disorder patients with a positive family history of major depression and decreased rapid eye movement latency may share pathophysiologic mechanisms with MDD.
IV. Prognosis and Course
A. Major Depressive Disorder
Ambiguities in the usage of terms such as remission, relapse, and recurrence have made definitive comment about the prognosis and course of MDD difficult. Some generalizations from the literature are as follows: (a) about 50% of patients who recover from an initial episode of MDD will have at least one subsequent episode; (b) patients with two or more past episodes of MDD have roughly a 75% chance of another recurrence; and (c) although a given episode of MDD usually responds well to short-term treatment, about 30% will show only partial recovery and about 20% will have a chronic course. Historical, psychosocial, and biologic risk factors for relapse or recurrence of MDD are summarized in Table 18.3.
B. Primary Dysthymic Disorder
This condition usually has an insidious onset and a chronic course, often evolving into a superimposed episode of MDD. Alternatively, dysthymic disorder may be seen as the residuum of an episode of MDD. Pure dysthymic disorder, which is uncomplicated by short-lived bouts of more severe affective symptomatology, appears to be uncommon in clinical practice. A small
P.246
percentage of dysthymic disorder patients will go on to develop hypomanic or manic episodes. The prognosis of secondary dysthymic disorder seems to depend on the course of the primary underlying disorder.
TABLE 18.3. RISK FACTORS FOR RECURRENCE OF MAJOR DEPRESSIVE DISORDER | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| |||||||||
V. Evaluation and Diagnosis
A. Adjunctive Psychologic Testing
A variety of objective and projective tests may support the diagnosis of MDD. On the Minnesota Multiphasic Personality Inventory, scale 2 (i.e., depression) is consistently elevated in patients with chronic MDD, and some evidence indicates that this elevation is correlated with an increased risk of suicide. On the Wechsler Adult Intelligence Scale-Revised, the classic depressive pattern is reflected by an overall performance score that is significantly lower than the verbal score. Depressed patients often give up on items or answer I don't know. On the Rorschach test, long reaction times to the cards are typical; chromatic color responses are diminished, as one might expect. On the thematic apperception test, patients typically give short stereotyped responses that amount to mere descriptions of the cards. Several useful scales for MDD and other mood disorders have been developed. The Zung Depression Scale is a self-report scale consisting of 20 items, each with a four-point severity rating. The Zung Depression Scale assesses affective, psychic, and somatic features of depression. The Beck Depression Inventory is also a self-report scale, but it may emphasize asthenia (lack of drive, pep, or energy) more than depression in some populations. Results using self-report scales depend heavily on the motivation of the patient and the tester. The Schedule for Affective Disorders and Schizophrenia developed by Spitzer et al. in 1978 is a highly detailed 78-page protocol that assesses both present and historic functioning. It is used primarily in research settings along with clinician-rated measures of morbidity, such as the Hamilton Depression Rating Scale and the Montgomery sberg Depression Rating Scale. In contrast, the Zung and Beck scales are most often used as screening devices.
B. Adjunctive Biologic Testing
Neuroendocrine studies. Despite the volumes of data amassed in this area, no blood test exists for MDD as yet. During the 1980s, the DST seemed promising as a simple test; subsequent research and clinical experience have substantially diminished this initial optimism. Abnormal DSTs (nonsuppression) may occur in a variety of medical and neuropsychiatric conditions, such as Cushing syndrome, pregnancy, nondepressed bulimia nervosa, Alzheimer disease, and alcoholism. That the DST can serve even as a validating criterion for MDD is by no means
P.247
clear. Nevertheless, an abnormal DST may be predictive of clinical course and outcome. More specifically, when an abnormal DST precedes the onset of an episode of MDD, its normalization may herald clinical recovery. A persistently abnormal DST despite apparent recovery can be an ominous sign and can be predictive of rapid relapse.Thyroid abnormalities, often of a subclinical nature, are seen in a substantial subgroup of depressed patients. The maximum thyroid-stimulating hormone response to thyroid-releasing hormone is diminished or blunted in about one-fourth of depressed patients; however, abnormal responses may also be seen in manic or mixed bipolar disorder patients. The thyroid-releasing hormone stimulation test may be helpful in distinguishing MDD from so-called reactive or neurotic depressions, many of which fall into the category of dysthymic disorder. These conditions typically do not show a blunted thyroid-stimulating hormone response. Other neuroendocrine probes for MDD are under investigation (e.g., growth hormone response to a variety of stimuli, such as clonidine and desipramine, may be diminished in MDD).
Other biologic studies. A number of studies using positron emission tomography or other imaging techniques suggest both functional and structural abnormalities in MDD. Positron emission tomographies of MDD patients may show a decreased ratio of caudate-to-hemisphere metabolism when compared with patients with bipolar disorder or with normal control subjects. This may suggest a role for the dopamine system in a subgroup of MDD patients. In a magnetic resonance imaging study of elderly patients with MDD, greater cortical and subcortical atrophy, as well as increased basal ganglia pathology, was seen in these patients compared with age-matched control subjects. Some functional neuroimaging data point to abnormal prefrontal cortical and limbic function in depression, suggesting that improved mood is associated with increases in prefrontal and anterior cingulate metabolism.
VI. Differential Diagnosis of Major Depressive Disorder and Dysthymic Disorder
A. Medical Conditions
A plethora of medical conditions may present with symptoms suggesting MDD, dysthymic disorder, or both. Some of the most common of these are shown in Table 18.4.
In a study of 755 patients seen by the psychiatric consultation-liaison service in one general hospital, nearly 40% of depressed patients had a diagnosis of organic mood disorder (i.e., secondary mood disorder). The most frequent causes or precipitants were stroke, Parkinson disease, lupus cerebritis, and human immunodeficiency virus infection. Hypothyroidism and multiple sclerosis also accounted for a substantial number of cases. Somewhat surprisingly, medications accounted for only about 7% of cases. As might be expected, propranolol and corticosteroids were implicated. Clues to a secondary etiology for an episode of MDD include onset after age 45, negative
P.248
family history of mood disorders, marked weight loss, accompanying signs of delirium or dementia, and absence of any evident psychosocial precipitant. (Note: The presence of a seemingly obvious precipitant such as the death of a spouse does not eliminate secondary organic factors as causes of the depression.)
TABLE 18.4. SOME MEDICAL CONDITIONS AND MEDICATIONS ASSOCIATED WITH DEPRESSIVE SYMPTOMS | |
|---|---|
|
Since the introduction of the cytokine immunomodulator interferon- 1a for the treatment of relapsing forms of multiple sclerosis, both depression and suicidal ideation and attempts that seem attributable to the treatment agent have been observed. This is a somewhat unusual finding because both the underlying disease and the treatment can produce the same psychiatric manifestations. Depression has also been found to be more common among patients being treated with interferon- 2b for malignant melanoma. An intriguing study has shown that pretreatment with the selective serotonin reuptake inhibitor (SSRI) paroxetine at about 30 mg per day reduces both the likelihood of a depressive episode and the severity of any depression that has occurred.
B. Other Psychiatric Disorders
A number of other psychiatric conditions may be mistaken for MDD or dysthymic disorder. Among the most important are normal bereavement and grief (see Chapter 16), various personality disorders (see Chapter 13), adjustment disorder with depressed mood, schizoaffective disorder (see Chapters 19 and 20), depression secondary to schizophrenia (see Chapter 20), and anxiety disorders (see Chapter 14) or somatoform disorders (see Chapter 4). Many patients with MDD will have comorbid axis I or II disorders. The chief differences between MDD and normal bereavement and grief are summarized in Table 16.1. However, these are, at best, crude guidelines; in some cultures, normal mourners may have a variety of depressive symptoms, including visual images of the deceased, persisting for more than a year after a loved one's death.
Among the personality disorders that may be confused with MDD or dysthymic disorder are borderline, narcissistic, avoidant, dependent, and obsessive-compulsive personality disorders. However, in such disorders, evidence of nearly lifelong impairment consistent with the supervening personality disorder criteria is usually seen. (Note: Sometimes, chronically depressed patients develop secondary alterations of personality that make differential diagnosis very difficult.)
Schizoaffective disorder is a poorly understood condition that is, in all likelihood, a rather heterogeneous collection of disorders. An apparent kinship seems to exist between the so-called bipolar type schizoaffective disorder and classic bipolar disorder. By contrast, the depressed type of schizoaffective disorder probably overlaps with patients with MDD and psychotic or delusional features, schizophrenia with intercurrent depression, and true schizoaffective disorder of a genetically distinct type. In practice, many patients diagnosed with schizoaffective disorder receive pharmacotherapies quite similar to those used in either bipolar disorder or MDD with psychotic features.
Secondary depression also presents complex diagnostic and treatment issues. Ideally, one tries to tease out the primary condition by carefully examining the longitudinal development of symptoms. For example, did the patient first develop panic attacks, followed months later by symptoms of MDD, or did the reverse occur? Similarly, a depressed patient with an antecedent history of deteriorating psychosocial function, ideas of reference, auditory hallucinations, and thought process abnormalities is most likely suffering from schizophrenia with secondary depression, a condition with different treatment implications than those for MDD with psychotic features. Such temporal distinctions are not always possible, and treatment often proceeds empirically and is directed against specific target symptoms. The relationship between depressive and anxiety disorders is discussed in section VIII.A.
P.249
VII. Treatment of Major Depressive Disorder and Dysthymic Disorder
A. Psychosocial Approaches
Some have said that there are as many psychotherapies for depression as there are psychotherapists. However, most approaches fall into one of the following four main categories: psychodynamic, experiential-expressive, cognitive-behavior, or interpersonal. These approaches are not mutually exclusive, nor do they necessarily rule out conjoint pharmacotherapy. No single approach has been demonstrated to be superior to the others, although cognitive-behavior and interpersonal approaches have been subjected to, and bolstered by, a number of reasonably well-controlled outcome studies. Although some writers distinguish cognitive from behavioral approaches, the former quite often include elements of the latter (e.g., cognitive therapists often give patients homework assignments or encourage in vivo exposure to troublesome situations). The converse, however, is not necessarily true.
Psychodynamic therapies. Belak elucidated the basic principles for brief psychodynamically oriented therapy for nonpsychotic depression. His approach, however, often evolves into longer term treatment. The following ten main problem areas are considered:
Self-esteem;
Punitive superego demands;
Aggression against the self;
Feelings of loss;
Feelings of anger and disappointment;
Feelings of having been deceived in relationships;
Unmet dependency needs ( oral demands );
Narcissism;
Denial of hidden rage;
Overall object relations, including the patient's transference to the therapist and others.
Classically conducted psychoanalysis is not appropriate for the treatment of severe depression, especially when psychotic features are present.
Experiential-expressive therapies. These therapies for depression have their roots in the work of Rogers, Maslow, and Peris. The emphasis is on the release of pent-up emotions, a here-and-now focus, and an empathic understanding on the part of the therapist. In general, little emphasis is placed on probing early experiences or the unconscious. Few well-controlled studies supporting this approach in the treatment of MDD have been conducted.
Cognitive-behavior therapies. Pioneered by Beck and Ellis, these therapies focus on the preconscious irrational ideas often espoused by depressed patients, as noted in section III.A. Treatment manuals for cognitive therapy have been developed that facilitate comparative outcome studies, if not therapeutic flexibility. Most studies of depression (variously defined) have shown cognitive or cognitive-behavioral approaches to be superior to minimal-treatment or no-treatment control subjects and perhaps to some psychodynamic approaches. A number of studies have found cognitive therapy to be comparable to or even superior to pharmacotherapy, but most of these studies have methodologic flaws, including generally inadequate pharmacotherapy and patient selection that excludes those with marked to severe pathology.
Interpersonal therapy. Developed by Klerman et al., interpersonal therapy assumes that depression occurs in an interpersonal context. Treatment is directed at helping the patient deal more effectively with current interpersonal problems, improve social functioning, communicate more effectively, and express painful affects. Little emphasis is given to early developmental experiences, unconscious mechanisms, or personality reconstruction. A manual of interpersonal therapy is also
P.250
available. A few controlled studies have shown interpersonal therapy to be comparable with pharmacotherapy for certain symptom areas such as social functioning.Neurovegetative complaints appear to be more responsive to properly chosen pharmacotherapies than to any of the currently practiced psychosocial therapies. A number of studies suggest that the optimal treatment of MDD involves the combination of pharmacotherapy with an effective form of psychotherapy delivered by a skilled clinician.
B. Somatic Approaches
Tricyclic antidepressants (TCAs). TCAs were the mainstay in the treatment of MDD for over 30 years. Despite the introduction of many new and improved medications, none has shown consistently superior efficacy to the original TCAs. Nevertheless, few clinicians would claim that TCAs are the drugs of first choice for MDD in most depressed patients, except for when cost issues are placed ahead of overall tolerability. SSRIs, which primarily block the presynaptic 5-HT transporter and generally lack direct effects on other receptors, are the most frequently used of the alternatives to TCAs; they appear to offer some advantages in their patterns and intensity of side effects when compared with the TCAs. Given the moderately improved side-effect profile and greatly reduced toxicity of newer antidepressants, today most clinicians begin a depressed patient on either an SSRI or one of the newer non-TCA agents (see section VII.B.3). However, because the TCAs are used in refractory MDD, usually as adjunctive agents, this group of medications is considered here in detail.
The TCAs and the some of their other pharmacologic properties are summarized in Table 18.5. These agents may be viewed along a continuum of additional effects (usually called side effects) that may help guide the clinician in choosing the agent. Often, these predictable side-effect profiles are what determine the choice of TCA, because all of these agents appear to be equally effective on a population basis in the treatment of MDD. This means that approximately equal efficacy is observed when these agents are prescribed to randomly chosen groups of patients; individual patients, however, may respond preferentially to a specific TCA. In addition to the patient's specific symptom pattern, other factors involved in selecting a particular TCA may include the patient's previous response to TCAs and the responses of any affected family members to TCAs. Despite much hope and effort, little evidence exists
P.251
to indicate that the unique neurochemical profile of a given TCA (e.g., the degree to which it inhibits presynaptic transport of NE, 5-HT, or dopamine) predicts anything about its efficacy in a specific patient.TABLE 18.5. COMPARATIVE QUALITIES OF SOME TRICYCLIC ANTIDEPRESSANTS
Anticholinergic Sedative Cardiovasculara Tertiary amines Amitriptyline 4 4 3 Imipramine 2 2.5 2 Doxepin 2.5 4 2.5 Clomipramine 2.5 3 2.5 Trimipramine 2.5 3 2 Secondary amines Desipramine 1 1 1.5 Nortriptyline 1.5 2 1.5 Protriptyline 3 0.5 1.5 Numbers are on a scale of 0 to 4, representing the following: 0, none; 1, minimal; 2, moderate; 3, substantial; 4, strong or marked. a Cardiovascular represents an estimated proportional summation of hypotension, tachycardia, and conduction abnormalities. All values are approximations based on in vitro and clinical data. The neurochemical basis for TCA side effects is not precisely known, but some of the following generalizations derived from in vitro and clinical studies can be made: (a) tertiary amine TCAs (e.g., amitriptyline, imipramine) tend to have greater sedative, anticholinergic, and hypotensive effects than do secondary amines (e.g., desipramine, nortriptyline); (b) the sedative and weight-promoting effects appear to be related to their affinity for the H1 (histamine) receptor and perhaps to effects on the 5-HT2C receptor; (c) their orthostatic hypotensive propensity is modestly related to their affinity for the 1-adrenergic receptor; (d) cardiac conduction abnormalities may be related to the presence and amount of hydroxy metabolites, especially in the elderly; and (e) sexual dysfunction, although extremely complex, may be mediated by their 5-HT properties. Considering the following four prototypical TCAs may be helpful for anchoring oneself: desipramine, amitriptyline, nortriptyline, and imipramine.
Desipramine is the least sedating and the least anticholinergic of the TCAs. Indeed, some patients may complain of agitation or insomnia with desipramine, at least during the first few weeks of treatment. Patients with depression and accompanying panic attacks seem especially sensitive to these effects. Desipramine, like all the TCAs, has some cardiovascular side effects stemming from at least the following three mechanisms: -adrenergic receptor antagonism leading to postural hypotension; vagolytic effects leading to tachycardia; and quinidine-like properties, which are shared with all class 1A antiarrhythmic agents, leading to various conduction abnormalities. Among the TCAs, however, desipramine is the least likely to provoke anticholinergic effects, such as dry mouth, constipation, blurred vision, and urinary retention. At therapeutic concentrations, its cardiovascular effects are relatively benign, although it may be associated with a higher than expected frequency of falls. Desipramine is not likely to be the TCA of choice for the extremely agitated patient with severe insomnia; for such a patient, a more sedating agent might be subjectively preferable, at least during the initial weeks of treatment. However, no compelling evidence exists indicating that the long-term outcome depends on matching the agent to the symptom picture; even in agitated insomniac patients, remission would be expected with desipramine in just as many cases as if amitriptyline were used. Presumably, a period of 3 to 6 weeks may be required for any TCA to modify the pathology underlying the depression; any earlier improvement may be due to nonspecific sedating effects.
Amitriptyline, the most anticholinergic and perhaps the most sedating TCA, is on the other end of the side-effects continuum. (Note: Some data suggest that doxepin and trimipramine are comparable in their sedating effects.) Amitriptyline is often associated with peripheral anticholinergic complaints, as well as with some central complaints, such as confusion, memory impairment, and delirium. In comparison with other TCAs, amitriptyline has a high incidence of cardiovascular effects, including postural hypotension. Sexual side effects, such as retrograde ejaculation, impotence, and anorgasmia, are also comparatively more frequent, although these can occur with any TCA (even though SSRIs cause even more sexual side effects than does amitriptyline). Amitriptyline would rarely be the TCA of choice, for example, in an elderly patient with bladder dysfunction, first-degree atrioventricular block, mild dementia, and psychomotor slowing or lethargy. Amitriptyline might be selected for an
P.252
extremely agitated younger adult patient with severe initial (sleep-onset) insomnia; however, even in such a patient, the anticholinergic side effects of this agent may prohibitively limit the necessary dosage increases. Despite these putative limitations, some primary care physicians use low-dose amitriptyline regimens because its sedative properties are seen as beneficial for any concomitant insomnia, agitation, or anxiety. Because it also has some effects on somatic pain and in the prevention of migraine, amitriptyline is popular among some clinicians.TABLE 18.6. COMPARATIVE SIDE EFFECTS OF SOME NON-MONOAMINE OXIDASE INHIBITOR, NONTRICYCLIC, OR HETEROCYCLIC ANTIDEPRESSANTS
Anticholinergic Sedative Cardiovascular Maprotiline 1.5 2.5 1 Trazodone 1 4 2 Fluoxetinea +/- 1 +/- Paroxetinea 0.5 1 +/- Sertraline +/- 1 +/- Bupropiona +/- 1 +/- Venlafaxinea +/- 2 1 Mirtazapine 1 3.5 1 Nefazodone +/- 2 1 Numbers are on a scale of 0 to 4, representing the following: 0, none; 1, minimal; 2, moderate; 3, substantial; 4, strong or marked; +/-, negligible. aThese agents also come in extended-release, sustained-release, or controlled-release formulations. Pharmacokinetic differences from the immediate release product result in reduced adverse effects for some patients. Nortriptyline is in the middle of the continuum of TCA side effects. Although nortriptyline is an active metabolite of amitriptyline, it is less anticholinergic and less sedating than amitriptyline; it is also less agitating and activating than desipramine. In this sense, nortriptyline is seen by some clinicians as the TCA for the average patient with MDD. It also has a fairly benign cardiovascular profile, and it is often recommended for the elderly patient with MDD. Even in elderly patients with mild conduction abnormalities, nortriptyline usually does not cause significant orthostatic (postural) hypotension. However, dosing with nortriptyline is more complicated than with other TCAs, because it is unique among the TCAs in having a fairly well-defined therapeutic window (i.e., in its concentration-response curve, levels above and below which it is less likely to be efficacious do exist). Most studies support 50 to 150 ng per mL as the optimal range for plasma levels of nortriptyline.
Imipramine is often chosen as an effective prototypical comparator TCA in research trials. As the first TCA introduced in the mid-1950s, it has been subjected to the most scrutiny. Its side-effect profile is midway between that of amitriptyline and that of nortriptyline. It is slightly less sedating and less anticholinergic than amitriptyline and is slightly more so than nortriptyline. Postural hypotension is not uncommon with imipramine; its other cardiovascular effects are less severe than those seen with amitriptyline.
TCAs may have additional side effects, such as tremor; excessive sweating; edema; skin rash; exacerbation of narrow-angle glaucoma; and behavioral complications, such as mania or psychosis. TCAs lower the seizure threshold, so they must be used cautiously in patients with a family or personal history of seizure disorder. (Note: The cumulative yearly incidence of TCA-related seizures is only about 0.3%.) TCAs can interact with a variety of other medications (see Chapter 29). Finally, TCAs may be very toxic when taken in overdoses; for example, even a 3-week supply (3,000 mg) of amitriptyline may be fatal if it is ingested all at once (see Chapter 3).
Some guidelines for prescribing TCAs
Principles. A few principles involved in the prescription and monitoring of the TCAs deserve emphasis.
Dose range. For the prototypical patient, a total daily dose of 175 to 300 mg for imipramine or its equivalent is usually required. One rule of thumb for imipramine is to attempt to reach 3.5 mg per kg per day, unless significant improvement or side effects occur at lower doses.
Plasma levels. Many clinicians attempt to use plasma levels to guide dosing, even though therapeutic ranges have not been unequivocally established. Once enough time has elapsed to reach steady state, a level in the putative therapeutic range may provide a degree of assurance that the drug was ingested, absorbed, and metabolized as expected. Modest evidence does indicate that most TCAs need to reach plasma levels of at least 100 ng per mL of the parent compound to achieve a full therapeutic
P.253
effect. Some patients may respond at lower concentrations, whereas others may not respond until 250 to 300 ng per mL has been reached or slightly exceeded (e.g., patients on imipramine, desipramine, or doxepin). Nortriptyline, with its putative therapeutic window, is an exception. Amitriptyline, which is partially metabolized to nortriptyline, may have an upper limit of effectiveness in some patients.Monitoring for side effects. Excessively high plasma TCA levels may be associated with increased adverse reactions, such as seizures and cardiotoxicity. With respect to the latter, obtaining baseline and follow-up electrocardiograms is a good practice in patients over the age of 50 or in those with a known history of cardiovascular disease. Because certain conduction abnormalities in children (e.g., congenitally long QT intervals [see Chapter 21]) and young adults may be subclinical, a few clinicians recommend obtaining a baseline electrocardiogram on all patients placed on TCAs. Certainly patients with first-degree or second-degree atrioventricular block should be monitored closely and perhaps should be treated with a non-TCA. Checking the patient's vital signs periodically is also a good medical practice.
Duration. TCAs usually should be prescribed at tolerated therapeutic doses for a minimum of 6, and perhaps up to 18, months after recovery from an episode of MDD; the old practice of tapering down to a maintenance dose as soon as the patient recovers has been linked to higher relapse rates.
Side effects. Treating TCA-related side effects can be difficult; sometimes, a difference of 10 to 20 mg in the daily TCA dose can markedly affect the patient's tolerance of the drug. For any TCA-related side effect, dose reduction is a reasonable first step. Unfortunately, some patients are so sensitive to the anticholinergic, hypotensive, or libidolytic effects of the TCA that dose reduction alone will not suffice. Dry mouth, constipation, and other anticholinergic side effects may be helped by advising the patient to increase water intake and dietary fiber. Urinary retention will sometimes require the addition of bethanechol (50 to 150 mg per day), which may also ameliorate dry mouth and constipation to some degree, to the therapeutic regimen. (Note: The clinician must rule out bladder obstruction before prescribing this agent.) Stool softeners, such as docusate sodium (100 mg, once to three times per day), may be helpful. Some clinicians recommend the use of fluoride lozenges for dry mouth. Postural hypotension may respond to a variety of interventions, including salt tablets, modest amounts of caffeine, or pressure (TEDS) stockings. In resistant cases, the mineralocorticoid fludrocortisone may be helpful in doses of about 0.4 mg per day.
Duration of MDD (recurrence versus relapse). The term relapse generally implies a worsening of an ongoing episode of depression; recurrence refers to the occurrence of a new episode. An arbitrary but common convention holds that worsening that occurs within 6 months after remission constitutes a relapse; worsening occurring thereafter is considered a recurrence. The highest rate of relapse occurs in the first few months after remission from an episode of MDD; the probability of relapse declines steadily during the course of the prospective follow-up. Patients with residual symptoms at the time of withdrawal of antidepressant medication have a higher rate of relapse than do patients who manifest no such symptoms. In terms of interval-specific probabilities of recovery from MDD, about 60% of patients will recover after 6 months. Of those not recovered by 6 months, about 30% can be expected to recover within
P.254
the next 6 months. Of those not recovered by 1 year, about 15% will recover in the next 6 months. Among those not recovered by 18 months, only 10% will remit in the next 6 months. Thus, the likelihood of recovery decreases substantially the longer that a patient remains depressed after initially entering treatment. The reader should note that most data of this type are derived from university medical centers; therefore, they may not generalize to other populations of depressed patients. Some evidence suggests that the duration of MDD episodes tends to remain constant or to lengthen slightly with each recurrence; the length of the preceding episode may serve as the lower estimate for the duration of the next one.Prevention of relapse and recurrence: the role of lithium. Results from a National Institute of Mental Health collaborative study suggest that discontinuation of antidepressant therapy is safe only after the patient has been free of significant symptoms for at least 16 to 20 weeks and that focusing on mild and severe symptoms is critical in reaching this decision. The National Institute of Mental Health study, based on its inclusion criteria, also incorporated patients with bipolar disorder. Both TCAs and lithium are effective in preventing recurrence of MDD, but TCAs appear to be more effective, especially for the prevention of severe episodes. Between 10% and 15% of patients initially diagnosed as having MDD will have a subsequent manic episode. When a high suspicion of latent or undeclared bipolar disorder is present, some clinicians advocate the addition of lithium to the TCA regimen. Factors suggesting a higher likelihood of bipolar disorder include an early age at onset of depression (under 25), a family history of bipolar disorder, a family history of affective disorder in consecutive generations, and the presence of atypical signs and symptoms.
Non monoamine oxidase inhibitors (MAOIs), non-TCAs ( heterocyclics ). This nongroup, which is sometimes called the heterocyclics, includes a wide array of agents, some of which are not cyclic in structure. Included in this grouping are the infrequently used tetracyclic maprotiline, the triazolopyridines trazodone and nefazodone, and the aminoketone bupropion. Their properties are summarized in Table 18.6.
P.255
Maprotiline, a compound that primarily inhibits the presynaptic transport of NE, appears to be as effective as the TCAs, but it also seems to have a somewhat higher incidence of seizures and skin rashes. It is less sedating than doxepin and also somewhat less anticholinergic. Seizures are especially common with maprotiline overdose. In general, maprotiline has few advantages over the TCAs or SSRIs, and it is infrequently prescribed as a first-line agent.
Trazodone is a proserotonergic antidepressant that is also an antagonist at several 5-HT1 and 5-HT2 receptor subtypes. Its primary side effects are sedation, dizziness, and gastrointestinal complaints. Even though trazodone is not strongly anticholinergic in vitro, some patients complain of dry mouth, which perhaps is a result of its 1-adrenergic receptor antagonist actions. Unlike some TCAs, trazodone usually does not promote substantial weight gain. Although it does not typically cause cardiac conduction abnormalities, trazodone has been linked with ventricular arrhythmias in a few instances. Nevertheless, trazodone appears to be significantly safer than the TCAs when it is taken in overdose. One rare but potentially dangerous side effect is priapism. Although this may occur in only about 1 in 7,000 men who take trazodone, permanent damage to the penis can occur when medical or surgical intervention is not given promptly. Other sexual side effects, including both anorgasmia and increased libido, have also been reported. Some studies suggest that trazodone is less consistently effective than are the standard TCAs. One possible explanation for this is the formation of significant amounts of its metabolite metachlorophenylpiperazine when trazodone doses exceed about 250 mg per day; metachlorophenylpiperazine has anxiogenic properties that may undermine the parent compound's antidepressant effect.
Trazodone is most frequently prescribed as a piggyback medication for sleep-onset insomnia in patients taking SSRIs or MAOIs. Evidence indicating that trazodone may increase deep sleep while preserving normal sleep architecture in healthy nondepressed adults does exist. Trazodone is also finding occasional use as an antiaggression agent in agitated demented patients.
Nefazodone is structurally similar to trazodone, but it was developed to have less 1-adrenergic receptor antagonism. It also lacks clinically significant anticholinergic and antihistaminic effects. It blocks presynaptic 5-HT transport, and it is a postsynaptic 5-HT2A receptor antagonist. It is rapidly absorbed and has a high first-pass metabolism involving cytochrome CYP 3A4 oxidative pathways for which it is both a substrate and an inhibitor. The parent compound has a half-life of 2 to 4 hours; it has at least three metabolites that contribute to its clinical actions, including small amounts of metachlorophenylpiperazine. Nefazodone appears to be a broadly effective antidepressant.
In contrast to other antidepressants, nefazodone produces minimal or no suppression of rapid eye movement sleep; it does not appear to cause priapism, even though this unwanted effect is associated with trazodone. Side effects, such as nausea, dry mouth, somnolence, dizziness, and constipation, have been observed, but they occur less often than with the TCAs. Sexual side effects are less common than they are with SSRIs. That a black box warning was added to the labeling for nefazodone because of rare cases of life-threatening hepatic failure should be noted. (Note: The estimated rate is 1 per 250,000 to 300,000 patient-years resulting in transplantation or death; the true rate is unknown.) Twice a day dosing is recommended (100 mg, twice a day, for the initial dose; reaching 400 mg per day after 1 week; 600 mg per day is often considered its maximum dose).
Bupropion is a nonsedating antidepressant that appears to inhibit the presynaptic dopamine transporter. It bears some structural and functional similarity to amphetamine. Bupropion has little, if any, anticholinergic activity; it does not promote weight gain; and it appears to have few cardiovascular side effects. It seems to be well tolerated in elderly patients with heart disease; in one study, it did not prolong cardiac conduction, exacerbate ventricular arrhythmias, or induce higher degrees of atrioventricular block. Lower doses of bupropion are usually indicated for the elderly, particularly to avoid the mild confusional states that sometimes occur with its use. Sexual dysfunction, an unwanted effect that is fairly common with SSRIs, is not common with bupropion; its prodopaminergic action may have a potentiating effect on erectile function. When bupropion was first studied, it appeared to have a more pronounced effect on the seizure threshold than the TCAs did. The usual figure quoted was a cumulative yearly seizure incidence of about 0.4% for bupropion versus that of about 0.2% for TCAs. (Note: In doses above 600 mg per day, the bupropion rate rose to about 0.6%.) However, a 102-site study of bupropion at doses of up to 450 mg per day involving more than 3,000 patients showed a seizure rate of 0.24% during an 8-week treatment phase and a rate of 0.4% for the entire study period. The overall seizure rate of 0.36% was interpreted as comparable with other antidepressants. The daily dose of bupropion should not exceed 450 mg, and it should be administered in divided doses of no more than 150 mg. A slow-release form of bupropion (bupropion-SR) may be administered on a twice daily basis; at steady-state, 150 mg twice a day of the slow-release form is bioequivalent to 100 mg three times a day of regular bupropion. The seizure risk for the slow-release preparation appears to be about 0.1% in the dosage range of 100 to 300 mg per day. (Note: Some data suggest that maprotiline [at greater than 150 mg per day], amoxapine [at greater than 300 mg per day], and clomipramine [at greater than 250 mg per day] may also enhance seizure risk.)
Because bupropion is sometimes given along with SSRIs, any inhibition by the SSRI of the hepatic clearance of bupropion must be considered, and appropriate dosage reductions must be made. Because bupropion appears to potentiate dopamine function, one might expect it to induce psychosis in predisposed individuals; case reports support this possibility, but no convincing evidence exists to indicate that bupropion is more likely to induce psychosis than is any other antidepressant. Some reports suggest that secondary brain syndromes or psychosis associated with bupropion tend to occur in patients with bipolar disorder or in those with premorbid psychotic features. Some data suggest that plasma levels of bupropion in the range of 10 to 19 ng per mL yield better responses than levels above 30 ng per mL.
Buspirone, an anxiolytic that first introduced a group of 5-HT1A partial agonists known as the azapirones, may also have antidepressant properties. In doses of 40 to 60 mg per day, buspirone appears to be safe and effective when compared with a placebo.
Venlafaxine, a phenethylamine derivative, is a racemic mixture with the active metabolite, O-desmethylvenlafaxine. Together they inhibit presynaptic transporter mechanisms for both NE and 5-HT (i.e., venlafaxine is a reuptake blocker for both neurotransmitters). It is relatively free of antimuscarinic, antihistaminic, and 1-adrenergic receptor antagonist properties. Venlafaxine has comparable efficacy with TCAs and SSRIs at doses of 75 to 375 mg per day; a modal dose is 250 mg per day in tablet form. Nausea, drowsiness, dizziness, dry mouth, and sweating are its most common side effects; all of
P.257
these probably affect between 10% and 35% of patients. However, an extended-release capsule formulation, venlafaxine XR, is better tolerated and is now more widely used than the immediate release formulation. Elevated blood pressure that is probably related to its presynaptic adrenergic effects may be a concern for some patients (e.g., when dosage exceeds 200 mg per day [or lower in elderly patients]).Mirtazapine is a novel agent that effectively facilitates both noradrenergic and serotonergic transmission. It antagonizes the 2-adrenergic receptors that normally reduce NE outflow; thus, it releases the noradrenergic neuron from its autoreceptor, leading to increased NE outflow. By also antagonizing 2-adrenergic heteroreceptors located on serotonergic nerve terminals, it leads to increased 5-HT output as well. Because mirtazapine blocks 5-HT2A/C and 5-HT3 receptors, its serotonergic effect may be more selectively directed at 5-HT1A receptors, the activation of which is thought to have antidepressant and anxiolytic effects. Moreover, the lack of 5-HT2A/C and 5-HT3 activation seems to reduce the typical serotonergic side effects, such as sexual dysfunction and gastrointestinal complaints (its 5-HT2C receptor antagonism may, however, contribute to weight gain in some patients). The main side effects associated with mirtazapine are somnolence, increased appetite, weight gain, and dizziness. Some clinicians advocate beginning with 30 mg per day (versus 15 mg per day), based on the theory that the increased noradrenergic effects will counteract some of mirtazapine's sedating antihistaminic effects. However, patients react quite individually to this agent, and, for many, 15 mg at bedtime is a reasonable initial regimen. Premarketing reports of agranulocytosis associated with mirtazapine have not proved to be a significant clinical problem.
Duloxetine is a recent addition to the non-SSRI, non-TCA pool of antidepressants. It affects reuptake of both 5-HT and NE and lacks affinity for muscarinic and histaminergic receptors. Some data suggest that it has higher binding affinity for these transporters than does venlafaxine. Clinical trials using doses in the range of 20 to 60 mg twice daily suggest that its efficacy is greater than that of a placebo in nonpsychotic MDD. More experience is needed with this agent to assess its place in the current armamentarium of antidepressants.
Reboxetine is a selective noradrenergic agent that is awaiting release. Although its comparative advantages are not yet clear, reboxetine is expected to be useful in patients in whom serotonergic agents either have been ineffective or are poorly tolerated (e.g., those with SSRI-related sexual dysfunction or gastrointestinal complaints). On the other hand, one might anticipate less robust effects for comorbid conditions, such as obsessive-compulsive disorder, premenstrual dysphoric disorder, or other conditions, that appear to benefit from a proserotonergic agent.
P.256
SSRIs. At the present time, SSRIs are the most commonly used and most widely accepted antidepressants in the United States. Although some clinicians have individual preferences for particular agents and some patients may tolerate one SSRI better than another, the only study that examined their comparable effectiveness in primary care settings found them to be quite comparable. This study examined 573 adults who were randomly assigned to fluoxetine, paroxetine, or sertraline. These SSRIs are relatively specific inhibitors of the presynaptic 5-HT transporter; this group also currently includes citalopram. In addition, the S enantiomer of citalopram, escitalopram, is also marketed in the United States for depression. Another SSRI, fluvoxamine, is also marketed in the United States, but it does not carry an indication for MDD.
P.258
The pharmacokinetic properties and dosages of these agents are shown in Table 18.7.All the SSRIs have a similar side-effect profile as follows: gastrointestinal distress, loose stools or diarrhea, sexual dysfunction (in men, mainly ejaculatory delay), tremor, and increased sweating. Dry mouth and constipation are somewhat more common with paroxetine. Either excitation or somnolence may occur; in one study, sertraline-associated insomnia or somnolence occurred in 17.5% and 14.5% of patients, respectively. However, these values were not significantly greater than those that were seen with the placebo. It appears that citalopram, sertraline, and paroxetine may be somewhat less likely than fluoxetine to cause psychomotor agitation. SSRIs also differ in their hepatic microsomal enzyme-inhibiting effects (see Chapter 29). The SSRIs should not be considered any more homogeneous as a group than the TCAs are. They may even have different effects on certain aspects of central nervous system function (e.g., paroxetine's effect on the sleep electroencephalogram is opposite to those of fluoxetine and fluvoxamine). Paroxetine appears to be more likely than the other SSRIs to produce a discontinuation syndrome in some patients after abrupt cessation of its use. Because other SSRIs may also cause this withdrawal problem, tapering is a better strategy than is abrupt withdrawal.
The SSRIs may be especially useful in patients who cannot tolerate TCA side effects, in the elderly, and perhaps in depressed patients with panic attacks or atypical features. They are far less toxic than the TCAs in overdose, and preliminary evidence suggests that at least one of the SSRIs, paroxetine, is less likely than clomipramine, a TCA mainly used in the treatment of obsessive-compulsive disorder (see Chapter 6); amitriptyline; or imipramine to induce mania in patients with bipolar disorder. Recent evidence has cast doubt on whether the SSRIs as a group are less likely than other antidepressants to induce switching in bipolar patients.
Fluoxetine has evoked more interest and emotion than has any other antidepressant in recent years. Fluoxetine and other SSRIs appear to be as effective as TCAs in controlled trials in outpatients with MDD. Fluoxetine appears to be associated with less weight gain than that seen with the TCAs, MAOIs, and paroxetine. Some patients may even lose 2 or 3 pounds during the first few months of taking fluoxetine. However, some patients taking any SSRI for 6 months or more report
P.259
bothersome weight gain. When premorbid weight is used as the baseline, however, many patients taking fluoxetine (and perhaps other SSRIs) appear to be simply regaining the weight lost during their depressive episode. Fluoxetine and other SSRIs can cause not only sexual dysfunction in men but also anorgasmia in women (the latter complaint may be more prevalent than was initially reported; it may possibly affect as many as 20% to 50% of patients). Fluoxetine differs from other SSRIs in that it is marketed in capsules of 10 or 20 mg, rather than as tablets, and in a liquid preparation. A preparation is available for the treatment of premenstrual dysphoric disorder. (Note: This formulation is called Sarafem rather than Prozac.) Fluoxetine is also the first SSRI to be available as a generic.TABLE 18.7. SOME SELECTIVE SEROTONIN REUPTAKE INHIBITOR PHARMACOKINETIC AND DOSAGE DATA
Mean T Active Metabolite Daily Dose Rangea (mg) Active Metabolite T Citalopram 1.5 d 20 40 Escitalopram 1.5 d 10 30 Fluoxetineb 2 3 d Norfluoxetine 5 80 5 9 d Fluvoxamine 15 h 25 300 Paroxetinec 1 d 20 50 Sertraline 1 d Desmethylsertraline 25 250 2 4 d Abbreviation: T , terminal half-life. aBased on manufacturer's recommendation and the authors' clinical experience; activity may be prolonged in the elderly; to 1/3 of the usual adult dosage may be indicated for some agents in the elderly. The elimination half-life (T ) may also be prolonged for some agents in patients with hepatic disease.
bAn extended-release once-a-week formulation (90 mg/wk) is available.
cA controlled-release once-a-day formulation is available.Titration of the dose may be particularly important for patients either with panic attacks alone or with depressive symptoms, because some of these patients may be exquisitely sensitive to the activating effects of SSRIs. One such SSRI side effect, akathisia, may be associated with considerable distress and even suicidal ideation. Contrary to exaggerated stories in the lay press, no convincing evidence indicates that fluoxetine or any other SSRI is more likely to be associated with an increased incidence of suicide when compared with other antidepressants or a placebo. (Note: When TCAs are ingested during an intentional overdose, there is a higher rate of death.) Some patients, particularly the elderly, may do best on doses of 5 to 10 mg per day of fluoxetine or its equivalent. Some clinicians use initial doses of 1 to 2.5 mg per day of fluoxetine or its equivalent in depressed patients with concomitant panic attacks. On the other hand, other patients clearly fare best on 40 to 80 mg per day of fluoxetine or its equivalent, and they tolerate these amounts without difficulty. Because SSRIs are sometimes activating and as they are capable of causing insomnia early in treatment, prescribing many of them as a single dose in the morning is a common practice. For sertraline, morning or evening dosing appears comparable. Given the long half-life of fluoxetine the parent compound and its principal metabolite norfluoxetine have elimination half-lives of 1 to 3 days and 7 to 9 days, respectively intermittent dosing (i.e., every 2 or 3 days) may make pharmacokinetic sense for some patients. This also raises the issue of the clinical need for a once a week formulation. A 90 mg, once weekly fluoxetine capsule that is available appears to be clinically comparable for most patients on the immediate formulation at 20 mg per day.
A controlled-release formulation of paroxetine (paroxetine CR) is available in the following three strengths: 12.5 mg, 25 mg, and 37.5 mg. It uses a polymeric matrix to release the tablet's contents over a 4-hour to 5-hour time period. Its enteric coating delays the onset of drug release until the tablets have reached the small intestine. Limited evidence suggests that the controlled-release formulation may be associated with less nausea in the first week of treatment.
As was noted earlier, some patients who experience SSRI-associated insomnia may benefit from the addition of a small dose of trazodone or doxepin at bedtime. Trimipramine, another highly sedating TCA, can also be used. SSRIs should not be coadministered with MAOIs because a serotonin syndrome can result from the combination; a 5-week washout of fluoxetine is suggested before initiating MAOI therapy, whereas 2 weeks should be sufficient for the other SSRIs.
MAOIs. For many years, the MAOIs were shunned by some psychiatrists as either dangerous or ineffective. In actuality, the MAOIs can be both safe and effective when used in the right way and for the right patient. Essentially, the currently available MAOIs are most appropriate for patients with atypical features and for those who have not responded to the classic antidepressants. AD is discussed in more detail in section VIII.A;
P.260
it is a mood disorder characterized by depression admixed with prominent anxiety, especially phobic or panic symptoms. Reversed neurovegetative signs are particularly common. The latter include increased appetite and weight gain, hypersomnia, and reversed diurnal variation in which mood worsens in the evening rather than in the morning. Notwithstanding the utility of MAOIs in AD, many patients with more typical depression may also respond to these agents.As Table 18.8 shows, MAOIs may be classified in several ways. The mitochondrial enzyme monamine oxidase (MAO) catabolizes catecholamines and serotonin. MAO occurs in two forms, A and B; and MAOIs may be more or less selective for either. MAO-A acts most selectively on the substrates 5-HT and NE, and it is found mainly in the gut and liver, as well as in the brain. MAO-B acts most selectively on phenylethylamine and benzylamine and predominates in the brain. Dopamine is a substrate for MAO-B, as well as for MAO-A. All currently marketed MAOIs (phenelzine, tranylcypromine, and isocarboxazid) are nonselective.
So-called irreversible MAOIs, such as phenelzine, isocarboxazid, and L-deprenyl (more recently renamed selegiline), bind tightly to their target enzyme, and they are not readily displaced. Reversible MAOIs are readily displaced by tyramine and other pressor amines found in various foods and beverages. The clinical importance of selective or reversible MAOIs lies mainly in their reduced likelihood of causing the cheese reaction in response to exogenous tyramine or other pressor amines. L-Deprenyl is relatively MAO-B selective at low concentrations, typically 10 mg per day; and it is thought to act primarily in the brain. (Note: L-Deprenyl is metabolized to methamphetamine and amphetamine.) At low concentrations, it does not interfere with MAO-A's metabolism of tyramine in the gastrointestinal tract. The reversible MAO-A inhibitors moclobemide and brofaromine (not yet marketed in the United States) appear to be particularly safe in the event of tyramine ingestion. They are displaced from the enzyme by the tyramine; the tyramine is then metabolized.
These theoretic issues notwithstanding, the MAOIs that are currently available and are indicated for use in depression are the three nonselective inhibitors phenelzine, isocarboxazid, and tranylcypromine. (Note: Tranylcypromine is a partially reversible inhibitor; MAO activity returns in about 7 to 10 days, rather than the 2 to 3 weeks needed for phenelzine or isocarboxazid.) The usual starting doses for these agents are as follows: tranylcypromine, 10 to 20 mg per day; phenelzine, 15 to 30 mg per day; and isocarboxazid, 20 to 30 mg per day. Elderly patients may require downward adjustment, as may others
P.261
who encounter side effects even at these low starting doses. Up to 6 weeks may be required for a robust clinical response to occur; the onset of action may be more rapid with tranylcypromine. The upper daily dosage ranges for tranylcypromine, phenelzine, and isocarboxazid are 60, 90, and 50 mg, respectively. Higher doses are occasionally required for some resistant patients.TABLE 18.8. SOME MONOAMINE OXIDASE INHIBITORS
Chemical Class Specificity Reversibility Phenelzine Hydrazine Nonselective Irreversible Tranylcypromine Nonhydrazine Nonselective Irreversiblea Isocarboxazid Hydrazine Nonselective Irreversible Selegiline (L-deprenyl) Nonhydrazine MAO-Bb Irreversible Meclobemidec Nonhydrazine MAO-A Reversible Brofarominec Nonhydrazine MAO-A Reversible aUsually reversible after 7 10 days rather than after 2 3 weeks.
bAt greater than 10 mg/d, selegiline usually loses its specificity.
cNot currently marketed in the United States.Abbreviation: MAO, monoamine oxidase. Aside from the cheese reaction, the main unwanted effects of MAOIs are insomnia, orthostatic hypotension, weight gain, sexual dysfunction, dry mouth, constipation, delayed micturition, nausea, and swelling or edema. Daytime drowsiness is occasionally reported, as are myoclonic twitches, sweating, chills, flushing, and restlessness. Disinhibition or hypomania may be seen, but the MAOIs appear to be less likely than the TCAs to precipitate mania in patients with bipolar disorder. A relatively infrequent neuropathy secondary to MAOI-induced vitamin B6 deficiency has been reported. This condition is readily treated with oral vitamin B6 (pyridoxine), 50 to 100 mg per day. Although hepatic toxicity is uncommon, it may be seen with the hydrazine-type MAOIs (i.e., phenelzine and isocarboxazid). Cardiac conduction abnormalities are relatively rare with the MAOIs, which gives them an edge over the TCAs. Some clinicians prefer MAOIs for elderly patients who have various degrees of atrioventricular block. Nonadherence and dropout are fairly common with the MAOIs, probably due to their high frequency of side effects. With phenelzine, insomnia, daytime somnolence, dry mouth, impaired sexual function, and orthostatic hypotension may be seen in up to 20% of patients. Both preparing the patient for such side effects and treating those that occur are important. Strategies for side effects treatment should begin with dosage reduction; other techniques are summarized in Table 18.9. Time alone may be sufficient to reduce some unwanted effects (e.g., phenelzine-induced anorgasmia may diminish after 2 to 4 months).
The most feared event with the MAOIs is the hypertensive response triggered by tyramine ingestion or the concomitant administration of certain sympathomimetic agents (e.g., indirect-acting drugs, such as ephedrine or L-dopa). In clinical practice, serious reactions of this type are infrequent. Although this habit is not to be encouraged, many patients test the dietary restrictions by ingesting one of the prohibited foods; they report either no reaction or only a bad but tolerable
P.262
headache. Moreover, about 30% of hypertensive crises in MAOI-treated patients cannot be ascribed to antecedent dietary factors; some reactions have no clear precipitating factor. Nevertheless, severe hypertensive crises that, on rare occasions, lead to death have been reported; providing the patient with a clear but realistic set of precautions is important. Prototypical dietary guidelines are summarized in Table 18.10. Probably the most important food items to avoid are those whose taste is enhanced by aging or fermentation (e.g., aged cheese of any type) and any spoiled foods. Patients should be warned that the tyramine content of foods varies from sample to sample, country to country, and even region to region. If a suspicious food is to be eaten, a small amount should be tried first, and the patient should then wait at least 2 hours before consuming any more. If a hypertensive episode does occur, it will usually start within 2 hours after tyramine ingestion; it will begin as a severe throbbing temporooccipital headache that is accompanied by marked palpitations and is followed by flushing, sweating, nausea, and vomiting. If untreated, this state may progress to cerebral hemorrhage. (Note: Young adults with occult circle of Willis aneurysms may be at particular risk.) Informing the patient of this reaction and advising him or her to seek immediate attention at an emergency department if such a reaction occurs are crucial.TABLE 18.9. TREATING SELECTED MONOAMINE OXIDASE INHIBITORS' SIDE EFFECTS
Side Effect Some Treatment Strategies Hypotension, dizziness Increased salt intake; TEDS stockings; fludrocortisone (0.2 mg per day); T3 or T4; tranylcypromine least associated with postural hypotension Insomnia Addition of trazodone (25 50 mg h.s.); benzodiazepines; paradoxical bedtime dosing; for tranylcypromine, consider less-stimulating MAOI Weight gain Dietary counseling (e.g., decreasing sweets, fats); increased daily aerobic exercise Sexual dysfunction For delayed ejaculation or anorgasmia, try switching to another MAOI after 1 to 2 week washout; bethanechol or cyproheptadine before intercourse Abbreviations: h.s. at bedtime; MAOI, monoamine oxidase inhibitor; T3, triiodothyronine; T4, thyroxine; TEDS, pressure. Some clinicians provide patients with one or two tablets of the calcium channel blocker nifedipine, 10 mg, to bite and then swallow in the event of a cheese reaction. Although it is usually effective, nifedipine can provoke excessive hypotension in some patients. A reasonable plan may be to advise the patient to take the nifedipine and then to call 911
P.263
or to proceed immediately to an emergency department. Phentolamine (typically 5 mg either intramuscularly or intravenously) or phenoxybenzamine (orally or parenterally) are the usual drugs administered in emergency departments (see Chapter 3).TABLE 18.10. SOME DIETARY RESTRICTIONS WITH MONOAMINE OXIDASE INHIBITORS
Foods to Avoid Usually Safe Foods Foods Unlikely to Pose Problems When Eaten in Moderation Aged or matured cheeses; this includes cooked foods containing these cheeses
Beer (especially tap and unpasteurized), red wine, sherry, liqueurs
Fermented (dry) sausage, beef or chicken liver, smoked or pickled fish, caviar
Canned or overripe figs, whole bananas, banana peel fiber
Fava or broad bean pods
Yeast or protein extracts (e.g., marmite), some fermented soy products, sauerkrautCottage, cream, ricotta, or farmer's cheeses
Vodka, gin, dry white wines
All fresh meats or fish
Banana pulp
Shelled beansYogurt or sour cream that has been properly refrigerated
Other alcoholic drinks, including bottled or canned beer if the quantity is limited to less than 12 oz/d and the beer is not consumed rapidly
Other fruit, if not overripe
Avocado, New Zealand spinach
Some soy sauces, yeast-containing breads, soy milk
Chocolate, caffeine-containing drinksFrom Folks D. Monoamine oxidase inhibitors: reappraisal of dietary considerations. J Clin Psychopharmacol 1983;3:249 252, with permission. Patients taking selegiline in doses of no more than 10 mg per day do not need to adhere to the MAOI diet ; however, the main active metabolites of L-deprenyl, amphetamine and methamphetamine may have dopaminergic actions of their own, including some pressor effects. In addition to the prohibited foods, numerous drugs and medications can interact adversely with MAOIs. The most severe drug drug reaction is that which occurs between MAOIs and meperidine; a single dose of the latter can provoke fatal hyperthermia. (Note: Dextromethorphan can also provoke this reaction, and it should be avoided.) In general, sympathomimetic agents, such as phenylephrine, ephedrine, and phenylpropanolamine, should be avoided; antihistamines are usually safe in moderation. The adverse interaction between the SSRIs and MAOIs has already been noted. Buspirone also has proserotonergic properties; it probably should not be combined with MAOIs. A washout of at least 2 weeks should occur before changing from one MAOI to another, because hypertensive reactions have been observed with abrupt switching. An up-to-date manufacturer's product brochure should be consulted for a complete list of drug drug interactions for a particular MAOI.
Treatment of MDD with psychotic features. Despite early controversy over the effect of TCAs in psychotic depression, most studies to date support the conclusion that TCAs alone (and by inference, heterocyclics alone) are not especially effective when psychotic features, such as somatic delusions, are present. Most studies suggest that only about 25% to 30% of patients with psychotic features respond well to TCAs alone, compared with the typical response rates of 65% to 70% in nonpsychotic MDD patients. MAOIs or SSRIs given alone produce similar outcomes. The best pharmacologic approach to psychotically depressed patients is probably a combination of an atypical antipsychotic agent and an antidepressant, either an SSRI (e.g., sertraline, citalopram) or a non-TCA, non-SSRI agent (e.g., venlafaxine). Although few controlled studies of such combined regimens have been conducted, one report found the combination of citalopram (20 to 40 mg per day) and haloperidol (4 to 9 mg per day) to be useful in seven patients with MDD with psychotic features. Preliminary evidence suggests that atypical antipsychotic agent monotherapy (eg., risperidone, quetiapine, olanzapine) may be useful in some patients. However, electroconvulsive therapy (see Chapter 24) may be the treatment of choice.
The clinician should keep in mind that some TCAs or SSRIs and some antipsychotic agents may inhibit each other's metabolism, thus resulting in higher than expected levels of each. Lower doses of each agent may sometimes be necessary to avoid anticholinergic or parkinsonian side effects. Some clinicians have had good results with thiothixene (10 to 30 mg per day) and nortriptyline (at doses producing nortriptyline plasma levels of about 80 ng per mL). Some evidence suggests that atypical antipsychotic agents may be useful in patients with schizoaffective illness, some of whom are psychotically depressed.
Valproic acid, carbamazepine, and lithium may all be useful in some patients with MDD with psychotic features, perhaps primarily in those who have bipolar features.
Refractory MDD and augmentation. Refractory or treatment-resistant MDD may be related to inadequate antidepressant dosing,
P.264
duration of treatment, or plasma levels or some combination of these factors. In some cases, plasma antidepressant levels may be within the putative therapeutic range, but they are inadequate for the particular patient. Although specific blood levels of non-TCAs (e.g., SSRIs) are less clearly linked with therapeutic effects, a negligible SSRI blood level may point to noncompliance, poor drug absorption, or hypermetabolism. Once dosing problems have been considered, the clinician should next reconsider the diagnosis (i.e., has a concomitant or comorbid medical, substance abuse, or personality disorder been overlooked?). The patient who is truly refractory to antidepressant monotherapy may benefit from a number of potentiating or switching strategies, depending on whether the patient has had a partial response or no response to the agent. In general, complete nonresponders to an adequate trial of one agent should be switched to a second agent. Although most clinicians seem to favor switching to a new chemical class (e.g., from an SSRI to bupropion), some consider changing from one SSRI to another to be a rational alternative. Indeed, response to one SSRI does not necessarily predict the response to another. In contrast, partial responders may benefit from augmentation of the original agent. Most clinicians wait about 4 weeks before beginning an augmentation strategy. For details on switching and augmentation strategies, the reader is referred to a useful supplement of the Journal of Clinical Psychiatry (2000;61[Suppl. 12]). The patient who is truly refractory to TCAs given alone may benefit from a number of potentiating strategies (Table 18.11).Probably the most robust of these approaches is the addition of lithium to a TCA or an SSRI. (Note: Most augmentation strategies noted for TCAs may also apply to SSRIs [see sections VII.B.3 and VII.B.4].) Clinical improvement may occur within 2 to 3 days, even at subtherapeutic lithium doses (e.g., 600 mg per day). Other patients may require several weeks at conventional levels of lithium (0.5 to 0.9 mEq per L) for improvement.
Thyroid hormone may also be used to potentiate TCAs; for this purpose, triiodothyronine is considered by many clinicians to be superior to thyroxine. However, few head-to-head comparisons of the use of thyroxine with that of triiodothyronine for this purpose have been conducted. The addition of stimulants such as methylphenidate or pemoline may also be helpful. (Note: Tolerance to stimulants may develop in some patients after a few weeks; some clinicians prefer to taper and discontinue the stimulant once improvement has occurred.) The addition of an MAOI to a TCA, which was long regarded by many clinicians as taboo, may have significant benefits for some patients. Unfortunately, research data demonstrating the efficacy and safety of this combination are limited. The risk of TCA MAOI combinations is primarily that of
P.265
inducing a serotonin syndrome of hyperthermia, delirium, and myoclonus, not of the occurrence of hypertension (i.e., the cheese reaction associated with excess tyramine). Limited evidence implicates imipramine as being the drug that is most likely to cause problems when used in combination with an MAOI. The best approach to combination therapy is to begin the TCA and MAOI together, using small doses of each, or to add a small amount of the MAOI (e.g., 5 mg per day of tranylcypromine) to ongoing TCA therapy.TABLE 18.11. SOME WAYS TO POTENTIATE THE THERAPEUTIC ACTIONS OF TRICYCLIC ANTIDEPRESSANTSa
Concomitant use of one or more of the following:
Lithium
Thyroid hormone
Antipsychotic agents, generally in the presence of delusional features
Monamine oxidase inhibitors (with extreme caution)
Selective serotonin reuptake inhibitorsaPotentiation should always be selective, and it should be undertaken with caution to avoid potentially harmful drug interactions or enhanced side effects. Some clinicians suggest that small doses of TCAs (e.g., desipramine, 10 to 30 mg per day) can be used as an augmentation strategy to enhance the efficacy of SSRIs in the treatment of MDD. The potential for drug interactions exists with these combinations; particular caution is indicated when desipramine and paroxetine are combined (see Chapter 29).
Antidepressants for dysthymic disorder. Few placebo-controlled studies of antidepressant therapy in pure dysthymic disorder have been conducted. As was noted earlier, dysthymic disorder often coexists with MDD or other axis I disorders. To date, studies examining various forms of chronic low-grade depression generally suggest that antidepressants are helpful. Some patients with dysthymic disorder overlap with atypical depressive patients. Others may have a characterologic (axis II) overlay that reduces the effectiveness of antidepressants; however, one should never assume that chronically depressed individuals are just axis II and that therefore they are unworthy of a vigorous antidepressant trial. Furthermore, some data suggest that some dysthymic patients may require more than 3 months of treatment before they show an adequate response to the SSRI.
VIII. Atypical, Seasonal, and Other Variant Forms of Depression
A. Diagnostic Considerations
The term atypical depression has been used in a variety of ways, and it may be misleading in an important sense patients with putatively atypical features are actually common in clinical practice. However, such patients share many clinical features with a variety of diagnostic groups, including patients with panic disorder, SAD, bipolar II disorder (i.e., MDD with hypomanic episodes), and borderline personality disorder (BPD). One possible set of criteria for AD is given in Table 18.12. Other classifications emphasize the features of reactive mood disturbance (nonautonomy); hyperphagia or binge eating (especially on carbohydrates or sweets); hypersomnolence; and interpersonal difficulties, such as rejection and criticism
P.266
sensitivity. Many of these features are found in patients with hysteroid dysphoria (HD), a group described by Liebowitz and Klein. HD patients tend to be attention junkies who have exquisitely fragile self-esteem, as well as a tendency to overeat and oversleep when depressed. This picture is strikingly similar to the winter-depressive category of SAD described by Rosenthal et al. These latter patients, in turn, are often women with bipolar II disorder.
TABLE 18.12. A PROTOTYPICAL CRITERIA SET FOR ATYPICAL DEPRESSIONa | ||
|---|---|---|
|
Given the diverse criteria for AD, its prevalence is difficult to establish. The prevalence of HD features, one possible prototypical AD subset, may be high. One study found that 6 of 18 depressed inpatients met HD criteria; most of the patients meeting HD criteria also met criteria for BPD, thus suggesting some overlap between AD, HD, and BPD. (Note: These data could also suggest that HD and BPD features may alter the threshold for hospitalization of depressed patients in some settings.) In another study of AD patients helped by isocarboxazid, the author concluded that his patient sample was typical of depressed outpatients.
The prevalence of SAD is estimated to be 5% for the general population; it may be as high as 38% in patients with recurring depression. Some data suggest that 20% to 25% of bipolar disorder patients may show seasonal mood fluctuations. About 75% of SAD patients meet the criteria for bipolar II disorder, and 15% to 20% meet the criteria for bipolar disorder. The hyperphagia (often with carbohydrate craving) and hypersomnolence seen in winter-type SAD suggest some commonality with AD. One population study in Fairbanks, Alaska suggested that about 9% and 19% of the sample met criteria for SAD or subsyndromal SAD, respectively, and concluded that one in four persons is affected by the seasonal changes found at that latitude.
One could view the aggregate of patient subtypes discussed in this section as being subsumed by a Venn diagram of three overlapping circles labeled AD, SAD, and bipolar II disorder. Many patients with low lethality borderline features (e.g., mood lability, disturbed interpersonal relationships, binge eating, and stimulant abuse) will also overlap with these patient types. Many neurotic features, including obsessive-compulsive, hysterical conversion, and hypochondriacal traits, have also been viewed as atypical or masked forms of depression. Of interest is the fact that the International and Statistical Classification of Diseases and Related Health Problems, tenth revision, contains the category mixed anxiety and depressive disorder. Patients with this diagnosis are primarily seen in primary care settings.
The relationships among anxiety, depression, and AD continue to be discussed. Despite comorbidity and overlapping symptoms, reasonable evidence exists to justify separating these disorders. Symptoms such as depressed mood, early morning awakening, and psychomotor retardation delineate patients whose future decompensations will most likely meet MDD criteria, whereas compulsive features, panic attacks, and agoraphobia are more closely associated with anxiety disorders.
B. Treatment Considerations
The heterogeneity of AD precludes any single drug of choice. A patient who has hypersomnia, hyperphagia, and carbohydrate craving when depressed but who also has hypomanic periods may do best on lithium, which perhaps may be combined with an MAOI, SSRI, or bupropion during the individual's depressed phases. Bupropion and tranylcypromine may have particular value in winter-type SAD, according to some authors. A patient with true SAD might benefit from phototherapy, although various antidepressants and lithium have been used successfully in SAD. A patient with BPD might benefit from any one of several pharmacologic approaches, including the use of low-dose antipsychotic medication (see Chapter 13). AD remains a diagnosis in search of a pharmacospecific treatment.
Given these complexities, how should the clinician proceed when faced by an AD patient who does not appear to fit into either the bipolar II disorder or SAD subtype? The presence of anxiety or panic attacks does not consistently
P.267
predict a good response to an MAOI, such as phenelzine, or to an SSRI (as compared with imipramine or a placebo). Preferential response to an MAOI or SSRI seems to be more closely related to the presence of mood reactivity, reversed neurovegetative patterns, or hysteroid dysphoric features. Because weight gain is often a problem in such patients, some clinicians choose tranylcypromine rather than phenelzine, even though most of the studies to date have used phenelzine. In AD patients who lack panic attacks or features of HD, looking closely at the symptom dimensions can be helpful (see Chapter 13). For the lethargic hypersomniac group, treatment with a nonsedating antidepressant, such as desipramine, protriptyline, or bupropion, is reasonable, although controlled trials with these are lacking. Some data suggest that mixed anxious-depressed patients or those with psychosomatic complaints may also respond well to bupropion. Bupropion does not seem to be useful, however, for depressed patients with concomitant panic attacks. The 5-HTIA partial agonist buspirone may provide relief to some patients with MDD with significant concomitant anxiety (i.e., with Hamilton Anxiety Rating Scale scores greater than 17). However, in one such study, buspirone (40 to 60 mg per day) worked best in patients with MDD with melancholia or in those who were more symptomatic. Because AD patients are often conceptualized as nonmelancholic (e.g., lacking profound weight loss, autonomy of mood, mood worsening in the morning), this study could also be construed as being unsupportive of buspirone's efficacy in AD. On the other hand, if one conceptualizes AD as a form of anxious depression, these data could support the usefulness of buspirone. Buspirone was also considered effective in patients suffering from irritability; mild secondary symptoms of depression; and interpersonal sensitivity symptoms consistent with, but not specific for, AD.
Some AD patients with impulsive features and histories of substance abuse resemble patients with borderline personality disorder (BPD). Some evidence suggests that fluoxetine and other SSRIs may be helpful in BPD (see Chapter 13). That some AD patients with borderline features would also respond well to fluoxetine or other SSRIs is plausible. Some clinicians support this use, but no controlled studies using well-defined AD patients have been published. A recent multicenter study of 395 outpatients with MDD that remitted with fluoxetine treatment failed to find evidence of fluoxetine's efficacy in maintaining remission for patients with atypical symptoms. This study's authors concluded that fluoxetine may not be as effective as the MAOIs for the treatment of AD.
Patients with SAD probably constitute a special subgroup of patients with AD. Phototherapy using bright full-spectrum artificial light is often considered the treatment of choice, but controlled studies versus MAOIs, SSRIs, bupropion, or other agents have not been published. Despite theories linking SAD with alterations in melatonin metabolism or sleep wake cycle shifts, no treatment specifically addressing these putative abnormalities has been demonstrated to be superior to the standard treatment alternatives. Whether the timing of phototherapy (e.g., morning or midday) makes any difference is also not clear; patients appear to do equally well with either. Many SAD patients with clear bipolar features merit a trial on lithium, possibly in combination with phototherapy. Such patients would also be expected to do better with MAOIs, SSRIs, or bupropion than with the TCAs because these appear less likely to induce mania.
IX. Depression in the Elderly
Faced with a variety of socioeconomic stressors, losses, health problems, and age-associated changes in organ system functioning (e.g., increased brain MAO activity, increased subcortical white matter pathology, and decreased brain choline acetyltransferase activity), the elderly may be at particular risk for depressive symptomatology. Data on age-related prevalence rates for MDD, however, are ambiguous. Prevalence rates vary from 1% to 10% for persons over 65 years of age. Recent data suggest that, in age cohorts over the age of 65,
P.268
prevalence rates of 1% to 2% are the norm, a rate much lower than that seen in younger groups. Suicide rates, however, are higher in the elderly (45 per 100,000) compared with the general population (13 per 100,000) (see Chapter 17). The prevalence of syndromes other than MDD, such as mixed anxiety-depression and dysphoric-dysthymic states, is higher in the elderly; and they are estimated to affect about 25% of persons over age 65. Depression in the elderly may also be more difficult to detect, and masked depressions are not infrequent (i.e., depression presenting as apathy or somatic complaints). The elderly are said to complain less of guilt and to have more depression-based cognitive decrements (see Chapter 5), compared with that seen in younger adults. Some depressed elderly are written off as mere hypochondriacs. Many medications taken by older persons may cause secondary depressions, and hypothyroidism and stroke-related depressions are more common among the elderly.
Treatment of depression in the elderly requires a biopsychosocial approach. Family involvement, when it is possible, is important; and short-term cognitive behavioral psychotherapies can be useful. Despite much commentary to the contrary, no single antidepressant of choice exists for the elderly. As was discussed earlier (see section VII.B), the classical tertiary amine TCAs have unwanted properties that can be particularly problematic for the elderly (e.g., constipation, urinary retention, blurred vision, and postural hypotension). The non-TCA agents, such as bupropion, the MAOIs, or the SSRIs, may be better tolerated than the tertiary amine TCAs; unfortunately, they have not yet been evaluated in large-sample trials in the elderly. Trazodone has some popularity, but it may be associated with hypotension and rarely with ventricular arrhythmias. Studies of SSRIs in depressed cardiac patients, most of whom are elderly, have generally been favorable. Although mirtazapine-related orthostatic hypotension is unusual, somnolence and dizziness are common, and they merit conservative dosing in debilitated elderly patients. On the other hand, some medically ill elderly patients with significant weight loss might benefit from mirtazepine's weight-promoting effects. Many clinicians claim a preference for nortriptyline or an MAOI, such as phenelzine or tranylcypromine. Stimulants, such as methylphenidate, may be effective, especially for short-term use.
When prescribing drugs for the elderly, the wise approach is to start low and go slow. One-third to one-half of the typical adult dose is generally used. (Note: Individual variation in dose requirements is significant in the elderly; plasma levels can be useful when available, particularly because other agents taken by the elderly may affect drug levels.) For the most part, elderly and younger adult patients appear to require comparable antidepressant blood levels; however, the elderly may achieve these same levels at lower doses. On the other hand, at least one study examining nortriptyline doses and plasma levels in frail elderly nursing home residents suggested that the optimum plasma level for these patients may be somewhat lower than that of younger populations (42 to 111 ng per mL versus 50 to 150 ng per mL).
X. Herbal or Over-the-Counter Antidepressants
A great deal of interest for these diverse agents has been expressed, but, unfortunately, little systematic research has been conducted. Herbal and over- the-counter agents touted as antidepressants include, but are not limited to, S-adenosyl-L-methionine (SAMe), dehydroepiandrosterone, inositol, St. John's wort (Hypericum perforatum), various amino acids and vitamins, and omega-3 fatty acids. Although a complete review of this topic is beyond the scope of this chapter, the reader is directed to the recent reviews referenced. In brief, only modest and sometimes controversial evidence exists for antidepressant effects with SAM-e, St. John's wort, and perhaps dehydroepiandrosterone and inositol. Few rigorously designed, long-term, controlled studies of these agents are available. The evidence to date suggests a possible role for SAM-e and St. John's wort in mild nonpsychotic MDD. How well these agents work compared with newer antidepressants used at clinically meaningful doses is still not clear. Furthermore, each of these nonprescription agents contrary to the popular
P.269
image of organic or natural remedies may have significant neuropsychiatric side effects, including the induction of mania or other systemic effects (e.g., the induction of the gastrointestinal track transporter protein p-glycoprotein by St. John's wort).
XI. The Depressed Suicidal Patient
The depressed patient who is suicidal is a medical emergency. Chapter 17 discusses the evaluation and treatment of the suicidal patient.
XII. National Depressive and Manic Depressive Association (NDMDA)
The NDMDA is an important lay (consumer) organization whose membership is drawn from patients, family members, and interested professionals. They provide support and information to patients and their families, and they actively advocate for improved care, education, and research. NDMDA, which is based in Chicago, can be reached at 1-800-82-NDMDA.
ADDITIONAL READING
Akiskal HS. Dysthymic disorder: psychopathology of proposed chronic depressive subtypes. Am J Psychiatry 1983;140:11 20.
Akiskal HS. New insights into the nature and heterogeneity of mood disorders. J Clin Psychiatry 1989;50:6 10.
Amsterdam J, ed. Refractory depression. New York: Raven Press, 1991.
Angst J. The course of monopolar depression and bipolar psychoses. Psychiatr Neurol Neurochir 1973;76:489 500.
Bonomo V, Fogliani AM. Citalopram and haloperidol for psychotic depression. Am J Psychiatry 2000;157:1706 1707.
Boulenger JP, Laval e YJ. Mixed anxiety and depression: diagnostic issues. J Clin Psychiatry 1993;54:3 8.
Bruce ML, Takeuchi DT, Leaf PJ. Poverty and psychiatric status: longitudinal evidence from the New Haven Epidemiologic Catchment Area Study. Arch Gen Psychiatry 1991;48:470 474.
Bymaster FP, Dreshfield-Ahmad LJ, Threlkeld PG, et al. Comparative affinities of duloxetine and venlafaxine for serotonin and norepinephrine transporters in vitro and in vivo, human serotonin receptor subtypes, and other neuronal receptors. Neuropsychopharmacology 2001;25:871 880.
Caine ED, Lyness JM, King DA. Reconsidering depression in the elderly. Am J Geriatr Psychiatry 1993;1:4 20.
Cornelius JR, Soloff PH, Perel JM, et al. A preliminary trial of fluoxetine in refractory borderline patients. J Clin Psychopharmacol 1991;11:116 120.
Danish University Antidepressant Group. Paroxetine: a selective serotonin reuptake inhibitor showing better tolerance, but weaker antidepressant effect than clomipramine in a controlled multicenter study. J Affect Disord 1990;18:289 299.
Demitrack MA. Can monoamine-based therapies be improved? J Clin Psychiatry 2002;63:14 18.
Dewan MJ, Pies RW. The difficult-to-treat patient with depression. In: Dewan M, Pies R, eds. The difficult to treat psychiatric patient. Washington, D.C.: American Psychiatric Press, 2001:81 114.
Duman RS. Novel therapeutic approaches beyond the serotonin receptor. Biol Psychiatry 1998;44:324 335.
Fabre LF. Buspirone in the management of major depression: a placebo-controlled comparison. J Clin Psychiatry 1990;51:55 61.
Fogelson DL, Bystritsky A, Pasnau R. Bupropion in the treatment of bipolar disorders: the same old story? J Clin Psychiatry 1992;53:443 446.
Goodnick P. Blood levels and acute response to bupropion. Am J Psychiatry 1992; 149:399 400.
Harnett DS. The difficult to treat psychiatric patient with co-morbid medical illness. In: Dewan M, Pies R, eds. The difficult to treat psychiatric patient. Washington, D.C.: American Psychiatric Press, 2001:325 358.
P.270
Janicak PG, Pandey GN, Davis JM, et al. Response of psychotic and nonpsychotic depression to phenelzine. Am J Psychiatry 1988;145:93 95.
Joffe R. Triiodothyronine potentiation of fluoxetine in depressed patients. Can J Psychiatry 1992;37:48 50.
Kocsis JH, Klein DN. Diagnosis and treatment of chronic depression. New York: Guilford Press, 1995.
Kroenke K, West SL, Swindle R, et al. Similar effectiveness of paroxetine, fluoxetine, and sertraline in primary care. JAMA 2001;286:2947 2955.
Liebowitz MR, Klein DF. Hysteroid dysphoria. Psychiatric Clin North Am 1979; 2:555 575.
Mann JJ, Aarons SF, Frances AJ, Brown RD. Studies of selective and reversible monoamine oxidase inhibitors. J Clin Psychiatry 1984;45:62 66.
McGrath PJ, Stewart JW, Petkova E, et al. Predictors of relapse during fluoxetine continuation or maintenance treatment of major depression. J Clin Psychiatry 2000; 61:518 524.
Musselman DL, Lawson DH, Gumnick JF, et al. Paroxetine for the prevention of depression induced by high-dose interferon alpha. N Engl J Med 2001;344:961 966.
National Institute of Mental Health. The impact of mental illness on society. NIH publication no. 01-4586. Bethesda, MD: National Institute of Mental Health, 2001.
Nelson JC. Augmentation strategies in depression 2000. J Clin Psychiatry 2000; 61:13 19.
Olfson M, Marcus SC, Druss B, et al. National trends in the outpatient treatment of depression. JAMA 2002;287:203 209.
Osser DN. A systematic approach to the classification and pharmacology of nonpsychotic major depression and dysthymia. J Clin Psychopharmacol 1993;13:133 144.
Pies R. Atypical depression. In: Tupin JP, Shader RI, Harnett DS, eds. Handbook of clinical psychopharmacology, 2nd ed. Northvale, NJ: Aronson, 1988:329 356.
Pies R. The diagnosis and treatment of subclinical hypothyroid states in depressed patients. Gen Hosp Psychiatry 1997;19:344 354.
Pitsikas N. Duloxetine Eli Lilly & Co. Curr Opin Investig Drugs 2001;1:116 121.
Quitkin FM, Harrison W, Stewart JW, et al. Response to phenelzine and imipramine in placebo nonresponders with atypical depression. Arch Gen Psychiatry 1991; 48:319 323.
Rosenthal NE, Sack DA, Gillin JC, et al. Seasonal affective disorder. Arch Gen Psychiatry 1984;41:72 80.
Rundell JR, Wise MG. Causes of organic mood disorder. J Neuropsych Clin Neurosci 1989;1:396 400.
Salzman C, Schneider L, Lebowitz B. Antidepressant treatment of very old patients. Am J Geriatr Psychiatry 1993;1:21 29.
Shader RI. Current clinical applications of monoamine oxidase inhibitors: issues and concerns. In: Shader RI, ed. MAOI therapy. New York: Audio Visual Medical Marketing, 1988:5 30.
Shader RI, Fogelman SM, Greenblatt DJ. Epiphenomenal, causal, or correlational more on the mechanism(s) of action of antidepressants. J Clin Psychopharmacol 1998;18:265 267.
Thase ME. Relapse and recurrence in unipolar major depression: short-term and long-term approaches. J Clin Psychiatry 1990;51:51 57.
Thase ME, Entsuah AR, Rudolph RL, et al. Remission rates during treatment with venlafaxine or selective serotonin reuptake inhibitors. Br J Psychiatry 2001; 178:234 241.
Thase ME, Greenhouse JB, Frank E, et al. Treatment of major depression with psychotherapy or psychotherapy-pharmacotherapy combinations. Arch Gen Psychiatry 1997;54:1009 1015.
Waalinder J, Feighner JP. Novel selective serotonin reuptake inhibitors, part 1. (Academic highlights). J Clin Psychiatry 1992;53:107 112.
Wong AH, Smith M, Boon HS. Herbal remedies in psychiatric practice. Arch Gen Psychiatry 1998;55:1033 1043.
Zajecka JM. Clinical issues in the long-term treatment with antidepressants. J Clin Psychiatry 2000;61:20 25.
EAN: 2147483647
Pages: 37
- ERP System Acquisition: A Process Model and Results From an Austrian Survey
- Enterprise Application Integration: New Solutions for a Solved Problem or a Challenging Research Field?
- The Effects of an Enterprise Resource Planning System (ERP) Implementation on Job Characteristics – A Study using the Hackman and Oldham Job Characteristics Model
- Distributed Data Warehouse for Geo-spatial Services
- Intrinsic and Contextual Data Quality: The Effect of Media and Personal Involvement