Making the Diagnosis and Choosing the Treatment
The initial choice of psychotherapy, medication, the combination of psychotherapy and medication, or ECT depends in part on the diagnosis of the mood disorder. With a prominent history of chronicity or a history of recurrences, the case for maintenance treatment is fairly clear. At this time, medications, by and large, are the maintenance treatments with established efficacy. For example, for more recurrent forms of major depressive disorder (there is a minor controversy as to whether or not recurrence refers to the occurrence of two or three lifetime episodes [Coppen et al. 1986; Depression Guideline Panel 1993; NIMH/NIH Consensus Development Panel 1985]), the likelihood of needing medication to prevent a recurrence is high. Consequently, medication is an essential element of long-term management of these recurrent mood disorders. For particularly long-standing dysthymic disorders, which are often complicated by recurrent episodes of major depression, there is now growing evidence that maintenance medication effectively prevents recurrences. If maintenance treatment is likely, initial treatments with fewer long-term side effects (e.g., sedation, weight gain) are preferred.
Which of the various medications to choose may be informed by the picture of the presenting symptoms. For example, psychotic depression usually requires both an antidepressant medication and an antipsychotic, because the combination of tricyclic and antipsychotic medication is more effective than a tricyclic antidepressant (TCA) alone. Alternatively, ECT is useful for psychotic depression, either as a first-line treatment or in patients for whom medication is not effective. One study demonstrated that amoxapine is effective in psychotic depression. For severe depressive or manic episodes with catatonic symptom features, acute treatment may entail a benzodiazepine such as lorazepam and/or ECT depending on the clinical circumstances (for a review, see Fink 1996).
The evidence for the efficacy of medication alone in more severe depressions is very clear, whereas psychotherapy alone has not been as well studied. For outpatients with endogenous or melancholic symptom features, psychotherapy alone appears not to be as predictably effective as medication. However, a recent study revealed that cognitive therapy was as effective as phenelzine (an MAOI) in depressions with atypical symptom features. There is also some suggestion that the selective serotonin reuptake inhibitors (SSRIs) (Pande et al. 1996) and bupropion (Goodnick et al. 1998; Nierenberg et al. 1998) also have efficacy. The evidence for the efficacy of ECT with mild, nonmelancholic forms of depression is lacking.
Comorbidity also affects initial treatment selection. Depending on the nature of other concurrent psychiatric disorders, the initial target of treatment may be either the mood disorder or the concurrent nonmood disorder. For example, effective treatment of obsessive-compulsive disorder (OCD) with depression usually results in remission of the depression. Similar findings have been reported for bulimia nervosa (Depression Guideline Panel 1993). In such instances, however, the nonmood disorder dictates the choice of treatment. Similarly, when major depressive disorder co-occurs with panic disorder or an eating disorder, medications that have established efficacy in both conditions are most likely preferred initially. For example, clomipramine or an SSRI may be preferred when both OCD and major depressive disorder are present.
Co-occurring substance abuse raises the possibility of a substance-induced mood disorder, which must be further evaluated by studying the patient’s history and requesting the patient to begin a prospective period of 2-6 weeks of abstinence depending on the substance, chronicity, and degree of prior abuse. If the depressive symptoms or disorder is substance induced, abstinence results in symptom remission. For patients who continue with clinically significant depressive symptoms after 2-6 weeks of abstinence, an independent mood disorder is diagnosed and should receive immediate treatment.
Not infrequently, personality disorders (Axis II) accompany mood disorders. The diagnosis of Axis II disorders, however, must be tentative during the depressed state, since the state itself affects both patient reporting and clinician perceptions as to the type and severity of the associated Axis II disorder (Corruble et al. 1996). It is also important not to diagnose an Axis II disorder (with recurrent major depression) without strongly considering whether or not the patient has recurrent major depression with incomplete interepisode recovery, which also appears to be an Axis II condition. Both the objectives and strategies of treatment are different in each case, as will be discussed later in this section.
The presence of a personality disorder is not a contraindication to directly treating the formal mood disorder, although the Axis II disorder may prolong the time to acute-phase treatment response, interfere with adherence, or even preclude complete symptom remission. Thus, in general, the presence of Axis II disorders suggests a less optimistic prognosis than would be the case were such disorders absent. Furthermore, there is some circumstantial evidence indicating that Axis II disorders represent risk factors for subsequent relapse or recurrence.
Axis II disorders affect both strategic planning and treatment tactics used. Strategic planning may call for the initial use of the combination of psychotherapy and medication when an Axis II disorder is present, or the addition of psychotherapy if maximal symptom reduction with medication has been obtained but some symptoms or psychosocial impairment remain. That is, once the symptoms of the major depressive episode have largely subsided, practitioners may wish to add psychotherapy to directly address the personality disorder, although evidence for the efficacy of this approach rests largely on clinical experience, as opposed to data from randomized controlled trials. On the other hand, several recent studies (Fava et al. 1998a; 1998b; Paykel et al. 1999) indicate that cognitive-behavioral therapy to remove residual depressive symptoms following a response without remission leads to a far better longer-term prognosis (see Rush and Thase 1999 for a review of the indications for psychotherapy).
A number of tactical issues are also raised in the presence of an Axis II disorder with depression. Ensuring adherence, establishing a therapeutic alliance, and conducting the long-term management of patients with Axis II disorders may tax the therapeutic skills of the clinician, whether or not medication is a primary treatment. In addition, the presence of an Axis II disorder appears to be associated with a slower and/or less complete response to medications, as well as to time-limited psychotherapies.
Mood disorders are not uncommon accompaniments of general medical conditions. In fact, general medical conditions are among the common risk factors for developing a mood disorder. Recent reviews indicate that between 15% and 25% of patients with myocardial infarctions (within 1 year of the incident) (Frasure-Smith et al. 1995), as well as between 15% and 25% of patients with cancer, diabetes (Carney 1998), or stroke (R. G. Robinson 1998), for example, will develop a major depressive episode, which can be both persistent and disabling. Recent evidence also indicates that the presence of a major depressive episode is associated with an increase in the morbidity and mortality of the associated general medical condition.
Thus, although treatment of these cases of depression is clearly indicated, the treatment strategies and tactics are more complex than when depression occurs alone. The basic principles that guide treatment of depression in the absence of general medical conditions generally apply. However, choosing among the various acute-phase options when major depression occurs with another general medical condition is influenced by factors such as prior response to antidepressant treatments, the relative medical safety of medications (as opposed to psychotherapy), and clinical judgments and empirical evidence about whether psychotherapeutic methods may be effective.
Choosing among medications is affected by drug interactions, the pharmacological profile of the antidepressant in relation to the general medical condition, drug dosing requirements, and side-effect profiles. For example, when concurrent general medical conditions are present, the higher side-effect rates often encountered with TCAs must be considered (Popkin et al. 1985). Demoralization is often a problem in this patient population, which may explain why stimulants appear to be useful (in open trials), even if they are not generally effective treatments for major depression.
The presence of complex, ongoing significantly stressful life events or social contextual issues is often profoundly disturbing to patients. Their presence or absence, however, should not influence the decision as to whether or not medication is used. It is often the case that patients with major depressive episodes who obtain symptom reduction with medication are then less disabled from the mood disorder and are better able to manage these complex life circumstances. On the other hand, the presence of such chronic, disturbing life circumstances (e.g., chronic marital discord, spousal abuse) logically argues for stronger consideration of combined treatment, either initially or sequenced, to obtain both complete symptom remission and psychosocial restoration or subsequently to develop a better prognosis (Fava et al. 1998a; 1988b).
Declaring Initial Treatment Failure
It is essential that the treatment selected be provided in an optimal fashion (e.g., duration and dosage for medication, adherence to recommended procedures for psychotherapy) to determine whether the treatment is both tolerated and effective. There is a tension involved in implementing any (initial) treatment: to provide a robust treatment trial for sufficient time to determine if it is effective without unnecessarily prolonging an ineffective treatment and thus failing to provide the patient with a new treatment strategy in a timely fashion. To ensure that the initial trial is adequately conducted, patient adherence to the treatment is critical. Poor adherence is probably responsible for more unsuccessful treatment attempts than is the medication selected.
Medication dosage obviously affects both clinical outcome and the number and severity of side effects. Some patients metabolize certain drugs more rapidly or more slowly than do others. Slow metabolizers, especially in the case of the more anticholinergic TCAs, encounter more side effects earlier, and at lower dosages. Fast metabolizers may experience virtually no side effects, even with rather large dosages. Overly aggressive dosing with associated elevated blood levels may cause arrhythmias, seizures, or delirium. However, side effects, especially with desipramine and nortriptyline, are not good indicators of blood levels. Indeed, orthostatic hypotension can occur even with low blood levels. Such factors emphasize the value of a therapeutic blood level to determine optimum dosing strategies, especially in medically fragile patients, when TCAs are used (Depression Guideline Panel 1993).
Regarding the length of the treatment trial, growing evidence indicates that acute-phase medication trials should last at least 6 weeks to determine whether the medication produces a benefit, and 8-10 weeks to define the full extent of symptom reduction (e.g., Frank et al. 1990; Schweitzer et al. 1990). Most (two-thirds to three-fourths), but not all, patients who ultimately do respond will show at least some benefit by 4 weeks if the dosage is adequate. Few analyses have examined the alternative question: When can one decide that further exposure to the treatment is unlikely to result in sufficient benefit? Clinical impression and several recent reports (e.g., Nierenberg 1995; Wexler and Nelson 1993) suggest that if the patient has not had at least a 20%-25% reduction in symptom severity by 4 weeks, then the likelihood of a response at 8 weeks is about 20% (assuming that adequate dosages were used).
Selecting Second Treatment Options
Should the first treatment fail, either because of intolerance or because of lack of efficacy, a strategic decision regarding the second treatment is indicated. The unsatisfactory response to an initial treatment trial (medication or psychotherapy) also calls for a diagnostic reevaluation. In some cases an occult general medical or substance abuse condition may be found with re-interviewing before proceeding to a second antidepressant treatment.
For patients who have received medication initially, common options include 1) adjusting the dose; 2) extending the period of the initial trial; 3) switching to an alternative treatment (alternative medication or psychotherapy); 4) augmenting the initial treatment with an additional, second treatment (e.g., initial treatment with psychotherapy augmented with medication or vice versa); and 5) obtaining a second opinion. Factors to consider for dose elevation include 1) no therapeutic effect at substantial therapeutic oral doses, 2) a history consistent with rapid drug metabolism, and 3) therapeutic blood levels indicating low circulating medication levels. Note, however, that blood levels, as yet, are not strongly related to antidepressant outcomes for most antidepressants, save for desipramine, imipramine, and nortriptyline (American Psychiatric Association Task Force on the Use of Laboratory Tests in Psychiatry 1985). Extending the initial trial further is indicated if 1) the initial trial is less than 6 weeks, 2) there is a partial response by 6 weeks, or 3) prior medication trials have been unsuccessful and shorter than 6 weeks.
Similarly, for psychotherapy, partial response by week 6 argues for further extending the trial period. Minimal or no response by 8 weeks is often indicative of an ultimately unsatisfactory outcome, at least with psychotherapies in time-limited formats. Extending a trial of light therapy beyond 3 weeks in patients who have not responded has not been evaluated. Clinical impression suggests that extending ECT beyond 10 treatments when no response at all has occurred within the 10 sessions is not recommended. However, careful studies of this issue are lacking.
The choice between switching from the initial treatment to a new treatment and augmenting the first treatment rests on the philosophy guiding the clinician and on the patient’s treatment history, as well as on the economics and timing involved. The best-documented augmentation strategies involve inexpensive medicines (e.g., lithium or thyroid hormones), and response, if it occurs, is often within 2-4 weeks. On the other hand, a switching strategy entails the need to wait longer than 2 weeks to attain a full effect and, in some cases, the need for a washout period for safety reasons (e.g., switching from fluoxetine to an MAOI). Conversely, it is not clear how long augmentation should be continued, and lithium augmentation entails some expense and inconvenience (i.e., blood level monitoring).
If the initial trial is the patient’s first treatment and the practitioner finds medical or economic reasons for favoring a single treatment, switching rather than augmenting is preferred. On the other hand, augmentation strategies, particularly the use of two different medications, seem to be effective in patients who have undergone one or more failed well-conducted single-medication trials. Thus, switching may be preferable for patients with one or two prior treatment attempts, whereas augmentation may be preferable for patients who have not responded to multiple (e.g., three or more) single-treatment trials of medication. Several reviews indicate that if the initial medication is ineffective or cannot be tolerated, a reasonable step in primary care is to switch medication classes. However, open trials have revealed some benefit to a switch within the SSRI class. In psychiatric settings, augmentation may be more common, because more patients in these settings have not responded to prior treatments.
The value of augmenting medication with psychotherapy is only now being evaluated. Two recent studies (Fava et al. 1998b; Paykel et al. 1999) tested the notion that residual symptoms, after some benefit but not full remission with medication, if addressed with psychotherapy would lead to less disability and a better prognosis in recurrent depression. Indeed, cognitive therapy has such an effect. However, further research on this phenomenon is needed.
Revision date: July 3, 2011
Last revised: by Dave R. Roger, M.D.