Obviously, one must consider the wide variety of possible medical causes of depression that require proper diagnosis to determine an appropriate intervention (Table). However, patients with TRD have high rates of comorbid psychiatric disorders as well; 35% of the depressed patients in STAR*D had at least 1 additional Axis I disorder.2 The most important assessment considerations include comorbid anxiety disorders, pain, insomnia, drug abuse, bipolar II disorder, attention deficit disorder, and partial response.
Of the 2876 patients with major depression in STAR*D who were eligible for analysis, the rates of comorbid anxiety disorders ranged from 12% (agoraphobia) to 31% (social anxiety); 53% of the patients qualified for anxious depression (major depression with high levels of anxiety symptoms).2 Patients with anxious depression and a variety of concurrent anxiety disorders had significantly lower remission rates.2 There was a relatively linear decline in remission rates with increasing anxiety symptom scores in patients with anxious depression; these persons also had relatively worse adverse effects and more serious adverse events with medication.14 Comorbid social anxiety disorder, in particular, has been associated with a more malignant course of depression.15
Pain symptoms are common in depressed persons and are among the symptoms least relieved by treatment.16 The likelihood of a poor response to antidepressants increases with increased pain severity. A greater than 50% reduction in pain symptoms early on improved the chance of remission in depressed patients treated with duloxetine(Drug information on duloxetine).17,18
Insomnia, a problem often seen in this patient population, is a risk factor for depression, and should be aggressively treated.19 A recent double-blind, clinical trial found that the addition of a hypnotic (eszopiclone) to fluoxetine(Drug information on fluoxetine) therapy lead to significantly higher response and remission rates at the end of the study than did the addition of placebo.20
Almost 30% of patients with major depression in the STAR*D also abused alcohol(Drug information on alcohol) and/or drugs.21,22 Another large sample found that those with comorbid depression and alcohol abuse were significantly younger at the onset of their first depressive episode, first psychiatric hospitalization, and first suicide attempt than were depressed persons who did not abuse alcohol.23 This comorbidity also led to a greater number of depressive episodes and suicide attempts.
In the STAR*D study there was a strong trend toward lower rates of remission in response to citalopram(Drug information on citalopram) as measured by the Quick Inventory of Depressive Symptomatology–Self Report (QIDS-SR) rating scale. This trend approached, but did not quite reach, statistical significance by the QIDS-SR definition.2
What seasoned clinician has not worked with a chronically depressed patient who manifests symptoms of hypomania or reports such episodes for the first time after months or years of treatment? Findings from a long-term prospective study indicate that depressive symptoms were present during 50% of the weeks of follow-up in a sample of 86 patients with bipolar II disorder; such symptoms were by far the most predominant.24 The diagnosis of bipolar II disorder can be difficult to make and, as a result, bipolar disorder is still often missed.25,26 The role of antidepressants (particularly as monotherapy) is especially controversial in bipolar II disorder, because of concerns about antidepressant-induced mania or hypomania.27,28
Estimates of the prevalence of comorbid depressive disorders in first-rate studies of children with ADHD have ranged from 9% to 38%, and elevated rates of major depression in adults with ADHD have been reported as well.29,30 Adults with major depression and ADHD were more likely to present for treatment than those with ADHD alone.31
In recent years, there has been growing recognition that getting better is not good enough. In treating depression, the goal is to become well, (ie, remission). Partial responders remain significantly functionally impaired, and an increased risk of relapse of depression among these patients has been shown in STAR*D as well as other studies.32-34
We need to better understand the mechanism of action of antidepressants. This knowledge will help guide clinicians as they switch and/or combine agents. It is hoped that improved methods of classifying patients will be developed to improve the success rates of treatments. Improved study designs (such as that used in the aripiprazole(Drug information on aripiprazole) trials in which the placebo was given on a single-blind basis to patients from the beginning of the study) may help minimize the placebo response.32,33
In addition to the need for novel compounds with new mechanisms of action (as well as a better understanding of the ones we have), a major innovation will be to move from the paradigm of average response to one of personalized medicine. Average response refers to the current practice of treating depression based on evidence that a drug benefits a large population. For example, if compound X is shown to be effective in 60% of individuals with major depression, all patients with major depression are therefore considered to be candidates for treatment with the drug. This often turns out to be a time-consuming and inefficient process.
Personalized medicine would use the results from genetic analysis or other testing methods (such as frontal quantitative electroencephalography) to predict an individual’s response to individual antidepressant agents.34 Some of the most provocative studies thus far have involved gene-based variations in the CYP450 system, particularly CYP 2D6.35 However, recent genetic data, much of it from the STAR*D study, have been analyzed to show associations between specific genetic variations and both treatment response and adverse events. Thus, variations in the genes encoding the 5HT2A receptor, GRIK4 (a subtype of glutamate receptor), and the ABCB1 transporter (a gene involved in encoding transporters across the blood-brain barrier) have shown associations with antidepressant response.36-38 Variations in the serotonin transporter gene solute carrier family 6, member 4 (SLC6A4) promoter were associated with citalopram-induced side effects.39 Markers within GRIK2 and GRIA3 (other genes encoding glutamate receptors) were significantly associated with the emergence of suicidal ideation with citalopram in STAR*D.40
Such findings raise the hope not only of more precise selection of antidepressant treatments and a better ability to predict tolerability but also offer new targets for antidepressant drug development. Given the heterogeneity of the molecular targets that have already been identified, it is little wonder that the proper diagnosis and management of TRD is such a complex and sometimes frustrating endeavor for both the patient and the clinician—but so fulfilling when one is successful.
In this Special Report: