Diagnosis: On Axis 1, it is worth considering whether the diagnosis of MDD is correct. One disorder frequently misdiagnosed as MDD is Bipolar Disorder (BD) – Depressive or Mixed Type. If the condition is a BD, standard antidepressants might only not help, but may actually do more harm than good. In such situations, mood stabilizers, perhaps particularly lamotrigine and/or atypical anti-psychotics may be indicated. It is also worth considering whether particular clinical features of MDD, suggesting specific treatments, might be present; for example, atypical features may respond best to monoamine oxidase inhibitors, psychotic features to antipsychotic augmentation, and seasonal features to Bright Light. Another Axis I consideration is whether alcohol or other substance use disorders are themselves driving the depression and/or clouding treatment response, and whether they need to be addressed before the MDD can be addressed.
On Axis II, patients with substantial personality issues may not be fully compliant with medications and may benefit from a psychosocial approach. Similarly, it is important to consider whether a General Medical Condition (Axis III), such as a metabolic or infectious disorder, requires active intervention. Finally, patients with severe psychosocial stressors (Axis IV) may not respond as quickly or as robustly as those who are stress free and may benefit from psychosocial approaches aimed at alleviating the stress or shoring their coping strategies.
Dose: All too often clinicians underdose MDD and mistake undertreatment for non-response. In general, at least 3–4 weeks at the highest dose recommended in any pharmacologic agent’s package insert would be considered an adequate trial (see duration below). For example, many clinicians consider 20 mg of citalopram, the starting dose, to also be an adequate dose. In the recently completed STAR*D trial, however, the mean dose needed for remission was over 40 mg, and many patients required 60 mg (1).
Duration: Similar to the issue of dosing, patients often are considered nonreponders before they have been on the medication long enough to have had an adequate trial. In STAR*D for example, only about 30% of those who remitted with citalopram had remitted by four weeks; another 30% of those who remitted didn’t begin to remit until after 10 or more weeks (1). These “delayed” remitters may have been considered “difficult to treat” if the medications had been discontinued prematurely.
Different Approaches: The effectiveness of antidepressant medications can be optimized by augmenting with non-pharmacologic or non-traditional interventions. For example, bright light treatment may speed or enhance response even in patients with non-seasonal MDD; medication plus cognitive-behavioral or interpersonal psychotherapy has been found in several studies to be more effective than medications alone, particularly in patients with chronic, severe, complicated or highly comorbid MDD; exercise has been shown in more than one study to have some antidepressant properties in patients if and when their depression allows them to work out; electroconvulsive treatment (ECT) should be considered for severe MDD after 4 or more failures with medication and medication augmentation trials, or sooner for MDD with significant suicidality or psychotic features; and vagal nerve stimulation (VNS) has been approved for MDD failing to respond to 5 or more adequately delivered trials.
Drugs: Once the diagnosis is established, the dose and duration of treatment have been optimized, adherence assured and nonpharmacologic approaches maximized, the choices are to switch medications or augment/combine. Most clinicians favor switching to another antidepressant if response is minimal and/or side effects problematic. Overall, only about 1 out of 4 patients switched to a second antidepressant after nonresponse or inability to tolerate an initial trial of a SSRI will remit to another monotherapy, whether it be with another SSRI, a dual reuptake inhibitor (e.g., venlafaxine) or a catecholinergic agent (e.g., bupropion).
In the absence of compelling data to guide treatment decisions, many clinicians use side effects and/or comorbid conditions to help select the next agent. Side effect considerations need to be individualized. For example, an adolescent male with MDD and premature ejaculation may be very adherent to a medication which also helps him delay ejaculation; an SSRI or SNRI would be good choices. On the other hand, an older male may prefer depression to the inability to ejaculate; for him, bupropion might be an ideal choice. Similarly, mirtazapine would be a poor choice for a woman who fears weight gain or somnolence, but an ideal choice for a cachectic cancer or AIDS patient with MDD. A co-occurring eating disorder or pre-menstrual worsening suggests switching to another SSRI; anxiety disorders suggest an SSRI or SNRI, pain an SNRI (duloxetine), smoking, attention-deficit disorder or obesity—bupropion.
Most clinicians favour augmenting (adding a non-antidepressant medication) or combining (adding another antidepressant) for patients who have had a partial or full response but nonremission to an initial monotherapy, especially where side effects have not been substantial. Although RCTs have not found one augmentation strategy to be preferable to others, some clinicians favor augmenting with lithium or lamotrigine where there is a bipolar phenotype (e.g., family history, early onset, psychotic or atypical features, irritability or agitation); thyroid for patients with borderline high thyroid-stimulating hormone (TSH) or symptoms consistent with hypothyroidism; estrogen for peri or post-menopausal women; stimulants for physically ill, unmotivated or lethargic patients; and buspirone for anxious patients. Atypical antipsychotics are increasingly being used to augment antidepressants for a range of depressed patients including those with bipolar spectrum disorders, MDD with psychotic features, and those with psychomotor agitation. Combining SSRIs with bupropion may be the most widely used combination strategy (4), although SSRIs plus SNRIs and SNRIs plus mirtazapine also have their proponents.
Determination : There are some patients who, despite our best efforts, fail to remit or even respond. Most of us have at least a few such patients in our practices. It is important that these patients not be blamed for their illness (“if only you tried harder, were less resistant, took better care of yourself or stopped seeing so many other doctors”); they are not refractory, their illness is. Rather, it helps to appreciate how important our continued care and attention is to such patients, that we may be their life-line in what is an incredibly painful and demoralizing disorder, and that our determination to keep working with them and continue trying new strategies may help them eventually turn the corner. Often, it does.