Pharmacotherapy

Medications

The acute treatment phase is focused on treating the patient to remission. Acute therapy typically lasts 6-12 weeks but should last until remission is reached (American Psychiatric Association, 2013; Diagnostic and Statistical Manual of Mental Disorders, 5th Edition).

Definition: Full remission is defined as a two-month period devoid of major depressive signs and symptoms (American Psychiatric Association, 2013; Diagnostic and Statistical Manual of Mental Disorders, 5th Edition).

For antidepressant medications, treatment adherence and achieving clinical goals are more important than the specific medication selected. Successful treatment often involves dosage adjustments and/or trial of a different medication to maximize response and minimize side effects (American Psychiatric Association, 2010).

Selection of an antidepressant medication

The overall effectiveness of antidepressant medications is generally equivalent between and within classes of medications (American Psychiatric Association, 2010). However, there are distinct differences in individual patient response to and side effects caused by the classes of medications and individual agents.

Antidepressant drug selection should be based on:

The patient's and family history of response to previous antidepressant medications (if any)
Patient preferences
Side effect profile (e.g., sedating, activating, weight gain, impact on sex life)
Antidepressant medications with anticholinergic side effects contribute to dry mouth/xerostoma, caries, gingivitis and periodontal disease (Tschoppe, 2010; Shinkai, 2006). This risk should be discussed with patients prior to initiation of these medications.
Safety in overdose (e.g., 10 days of a TCA can be a lethal overdose)
Availability and costs
Drug-drug interactions
Positive or negative impacts on the patient's comorbid psychiatric or medical conditions (for example, smoking cessation, ADHD or anxiety). See Annotation #4a, "Is the Presence of Substance Use Disorder or Psychiatric Comorbidity Suspected?" for more information on psychiatric comorbidities.

Medications and Genomics. The genetic differences in the metabolism of certain medications including antidepressants can be determined by cytochrome P450 genetic testing. This testing may identify individual patients who may be more sensitive to serious adverse reactions or medications with narrow therapeutic windows of specific medications. However, the clinical significance and applicability to daily clinical practice has not yet been established (Narasimhan, 2012; Porcelli, 2011).

Is medication needed? A meta-analysis of efficacy of acute (three-month) treatment with antidepressants (Fournier, 2010) for depression found the magnitude of benefit from antidepressant medications increased with the severity of a patient’s depressive symptoms. If a patient’s initial symptoms are minimal or qualified as mild or moderate depression, symptom benefits from antidepressants may not significantly differ from placebo. They suggested that for short-term and less-severe patients, behavioral activation plus lifestyle modifications may be enough. For patients with very severe depression, the authors found the benefit of medications over placebo is substantial (Fournier, 2010).

Classes of Medications

Selective serotonin reuptake inhibitors (SSRIs) and other antidepressants

SSRIs and other second-generation antidepressants such as venlafaxine, duloxetine, desvenlafaxine, mirtazapine, bupropion, levomilnacipran, vilazolone and vortioxetine are frequently recommended as first-line antidepressant treatment options due to the quality and quantity of published data, relative tolerability of side effects compared to TCAs and MAOIs, and their overall relative safety (American Psychiatric Association, 2010; Trivedi, 2001). They generally lack the common adverse reactions (anticholinergic, sedative effects) of the tricyclics antidepressants and cause fewer problems when taken in overdose. However, they may cause headache, nervousness, insomnia and sexual side effects (von Wolff, 2013). Newer antidepressant agents such as vilazodone, vortioxetine and levomilnacipran are available, but data on their long-term use is limited. They may also be more expensive or not routinely covered by insurance plans for some patients.

The current evidence does not support the choice of one second-generation antidepressant over another due to differences in efficacy or effectiveness. The choice of medication may depend on onset of action and adverse events (Gartlehner, 2008).

Secondary amine tricyclics (TCAs)

The literature clearly supports the effectiveness of tricyclics. Because of associated side effects, TCAs are used less frequently as first-line agents.

Secondary (nortriptyline) amine tricyclics cause less orthostatic hypotension and sedation than do tertiary (amitriptyline) amine tricyclics.

These medications should be monitored cautiously in patients with heart problems, or in patients with potential for drug interactions. Monitoring blood levels and EKG may be advised.

Monoamine oxidase inhibitors (MAOIs)

MAOIs, in general, should be restricted for patients who do not respond to other treatments, because of the potential for serious side effects and the necessity of dietary restrictions. Patients who have major depressive disorders with atypical features are one group for whom several studies suggest MAOIs may be particularly effective. However, in clinical practice, many psychiatrists start with SSRIs in such patients because of the more favorable adverse effect profile. Consider a dietary and/or psychiatry consultation if prescribing MAOIs.

Atypical antipsychotics

The atypical antipsychotics are recommended by the American Psychiatric Association as second-line augmentation options (American Psychiatric Association, 2010). Three atypical antipsychotics have been approved for the adjunctive treatment of major depressive disorder: aripiprazole, quetiapine and the combination of olanzapine and fluoxetine. There is some evidence regarding the use of quetiapine as monotherapy for the treatment of major depression (Zhornitsky, 2011). Unfortunately, the adverse effects of atypical antipsychotics may concern some patients (Wright, 2013). In a review of three randomized, placebo-controlled studies of quetiapine extended-release monotherapy in adults with major depressive disorder, the authors found it effective in response and remission of symptoms of depression. However, quetiapine was associated with a higher rate of side effects compared to placebo (Maneeton, 2012). When used as part of an augmentation strategy, doses should be individualized and safety (and efficacy) should be frequently reassessed (Wright, 2013).

Adherence, Patient Interaction and Monitoring

Adherence is important. For antidepressant medications, adherence to a therapeutic dose and meeting clinical goals are more important than the specific drug selected. Successful treatment often involves dosage adjustments and/or trial of a different medication at some point to maximize response and minimize side effects (American Psychiatric Association, 2010).

Key messages for patients using antidepressant therapy. When antidepressant therapy is prescribed, the following key messages should be highlighted to support medication adherence and completion of treatment goals:

Side effects from medication often precede therapeutic benefit and typically recede over time. It is important to expect side effects prior to symptom benefit.
Most people should be on medication at least 6-12 months after adequate response to symptoms (American Psychiatric Association, 2010).
Patients may have symptom improvement after two weeks but will need a longer length of time for full response and remission.
Take the medication as prescribed, even after you feel better. Premature discontinuation of antidepressant treatment has been associated with a 77% increase in the risk of relapse/recurrence of symptoms (Melfi, 1998). The probability of recurrence of depressive symptoms was found to be 25% after one year, 42% after two years, and 60% after five years in one study (Solomon, 2000). Each episode of recurrence increased the risk of subsequent episodes by 16% (Solomon, 2000).
Do not stop taking the medication without calling your clinician. Side effects often can be managed by changes in the dosage or dosage schedule.

Adherence strategies. Consider increasing education, engagement and follow-up for patients who are at higher risk for not adhering to treatment. For antidepressant treatment, this includes patients who are newly diagnosed with depression, in the midst of their first depression, or who have lapsed in the middle of a previous course of treatment (Vanelli, 2008). In addition to medication monitoring, clinical management of patients placed on antidepressants should include the clinician's support and reassurance.

Duration of Pharmacotherapy

Without long-term antidepressant treatment, major depressive relapses and recurrences occur in 50-80% of patients (American Psychiatric Association, 2010). Double-blind discontinuation studies reveal that antidepressants decrease the risk of relapse and recurrence; such studies have repeatedly shown antidepressants to be more efficacious than placebo substitution.

The dose of antidepressant medication that leads to satisfactory acute therapeutic response should be maintained during long-term treatment to reduce the risk for relapse and recurrence of depression (Sonawalla, 2001; Flint, 2000; Frank, 1993).

Discuss with the patient the specific side effect profiles, costs and benefits of different antidepressants, including generics. Cost implications for patients need to be discussed between clinician and patient.

There is no difference between brand versus generic medications based on adverse clinical outcomes.

When considering how long to continue medication after the remission of acute symptoms, two issues need to be considered: maintenance and prophylactic treatment. Patients who require several medication changes to achieve remission of an acute major depressive episode have a higher rate of relapse and a shorter period of time until relapse in comparison with patients who require fewer medication changes to achieve remission (Rush, 2006).

A recent review of maintenance trials for patients with major depression found continuing medication treatment after initial response to an antidepressant was successful in significantly reducing relapse rates compared with placebo. Mean relapse rates were 18% for antidepressants and 37% for placebo (Borges, 2014).

Significant data support the efficacy of antidepressants in preventing the recurrence of a major depressive episode. Although more research needs to be conducted, findings indicate that patients who are at highest risk of future episodes have had multiple prior episodes or were older at the time of the initial episode (Keller, 1998). These patients are candidates for long-term or lifetime prophylactic treatment. Analysis suggests that recurrence rates are reduced by 70% when patients are maintained on antidepressants for three years following their previous episode (average recurrence on placebo is 41% versus 18% on active treatment) (Hirschfeld, 2001; Greden, 1993).

Depression Medication Treatment Duration Based on Episode

(Sources: National Institute for Health and Care Excellence, 2012; American Psychiatric Association, 2010; Hirschfield, 2001; Keller, 1998; Greden, 1993)

Discontinuation of Pharmacotherapy

Premature treatment discontinuation can be triggered by a number of factors, including lack of adequate education about the disease, failure on the part of either physician or the patient to establish goals for follow-up, psychosocial factors and adverse side effects. Appropriate ongoing collaborative care for depression can increase remission rates to as much as 76% by 24 months (Rost, 2002; Schoenbaum, 2002).

Complicating factors are those situations where evidence either shows or suggests higher rates of recurrence after stopping antidepressants. Such factors include:

pre-existing persistent depressive disorder,
inability to achieve remission, and
recurrence of symptoms in response to previously attempted lowering dose or discontinuation of pharmacotherapy.

(Paykel, 1995)

If discontinuation of treatment is thought to be appropriate or necessary despite the known risks, a plan of action should be in place for prompt intervention if relapse occurs (Greden, 1993).

In general, it is recommended that the dose be tapered over a period of weeks to several months when discontinuing an antidepressant. (Note that this approach is only feasible when the starting dose is lower than the therapeutic dose.)

The various existing antidepressants exhibit a wide array of half-lives and therapeutic dose ranges. Therefore, a discussion of detailed discontinuation strategies is beyond the scope of this guideline.

Risks

Clinicians should be alert for worsening of symptoms. Health care clinicians should carefully evaluate their patients in whom depression persistently worsen, or emergent suicidality is severe or abrupt in onset, or was not part of the presenting symptoms. Reassessment is required to determine what intervention, including discontinuing or modifying the current drug therapy, is indicated.

The clinician should instruct the patient and the patient's caregiver to be alert for the emergence of agitation, irritability and other symptoms. The emergence of suicidality and worsening depression should be closely monitored and reported immediately to the clinician.

Risks Related to Special Populations

Children, adolescents and young adults

The U.S. Food and Drug Administration has requested manufacturers of antidepressants include a warning statement regarding antidepressants increasing the risk compared to placebo of suicidal thinking and behavior (suicidality) in children, adolescents and young adults. The full warning statement can be found at http://www.fda.gov/Drugs/DrugSafety/InformationbyDrugClass/UCM096273.

Be alert for worsening of symptoms. Health care clinicians should carefully evaluate their patients in whom depression persistently worsen, or emergent suicidality is severe or abrupt in onset, or was not part of the presenting symptoms. Reassessment is required to determine what intervention, including discontinuing or modifying the current drug therapy, is indicated.

Elderly patients

Because of the potential for decreased renal and hepatic function, and also for concomitant diseases and medications, the elderly are at higher risk of significant side effects or drug interactions with antidepressant medications. For elderly patients with moderate to severe depression, TCAs such as nortriptyline continue to be regarded as the most effective treatment (Alpert, 2003; Gastó, 2003). Consider starting at the lowest possible dose and increasing slowly to effective dose or until side effects appear. Tertiary amine tricyclics should generally be avoided in elderly patients because of the high incidence of orthostatic hypotension, sedation, cognitive problems and cardiac effects with these agents.

Risks of Medication Interactions

Many antidepressant agents have clinically significant drug interactions, particularly those agents that undergo cytochrome P450 enzymatic metabolism in the liver. A complete discussion of this topic is beyond the scope of this guideline. Practitioners are advised to consult references such as the Physician's Desk Reference, American Hospital Formulary Service, Epocrates or Micromedex for more information about drug interactions with specific agents, and to assess the significance of the interaction prior to prescribing antidepressants.

Risks Associated with Specific Medications

Citalopram warning

In 2011, the Food and Drug Administration (FDA) published a "Medwatch" drug safety alert regarding the potential risk of abnormal heart rhythms associated with citalopram doses greater than 40 mg a day due to concerns about prolonged QT interval prolongation and the risk for torsades de pointes. Prescribers were initially told to avoid using citalopram doses higher than 40 mg and discouraged from using it at all in patients with congenital long QT syndrome, bradyarrhythmias, congestive heart failure, or at risk for developing hypokalemia or hypomagnesemia. In March 2012 this was revised by downgrading the warning from "contraindicated" to "not recommended" for patients with congenital long QT syndrome because patients with this condition have few viable alternative treatments. Ongoing monitoring was suggested, a maximum dose of 20 mg/day was recommended for age > 60, and discontinuation was recommended when QTc > 500ms.

A recent review of Veterans Health Administration patients who were prescribed citalopram between 2004 and 2009 (N=618,450) found daily doses of citalopram greater than 40 mg a day were associated with lower risks of ventricular arrhythmias, all-cause mortality and non-cardiac mortality compared with lower doses of citalopram. Overall, no increased risks of cardiac mortality were observed. These results were similar when compared with a cohort of patients prescribed sertraline (N=365,898) during the same time period (Zivin, 2013).

Serotonin syndrome

Serotonin syndrome is a potentially life-threatening, pharmacodynamic drug interaction resulting in excessive nervous system levels of serotonin. Patients experiencing this reaction may present with mental status changes such as anxiety, confusion, delirium or coma. Autonomic symptoms may include tachycardia, labile blood pressure and hyperthermia. Muscle rigidity, ataxia, tremor, myoclonus and other neurologic symptoms are also common.

Serotonin syndrome has often been inaccurately reported and erroneously attributed to various serotonergic medications. Specific diagnostic criteria have been developed to assist prescribers in the diagnosis of the "toxidrome" (Evans, 2010; Gillman, 2006). Rather than an idiosyncratic reaction, serotonin syndrome or serotonin toxicity is the result of drug-induced elevations of intrasynaptic serotonin. Not all serotonergic agents are capable of producing the intrasynaptic elevation of serotonin associated with true serotonin toxicity (Gillman, 2006).

The primary criterion for an accurate diagnosis and risk assessment is recent exposure to a serotonergic agent or combination of agents able to produce significant elevations of synaptic serotonin. According to the Hunter Area Toxicology Service (HATS) data, the higher levels of intrasynaptic serotonin caused by combinations of MAOIs with an SSRI are likely to cause hyperpyrexia and death. The combinations of clomipramine, imipramine or venlafaxine with an MAOI have also been associated with fatalities (Gillman, 2006).

In 2006, the FDA issued a warning about the life-threatening risk of combining SSRIs with triptans (for the treatment of migraine headaches). The warning included 29 case reports. Most of the case reports were incomplete and often did not meet established diagnostic criteria for serotonin syndrome. Current evidence does not support limiting the use of triptans with SSRIs or SNRIs (Evans, 2010; Gillman, 2010; Wenzel, 2008).

Integrative Medicine

While there are many integrative treatments available, our discussion highlights some of the types of treatments. They include acupuncture, yoga, herbs and dietary supplements. They were selected because they are evidence based and/or more commonly utilized.

Acupuncture

There is considered to be high-level evidence to support the use of acupuncture during pregnancy for the treatment of depressive episodes (Sniezek, 2013). An open, parallel-arm, randomized study showed acupuncture to result in equal efficacy in comparison to counseling with a significant reduction in depressive symptoms for both in comparison to usual care (MacPherson, 2013). Existing meta-analyses and systematic reviews vary with respect to acupuncture protocol (manual, electroacupuncture or sham), methodological soundness and efficacy results (Freeman, 2010). Both sham and active acupuncture participants generally report symptomatic depression improvement (Freeman, 2010). Serious adverse events from acupuncture are very uncommon, which may appeal to those who seek to avoid side effects associated with traditional treatments (e.g., medication side effects).

Yoga

Yoga has been shown to be effective as an adjunctive treatment to decrease symptom severity (Ravindran, 2009). It has yet to be determined what aspects of yoga are responsible for any potential depressive symptom improvements (Louie, 2014).

Tai Chi

Limited evidence suggests that Tai Chi may be effective for psychological well-being measures that include depressive symptoms (Wang, 2014). There is yet insufficient evidence, though, to recommend Tai Chi for the treatment of depressive episodes.

Herbals and dietary supplements

Caution: Many drugs interact with St. John's Wort, including other antidepressants, warfarin, oral contraceptives, antiretroviral, anti-cancer and anti-rejection drugs. Care should be taken to ask all patients what medications they are taking, including over-the-counter and supplements, to avoid these interactions.

Herbal products and nutritional supplements are not evaluated or regulated by the U.S. Food and Drug Administration for safety, efficacy or bioavailability.

St. John's Wort and Sam-E. In a meta-analysis (Morgan, 2008), S-adenosylethione (Sam-E) and hypericum perforatum (St. John's Wort) were found to have indications for mild to moderate depression but not major depression. Sam-E and St. John's Wort should not be taken in combination with other antidepressant medications. A Cochrane meta-analysis concluded that there is insufficient evidence to recommend the use of St. John's Wort in the treatment of major depression. Research is limited by lack of large scale RCTs and high risk of bias in the majority of trials meeting inclusion criteria for the meta-analysis (Smith, 2010; Linde, 2008).

Saffron (Crocus sativus L.)

A recent meta-analysis show potential efficacy of saffron in comparison to placebo, with possible equal efficacy to fluoxetine and imipramine. The studies included were small and only studied the acute phase of treatment (Hausenblas, 2013). It will require further study to adequately assess the acute efficacy, as well as establish the safety, tolerability and durability of efficacy in the continuous and maintenance phases of therapy.

Possible link between deficiency and depression. A number of researchers have published studies and review articles regarding an increased risk of depression in patients with low levels of zinc, omega-3 fatty acid or magnesium. Unfortunately, studies on appropriate supplementation of these dietary aides are often inconsistent in their design and results. While the replacement of zinc, omega-3 fatty acid and magnesium in patients with known deficiencies and who have major depression is often recommended, the exact dosages and durations of supplementation are not known (Appleton, 2010; Siwek, 2010; Colangelo, 2009).

Omega-3 fatty acid not helpful as treatment. A recent meta-analysis of randomized, placebo-controlled trials of omega-3 fatty acid (FA) in the treatment of major depressive disorder was designed to analyze the efficacy of omega-3 FAs in the treatment of MDD and to examine possible sources of heterogeneity between trials. The meta-analysis demonstrated no significant benefit of omega-3 FA treatment compared with placebo and significant heterogeneity in study design, as well as publication bias (Bloch, 2012).

Vitamin D. At this time, there is insufficient evidence on the antidepressant effects of vitamin D (Shaffer, 2014; Li, 2014; Thacher, 2011).

Medical cannabis

There is insufficient evidence to support use of cannabis in treatment of depression. A systematic review and meta-analysis of randomized, clinical trials of cannabinoids for the treatment of several indications, including depression, was published in June 2015. No studies of cannabinoids for the treatment of depression met the authors' inclusion criteria. The authors evaluated five studies of other primary indications (chronic pain and spasticity in multiple sclerosis) that reported depression as an outcome measure. Three of the studies found no difference between cannabinoids (dronabinol and nabiximols) in depression outcomes, compared with placebo. The majority of these studies were found to have a high risk of bias (Whiting, 2015).

Special Populations

See Appendix D, "Special Populations," for more information regarding treatment for the following conditions/populations, as applicable: 1) cardiovascular and cerebrovascular disease; 2) diabetes; 3) chronic pain; 4) geriatrics [includes dementia/cognitive impairment]; and 5) pregnant and postpartum women.