CONTINUATION PERIOD

This period usually lasts five to eight months after the end of the acute treatment period. The goal at this phase is the prevention of relapse. There is a high risk of relapse if treatment is discontinued after the acute treatment phase, with rates of 15% after six months and 22% after 12 months. The two best predictors of relapse are a high number of previous depressive episodes (greater than three predicted relapses) and underlying dysthymic disorder.

Once a patient has responded to medication, treatment should be continued for a minimum of four to six months from the point of initial response. This period should be lengthened for the patient with a history of longer depressive episodes.

In the past, it was suggested that, on achievement of euthymia, doses could be reduced. However, it is more likely that levels similar to those needed at the acute stage of treatment will be required throughout the continuation period.

MAINTENANCE PERIOD

The goal of the maintenance period is to prevent the recurrence of depression. There are a number of reasons to consider long-term prophylactic therapy for depression rather than medication withdrawal:

1. Depression is a lifelong disease, with recurrence being the norm rather than the exception

2. As the number of acute episodes increases, the risk of future episodes increases as well, and the interval between episodes shortens

3. Each subsequent episode carries a higher morbidity and disability

4. There is a fear that treatment response may decrease with an increasing number of depressive episodes

Several factors influence the decision to maintain long-term prophylaxis for depression, including the seriousness of previous episodes, the severity of impairment caused by such episodes, the degree of response to previous treatments, and the ability of the patient to tolerate the drug. Central in the decision process is the concept of recurrent depression: that some patients are more likely than others to have a recurrence of the disease. Three previous episodes of depression make recurrent depression likely.

The best predictors of the likelihood of recurrence appear to be older age of onset and number of episodes. Long-term maintenance is the treatment of choice for the following groups of patients:

1. 50 years old or more at the time of the first depressive episode

2. 40 years old or more at first episode and have had at least one subsequent recurrence

3. Anyone who has had more than three episodes

The recommended length of maintenance therapy varies from five years of treatment to indefinite continuation. Recent studies with both acute major depression and recurrent depression have shown significantly higher relapse rates 2-3 years post-discontinuation compared with patients on maintenance treatment.

Regarding choice of agent, there are no rigorous studies comparing different antidepressants during the maintenance period. It is usually assumed that the same agent used in the acute and continuation period will also be the preferred one in the maintenance period.

Equally important in preventing recurrence of depression is the problem of maintaining adherence to medication long after the acute episode has resolved. Proper education and support will help with compliance. Toleration of side effects is important; patients are more likely to comply with agents that have more favorable side effect profiles. selective serotonin reuptake inhibitors (SSRIs) or bupropion (sustained or extended release forms) are generally the best-tolerated antidepressants.

Although lower doses for prophylaxis have been recommended, there are few data to support this contention. Even though lower doses may increase compliance, full doses should be used until new information indicates otherwise.

DISCONTINUATION OF TREATMENT

Before discontinuing treatment with an antidepressant it is important to remember that depression is often a lifelong disease with a chronic course. One should always weigh the benefits of discontinuation against the risks of recurrent depression. Patients with a single episode of acute depression and who have an onset before age 50 are the best candidates for discontinuation.

For tricyclic antidepressants (TCAs), the usual strategy is to taper the dose at a rate of 25 to 50 mg/day every 2 to 3 days. Too rapid a decrease in dose may produce symptoms of cholinergic ‘rebound’ or supersensitivity (nausea, vomiting, cramps), other signs of autonomic hyperactivity (diaphoresis, anxiety, agitation, headache), and insomnia as early as 48 hours or as late as 2 weeks after discontinuation. These early symptoms may be mistaken for relapse of depression.

Monoamine oxidase inhibitors (MAOIs) may also have a withdrawal syndrome on abrupt cessation of treatment, including symptoms of psychosis; however, this is rarer than that seen with the TCAs.

Recommendations for discontinuing selective serotonin reuptake inhibitors (SSRIs) depend on the particular drug. Fluoxetine has a long half-life and abrupt withdrawal should be permissible. Shorter half-life drugs, such as sertraline, citalopram, escitalopram, and paroxetine, may require a 7-10-day taper. The withdrawal syndrome with shorter-acting agents includes symptoms of fatigue, insomnia, abdominal distress, and influenza-like symptoms. The same may be true for venlafaxine and duloxetine.

A first episode of depression has a high risk of recurrence and the risk is higher in patients who have only a partial response to treatment. After discontinuation the goal is to enable early intervention if symptoms recur. The patient should be educated to recognize the symptoms of depression and seek help at an early stage.

In patients showing an inadequate response after a reasonable time, the decision is either to continue with the same medication and augment with an additional agent, or to switch medications altogether, depending on whether the patient has shown any response to the current strategy. Partial responders may be more likely to benefit from treatment augmentations, whereas patients who show no response or worsen during treatment warrant a new agent.

Antidepressant Augmentation

Typical augmentation strategies shown in Table: Augmentation Strategies include the addition of lithium carbonate, thyroid hormone, a stimulant, an atypical antipsychotic, or by the use of antidepressant combinations.

Table: Augmentation Strategies

† Inadequate data.

LITHIUM AUGMENTATION

Lithium has been used successfully with most antidepressants, including tricyclic antidepressants (TCAs), serotonin reuptake inhibitors, and bupropion, with response rates as high as 65%. The blood level of lithium necessary for adjunctive use has not been well established. It is probably best to start at a low dose (300 mg b.i.d) and increase to a therapeutic blood level (0.8 to 1.2 mEq/L) if there is no response. It may take three to six weeks for augmenting effect. If augmentation is successful, it should be continued throughout the acute phase of treatment. This strategy can be limited by side effects and lithium’s narrow therapeutic index. It is associated with hypothyroidism, tremor, increased thirst, increased urination, nausea, weight gain, and acne. Patients should be cautioned to avoid dehydration, which can precipitate toxic lithium blood levels.

THYROID HORMONE AUGMENTATION

Starting dose is 25 µg/day of triiodothyronine, which can be increased to 50 µg/day in a week if there is no response. The trial should continue for at least three weeks. Reported response rates for thyroid augmentation are lower than those for lithium augmentation (25%). As with lithium augmentation, if there is a positive response, it occurs relatively early.

STIMULANT AUGMENTATION

Methylphenidate and dextroamphetamine at dosages between 5 and 20 mg have been used for antidepressant augmentation, but there are few systematic data regarding the proper dose or length of treatment for this potential use.

ANTIDEPRESSANT DRUG AUGMENTATION

Second-generation antipsychotic drugs are increasingly being used adjunctively with antidepressants for residual or treatment-resistant symptoms. APDs are generally used within their therapeutic range.

COMBINED ANTIDEPRESSANT THERAPY

Open trials have supported the use of combined therapy of a TCA and a serotonin reuptake inhibitor in patients for whom either class alone has failed. When antidepressants are combined, it is important to remember that the serotonin reuptake inhibitors can potentiate TCA levels, and this should be monitored carefully. Monoamine oxidase inhibitors (MAOIs) have also been used in combination with tricyclic antidepressants (TCAs), although this should be monitored closely given the risk of potential toxic interactions. Given the risk of a serotonin syndrome, MAOIs should not be combined with serotonin reuptake inhibitors. Bupropion is frequently added to an SSRI to enhance efficacy or to alleviate side effects such as decreased libido, fatigue, and sedation.

NONPHARMACOLOGICAL APPROACHES

A number of nonpharmacological augmentation options exist, particularly the concomitant use of psychotherapy. The combination of psychotherapy and medication may offer benefits that either therapy alone cannot offer, including additional efficacy as well as prevention of relapse. This has been shown to some degree for both cognitive-behavioral therapy and interpersonal therapy.

Changing to a New Agent

Patient showing no response or whose condition deteriorates during therapy should trial an alternative single agent. Recent studies support the belief that it is best to switch to an agent of a different class, and approximately 50% of patients unresponsive to a first trial respond to an antidepressant of a different class.

When the switch involves an MAOI, sufficient time must be given for medication clearance. Although seldom used, monoamine oxidase inhibitors (MAOIs) may be very effective in patients not responsive to other classes of antidepressants. Generally, 10 to 14 days is required for clearance of tricyclic antidepressants (TCAs) and MAOIs. Fluoxetine requires a much longer period — 6 weeks — whereas sertraline and paroxetine require about two weeks when switching to an MAOI. Another alternative is changing from an SSRI medication to an SNRI. There is some evidence that serotonin-norepinephrine reuptake inhibitors (SNRIs) may lead to higher rates of response and remission than selective serotonin reuptake inhibitors (SSRIs).

ECT ECT and other forms of stimulant therapies (e.g., vagal nerve stimulation) should be considered for patients who are non-responsive to pharmaco- and psychotherapies.

Starting doses, titration schemes, and target doses for commonly-used tricyclic antidepressants (TCAs), monoamine oxidase inhibitors (MAOIs), selective serotonin reuptake inhibitors (SSRIs), and other agents are shown in Table: Starting Dose, Titration Steps, and Target Doses for Common Antidepressants. Once medication is initiated, it should be gradually increased to therapeutic levels by titration.

TCAs are usually started at a relatively low dose; these low doses are preferred in elderly patients. In the frail elderly, further dose reductions may be needed — 50% or less of the usual starting dose.

The usual treatment plan with SSRIs is to start a patient at the lowest effective dose and increase as indicated by clinical response. An increased response in usually seen with increased dose, however the dropout rate due to side effects also increases. For children, adolescents, the elderly, and patients who find medications generally difficult to tolerate, 50% reductions in the starting doses shown in Table: Starting Dose, Titration Steps, and Target Doses for Common Antidepressants are acceptable. During titration, SSRIs may temporarily increase anxiety, insomnia, or both. These side effects can be alleviated by dose reduction, short-term use of benzodiazepines for anxiety, and the addition of either hypnotics or trazodone for sleep.

With both TCAs and SSRIs there is a significant delay between initiation of medication and response; there is no reason to believe that increasing the dose beyond the therapeutic range hastens response.

Bupropion (immediate and sustained release forms) and trazodone require divided doses. In the elderly, a usual starting dose of bupropion is 100 mg/day of the sustained release preparation, then increased to 100 mg b.i.d. Single daily dosing is possible using the extended release form of bupropion.

Table: Starting Dose, Titration Steps, and Target Doses for Common Antidepressants

Therapeutic Drug Monitoring

Although blood levels are available for many antidepressants, those for imipramine, desipramine, and nortriptyline have been best established. Imipramine and desipramine appear to have a curvilinear dose-response curve with an optimal range of 150 to 300 ng/mL. Nortriptyline appears to have a therapeutic window in the range of 50 to 150 ng/mL, usually reached by doses ranging from 50-100 mg/day. These blood levels are nominal, as some patients do respond above or below these ranges, and blood level monitoring should not be a substitute for clinical observation.

Drug levels have not been well established for the monoamine oxidase inhibitors (MAOIs) and the serotonin reuptake inhibitors. For the latter compounds there are numerous speculations about possible dose-response curves, including linear and “therapeutic window” models.

Early, or Pre-response, Period

Patients should initially be followed weekly to judge their response to treatment and manage the side effects of the various medications; given realistic time and economic constraints, such contacts may at times be by telephone.

Side effects are a primary reason for treatment non-adherence, and patients may be more inclined at this early phase to simply discontinue medication rather than to first discuss it with their physician. It is inevitable that patients will experience some side effects from their medications. Patients are able to tolerate side effects better when they are framed in a positive light — as a sign that the drug is present in their system.

Strategies for reduction of side effects include:

• Selective serotonin reuptake inhibitors (SSRIs), through scored tablets or liquid preparations, can be begun at 25-50% of normal starting dose in patients particularly sensitive to side effects. Panic disorder patients often fit into this category

• Tricyclic antidepressants (TCAs), although often given in a single-daily dose, could be divided across the day to minimize dose-related side effects

• Taking medications with food may decrease nausea

• Reminding patients that sedating medications should be taken in the evening, and activating ones in the morning

• Flexibility in time of dosing: a significant percentage of patients on SSRIs experience sedation rather than insomnia, warranting a change to evening dosing

Should intolerable side effects warrant medication change, select a medication in the same class as the first but with a different side effect profile; patients can have variable reactions to different SSRIs despite their apparent similarity. For instance, paroxetine is more sedative, and sertraline may cause more gastrointestinal distress than other SSRIs.

Response, or Acute Treatment, Period

This period overlaps with the initial phase of treatment and continues until response is achieved, usually two to four months. The goal during this phase is to control the present symptoms of depression. It is important to differentiate between partial and complete response; complete response implies total recovery from all symptoms of depression, whereas a partial response is usually defined as a reduction in symptoms.

The time to response varies between patients. Few patients show a significant response before two weeks. The usual range for response is three to four weeks; however, it can take six weeks or longer. For patients who complete a satisfactory treatment regimen, the response rate for antidepressants is about 60 to 70%, although some of these responses will be partial. Response rates may be as high as 80% with antidepressants when an adequate dose is given for an adequate time.

Both longitudinal and cross-sectional factors should be considered when selecting an antidepressant for major depression:

• What were the course, duration, and severity of any previous episodes of depression?

• Is there a history of antidepressant response?

• How well was the antidepressant tolerated?

• Is there a family member with a history of antidepressant response; if so, to which medication?

• If there is a history of antidepressant failure, were the trials of an adequate dose and duration?

• Are melancholic, atypical, or psychotic features present?

• Does the patient have a history of sensitivity to anticholinergic, histaminic, α-adrenergic, serotonergic, or noradrenergic side effects?

• Does the patient have suicidal features or a history of impulsivity that may increase the risk for potential overdose?

• Does the patient have a history of a cardiac conduction delay or recent myocardial infarction (which would contraindicate the use of TCAs)?

• Does the patient have a history of symptoms suggestive of mania (or more minor variants of manic episodes)? If so, avoid tricyclic antidepressants (TCAs) and monoamine oxidase inhibitors (MAOIs)

• Does the patient presently take or need sympathomimetics (which would contraindicate the use of MAOIs)?

• If an antidepressant trial has failed, was the patient a partial responder or a nonresponder?

• Were any augmentation strategies employed?

Special Considerations in the Selection of an Antidepressant

The following factors should be considered when selecting an antidepressant agent.

Gender Women may have slower gastrointestinal absorption than men due to less gastric acid and slower gastric emptying. A woman’s volume of distribution differs as well, given her increased ratio of adipose tissue to lean body mass. Water retention associated with the menstrual cycle may also affect the volume of distribution. Oral contraceptives can alter the hepatic metabolism of tricyclic antidepressants (TCAs).

Ethnic and Racial Factors African-Americans may be more likely than those of European descent to be slow metabolizers of antidepressant due to possible genetic differences in metabolic enzyme expression; a growing number of studies suggest that African-Americans will have higher plasma levels per dose of antidepressant. This has been primarily demonstrated with TCAs; data with selective serotonin reuptake inhibitors (SSRIs) are lacking. Some data suggest that Asians are slower to metabolize nortriptyline than other groups.

The greatest concern regarding race and ethnicity is suggested in a study of prescribing practices in Westchester County (New York) mental health clinics, which found that minorities were less likely than nonminorities to be offered antidepressant treatment, independent of diagnosis.

Age Is the patient elderly? If so, selective serotonin reuptake inhibitors (SSRIs) are indicated as half-lives and steady-state concentrations of these agents are only minimally affected by age (although paroxetine may be an exception to this).

Comorbidities Patients with renal impairment may require dose adjustment for many antidepressants; fluoxetine and sertraline appear to be the exception. Liver disease can increase levels of tricyclic antidepressants (TCAs) — blood levels should be monitored. SSRIs are extensively metabolized in the liver and lower doses should be given in hepatic-impaired patients. Nefazodone is not indicated due to association with hepatic injury.

Pregnant women and nursing mothers Clinicians must consider whether the potential benefits of treatment justify the potential risk to the fetus. The risks of untreated depression are well-described, and the risks of medication treatment during pregnancy and lactation are less well-known. In general, studies have not shown increases in fetal malformations from antidepressant exposure. Reports of withdrawal syndromes among neonates have led some experts to recommend tapering and discontinuing antidepressants 10 to 14 days before the mother’s due date. Antidepressants are excreted in varying degrees in breast milk, so breastfeeding in antidepressant-treated women should be done with caution.

Predictors of poor response In assessing the likelihood of success, the following predictors of poor response have been identified: neurovegetative symptoms such as hypersomnia, hyperphagia, psychomotor agitation, anxiety and irritability, personality features (such as neurotic, hypochondriacal, and hysterical traits), multiple prior episodes, delusions, and psychomotor agitation.

Continuation: Drug Selection and Initiation of Treatment for Major Depression. Treatment

All antidepressants appear to treat more than depressive disorders. Particularly consistent have been data showing their utility for anxiety disorders; they have begun to supplant the sedative/hypnotics for these conditions. Many other psychiatric and medical disorders have all been successfully treated with these agents.

Panic Disorder (PD)

Selective serotonin reuptake inhibitors (SSRIs) are considered the first-line treatment for anxiety disorders, with good evidence of efficacy in panic disorder. Patients are begun at low doses (e.g., 5mg of fluoxetine), and increased slowly to an effective dose to minimize potential side effects. There is also strong evidence for the use of tricyclic antidepressants (TCAs) and moderate evidences for the use of monoamine oxidase inhibitors (MAOIs).

Obsessive-Compulsive Disorder (OCD)

Selective serotonin reuptake inhibitors (SSRIs) show strong evidence of effectiveness in the treatment of obsessive-compulsive disorder and are considered the first-line treatment option in this disorder; all agents appear equally effective. The recommended starting dose is the same as that for depression, although higher doses (60-80 mg of fluoxetine) may be required for adequate response. There is evidence for slow but continued improvement in symptoms for many months after initiation of treatment; medication should therefore be trialed for up to four months if the patient shows a partial response. There is also strong evidence for the use of the TCA clomipramide in this disorder. Augmentation with lithium may be beneficial in partial responders.

Generalized Anxiety Disorder (GAD)

Several agents have shown efficacy in this disorder. Good evidence exists for the use of selective serotonin reuptake inhibitors (SSRIs), venlafaxine, nefazodone and mirtazapine. Both tricyclic antidepressants (TCAs) and monoamine oxidase inhibitors (MAOIs) have also shown moderate effectiveness; MAOIs also show efficacy in other anxiety disorders.

Social Phobias and Posttraumatic Stress Disorder (PTSD)

Selective serotonin reuptake inhibitors (SSRIs) and venlafaxine have shown efficacy in social phobia (or social anxiety disorder); this condition may also be at least partially responsive to TCA medications. SSRIs have shown effectiveness in posttraumatic stress disorder (PTSD).

Bulimia

Selective serotonin reuptake inhibitors (SSRIs) are commonly used in the treatment of bulimia nervosa. As with obsessive-compulsive disorder (OCD), there appears to be a dose-response effect, with larger doses often required.

There is strong evidence for the use of tricyclic antidepressants (TCAs), either alone or in combination with cognitive therapy, to decrease the binging and purging of bulimic patients. Monoamine oxidase inhibitors (MAOIs) have also been reported successful in the treatment of bulimia, although the many dietary restrictions have made physicians reluctant to prescribe these agents.

Anorexia Nervosa

The American Psychiatric Association Practice Guidelines for Eating Disorders states that medications should not be used routinely for anorexia nervosa. They should be considered after weight gain and for persistent depression. A few published controlled studies show unimpressive results, and bupropion is contraindicated because of elevated seizure risk in patients with eating disorders.

Body Dysmorphic Disorder (BDD)

Selective serotonin reuptake inhibitors (SSRIs) have also shown efficacy in the treatment of BDD. It may be at least partially responsive to TCA medications, particularly those selective for serotonin. Moderate evidence exists for the use of monoamine oxidase inhibitors (MAOIs).

Premenstrual Dysphoria (PMDD)

PMDD is a chronic cyclical disorder in which serotoninergic function is reduced during the luteal phase. The treatment of choice is a SSRI; either sertraline or modified-release paroxetine, taken daily or only during the luteal phase. Treatment with sertraline should be initiated at a dose of 50 mg/day and if necessary increased in increments of 50 mg/day at each menstrual cycle to a maximum of 150 mg/day if taken every day or 100 mg/day if taken during the luteal phase only. The initial dose of modified-release paroxetine is 12.5 mg/day increasing to 25 mg/day after one week if necessary.

Intermittent dosing is usually as effective as continuous administration. Beneficial effects are seen within one to two days and response rates for improvements in mood and physical symptoms are about 50 to 60%. Evidence of long-term efficacy is lacking. Specialists recommend continuing treatment until the menopause as there is evidence that stopping treatment precipitates recurrence. Nevertheless, a trial period off treatment may be worthwhile after two years if supported by the patient.

Childhood Disorders

Tricyclic antidepressants (TCAs), especially imipramine, are indicated for the treatment of enuresis. Anxiety and phobias in children (such as separation anxiety or school phobia and ADHD) are responsive to TCAs.

Uses for selective serotonin reuptake inhibitors (SSRIs) in children include repetitive-type abnormalities associated with autism and mental retardation, ADHD (as an adjunct to methylphenidate), and chronic enuresis. Bupropion has been used successfully in the treatment of ADHD in both children and adults. It may, however, exacerbate tics in attention-deficit patients with concomitant Tourette’s syndrome.

Other Psychiatric Disorders

Other indications for selective serotonin reuptake inhibitors (SSRIs) may include depersonalization disorder, obsessive jealousy, pathological gambling, Tourette’s syndrome, hypochondriasis, and both paraphiliac and nonparaphiliac sexual disorders.

SSRIs have shown effectiveness in a number of more complex behavioral disorders such as obesity (particularly fluoxetine in higher doses), binge eating (sertraline), substance abuse, and to alleviate certain aggressive behaviors such as impulsivity and uncontrolled anger in adults, children, and the demented elderly.

Trazodone is frequently used as a sedative in the elderly; as it can cause or worsen orthostatic hypotension, blood pressure should be monitored when used in this group. In the demented elderly, trazodone is useful in treating behavioral disorders associated with dementia.

The sedative effect of trazodone makes it useful in weaning patients from benzodiazepines and other sedative drugs.

Nefazodone has shown efficacy in treating anxiety associated with major depression. Similarly, mirtazapine has shown efficacy in anxiety symptoms in general.

Other Medical Conditions

Migraine and cluster headaches have been responsive to both tricyclic antidepressants (TCAs) and selective serotonin reuptake inhibitors (SSRIs). Diabetic neuropathy and other pain syndromes such as facial pain, fibrositis, and arthritis have been responsive to both TCAs and SSRIs and the SNRI duloxe-tine. TCAs such as protriptylene have shown efficacy for sleep apnea. SSRIs have been used in the treatment of restless leg syndrome.

A full list of indications is shown in Table: Various Uses of Antidepressants.

Table: Various Uses of Antidepressants

The use of antidepressants in the treatment of depression remains the best-understood use of these medications; however, there is a growing list of other indications for antidepressants, including panic disorder (PD), obsessive-compulsive disorder (OCD), bulimia and posttraumatic stress disorder (PTSD). Many of these illnesses respond best to combination treatment modalities that include medication and various forms of psychotherapy.

The first antidepressant was the monoamine oxidase inhibitor (MAOI), iproniazid, initially licensed as an antituberculosis drug. This was soon followed by the tricyclic antidepressant (TCA), imipramine. Although both were developed in the early 1950s, they represented different paths of discovery: iproniazid was the result of clinical and laboratory collaboration, whereas imipramine’s introduction was largely based on clinical observation.

Within the past 25 years, similarly efficacious tricyclic and heterocyclic antidepressants have been developed but with varying side-effect profiles. A new direction in antidepressant pharmacology began in 1987 with the discovery and marketing of fluoxetine, the first antidepressant selective for serotonin reuptake blockade. Beyond selective serotonin reuptake inhibitors (SSRIs), the latest offerings in antidepressant pharmacology include agents that act at multiple neurotransmitter levels.

In evaluations of the many antidepressants available, the focus has generally been on the agent’s side-effect profile and ease of administration. Despite many efforts in this area, there is no conclusive evidence to demonstrate the clinical superiority of any one group of agents.

Mechanisms of Action

All known antidepressants affect monoamine neurotransmission.

Monoamine oxidase inhibitors (MAOIs) inhibit the activity of monoamine oxidase, resulting in a decrease in the breakdown of dopamine, serotonin, and norepinephrine in the synapse, thus increasing the amount of these neurotransmitters available for release and synaptic transmission. MAOIs are highly effective antidepressants and anxiolytics but are rarely used due to dietary restrictions that must be adhered to in order to avoid tyramine-induced hypertensive crises.

Tricyclic antidepressants (TCAs) block presynaptic reuptake of norepinephrine and serotonin to various degrees. Several are related by metabolism, thus the tertiary amines amitriptyline and imipramine are metabolized to the secondary amines nortriptyline and desipramine respectively. All act through reuptake inhibition and are generally selective for the norepinephrine transporter; several, however, have equal or greater affinity for the serotonin transporter. The TCAs were the drugs of choice for depression through the 1980s. Though effective, their nonselective actions on cholinergic, histaminergic, and presynaptic adrenergic receptors resulted in a number of side effects.

Selective serotonin reuptake inhibitors (SSRIs) were first introduced in the late 1980s, and rapidly eclipsed the tricyclic antidepressants (TCAs) as the drugs of choice for depression. There are subtle differences between compounds, mainly in terms of their half-life, their potency for reuptake inhibition, and their affinity for some other receptors. As SSRIs more selectively affect reuptake, with few effects on the adrenergic, histaminergic, and cholinergic systems, side effects have been reduced.

The selective norepinephrine reuptake inhibitors (NRIs) such as reboxetine (not available in the US) or atomoxetine share a similar mechanism with the SSRIs, but act on the norepinephrine transporter and have little affinity for the serotonin transporter.

Several other second-generation agents, sometimes referred to as “atypical antidepressants”, were introduced during the same period as the selective serotonin reuptake inhibitors (SSRIs). These include bupropion, which seems to exert primarily a dopaminergic effect, and trazodone, which is structurally related to the tricyclic antidepressants (TCAs) but has a primary serotonergic mechanism.

Of the serotonin receptors, 5-hydroxytryptamine type 1A (5-HT_1A) appears most related to the therapeutic effects of antidepressants, and this receptor influences norepinephrine, dopamine, acetylcholine, neuropeptides, other serotonin receptors, and probably beta-receptor downregulation.

The so-called third-generation antidepressants include serotonin-norepinephrine reuptake inhibitors (SNRIs — venlafaxine, milnacipran and duloxetine), mixed serotonin antagonist/reuptake inhibitors (nefazodone and trazodone) and the mixed serotonin/noradrenaline antagonist mirtazapine. SNRIs have equal affinity for the norepinephrine and serotonin transporter. Though, like several of the TCAs, venlafaxine has multiple receptor effects, it is relatively free of the anticholinergic and antihistaminic side effects that are common with the tricyclic antidepressants (TCAs).

Mixed serotonin antagonist/reuptake inhibitors such as nefazodone have multiple mechanisms of action, with both serotonin (as well as norepinephrine) transporter inhibition as well as antagonism of 5-HT_2A and alpha-₁-receptors. Trazodone is similar; however, its effects are somewhat less specific.

The mixed serotonin/noradrenaline antagonist mirtazapine is unique in that it appears to work primarily through specific receptor blockade of the alpha-₂-autoreceptors on presynaptic noradrenergic neurons, enhancing noradrenergic output. This class may exert a similar effect toward autoreceptors on serotonin neurons. Antagonism of 5-HT₂ and 5-HT₃ receptors may also concentrate the effect of serotonin on 5-HT_1A receptors. The antidepressants available in the US, their class, and relative costs are listed in Table Table: Antidepressants available in the United States.

Table: Antidepressants available in the United States

* Recommended dose derived from Lexi-comp online.

** Price index: (dollar cost of average dose/d) The price index is a rough calculation, assuming an average recommended dose for the treatment of major depression, and assuming that the most efficient dosing regimen, and least expensive tablet choice, is prescribed. When available, generic prices are used. Prices were derived from 2006 Cardinal Health wholesale price guide provided by the Pharmacy Department of the New York State Institute.

Pharmacokinetics

The pharmacokinetics of tricyclic antidepressants (TCAs) are complex, as reflected in the diversity of half-lives (10 to 40 hours). TCAs are primarily absorbed in the small intestine, reaching peak plasma levels two to six hours after oral administration. Absorption can be affected by changes in gut motility. The drugs are extensively metabolized in the liver on first pass through the portal system. They are lipophilic, have a large volume of distribution, and are highly protein-bound (85-95%). TCAs are metabolized to active metabolites in the liver by hepatic microsomal enzymes and excreted by the kidneys. The rate of metabolism can vary genetically; 7 to 9% of the white population are slow hydroxylators. There is a large range of elimination half-lives.

Tricyclic antidepressants as a class have a relatively narrow therapeutic index; there is a significant risk of toxicity with blood levels of only two to six times the therapeutic level. A 1-week supply can be fatal in overdose, as blood concentrations of greater than 1000 ng/dl are correlated with prolongation of the QRS complex and arrhythmias. TCAs are commonly ingested agents by which patients successfully commit suicide by overdose.

Monoamine oxidase inhibitors (MAOIs) are also well absorbed from the gastrointestinal tract. Their short half-life of one to two hours is not particularly relevant as they bind irreversibly with MAO. Thus, the activity of these drugs depends less on pharmacokinetics and more on the synthesis of new MAO to restore normal enzyme activity. This synthesis requires approximately two weeks. This class of drugs is little used due to their potentially dangerous interactions with sympathomimetics and foods containing tyramine.

Each of the six SSRI agents (fluoxetine, paroxetine, sertraline, fluvoxamine, citalopram and its S-enantiomer escitalopram) selectively block the reuptake of serotonin presynaptically, though each drug differs structurally from the others. As a result, these agents differ in their pharmacokinetic profiles. Many selective serotonin reuptake inhibitors (SSRIs) are, like the tricyclic antidepressants (TCAs), highly protein bound though the proportions of citalopram and escitalopram that are protein-bound are 80 and 56% respectively. In contrast, however, they have varying half-lives ranging from approximately 24 hours to several days.

All SSRIs are well absorbed and not generally affected by food administration, though sertraline is an exception to this rule — its blood level may be increased by food. All have large volumes of distribution and are extensively protein-bound. They are metabolized by hepatic microsomal enzymes and are potent inhibitors of these enzymes. The only serotonin reuptake inhibitor with an active metabolite is fluoxetine, whose metabolite norfluoxetine has a half-life of 7 to 15 days. Thus, it may take several months to achieve steady state with this agent. This is considerably longer than citalopram (half-life 1.5 days) or sertraline and paroxetine (half-lives 24 hours).

There is no correlation between half-life and time to onset. Drugs with shorter half-lives have an advantage in cases where rapid elimination is desired (e.g., in the case of an allergic reaction). Drugs with a longer half-life may also have advantages: fluoxetine, for example, has been successfully given in a once-weekly dosing during the continuation phase of treatment and a once-weekly formulation of this drug is currently available. All serotonin reuptake inhibitors are eliminated in the urine as inactive metabolites. Both fluoxetine and paroxetine are capable of inhibiting their own clearance at clinically relevant doses. As such, they have nonlinear pharmacokinetics: changes in dose can produce proportionately large plasma levels.

Citalopram and its S-enantiomer escitalopram are the most recent selective serotonin reuptake inhibitors (SSRIs) to be introduced in the Untied States; The Food and Drug Administration (FDA) approved citalopram in 1998 and escitalopram in 2002. Of the available SSRIs, these are the most selective for serotonin receptor blockade; escitalopram is 100 times more potent than the R-enantiomer.

As with most other antidepressants, bupropion undergoes extensive first pass metabolism in the liver. The parent compound has a half-life of 10 to 12 hours, but has three active metabolites. One, threohydrobupropion, has a half-life of 35 hours and is relatively free in plasma (it is only 50% protein-bound). There is considerable individual variability in the levels of bupropion and its metabolites. Trazodone has a relatively short half-life of three to nine hours; as a result of this and its apparent lack of active metabolites plasma levels of trazodone can be quite variable, requiring divided dosing.

Venlafaxine has a short half-life (4 hours); however, it is available in an extended release formulation that allows once-daily dosing. It appears to have a dual effect, in which at lower doses it primarily acts on the serotonin transporter, and clinically significant norepinephrine reuptake inhibition is not seen until higher doses are used (150 mg/day and above).

Nefazodone has relatively low bioavailability, and a short half-life (2-8 hours), and thus it is usually given in twice-daily doses. Mirtazapine has a half-life of 13 to 34 hours.

Duloxetine is the second selective inhibitor of norepinephrine and serotonin to be introduced in the United States. Compared with venlafaxine it has relatively greater effect on 5-HT reuptake in vitro but the clinical significance of this difference is unclear. The half-life of duloxetine is 8-17 hours (mean 12 hours).

The selective serotonin reuptake inhibitors (SSRIs) are also, to varying degrees, potent inhibitors of the P450 hepatic enzyme system. The result is increased blood levels of concomitant agents that are also metabolized by this enzymatic system. Switching from an SSRI to another antide-pressant group illustrates the clinical relevance. For example, when a patient changes from fluoxetine to venlafaxine, the inhibition of the P450 2D6 isozyme will reduce venlafaxine metabolism for several weeks. The resulting increase in the venlafaxine blood level can hasten the development of side effects. The aware clinician can avoid this by starting with a lower than usual initial dose and titrating upward slowly.