Chapter 15: Drug Therapy of Depression and Anxiety Disorders

All drugs commonly used to treat depression share, at some level, primary effects on serotonergic or noradrenergic neurotransmitter systems (Shelton and Lester, 2006). In general, antidepressants enhance serotonergic or noradrenergic transmission, although the nature of this effect may change with chronic treatment (Shelton, 2000). The dependence of many current treatments on serotonergic or noradrenergic mechanisms is highlighted by clinical studies employing monoamine depletion strategies. For example, tryptophan depletion that acutely reduces 5-HT neurotransmission results in a relatively brisk relapse (within 5-10 hours) of the symptoms of depression in patients who had shown remission on an SSRI but not a NE reuptake inhibitor (Delgado et al., 1991). Conversely, catecholamine depletion by blocking the rate-limiting enzyme for the synthesis of NE and dopamine (DA) results in a relapse of symptoms in patients who had recently remitted on an NE reuptake inhibitor, but not an SSRI (Miller et al., 1996). Sites of interaction of antidepressant drugs with noradrenergic and serotonergic neurons are depicted in Figure 15–1.

Long-term effects of antidepressant drugs evoke adaptive or regulatory mechanisms that enhance the effectiveness of therapy. These responses include increased adrenergic or serotonergic receptor density or sensitivity, increased receptor-G protein coupling and cyclic nucleotide signaling, induction of neurotrophic factors, and increased neurogenesis in the hippocampus (Schmidt and Duman, 2007). Persistent antidepressant effects depend on the continued inhibition of 5-HT or NE transporters, or enhanced serotonergic or noradrenergic neurotransmission achieved by an alternative pharmacological mechanism. For example, chronic treatment with some antidepressants that interact directly with monoamine transporters (e.g., SSRIs, SNRIs, or NE reuptake inhibitors) reduces the expression and activity of 5-HT or NE transporters in the brain, which results in enhanced serotonergic or noradrenergic neurotransmission (Benmansour et al., 1999; Zhao et al., 2008). Compelling evidence suggests that sustained signaling via NE or 5-HT increases the expression of specific downstream gene products, particularly brain-derived neurotrophic factor (BDNF), which appears to be related to the ultimate mechanism of action of these drugs (Sen et al., 2008). Unfortunately, neither earlier theories of monoaminergic receptor down-regulation and altered signaling nor current theories of neurogenesis and modulation of neurotrophic factors has yet led to new antidepressant treatments. Glutamate, neurokinin, corticotropin releasing hormone receptors and cyclic nucleotide phosphodiesterases may be potential targets for the development of novel antidepressant drugs (O'Donnell and Zhang, 2004; Rakofsky et al., 2009; Witkin et al., 2007; Zarate et al., 2006).

Another important issue in the use of antidepressants is a phenomenon known as the "switch" from a depressed episode to a manic or hypomanic episode (Goldberg and Truman, 2003), a significant challenge in managing bipolar illness. For this reason, antidepressants are not recommended as monotherapy for bipolar illness. However, patients with bipolar illness may present with major depressive episodes early in the course of their illness. SSRIs and bupropion may be somewhat less likely to induce the switch from depression to mania than antidepressants from other pharmacological classes.

A controversial issue regarding the use of all antidepressants is their relationship to suicide (Mann et al., 2006). Data establishing a clear link between antidepressant treatment and suicide are lacking. In general, for reasons of safety, suicidal patients are not enrolled in clinical trials designed to seek approval for marketing a new medication. Thus, the FDA uses "suicidality" (i.e., suicidal ideation or suicide attempts and self-injurious behavior) as a proxy for risk of suicide. The FDA has issued a "black box" warning regarding the use of SSRIs and a number of other antidepressants in children and adolescents, particularly during the early phase of treatment, due to the possibility of an association between antidepressant treatment and suicide. However, there is strong epidemiological evidence that the rate of suicide has been decreasing since the time that SSRIs were first prescribed and then gained widespread usage; these data, however, cannot demonstrate a causal relationship. An analysis of health records of over 65,000 patients undergoing different types of pharmacological treatment for depression found no suggestion of increases in suicide or suicide attempts (Simon et al., 2006). Relevant to this point is the observation that an increase in suicide occurred for children and adolescents after regulatory authorities in the U.S. and Europe issued public health warnings about a possible association between antidepressants and suicidal ideation and acts, perhaps due to a reduction in the use of antidepressants in these patient populations after the announcement (Gibbons et al., 2007). Most clinicians agree that for seriously depressed patients, the risk of not being on an effective antidepressant drug outweighs the risk of being treated with one. Of course, clinical common sense also demands that special attention be paid to potentially suicidal patients regardless of their medication status. In addition, patients and family members should be warned to look for the exacerbation of symptoms such as insomnia, agitation, anxiety, or either the onset or worsening of suicidal thoughts and behaviors, particularly early in the course of therapy.

Monoamine Oxidase Inhibitors

The first class of drugs with relatively specific antidepressant effects were the MAOIs (Hollister, 1981). Iproniazid, which was developed initially for the treatment of tuberculosis, was found to have mood-elevating effects in patients with tuberculosis. Subsequently, iproniazid was shown to inhibit MAO, leading to the development of additional MAOIs including phenelzine, isocarboxazid, and tranylcypromine. As irreversible inhibitors of both MAO-A and MAO-B, these drugs have pronounced effects on the body's ability to metabolize endogenous monoamines (e.g., 5-HT, NE, and DA) and exogenous monoamines (e.g., tyramine). The protection of exogenous monoamines leads to significant drug and food interactions. More recently, reversible and selective inhibitors of MAO-A and MAO-B have been developed (Livingston and Livingston, 1996), and these have fewer side effects and fewer interactions with food and other drugs. The MAO-B selective inhibitor selegiline is used in the treatment of Parkinson disease; selective inhibitors of MAO-A, such as moclobemide, are effective antidepressants but are not approved for use in the U.S.

Tricyclic Antidepressant and Selective Reuptake Inhibitors

The initial development of the TCAs resulted from psychopharmacological characterization of a series of structural analogs that had been developed as potential antihistamines, sedatives, analgesics, and antiparkinson drugs (Hollister, 1981). One of the compounds, imipramine, which has a phenothiazine-like structure, modified behavior in animal models. Unlike the phenothiazines, imipramine had limited efficacy in schizophrenic patients, but improved symptoms of depression. Imipramine and related TCAs became the mainstay of drug treatment of depression until the later development of the SSRIs.

TCAs with a tertiary-amine side chain, including amitriptyline, doxepin, and imipramine, inhibit both norepinephrine and serotonin uptake, while clomipramine is somewhat selective for inhibition of serotonin uptake. Chemical modification of the TCA structure led to the earliest SSRI zimelidine which, while effective, was withdrawn from the market due to serious adverse effects. Fluoxetine and fluvoxamine were the first widely used SSRIs. At the same time, selective inhibitors of norepinephrine reuptake entered clinical development; while not approved for use in the U.S. for the treatment of depression, one norepinephrine reuptake inhibitor, atomoxetine, is used for the treatment of attention deficit hyperactivity disorder. Subsequent drug development efforts focused on serotonin and norepinephrine reuptake inhibitors (SNRIs), resulting in venlafaxine and duloxetine, which lack the complex receptor pharmacology exhibited by the TCAs.

SSRI treatment causes stimulation of 5-HT_1A and 5-HT₇ autoreceptors on cell bodies in the raphe nucleus and of 5-HT_1D autoreceptors on serotonergic terminals, and this reduces serotonin synthesis and release toward pre-drug levels. With repeated treatment with SSRIs, there is a gradual down-regulation and desensitization of these autoreceptor mechanisms. In addition, down-regulation of postsynaptic 5-HT_2A receptors may contribute to antidepressant efficacy directly or by influencing the function of noradrenergic and other neurons via serotonergic heteroreceptors. Other postsynaptic 5-HT receptors likely remain responsive to increased synaptic concentrations of 5-HT and contribute to the therapeutic effects of the SSRIs.

Later-developing effects of SSRI treatment also may be important in mediating ultimate therapeutic responses. These include sustained increases in cyclic AMP signaling and phosphorylation of the nuclear transcription factor CREB, as well as increases in the expression of trophic factors such as BDNF. In addition, SSRI treatment increases neurogenesis from progenitor cells in the dentate nucleus of the hippocampus and subventricular zone (Santarelli et al., 2003). In animals models, some behavioral effects of SSRIs depend on increased neurogenesis (probably via increased expression of BDNF and its receptor TrkB), suggesting a role for this mechanism in the antidepressant effects. Recent evidence indicates the presence of neural progenitor cells in the human hippocampus, providing some support for the relevance of this mechanism to the clinical situation (Manganas et al., 2007). Further, repeated treatment with SSRIs reduces the expression of SERT, resulting in reduced clearance of released 5-HT and increased serotonergic neurotransmission. These changes in transporter expression parallel behavioral changes observed in animal models, suggesting some role for this regulatory mechanism in the late-developing effects of SsRIs (Zhao et al., 2009). These persistent behavioral changes depend on increased serotonergic neurotransmission, similar to what has been demonstrated clinically using depletion strategies (Delgado et al., 1991).

Serotonin-Norepinephrine Reuptake Inhibitors

Many older TCAs block both SERT and NET, but at a high side effect burden. Four medications with a non-tricyclic structure that inhibit the reuptake of both 5-HT and norepinephrine have been approved for use in the U.S. for treatment of depression, anxiety disorders, and pain: venlafaxine and its demethylated metabolite, desvenlafaxine; duloxetine; and milnacipran (approved only for fibromyalgia pain in the U.S.). Off-label uses include stress urinary incontinence (duloxetine), autism, binge eating disorders, hot flashes, pain syndromes, premenstrual dysphoric disorders, and post-traumatic stress disorders (venlafaxine)

The rationale behind the development of these newer agents was that targeting both SERT and NET, analogous to the effects of some TCAs, might improve overall treatment response. Meta-analyses provide some support for this hypothesis (Entsuah et al., 2001). Specifically, the remission rate for venlafaxine appears slightly better than for SSRIs in head-to-head trials. However, many of the studies included in the meta-analyses used a daily venlafaxine dose of 150 mg, which would have modest effects on noradrenergic neurotransmission. Duloxetine, in addition to being approved for use in the treatment of depression and anxiety, also is used for treatment of fibromyalgia and neuropathic pain associated with peripheral neuropathy.

Mechanism of Action. SNRIs inhibit both SERT and NET (Table 15–2). Depending on the drug, the dose, and the potency at each site, SNRIs cause enhanced serotonergic and/or noradrenergic neurotransmission. Similar to the action of SSRIs, the initial inhibition of SERT induces activation of 5-HT_1A and 5-HT_1D autoreceptors. This action decreases serotonergic neurotransmission by a negative feedback mechanism until these serotonergic autoreceptors are desensitized. Then, the enhanced serotonin concentration in the synapse can interact with postsynaptic 5-HT receptors.

Serotonin Receptor Antagonists

Several antagonists of the 5-HT₂ family of receptors are effective antidepressants, although most agents of this class affect other receptor classes as well. The class includes two pairs of close structural analogues, trazodone and nefazodone, as well as mirtazapine (REMERON, others) and mianserin (not marketed in the U.S.).

The efficacy of trazodone may be somewhat more limited than the SSRIs; however, low doses of trazodone (50-100 mg) have been used widely both alone and concurrently with SSRIs or SNRIs to treat insomnia. When used to treat depression, trazodone typically is started at 150 mg/day in divided doses with 50 mg increments every 3-4 days. The maximally recommended dose is 400 mg/day for outpatients and 600 mg/day for inpatients.

Both mianserin and mirtazapine are quite sedating and are treatments of choice for some depressed patients with insomnia. The recommended initial dosing of mirtazapine is 15 mg/day with a maximal recommended dose of 45 mg/day. Since the t_1/2 is 16-30 hours, the recommended interval for dose changes is no less than 2 weeks.

Mechanism of Action. The most potent pharmacological effects of trazodone are blockade of the 5-HT₂ and α₁ adrenergic receptors. Trazodone also inhibits the serotonin transporter, but is markedly less potent for this action relative to its blockade of 5-HT_2A receptors. Similarly, the most potent pharmacological action of nefazodone also is the blockade of the 5-HT₂ family of receptors.

Both mirtazapine and mianserin potently block histamine H₁ receptors. They also have some affinity for α₂ adrenergic receptors, which is claimed to be related to their therapeutic efficacy, but this point is questionable. Their affinities for 5-HT_2A, 5-HT_2C, and 5-HT₃ receptors are high, though less so than for histamine H₁ receptors. Both of these drugs have been shown in double-blind, placebo-controlled studies to increase the antidepressant response when combined with SSRIs compared to the action of the SSRIs alone. The exact monoamine receptor accounting for the effects of mirtazapine and mianserin is not clear, although the data from clinical trials suggest that olanzapine, aripiprazole, and quetiapine enhance the therapeutic effects of SSRIs or SNRIs, an important hint that their unique capacity to block the 5-HT_2A receptor is the most potent shared pharmacological action amongst these antidepressant and antipsychotic drugs.

Bupropion

Bupropion (wellbutrin, others) is discussed separately because it appears to act via multiple mechanisms. It enhances both noradrenergic and dopaminergic neurotransmission via reuptake inhibition (Table 15–2); in addition, its mechanism of action may involve the presynaptic release of NE and DA (Foley et al., 2006). Bupropion is indicated for the treatment of depression, prevention of seasonal depressive disorder, and as a smoking cessation treatment (under the ZYBAN brand). Bupropion has effects on sleep EEG that are opposite those of most antidepressant drugs. Bupropion may improve symptoms of attention deficit hyperactivity disorder (ADHD) and has been used off-label for neuropathic pain and weight loss. Clinically, bupropion is widely used in combination with SSRIs to obtain a greater antidepressant response; however, there are very limited clinical data providing strong support for this practice.

Mechanism of Action. Bupropion appears to inhibit NET. It also blocks the DAT but its effects on this transporter are not particularly potent in animal studies. In addition, it has effects on VMAT2, the vesicular monoamine transporter (see Figure 8–6). The hydroxybupropion metabolite may contribute to the therapeutic effects of bupropion: this metabolite appears to have a similar pharmacology and is present in substantial levels.

Atypical Antipsychotics

In addition to their use in schizophrenia, bipolar depression, and major depression with psychotic disorders, atypical antipsychotics have gained further, off-label use for major depression without psychotic features (Jarema, 2007). In fact, both aripiprazole (abilify) added to SSRIs and SNRIs and a combination of olanzapine and the SSRI fluoxetine (symbyax) have been approved by the FDA for treatment-resistant major depression (i.e., following an inadequate response to at least two different antidepressants).

The initial recommended dose of aripiprazole is 2-5 mg/day with a recommended maximal dose of 15 mg/day following increments of no more than 5 mg/day every week. The olanzapine-fluoxetine combination is available in fixed-dose combinations of either 6 or 12 mg of olanzapine and 25 or 50 mg of fluoxetine. Quetiapine (seroquel) may have either primary antidepressant actions on its own or adjunctive benefit for treatment-resistant depression; it is used off-label for insomnia. Quetiapine also is currently under review by the FDA for additional indications in major depression and generalized anxiety disorder.

Mechanism of Action. The mechanism of action and adverse effects of the atypical antipsychotics are described in detail in Chapter 16. The side effects seen with atypical antipsychotics in patients with schizophrenia may not exactly translate to patients with major depression, due to a predisposition toward metabolic syndrome in patients with schizophrenia. The major risks of these agents are weight gain and metabolic syndrome, a greater problem for quetiapine and olanzapine than for aripiprazole.

Tricyclic Antidepressants

Due to their potential to cause serious side effects, the TCAs generally are not used as first-line drugs for the treatment of depression. Nevertheless, these drugs have established value for the treatment of major depression (Hollister, 1981). TCAs and first-generation antipsychotics are synergistic for the treatment of psychotic depression. Tertiary amine TCAs (e.g., doxepin, amitriptyline) have been used for many years in relatively low doses for treating insomnia. In addition, because of the role of norepinephrine and serotonin in pain transmission, these drugs are commonly used to treat a variety of pain conditions.

Mechanism of Action. The inspiration for the development of both SSRIs and SNRIs derived from the appreciation that a salient pharmacological action of TCAs is antagonism of serotonin and norepinephrine transporters (Table 15–2). The antidepressant action of TCAs was discovered during clinical trials in schizophrenic patients; imipramine had little effect on psychotic symptoms but had beneficial effects on depressive symptoms (Hollister, 1981). In addition to inhibiting NET somewhat selectively (desipramine, nortriptyline, protriptyline, amoxapine) or both SERT and NET (imipramine, amitriptyline), these drugs also block other receptors (H₁, 5-HT₂, α₁, and muscarinic). Given the superior activity of clomipramine over SSRIs, some combination of these additional pharmacological actions may contribute to the therapeutic effects of TCAs. One TCA, amoxapine, also is a dopaminergic receptor antagonist; its use, unlike that of other TCAs, poses some risk for the development of extrapyramidal side effects such as tardive dyskinesia.

Monoamine Oxidase Inhibitors

MAOIs have efficacy equivalent to that of the TCAs but are rarely used because of their toxicity and major drug and food interactions (Hollister, 1981). The MAOIs approved for treatment of depression include tranylcypromine (parnate, others), phenelzine (nardil), and isocarboxazid (marplan). Selegiline (emsam) is available as a transdermal patch and approved for use in the treatment of depression; transdermal delivery may reduce the risk for diet-associated hypertensive reactions (described later).

Mechanism of Action. The MAOIs nonselectively and irreversibly inhibit both MAO-A and MAO-B, which are located in the mitochondria and metabolize (inactivate) monoamines, including 5-HT and NE (Chapter 8). Selegiline inhibits MAO-B at lower doses, with effects on MAO-A at higher doses. Selegiline also is a reversible inhibitor of monoamine oxidase, which may reduce the potential for serious adverse drug and food interactions. While both MAO-A and MAO-B are involved in metabolizing 5-HT, only MAO-B is found in serotonergic neurons (Chapter 13).

The metabolism of most antidepressants is mediated by hepatic CYPs (see Table 15–3). Some antidepressants inhibit the clearance of other drugs by the CYP system, as discussed in the following section, and this possibility of drug interactions should be a significant factor in considering the choice of agents.

Selective Serotonin Reuptake Inhibitors. All of the SSRIs are orally active and possess elimination half-lives consistent with once-daily dosing (Hiemke and Hartter, 2000). In the case of fluoxetine, the combined action of the parent and the desmethyl metabolite norfluoxetine allows for a once weekly formulation (prozac weekly). CYP2D6 is involved in the metabolism of most SSRIs and the SSRIs are at least moderately potent inhibitors of this isoenzyme (Table 15–3). This creates a significant potential for drug interaction for post-menopausal women taking the breast cancer drug and estrogen antagonist, tamoxifen (Chapter 62); the parent molecule is converted to a more active metabolite by CYP2D6, and SSRIs may inhibit this activation and diminish the therapeutic activity of tamoxifen. Since venlafaxine and desvenlafaxine are weak inhibitors of CYP2D6, these antidepressants are not contraindicated in this clinical situation. However, care should be used in combining SSRIs with drugs that are metabolized by CYPs 1A2, 2D6, 2C9, and 3A4 (e.g., warfarin, tricyclic antidepressants, paclitaxel).

Serotonin-Norepinephrine Reuptake Inhibitors. Both immediate release and extended-release (tablet or capsule) preparations of venlafaxine (effexor xr, others) produces steady state-levels of drug in plasma within 3 days. The elimination half-lives for the parent venlafaxine and its active and major metabolite desmethyl venlafaxine are 5 and 11 hours, respectively. Desmethylvenlafaxine is eliminated by hepatic metabolism and by renal excretion. Venlafaxine dose reductions are suggested for patients with renal or hepatic impairment. Duloxetine has a t_1/2 of 12 hours. Duloxetine is not recommended for those with end-stage renal disease or hepatic insufficiency.

Serotonin Receptor Antagonists. Mirtazapine has an elimination t_1/2 of 16-30 hours. Thus, dose changes are suggested no more often than 1-2 weeks. Clearance of mirtazapine is decreased in the elderly and in patients with moderate to severe renal or hepatic impairment. Pharmacokinetics and adverse effects of mirtazapine may have an enantiomer-selective component (Brockmöller et al., 2007). Steady-state trazodone is observed within 3 days following either a twice or three times daily dosing regimen. Nefazodone has a t_1/2 of only 2-4 hours; its major metabolite hydroxynefazodone has a t_1/2 of 1.5-4 hours.

Bupropion. The terminal phase of bupropion elimination has a t_1/2 of 21 hours. The elimination of bupropion involves both hepatic and renal routes. Patients with severe hepatic cirrhosis should receive a maximum dose of 150 mg every other day while consideration for a decreased dose should also be made in cases of renal impairment.

Tricyclic Antidepressants. The TCAs, or their active metabolites, have plasma exposure half-lives of 8-80 hours; this makes once-daily dosing possible for most of the compounds (Rudorfer and Potter, 1999). Steady-state concentrations occur within several days to several weeks of beginning treatment. TCAs are largely eliminated by hepatic CYPs (see Table 15–3). Determination of plasma levels may be useful in identifying patients who appear to have toxic effects and may have excessively high levels of the drug, or those in whom lack of absorption or noncompliance is suspected. Dosage adjustments of TCAs are typically made according to patient's clinical response, not based on plasma levels. Nonetheless, monitoring the plasma exposure has an important relationship to treatment response: there is a relatively narrow therapeutic window, as discussed below.

About 7% of patients metabolize TCAs slowly due to a variant CYP2D6 isoenzyme, causing a 30-fold difference in plasma concentrations among different patients given the same TCA dose. To avoid toxicity in "slow metabolizers," plasma levels should be monitored and doses adjusted downward.

Monoamine Oxidase Inhibitors. MAOIs are metabolized by acetylation, although the end products are incompletely characterized. A significant portion of the population (50% of the Caucasian population and an even higher percentage among Asians) are "slow acetylators" and will exhibit elevated plasma levels. The nonselective MAOIs used in the treatment of depression are irreversible (sometimes called "suicide") inhibitors; thus, it takes up to 2 weeks for MAO activity to recover, even though the parent drug is excreted within 24 hours (Livingston and Livingston, 1996). Recovery of normal enzyme function is dependent on synthesis and transport of new MAO to monoaminergic nerve terminals. Despite this irreversible enzyme inhibition, MAOIs require daily dosing.

Selective Serotonin Reuptake Inhibitors. The SSRIs, unlike the TCAs, do not cause major cardiovascular side effects. The SSRIs are generally free of antimuscarinic side effects (dry mouth, urinary retention, confusion), do not block histamine or α adrenergic receptors, and are not sedating (Table 15–4). The favorable side effect profile of the SSRIs may lead to better patient compliance compared to that for the TCAs.

SSRIs are not free from side effects, however. Excessive stimulation of brain 5-HT₂ receptors may result in insomnia, increased anxiety, irritability, and decreased libido, effectively worsening prominent depressive symptoms. Excess activity at spinal 5-HT₂ receptors causes sexual side effects including erectile dysfunction, anorgasmia, and ejaculatory delay; these effects may be more prominent with paroxetine (Vaswani et al., 2003). Stimulation of 5-HT₃ receptors in the CNS and periphery contributes to GI effects, which are usually limited to nausea but may include diarrhea and emesis. Some patients experience an increase in anxiety, especially with the initial dosing of SSRIs. With continued treatment, some patients also report a dullness of intellectual abilities and concentration. In addition a phenomenon of a residual "flat affect" can occur in an otherwise successful treatment with SSRIs. A number of these side effects may be difficult to distinguish from symptoms of depression.

In general, there is not a strong relationship between SSRI serum concentrations and therapeutic efficacy. This finding is not surprising given that most antidepressant studies have been conducted at doses/plasma exposures which saturate the brain SERT, consistent with the flat dose or exposure/efficacy relationships observed. Thus, dosage adjustments are based more on evaluation of clinical response and management of side effects than on measurements of plasma drug concentrations.

Sudden withdrawal of antidepressants can precipitate a withdrawal syndrome. For SSRIs or SNRIs, the symptoms of withdrawal may include dizziness, headache, nervousness, nausea, and insomnia. This withdrawal syndrome appears most intense for paroxetine and venlafaxine compared to other antidepressants due to their relatively short half-lives and, in the case of paroxetine, lack of active metabolites. On the other hand, the active metabolite of fluoxetine, norfluoxetine, has such a long t_1/2 (1-2 weeks) that few patients experience any withdrawal symptoms when discontinuing fluoxetine.

Unlike the other SSRIs, paroxetine is associated with an increased risk of congenital cardiac malformations. Epidemiological data suggest that paroxetine might increase the risk of congenital cardiac malformations when administered in the first trimester of pregnancy. Venlafaxine also is associated with an increased risk of perinatal complications. Therefore, these drugs should not be used in pregnant women; careful consideration should be given to using these mediations in women of reproductive potential, and these subjects should be advised to avoid pregnancy while they are taking the drugs.

Serotonin-Norepinephrine Reuptake Inhibitors. The SNRIs have desirable safety advantages over the TCAs (Table 15–4). SNRIs have a side effect profile similar to that of the SSRIs, including nausea, constipation, insomnia, headaches, and sexual dysfunction. The immediate release formulation of venlafaxine can induce sustained diastolic hypertension (systolic blood pressure > 90 mm Hg at consecutive weekly visits) in 10-15% of patients at higher doses; this risk is reduced with the extended-release form. This effect of venlafaxine may not be associated simply with inhibition of NET, since duloxetine does not share this side effect.

Serotonin Receptor Antagonists. The main side effects of mirtazapine, seen in >10% of the patients in clinical trials, are somnolence, increased appetite, and weight gain. A rare side effect of mirtazapine is agranulocytosis. In the 2 patients having this side effect during the pre-marketing phase in nearly 2800 patients, bone marrow function recovered when mirtazapine treatment was discontinued. Trazodone use is associated with priapism in rare instances; this should be considered a medical emergency and surgical intervention may be required. Nefazodone was voluntarily withdrawn from the market in Europe and the U.S. after rare cases of liver failure were associated with its use. Generic nefazodone is still available in the U.S.

Bupropion. At doses higher than that recommended for depression (450 mg/day), the risk of seizures increases significantly. The use of extended release formulations often blunts the maximum concentration observed after dosing and minimizes the chance of reaching drug levels associated with an increased risk of seizures.

Tricyclic Antidepressants. TCAs are potent antagonists at histamine H₁ receptors; H₁ receptor antagonism contributes to the sedative effects of TCAs (Table 15–4). Antagonism of muscarinic acetylcholine receptors contributes to cognitive dulling as well as a range of adverse effects mediated by the parasympathetic nervous system (blurred vision, dry mouth, tachycardia, constipation, difficulty urinating). Some tolerance does occur for these anticholinergic effects, which are mitigated by titration strategies to reach therapeutic doses over a reasonable period of time. Antagonism of α₁ adrenergic receptors contributes to orthostatic hypotension and sedation. Weight gain is another side effect of this class of antidepressants.

TCAs also have quinidine-like effects on cardiac conduction that can be life threatening with overdose and limit the use of TCAs in patients with coronary heart disease. This is the primary reason that no more than a one-week supply should be provided to a new patient; even during maintenance treatment, only a very limited supply should be available to the patient at any given time. Like other antidepressant drugs, TCAs also lower the seizure threshold.

Monoamine Oxidase Inhibitors. Hypertensive crisis resulting from food or drug interactions is one of the life-threatening toxicities associated with use of the MAOIs. Foods containing tyramine are a contributing factor. MAO-A within the intestinal wall and MAO-A and MAO-B in the liver normally degrade dietary tyramine. However, when MAO-A is inhibited, the ingestion of certain aged cheeses, red wines, sauerkraut, fava beans, and a variety of other tyramine-containing foods leads to accumulation of tyramine in adrenergic nerve endings and neurotransmitter vesicles and induces norepinephrine and epinephrine release. The released catecholamines stimulate postsynaptic receptors in the periphery, increasing blood pressure to dangerous levels. These episodes can be reversed by antihypertensive medications. Even when the patient is highly vigilant, dietary indiscretions or use of prescription or over-the-counter medications that contain sympathomimetic compounds may occur, resulting in a potentially life-threatening elevation of blood pressure. In comparison to tranylcypromine and isocarboxazid, the selegiline transdermal patch is better tolerated and safer. Another serious, life-threatening issue with chronic administration of MAOIs is hepatotoxicity.

MAO-A inhibitors are efficacious in treating depression. However, MAO-B inhibitors such as selegiline (with oral formulations) are effective in treating depression only when given at doses that block both MAO-A and MAO-B. Thus, these data emphasize the importance of enhancing the synaptic availability of 5-HT and NE as important mediating events for many antidepressant drugs. While not available in the U.S., reversible inhibitors of MAO-A (RIMAs, such as moclobemide) have been developed. Since these drugs are selective for MAO-A, significant MAO-B activity remains. Further, since the inhibition of MAO-A by RIMAs is reversible and competitive, as concentrations of tyramine rise, enzyme inhibition is surmounted. Thus, RIMAs produce antidepressant effects with reduced risk of tyramine-induced hypertensive crisis.

Selective Serotonin Reuptake Inhibitors. Most antidepressants, including the SSRIs, exhibit drug-drug interactions based on their routes of metabolism CYPs. Paroxetine and, to a lesser degree, fluoxetine are potent inhibitors of CYP2D6 (Hiemke and Hartter, 2000). The other SSRIs, outside of fluvoxamine, are at least moderate inhibitors of CYP2D6. This inhibition can result in disproportionate increases in plasma concentrations of drugs metabolized by CYP2D6 when doses of these drugs are increased. Fluvoxamine directly inhibits CYP1A2 and CYP2C19; fluoxetine and fluvoxamine also inhibit CYP3A4. A prominent interaction is the increase in TCA exposure that may be observed during co-administration of TCAs and SSRIs.

Another important drug-drug interaction with SSRIs occurs via a pharmacodynamic mechanism. MAOIs enhance the effects of SSRIs due to inhibition of serotonin metabolism. Administration of these drugs together can produce synergistic increases in extracellular brain serotonin, leading to the serotonin syndrome. Symptoms of the serotonin syndrome include hyperthermia, muscle rigidity, myoclonus, tremors, autonomic instability, confusion, irritability, and agitation; this can progress toward coma and death. Other drugs that may induce the serotonin syndrome include substituted amphetamines such as methylenedioxymethamphetamine (Ecstasy), which directly releases serotonin from nerve terminals. The primary treatment is stopping all serotonergic drugs, administering nonselective serotonin antagonists, and supportive measures.

Since currently available MAOIs bind irreversibly to MAO and block the enzymatic metabolism of monoaminergic neurotransmitters, SSRIs should not be started until at least 14 days following discontinuation of treatment with an MAOI; this allows for synthesis of new MAO. For all SSRIs but fluoxetine, at least 14 days should pass prior to beginning treatment with an MAOI following the end of treatment with an SSRI. Since the active metabolite norfluoxetine has a t_1/2 of 1-2 weeks, at least 5 weeks should pass between stopping fluoxetine and beginning an MAOI.

Serotonin-Norepinephrine Reuptake Inhibitors. While 14 days are suggested to elapse from ending MAOI therapy and starting venlafaxine treatment, an interval of only 7 days after stopping venlafaxine is considered safe before beginning an MAOI. Duloxetine has a similar interval to initiation following MAOI therapy, but requires only a 5-day waiting period to begin MAOI treatment after ending duloxetine. Failure to observe these required waiting periods can result in the serotonin syndrome, as noted above for SSRIs.

Serotonin Receptor Antagonists. Trazodone dosing may need to be lowered when given together with drugs that inhibit CYP3A4. Mirtazapine is metabolized by CYPs 2D6, 1A2, and 3A4, but does not potently inhibit any of these isoenzymes. Trazodone and nefazodone are weak inhibitors of serotonin uptake and should not be administered with MAOIs due to concerns about the serotonin syndrome. However, it is unclear whether these drugs appreciably block brain SERT at doses used in the treatment of depression.

Bupropion. The major route of metabolism for bupropion is CYP2B6. While there does not appear to be any evidence for metabolism by CYP2D6 and this drug is frequently administered with SSRIs, the potential for interactions with drugs metabolized by CYP2D6 should be kept in mind until the safety of the combination is firmly established.

Tricyclic Antidepressants. Drugs that inhibit CYP2D6, such as SSRIs, may increase plasma exposures of TCAs. Other drugs that may act similarly are phenothiazine antipsychotic agents, type 1C anti-arrhythmic drugs, and other drugs with antimuscarinic, antihistaminic, and α adrenergic antagonistic effects. TCAs can potentiate the actions of sympathomimetic amines and should not be used concurrently with MAOIs or within 14 days of stopping MAOIs.

Monoamine Oxidase Inhibitors. A large number of drug-drug interactions lead to contraindications for simultaneous use with MAOIs. CNS depressants including meperidine and other narcotics, alcohol, and anesthetic agents should not be used with MAOIs. Meperidine and other opioid agonists in combination with MAOIs also induce the serotonin syndrome. As discussed earlier, SSRIs and SNRIs are contraindicated in patients on MAOIs, and vice versa, to avoid the serotonin syndrome. In general, other antidepressants such as TCAs and bupropion also should be avoided in patients taking an MAOI.

Anxiolytic Drugs

A variety of agents and drug classes provide anxiolytic effects. The primary treatments for anxiety-related disorders include the SSRIs, SNRIs, benzodiazepines, the azipirone buspirone, and beta adrenergic antagonists (Atack, 2003). Historically, TCAs, particularly clomipramine, and MAOIs have been used for the treatment of some anxiety-related disorders, but their use has been supplanted by drugs with lower toxicity. Specific issues relating to mechanism of action, adverse effects, pharmacokinetics, and drug interactions are discussed earlier and in Chapters 16 and 17.

The SSRIs and the SNRI venlafaxine (discussed earlier) are well tolerated with a reasonable side effect profile; in addition to their documented antidepressant activity, they also have have anxiolytic activity with chronic treatment. The benzodiazepines are effective anxiolytics as both acute and chronic treatment. There is concern regarding their use because of their potential for dependence and abuse as well as negative effects on cognition and memory. Buspirone, like the SSRIs, is effective following chronic treatment. In acts, at least in part, via the serotonergic system, where it is a partial agonist at 5-HT_1A receptors. Buspirone also has antagonistic effects at dopamine D₂ receptors, but the relationship between this effect and its clinical actions is uncertain. β adrenergic antagonists, particularly those with higher lipophilicity (e.g., propranolol and nadolol) are occasionally used for performance anxiety such as fear of public speaking; their use is limited due to significant side effects such as hypotension.

The antihistamine hydroxyzine and various sedative-hypnotic agents have been used as anxiolytics, but are generally not recommended because of their side effect profiles. Hydroxyzine, which produces short-term sedation, has been used in patients that cannot use other types of anxiolytics (e.g., those with a history of drug or alcohol abuse where benzodiazepines would be avoided). Chloral hydrate has been used for situational anxiety, but there is a narrow dose range where anxiolytic effects are observed in the absence of significant sedation, and, therefore, the use of chloral hydrate is not recommended.

Clinical Considerations with Anxiolytic Drugs. The choice of pharmacological treatment for anxiety is dictated by the specific anxiety-related disorders and the clinical need for acute anxiolytic effects (Millan, 2003). Among the commonly used anxiolytics, only the benzodiazepines and β adrenergic antagonists are effective acutely; the use of β adrenergic antagonists is generally limited to treatment of situational anxiety. Chronic treatment with SSRIs, SNRIs, and buspirone is required to produce and sustain anxiolytic effects. When an immediate anxiolytic effects is desired, benzodiazepines are typically selected.

Benzodiazepines, such as alprazolam, chlordiazepoxide, clonazepam, clorazepate, diazepam, lorazepam, and oxazepam, are effective in the treatment of generalized anxiety disorder, panic disorder, and situational anxiety. In addition to their anxiolytic effects, benzodiazepines produce sedative, hypnotic, anesthetic, anticonvulsant, and muscle relaxant effects. The benzodiazepines also impair cognitive performance and memory, adversely affect motor control, and potentiate the effects of other sedatives including alcohol. The anxiolytic effects of this class of drugs are mediated by allosteric interactions with the pentameric benzodiazepine-GABA_A receptor complex, in particular those GABA_A receptors comprised of α2, α3, and α5 subunits (Chapters 14 and 17). The primary effect of the anxiolytic benzodiazepines is to enhance the inhibitory effects of the neurotransmitter GABA.

One area of concern regarding the use of benzodiazepines in the treatment of anxiety is the potential for habituation, dependence, and abuse. Patients with certain personality disorders or a history of drug or alcohol abuse are particularly susceptible. However, the risk of dependence must be balanced with the need for treatment, since benzodiazepines are effective in both short- and long-term treatment of patients with sustained or recurring bouts of anxiety. Further, premature discontinuation of benzodiazepines, in the absence of other pharmacological treatment, results in a high rate of relapse. Withdrawal of benzodiazepines after chronic treatment, particularly those with short durations of action, can include increased anxiety and seizures. For this reason, it is important that discontinuation be carried out in a gradual manner.

Benzodiazepines cause many adverse effects, including sedation, mild memory impairments, decreased alertness, and slowed reaction time (which may lead to accidents). Memory problems can include visual-spatial deficits, but will manifest clinically in a variety of ways, including difficulty in word-finding. Occasionally, paradoxical reactions can occur with benzodiazepines such as increases in anxiety, sometimes reaching panic attack proportions. Other pathological reactions can include irritability, aggression, or behavioral disinhibition. Amnesic reactions (i.e., loss of memory for particular periods) can also occur. Benzodiazepines should not be used in pregnant women; there have been rare reports of craniofacial defects. In addition, benzodiazepines taken prior to delivery may result in sedated, under-responsive newborns and prolonged withdrawal reactions. In the elderly, benzodiazepines increase the risk for falls and must be used cautiously. These drugs are safer than classical sedative-hypnotics in overdosage and typically are fatal only if combined with other CNS depressants.

Benzodiazepines have at least some abuse potential, although their capacity for abuse is considerably below that of other classical sedative-hypnotic agents. When these agents are abused, it is generally in a multi-drug abuse pattern. In fact, the primary reason for misuse of these agents often is failed attempts to control anxiety. Tolerance to the anxiolytic effects develops with chronic administration, with the result that some patients escalate the dose of benzodiazepines over time. Ideally, benzodiazepines should be used for short periods of time and in conjunction with other medications (e.g., SSRIs) or evidence-based psychotherapies (e.g., cognitive behavioral therapy for anxiety disorders).

SSRIs and the SNRI venlafaxine are first-line treatments for most types of anxiety disorders, except when an acute drug effect is desired; fluvoxamine is approved only for obsessive-compulsive disorder. As for their antidepressant actions, the anxiolytic effects of these drugs become manifest following chronic treatment. Other drugs with actions on serotonergic neurotransmission, including trazodone, nefazodone, and mirtazapine, also are used in the treatment of anxiety disorders. Details regarding the pharmacology of these classes of were presented earlier.

Both SSRIs and SNRIs are beneficial in specific anxiety conditions such as generalized anxiety disorder, social phobias, obsessive-compulsive disorder, and panic disorder. These effects appear to be related to the capacity of serotonin to regulate the activity of brain structures such as amygdala and locus coeruleus that are thought to be involved in the genesis of anxiety. Interestingly, the SSRIs and SNRIs often will produce some increases in anxiety in the short-term that dissipate with time. Therefore, the maxim "start low and go slow" is indicated with anxious patients; however, many patients with anxiety disorders ultimately will require doses that are about the same as those required for the treatment of depression. Anxious patients appear to be particularly prone to severe discontinuation reactions with certain medications such as venlafaxine and paroxetine; therefore, slow off-tapering is required.

Buspirone is used in the treatment of generalized anxiety disorder (Goodman, 2004). Like the SSRIs, buspirone requires chronic treatment for effectiveness. Also, like the SSRIs, buspirone lacks many of the other pharmacological effects of the benzodiazepines: it is not an anticonvulsant, muscle relaxant, or sedative, and it does not impair psychomotor performance or result in dependence. Buspirone is primarily effective in the treatment of generalized anxiety disorder, but not for other anxiety disorders. In fact, patients with panic disorder often note an increase in anxiety acutely following initiation of buspirone treatment; this may be the result of the fact that buspirone causes increased firing rates of the locus coeruleus, which is thought to underlie part of the pathophysiology of panic disorder.

Mood and anxiety disorders are the most common psychiatric illnesses and are encountered frequently by clinicians in all disciplines. Depression represents a spectrum of conditions with a range of severity and a high frequency of comorbidities. Depressive conditions range from mild, self-limiting conditions to extremely severe illnesses that can include a high suicidal potential, psychosis, and serious functional impairment. Whereas the likelihood of receiving treatment for depression or anxiety has improved, problems remain with regard to duration, dosing, and adherence to treatment. Unfortunately, persons with depressive or anxiety disorders continue to suffer considerable delays in diagnosis and adequate treatment. Prominently used agents inhibit transmitter reuptake via SERT and NET.

Current antidepressant and anxiolytic treatment strategies are imperfect. Many patients have significant residual symptoms after pharmacotherapy. Fortunately, antidepressant/anxiolytic drug development is expanding beyond standard 5-HT and NE uptake inhibitors. New antidepressant medications include triple (5-HT, NE, and DA) reuptake inhibitors, agents combining inhibition of 5-HT reuptake and modulation of 5-HT receptors, antagonists of CNS peptides (e.g., corticotropin-releasing hormone-1 receptor and neurokinin-1), modulators of cyclic nucleotide signaling, σ₁ and σ₂ receptor ligands, melatonin receptors (particularly, combined 5-HT receptor antagonists/melatonin receptor agonists), and glutamate receptor antagonists (Zarate, Jr. et al., 2006).

Drugs that target NMDA, AMPA, and metabotropic glutamate receptors are very promising. Other targets, such as specific 5-HT receptor agonists, partial agonists, or antagonists, and GABA_A receptor agonists may prove useful in the future for anxiety. Modulating the endocannabinoid signaling system is also promising with regard to both antidepressant and anxiolytic actions. Finally, there is increasing evidence that some naturally occurring products such as S-adenosylmethionine, l-methylfolate, N-acetyl cysteine, and omega-3 fatty acids may be useful in depression and anxiety.