Preskorn.com Printed from: http://www.preskorn.com/books/omd_s7.html
Outpatient Management of Depression
7 - Considerations When Selecting an Antidepressant


As a basis for conceptualizing the pharmacology of the 22 different antidepressant options available in the United States, Chapter 6 presented a classification system which grouped all available antidepressants into eight mechanistically defined classes. This chapter will translate that basic pharmacology into five clinically relevant factors to be considered when selecting any drug for any patient:

These five factors are summarized by the mnemonic, STEPS. Each factor can be further subdivided as shown in Table 7.1. The discussion in this chapter will focus on each factor and present a summary of how each antidepressant group relates to that factor. A summary of the clinical pharmacology of each class of antidepressants as they relate to these five factors will be discussed in Chapter 8. A summary of how the various classes (and individual members within these classes) compare on these five factors is provided in Tables 7.2 through 7.9.

Safety

 

TABLE 7.1 — STEPS Criteria for Selecting an Antidepressant

Safety

  • Therapeutic index
  • Drug-drug interactions

    – Pharmacodynamics
    –Pharmacokinetics

Tolerability

Efficacy

  • Overall
  • Unique spectrum of activity
  • Rate of response
  • Maintenance and prophylaxis

Payment

Simplicity

  • Ease of administration

 
Adapted from: Preskorn SH. J Clin Psychiatry. 1994;55(suppl A):6-22, 23-24, 98-100.
TABLE 7.2 — STEPS Criteria for Mixed Reuptake and Neuroreceptor Antagonists (eg, Amitriptyline) Safety

Safety
• Therapeutic index: narrow. Serious toxicity can result from overdose, either acute ingestion or from gradual accumulation due to slow clearance
• Drug-drug interactions:
– Pharmacodynamic: prone to cause multiple types of such interactions due to their multiple mechanisms of action (Table 6.3)
– Pharmacokinetic: can be the target of such interactions but are not prone to cause them. Due to narrow therapeutic index, care should be taken when prescribing TCAs to patients on drugs known to inhibit CYP enzymes (Table 6.10)
Tolerability
• Poor due to numerous types of adverse effects mediated by the blockade of histamine receptors (eg, sedation), muscarinic acetylcholine receptors (eg, constipation), and alpha-1–adrenergic receptors (eg, orthostatic hypotension) (Tables 6.2, 6.3, and 6.7)
Efficacy
• Overall: highest response rates of any antidepressants. Clomipramine in particular has been found to be superior to two SSRIs, citalopram and paroxetine, in hospitalized patients with clinical depression
• Unique spectrum of activity: imipramine has been found effective in patients who have not benefited from treatment with the SSRI, sertraline
• Rate of response: 2 to 4 weeks
• Maintenance of response: established by controlled studies Payment
• Generic versions are available making these antidepressants the least expensive in terms of acquisition costs. These savings are likely offset by expenses arising from problems with their safety and tolerability (Table 7.10)
Simplicity
• Must titrate dose due to tolerability problems. Therapeutic drug monitoring should be done once early in treatment to guide dose adjustment. Therapeutic ranges have been established for most of these TCAs. Can be given once a day
Adapted from: Preskorn SH. J Clin Psychiatry. 1994;55(suppl A):6-22, 23-24, 98-100.
TABLE 7.3 — STEPS Criteria for Norepinephrine Selective Reuptake Inhibitors (eg, Desipramine)

Safety
• Therapeutic index: narrow. The only NSRIs available at present in the United States are secondary amine TCAs which have the same toxicity problems as tertiary amine TCAs (eg, amitriptyline). However, this safety problem is due to inhibition of fast sodium channels as opposed to norepinephrine reuptake inhibition. In fact, there are investigational, nontricyclic NSRIs (eg, reboxetine) which do not have a narrow therapeutic index in terms of cardiotoxicity
• Drug-drug interactions:
– Pharmacodynamic: limited at therapeutic concentrations to those produced by norepinephrine potentiation
– Pharmacokinetic: can be target of such interactions but not prone to cause them. Due to narrow therapeutic index, care should be taken when prescribing with drugs known to inhibit CYP enzymes (Table 6.10)
Tolerability
• Generally good. Adverse effects mediated by norepinephrine potentiation (Tables 6.2, 6.3, and 6.6)
Efficacy
• Overall: comparable to tertiary amine TCAs
• Unique spectrum of activity: have been found in several studies to be effective in approximately 50% of patients who have not benefited from treatment with a variety of SSRIs • Rate of response: 2 to 4 weeks
• Maintenance of response: established by controlled studies Payment
• Generic versions are available but cost almost as much as newer antidepressants (Table 7.10)
Simplicity
• Can start on an effective dose immediately. Therapeutic drug monitoring should be done once early in treatment to guide dose adjustment. Therapeutic ranges have been established for most of these TCAs. Can be given once a day
Abbreviations: NSRI, norepinephrine selective reuptake inhibitor; TCA, tricyclic antidepressant; CYP, cytochrome P450 enzyme; SSRI, serotonin selective reuptake inhibitor.
TABLE 7.4 — STEPS Criteria for Serotonin Selective Reuptake Inhibitors (eg, Sertraline)

Safety
• Therapeutic index: wide. No serious systemic toxicity demonstrated, even after substantial overdose
• Drug-drug interactions:
– Pharmacodynamic: serotonin syndrome may occur when used with other serotonin agonists. Can potentiate dopamine antagonists in terms of extrapyramidal effects (eg, motor restlessness)
– Pharmacokinetic: considerable differences exist among SSRIs in terms of their potential for decreasing the rate of oxidative metabolism of a variety of drugs by inhibiting CYP enzymes: fluoxetine, fluvoxamine > paroxetine > citalopram, sertraline (Table 6.10). No known clinically significant effect of other drugs on the clearance of SSRIs
Tolerability
• Good. Nausea and loose stools can occur early and are dose dependent but tolerance typically develops. Sexual dysfunction (eg, anorgasmia) can occur in approximately 30% of patients (Tables 6.2, 6.3, and 6.6) Efficacy
• Overall: equivalent to TCAs in outpatients
• Unique spectrum of activity: can work in TCAs nonresponders • Rate of response: 2 to 4 weeks
• Maintenance of response: evidence from controlled studies Payment
• Brand name only available; see discussion (Table 7.10)
Simplicity
• Ease of administration: fluoxetine, paroxetine, and sertraline can be started at an effective dose immediately. Nonlinear pharmacokinetics of paroxetine likely contribute to an increased risk of experiencing an SRI discontinuation syndrome after abrupt cessation. The long half-life of fluoxetine and active metabolite, norfluoxetine, means time to maximum effect and time to washout can take up to 2 months
Abbreviations: SSRI, serotonin selective reuptake inhibitor; CYP, cytochrome P450 enzyme; TCA, tricyclic antidepressant; SRI, serotonin reuptake inhibitor.
TABLE 7.5 — STEPS Criteria for Serotonin and Norepinephrine Reuptake Inhibitors (eg, Venlafaxine)

Safety
• Therapeutic index: wide. No serious systemic toxicity demonstrated. Dose-dependent hypertension going from 1% above placebo at doses £ 100 mg/day to 11% above placebo at doses ³ 300 mg/day. That fact is consistent with NE uptake inhibition occurring at higher doses
• Drug-drug interaction:
– Pharmacodynamic: same as SSRIs at low doses. NE-mediated interactions possible at higher doses
– Pharmacokinetic: has profile comparable to citalopram and sertraline in terms of no to minimal effects on most CYP enzymes (Table 6.10)
Tolerability
• Comparable to SSRIs at low dose; NE-mediated adverse effects at higher doses (Tables 6.2, 6.3, and 6.7)
Efficacy
• Overall: equivalent to TCAs. High-dose venlafaxine was superior to fluoxetine in double-blind study
• Unique spectrum of activity: possible efficacy in cases not responsive to TCAs or SSRIs
• Rate of response is a function of dose: 2 to 4 weeks at doses £ 150 mg/day and 4 to 7 days at doses of ³ 300 mg/day
• Maintenance of response: evidence from controlled studies
Payment
• Brand name only available; see discussion (Table 7.10)
Simplicity
• Ease of administration: can be started at an effective dosage (75 mg/day); once daily dosing with sustained-release formulation. Dose titration is not necessary for many patients but increasing the dose is a reasonable strategy if starting dose is ineffective
Abbreviations: NE, norepinephrine; SSRI, serotonin selective reuptake inhibitor; CYP, cytochrome P450 enzyme; TCA, tricyclic antidepressant.
TABLE 7.6 — STEPS Criteria for Serotonin-2A Receptor Antagonist and Weak Serotonin Reuptake Inhibitors (eg, Nefazodone)

Safety
• Therapeutic index: wide. No serious systemic toxicity due to acute overdose
• Drug-drug interactions:
– Pharmacodynamic: can interact with other agents that decrease arousal or impair cognitive performance. Can interact with adrenergic agents affecting blood pressure regulation. Complex interactions with other serotonin-active agents
– Pharmacokinetic: substantially inhibits CYP 3A which is responsible for 50% of known oxidative drug metabolism, including its own metabolism (ie, nonlinear pharmacokinetics or autoinhibition)
Tolerability
• Dizziness, drowsiness, and confusion are dose-dependent adverse effects; hence the need for dose titration (Tables 6.2, 6.3, and 6.7)
Efficacy
• Overall: equivalent to TCAs in outpatients
• Unique spectrum of activity: none demonstrated
• Rate of response: 2 to 4 weeks
• Maintenance of response: evidence from controlled studies Payment
• Generic version of trazodone available. Brand name only for nefazodone. The latter has clinically important pharmacologic advantages over the former (Table 7.10)
Simplicity
• Ease of administration: requires dose titration and divided daily dosing for optimal antidepressant effect
Safety
• Therapeutic index: wide. No serious systemic toxicity due to acute overdose
• Drug-drug interactions:
– Pharmacodynamic: can interact with other agents that decrease arousal or impair cognitive performance. Can interact with adrenergic agents affecting blood pressure regulation. Complex interactions with other serotonin-active agents
– Pharmacokinetic: substantially inhibits CYP 3A which is responsible for 50% of known oxidative drug metabolism, including its own metabolism (ie, nonlinear pharmacokinetics or autoinhibition)
Tolerability
• Dizziness, drowsiness, and confusion are dose-dependent adverse effects; hence the need for dose titration (Tables 6.2, 6.3, and 6.7)
Efficacy
• Overall: equivalent to TCAs in outpatients
• Unique spectrum of activity: none demonstrated
• Rate of response: 2 to 4 weeks
• Maintenance of response: evidence from controlled studies Payment
• Generic version of trazodone available. Brand name only for nefazodone. The latter has clinically important pharmacologic advantages over the former (Table 7.10)
Simplicity
• Ease of administration: requires dose titration and divided daily dosing for optimal antidepressant effect
Abbreviations: CYP, cytochrome P450 enzyme; TCA, tricyclic antidepressant.
TABLE 7.7 — STEPS Criteria for Serotonin (5-HT2A and 5-HT2C) and Norepinephrine (a-2) Receptor Antagonists (eg, Mirtazapine)

Safety
• Therapeutic index: wide in terms of acute drug overdose. There were three cases of agranulocytosis out of approximately 3000 patients in clinical trial program. That number was too small to establish with confidence the actual incidence or even whether there was a causal relation. Postmarketing experience has not indicated that toxicity is a major concern. Nevertheless, a white count should be obtained if a patient presents with signs of fever or infection
• Drug-drug interaction:
– Pharmacodynamic: can cause multiple types of such interactions due to multiple mechanisms of action (Table 6.3)
– Pharmacokinetic: unlikely to either be a victim or a cause of such interactions based on in vitro studies, but that should be confirmed by in vivo studies
Tolerability
• Primary problems are sedation due to histamine receptor blockade and weight gain due to 5-HT2C receptor blockade. These can be treatment limiting (Tables 6.2, 6.3, and 6.7). Dose titration may not help
Efficacy
• Overall: good; was superior to fluoxetine in double-blind study of hospitalized patients
• Unique spectrum of activity: worked in 54% of patients who had not benefited from treatment with amitriptyline. Open label studies have reported effectiveness in patients who had not benefited from SSRIs
• Rate of response: 2 to 4 weeks
• Maintenance of response: data limited to noncontrolled open label extension studies
Payment
• Brand name only available; see discussion (Table 7.10)
Simplicity
• Can start on an effective dose immediately. Sedation can be a problem, but starting with a lower dose may only decrease efficacy without improving tolerability, because mirtazapine is a more potent histamine receptor blocker than it is a serotonin and adrenergic receptor blocker. No rigorous data on whether higher doses increase either efficacy or tolerability
TABLE 7.8 — STEPS Criteria for Dopamine and Norepinephrine Reuptake Inhibitors (eg, Bupropion)

Safety
• Therapeutic index: narrow. The dose needed for efficacy (300-450 mg/day) is only 2 to 3 times less than the dose that causes seizures in 2% of patients. That is the reason the maximum recommended daily dose is 450 mg/day. Seizures can also occur following an acute overdose but are generally managed easily in a medical setting
• Drug-drug interaction:
– Pharmacodynamic: should potentiate and reduce the effects of other dopamine and norepinephrine agonists and antagonists, respectively
– Pharmacokinetic: can be affected by fluoxetine and probably others in a clinically significant way. Due to its narrow therapeutic index, care should be taken when prescribing bupropion to patients on drugs known to inhibit CYP enzymes. Case report data suggest that it can substantially inhibit CYP 2D6
Tolerability
• Good. Does not cause sexual dysfunction seen with antidepressants which are serotonin reuptake inhibitors (Tables 6.2, 6.3, and 6.7)
Efficacy
• Overall: probably less than TCAs • Unique spectrum of activity: can work in TCA nonresponders
• Rate of response: 2 to 4 weeks
• Maintenance of response: not demonstrated in formal studies
Payment
• Brand name only available; see discussion (Table 7.10)
Simplicity
• Ease of administration: requires divided daily dosing for antidepressant effect and dose titration to achieve efficacy and minimize seizure risk
Abbreviations: CYP, cytochrome P450 enzyme; TCA, tricyclic antidepressant.
TABLE 7.9 — STEPS Criteria for Monoamine Oxidase Inhibitors (eg, Tranylcypromine)

Safety
• Therapeutic index: narrow. Serious toxicity can result from acute overdose
• Drug-drug interaction:
– Pharmacodynamic: hypertensive crisis can result from coadministration with tyramine and sympathomimetic agents. Serotonin syndrome can result from coadministration with serotonin agonists (eg, serotonin uptake inhibitors) (Table 6.2)
– Pharmacokinetic: inhibits the oxidative enzyme, MAOI, but effects on CYP enzymes have not been studied. Not known to be affected by other drugs in a clinically significant way
Tolerability
• Generally good, especially if kept to effective minimum dose. Ironically, main tolerability problem is hypotension (Table 6.3)
Efficacy
• Overall: Probably less than TCAs
• Unique spectrum of activity: can work in TCA nonresponders
• Rate of response: 2 to 4 weeks
• Maintenance of response: not adequately tested
Payment
• Generic versions are available making these antidepressants the least expensive in terms of acquisition costs (Table 7.10). These savings are likely offset by expenses arising from problems with their safety and tolerability
Simplicity
• Ease of administration: typically administered in divided daily doses. Dose titration recommended to optimize efficacy and minimize hypotension
Abbreviations: MAOI, monoamine oxidase inhibitor; CYP, cytochrome P450 enzyme; TCA, tricyclic antidepressant.

Tertiaryamine tricyclic antidepressants (TATCAs) have a narrow acute therapeutic index due to their inhibition of sodium (Na+) fast channels at concentrations only 10 times higher than those needed to treat major depression.194 For this reason, an overdose of these drugs carries a serious risk of slowing intracardiac conduction to the point of inducing a fatal ventricular arrhythmia. This effect also occurs with secondary amine TCAs (eg, desipramine) even though these antidepressants are norepinephrine selective reuptake inhibitors (NSRIs) at usual therapeutic concentrations (Table 6.2). Other than TCAs, fatal overdose is not an issue with any other antidepressant when taken alone since they do not affect intracardiac conduction.

Some clinicians mistakenly believe that the arrhythmia caused by antidepressants such as desipramine is due to their ability to inhibit the neuronal uptake pump for norepinephrine. However, reboxetine (an investigational non-TCA NSRI) and venlafaxine (a serotonin and norepinephrine reuptake inhibitor [SNRI]) block norepinephrine uptake, but do not inhibit Na+ fast channels even at concentrations achieved following a moderate overdose. Arrhythmias generally do not occur following overdose of these norepinephrine-active antidepressants.41,73

Although not as serious as the cardiotoxicity of TCAs, there are some safety concerns with some newer antidepressants. For example, bupropion has a dose-dependent risk of seizures.57 At dosages of 450 mg/day, the seizure risk is 0.4%; that risk may be lower with the sustained-release formulations because of the blunting of peak levels. The risk of seizures due to bupropion increases substantially when its dosage exceeds 450 mg/day (the maximum recommended daily dosage).

Consistent with their pharmacology as indirect sympathomimetic agonists, NSRIs (eg, desipramine) and high-dosages of the SNRI venlafaxine (ie, > 225 mg/day) can cause dose-dependent increases in blood pressure.73,169,240 The magnitude of the increases in blood pressure produced by these drugs is generally modest, but in some cases, can be sufficient to warrant either discontinuation or the addition of an antihypertensive medication.

Finally, some newer antidepressants (eg, fluoxetine) at their usually effective antidepressant dose produce substantial inhibition of one or more drug metabolizing cytochrome P450 (CYP) enzymes (Table 6.10 and Figure 6.2). That action carries with it the liability for causing clinically significant and even fatal pharmacokinetically mediated drug-drug interactions (Chapter 10).

Antidepressants are among the most widely prescribed medications and are often taken on a long-term basis to prevent the recurrence of major depressive episodes. There is no evidence of serious long-term safety problems with any antidepressant, based on spontaneous reports to the Food and Drug Administration (FDA). The long-term safety of antidepressants has also been systemically evaluated during clinical trial development programs.103,169,178 The most rigorous data in this regard come from year-long, placebo-controlled relapse prevention studies.62,67,112,148

Other data come from open-label extension studies. In these studies, patients who have completed double-blind acute treatment studies can elect to continue on the medication until it is approved by the FDA. Given these two types of approaches, the safety of the drug will have typically been assessed in several hundred patients for up to 1 year by the time it is available in the United States.

The more sites of action affected by a drug, the more potential there is for pharmacodynamic interaction with other coprescribed drugs.192 In fact, the sites of action affected by a drug determine the specific type of pharmacodynamic drug-drug interaction(s) it can cause.

Tertiary amine TCAs, due to their multiple effects on different neural targets (Table 6.2), will cause the most types of pharmacodynamic drug-drug interactions.192 Specifically, these antidepressants can:

Using Tables 6.2 and 6.3, the clinician can determine which other antidepressants share the same potential to cause specific types of pharmacodynamic drug interactions as do the TATCAs.

The CYP enzyme-mediated drug-drug interactions are the most common type of pharmacokinetic drug-drug interaction.93,94 The clinician can use Table 6.10 to determine which antidepressants are most likely to cause such interactions (ie, be the perpetrator). Table 6.9 can be used to determine which drugs will be affected (ie, be the victim) by an alteration (ie, induction or inhibition) in the metabolic capacity of a specific CYP enzyme. Fluoxetine and fluvoxamine will cause the most types of CYP enzyme-mediated drug interactions because they inhibit more than one CYP enzyme to a substantial degree at their usually effective antidepressant dose (Table 6.10 and Figure 6.2).

Pharmacokinetic drug interactions can present in many different ways (ie, decreased tolerability, decreased efficacy, withdrawal syndromes, or increased toxicity) because they typically produce an effect comparable to a change in the dose of the affected drug (Chapter 10). The clinician may not realize that starting or stopping the inhibitor has effectively changed the functional dose of the affected or victim drug by changing its clearance. Hence, the change in efficacy or tolerability may incorrectly be attributed to a patient sensitivity problem or to an underlying medical problem. Such misdiagnosis can:

Tolerability

As with drug-drug interactions, the more sites of action that a drug affects, the more types of adverse effects it can produce. The clinician can use Tables 6.2 and 6.3 to predict the types of adverse effects of each class of antidepressant. Tables 6.4 through 6.7 contain frequency of specific adverse effects for various antidepressants based on double-blind, placebo-controlled studies. Chapter 11 discusses the possible management approaches to the most common nuisance adverse effects of specific antidepressants.

Although most of the adverse effects caused by antidepressants occur after the first dose, or at least within the first week of treatment, sexual dysfunction is an exception.145,147 This effect frequently does not come to clinical attention until several weeks or months into treatment. All antidepressants that produce substantial serotonin uptake inhibition can cause:

There is no universally effective treatment for the sexual dysfunction caused by serotonin uptake inhibitors, but potential useful "antidotes" are discussed in Chapter 11.

Other types of sexual dysfunction caused by antidepressants include:

Norepinephrine uptake inhibitors (eg, desipramine), bupropion, mirtazapine and nefazodone at dosages below 500 mg/day appear to have minimal risk of causing sexual adverse effects.103,151,171,213

Efficacy

As with many other illnesses, the treatment of clinical depression can be divided into several clinically important and distinct phases as follows:

There are several factors that need to be weighed when considering the induction of an acute remission as follows:

Clinical depression is a syndrome that will likely prove to be more than one illness when understood from the perspective of pathophysiology and/or etiology. The results of genetic and familial studies, biological marker studies, psychosocial studies and antidepressant clinical trials are all consistent with the conclusion that clinical depression is a heterogeneous group of disorders. That means that antidepressants with different mechanisms of action may treat different types of clinical depression (ie, have different overall acute efficacy and different spectra of efficacy).

Overall acute efficacy refers to how many patients with clinical depression will respond to a trial of a particular antidepressant. Assuming all other factors are equal (ie, safety, tolerability, cost and simplicity), the antidepressant of first choice would be the one that will produce a clinically meaningful, acute response in the largest number of patients.

Unique spectrum of efficacy refers to whether an antidepressant will work in a patient who has not benefited from a previous trial of another antidepressant. This issue is clinically important because virtually any antidepressant fails to produce an adequate response in one third to one half of patients suffering from clinical depression.

Rate of response is self-explanatory. This issue is more of interest to psychiatrists than to the primary-care practitioner, since the usual time course for response to an antidepressant (ie, approximately 2 weeks) does not typically pose a significant problem for outpatients with mild-to-moderate clinical depression.

Maintenance efficacy refers to whether there is evidence that continued treatment with the antidepressant is an effective strategy to prevent a relapse of the current episode. Since clinical depression is a recurrent condition in 50% or more of cases, prophylactic efficacy refers to whether there is evidence that indefinite treatment with an antidepressant in high-risk patients can prevent future episodes.

There is some evidence that drugs having more than one mechanism of antidepressant action may produce a higher percentage of response than do antidepressants with a single mechanism of action. Specifically, there are double-blind studies in which clomipramine (a TATCA), mirtazapine and venlafaxine produced a statistically significant better response rate than did an active control with a single mechanism of action (eg, serotonin uptake inhibition) in hospitalized patients with severe clinical depression.48,55,56,267 On the other hand, antidepressants with multiple mechanisms of action often cause a higher dropout rate due to an increased incidence of adverse effects. Thus, their increased efficacy can be offset by the increased dropout rate such that the net outcome is a "wash."

Most antidepressants take at least 2 weeks to demonstrate a higher response rate than the parallel placebo-controlled condition.103 In fact, it is advisable to not give up on an antidepressant or even to increase the dose of the single-mechanism-of-action antidepressant until a trial of at least 4 weeks on a stable dose has been given. In contrast, there is one study indicating that high-dosage venlafaxine (ie, > 300 mg/day) can produce a response in 20% of patients within 1 week and can statistically separate from placebo as soon as 4 days.169 These findings are consistent with combined inhibition of both serotonin and norepinephrine uptake pumps at such doses of venlafaxine (Figure 6.4).103 A similar rapid onset of antidepressant action has also been produced by using a combination of an NSRI (eg, desipramine) and a serotonin selection reuptake inhibitor (SSRI) (eg, sertraline).153,258

Of course, either of these approaches (ie, high dose venlafaxine or combined use of an NSRI and an SSRI) to increase efficacy by affecting these two mechanisms of action sets the stage for more types of adverse effects mediated by these two different mechanisms of action (Table 6.5). The increase in tolerability problems can more than offset the increase in efficacy, particularly in the primary-care setting where there is typically no need for a more rapid response than is achieved with conventional antidepressant treatment. Nevertheless, these strategies can be useful when dealing with a severely ill or hospitalized patient where time of response is a critical issue.

One advantage of having eight mechanistically different classes of antidepressants is the possibility that patients who are intolerant of or do not respond to one type of antidepressant may benefit from treatment with a mechanistically different type of antidepressant. While this possibility is intuitively plausible, this subject has received scant systematic study (Chapter 11). Most studies that have tested this possibility have not been scientifically rigorous, but were instead open label and hence subject to potential biases.

The best data on this topic come from a series of double-blind, crossover studies with NSRIs and SSRIs.245 Based on these studies, approximately 50% of patients who do not benefit from a trial of an SSRI will benefit when switched to an NSRI. The reverse is also true. This evidence is also fully consistent with the aforementioned data that combined treatment with an NSRI and an SSRI or with high-dose venlafaxine produces more responders than does treatment with either mechanism alone. While the available data are modest, there certainly is no evidence to suggest that a nonresponse to one type of antidepressant predicts nonresponse to a mechanistically different type of antidepressant.61,245

The FDA does not require proof of either maintenance or prophylactic efficacy to approve a drug as an antidepressant. Nevertheless, most approved newer antidepressants have been tested and shown to have sustained efficacy over the maintenance phase (ie, 6 months after the induction of an acute remission). The best studies involve open-label treatment for 2 to 3 months to induce an acute response and then double-blind, random assignment to either remain on the antidepressant or be switched to placebo.62,67,148 Approximately one fourth to one third more patients on placebo will relapse back into their depressive episode over a 6-month interval in comparison to those maintained on the antidepressant. These data are sufficiently compelling so that continued treatment with an antidepressant for 6 months after an acute response is recommended for all first-episode patients.4,61

Less well studied is the issue of prophylactic efficacy. The single best study has been done with the TATCA, imipramine.76 The follow-up treatment period for most patients in this study was 3 years, but some were followed for 5 years. The patients were selected for being at high risk for relapse. Participants were required to have had at least four episodes of depression (ie, the current one plus three prior episodes). The results provided compelling support for the benefit of prophylactic treatment with an antidepressant in such a high risk population-only 20% of the patients who remained on imipramine relapsed at the end of 3 years versus 95% of patients who were switched to placebo. A more recent 76-week, double-blind, randomized, parallel-group study demonstrated the superiority of sertraline over placebo in the maintenance treatment of patients with chronic major depression.112

Studies with other antidepressants have generally not lasted more than 1 year and have not been conducted in patients with such a high risk for recurrent episodes. Nevertheless, these studies have also demonstrated fewer recurrent episodes on drug versus a parallel, placebo-controlled condition.

Payment

This issue is complex, controversial, and not unique to antidepressants. Clinicians often find themselves between two competing forces:

The latter is generally concerned with the acquisition cost (ie, how much does the prescription cost) (Table 7.10). The former points out that the increased acquisition cost of the new medication is more than offset by other "hidden" costs of the older and often generic drugs.

For example, the acquisition cost of a TATCA is considerably less than all of the newer antidepressants, but there are other associated costs that should be taken into account when doing a cost comparison.40 These include but are not limited to:

Simplicity

An important issue when selecting an antidepressant is how easily can the clinician select an optimal dose for the patient; that is, determine the dose which has the greatest likelihood of producing a good antidepressant response and the least risk of causing either nuisance or serious adverse effects (Table 7.11).169 This issue is particularly important for the primary-care practitioner given time constraints in daily practice.

A related issue is how easy is it for the patient to take the medication as prescribed. An ideal drug would be one that:

Start At Effective Dose

Several of the antidepressants meet many of the requirements above. Most SSRIs (ie, fluoxetine, paroxetine, and sertraline, but not citalopram and fluvoxamine), mirtazapine, most NSRIs (eg, desipramine, nortriptyline), and venlafaxine can be started at an effective dose. A few additional comments are warranted with regard to mirtazapine. Sedation can be a problem in early treatment with mirtazapine.178 For this reason, some clinicians may consider starting with a lower dose (7.5 mg), but this strategy may not avoid sedation (the drug's most common effect [Table 6.7]) and may compromise efficacy.151

In contrast to the above antidepressants, bupropion, nefazodone, and TATCAs do require dose titration primarily to minimize adverse effects.103 In the case of bupropion, the goal of titration is to find the lowest effective dose to minimize the risk of seizures.57 In the case of nefazodone and TATCAs, the issue is minimizing nonserious but nevertheless nuisance and sometimes rate-limiting adverse effects.


TABLE 7.10 — Average Wholesale Price of Representative Doses of Antidepressants*
Generic/Trade Drug Name Strength (mg) Price ($/100)
Generic Trade
Tertiary Amine Tricyclic Antidepressants
Amitriptyline/Elavill 25 10 40
50 12 72
100 18 125
Doxepin/Sinequan 25 14 47
50 19 65
100 40 119
Imipramine/Tofranil 25 6 47
50 9 80
Selective Norepinephrine Reuptake Inhibitors  
Desipramine/Norpramin 25 25 65
50 50 122
100 100 204
Nortriptyline/Pamelor 25 80 100
50 150 190
Serotonin Selective Reuptake Inhibitors  
Citalopram/Celexa 20 NA 193
40 NA 202
Fluoxetine/Prozac 10 NA 230
20 NA 240
Fluvoxamine/Luvox 50 NA 206
100 NA 212
Paroxetine/Paxil 10 NA 190
20 NA 200
30 NA 210
40 NA 222
Sertraline/Zoloft 50 NA 180
100 NA 200
Serotonin and Norepineprhine Reuptake Inhibitors  
Venlafaxine-IR/Effexor IR 37.5 NA 105†
75 NA 115†
100 NA 122†
Venlafaxine-XR/Effexor XR 37.5 NA 194
75 NA 217
150 NA 237
Serotonin 2A Antagonists  
Nefazodone/Serzone 100, 150, 200, 250 NA 93†
Trazodone/Desyrel 50 30 150
100 40 250
300 NA 390
Serotonin (5-HT2A and 2C) and Adrenergic (-2) Antagonists  
Mirtazapine/Remeron 15 NA 195
30 NA 202
Dopamine and Norepinephrine Reuptake Inhibitors  
Bupropion-IR/Wellbutrin IR 75 NA 60†
100 NA 83†
Bupropion-SR/Wellbutrin SR 100 NA 115†
150 NA 117†
Monoamine Oxidase Inhibitors  
Phenelzine/Nardil 15 NA 42
Tranylcypromine/Parnate 10 NA 50
Abbreviations: NA, none available; IR, immediate release; SR, sustained release; XR, extended release.
* Information taken from 1998 Drug Topics Red Book and rounded to the nearest dollar. † Recommended dosing is twice or three times a day, whereas other drugs in this chart are given once a day. That difference will affect the daily cost of treatment (ie, how long 100 pills will last).
TABLE 7.11 — Summary of Package Insert Dosing Guidelines
Generic/Trade Drug Name Recommended Dose
Start/ Max (mg/day)		 Dosage Guidelines for Specific Patients
Children Adolescents Elderly Hepatic* Renal*
Mixed Reuptake Inhibitors and Neuroreceptor Blockersa,b  
Amitriptyline/Elavil 75/300c NA
Amoxapine/Ascendin 100/600c,d NR NA
Clomipramine/Anafranile 25/250c,f NA NA
Doxepin/Sinequan 75/300c NA NA NA
Imipramine/Tofranil 75/300cc NA
Norepinephrine Selective Reuptake Inhibitorsa,b  
Desipramine/Norpramin 100/300c NA
Maprotiline/Ludiomil 75/225c,f NA NA
Nortriptyline/Pamelor 50/150c NA NA
Serotonin Selective Reuptake Inhibitors  
Citalopram/Celexa 20/60 NA NA
Fluoxetine/Prozac 20/80 NA NA
Fluvoxamine/Luvoxe 50/300g NR NA
Paroxetine/Paxil 20/50 NA NA
Sertraline/Zoloft 50/200 Same Same
Serotonin and Norepinephrine Reuptake Inhibitors  
Venlafaxine-IR/Effexor IR 75g,h/375g NA NA NA
Venlafaxine-XR/Effexor XR 75h/375 NA NA NA
Serotonin (5-HT2A) Receptor Blockers and Weak Serotonin Uptake Inhibitors  
Nefazodone/Serzone 200g/600g NA NA
Trazodone/Dyserel 150g/600g NA NA
Serotonin (5-HT2A and 5-HT2C) and Norepinephrine Receptor Blockers  
Mirtazapine/Remeron 15/45 NA NA
Dopamine and Norepinephrine Reuptake Inhibitors  
Bupropion-IR/Wellbutrin IR 200g/450c,f,i NA NA
Bupropion-SR/Wellbutrin SR 150/400c,f,i NA NA
Monoamine Oxidase Inhibitors  
Phenelzine/Nardil 45g/90g NA NA
Tranylcypromine/Parnate 30g/60g NA NA
Abbreviations: NA, not available; NR, not recommended; IR, immediate release; XR, extended release; SR, sustained release.
* Impairment.
  • Starting dose may be given either as a once-a-day dose or on a divided schedule. Once an effective and tolerated dose has been established, it may be given on a once-a-day basis, but a divided dose may still be more prudent with a higher total dose and in patients who are elderly or debilitated. The maximum once-a-day dose of doxepin is 150 mg.
  • Usual dose may be given either as a once-a-day dose or on a divided schedule.
  • Therapeutic drug monitoring has been either demonstrated to increase the safe and efficacious use of this drug or theoretically should; demonstrated for amitriptyline, clomipramine, desipramine, imipramine, and nortriptyline. Theoretical for the rest, but has not been adequately studied.
  • Doses should exceed 400 mg/day only in hospitalized patients who do not have a history of seizures and who have not benefited from an adequate trial of 400 mg/day.
  • Not formally labeled by the FDA for the treatment of clinical depression but rather for obsessive-compulsive disorder; labeled for use as an antidepressant in other countries.
  • Maximum daily dose should not be exceeded due to an increased risk of seizures.
  • Dose should be given on a divided schedule (bid or tid).
  • For some patients, it may be desirable to start at half the dose for 4 to 7 days to improve tolerance, particularly in terms of nausea.
  • It is particularly important to administer in a manner most likely to minimize the risk of seizures. Dose increases should not exceed 100 mg/day in a 3-day period. Cautious dose titration can also minimize agitation, motor restlessness, and insomnia. Time between doses should be at least 4 hours for 100 mg IR doses, 6 hours for 150 mg IR doses, and 8 hours for SR doses. Increases above 300 mg/day should only be done in patients with no clinical effects after several weeks of treatment at 300 mg/day. Bupropion should be discontinued in patients who do not experience an adequate response after an adequate period on maximum recommended daily dose. Dosing in the elderly, the debilitated, and patients with hepatic and/or renal impairment has not been adequately studied so increased caution may be prudent.

Additional comments on dose titration: The package inserts for the following drugs indicate that they can be started at a dose which is usually effective to treat clinical depression: fluoxetine, mirtazapine, paroxetine, tranylcypromine, sertraline, venlafaxine. The following comments apply about the use of higher doses with these antidepressants. For fluoxetine, paroxetine, and sertraline: although fixed-dose studies in patients with clinical depression found no advantage on average to higher doses, an increase may be considered after several weeks on the starting dose if no clinical improvement has been observed. For mirtazapine, dose escalation should not be made at intervals of less than 1 to 2 weeks to adequately evaluate therapeutic response to a given dose. For tranylcypromine, improvement can be seen between 48 hours and 2 weeks of starting therapy; if not, dose increases in 10 mg/day increments may be made at intervals of 1 to 3 weeks.

The package inserts for the following drugs recommend starting at a lower than usually effective dose and titrate up to a dose which is usually effective to treat clinical depression in order to minimize tolerability or safety problems: amitriptyline, amoxapine, bupropion, citalopram, clomipramine, fluvoxamine, doxepin, imipramine, nefazodone, phenelzine, trazodone, and trimipramine. The following are additional comments about dose titration with these antidepressants: For the tricyclic antidepressants, the dose should be gradually increased during the first 2 weeks based on therapeutic drug monitoring and clinical assessment of efficacy and tolerability. For fluvoxamine, a lower than usually effective starting dose is recommended to improve tolerability. The dose should be increased every 4 to 7 days as tolerated until maximum therapeutic benefit is achieved. For citalopram, the starting dose is 20 mg/day with the recommendation to generally increase to 40 mg/day. While doses above 40 mg/day are not ordinarily recommended, some patients may require a dose of 60 mg/day. For nefazodone, a lower than usually effective starting dose is recommended to improve tolerability. Dose titration should occur in increments of 100 to 200 mg/day as determined by tolerability and the need for further clinical improvement. These incremental advances should be done using divided doses and at intervals of at least 1 week. It may be advisable to titrate up more slowly in elderly and debilitated patients. For phenelzine, a lower than usually effective starting dose is recommended to improve tolerability. Its dose should be increased to at least 60 mg/day at a fairly rapid pace consistent with good tolerability. For trazodone, the same comments apply as for nefazodone when trazodone is used as an antidepressant; however, it is now mainly used as a nonhabit-forming sedative given as a single bedtime dose of 50 to 200 mg as needed for sleep.

The next issue is how easily the optimal dose of the antidepressant can be determined. Therapeutic plasma level ranges have been established for most TCAs. Thus, the clinician can use therapeutic drug monitoring (TDM) to adjust the dosage to compensate for interindividual differences in clearance in order to optimize the likelihood of antidepressant efficacy while simultaneously avoiding toxicity.191 While increasing efficacy is an advantage of TDM with TCAs, the principal and compelling reason to use TDM with TCAs is to avoid toxicity in slow metabolizers. This is not an issue with any other class of antidepressants with the possible exception of bupropion.180

In the case of the SSRI class, there is no advantage to using higher than the usually effective minimum dose.170 This finding is based on the results of double-blind, fixed-dose studies which have found flat dose-response curves above the usually effective minimum dose for these antidepressants.170 This finding does not preclude the possibility that some patients may benefit from a higher dose of an SSRI (Chapter 11). Although beyond the scope of this book, paroxetine is the only SSRI for which there is substantial evidence that higher doses are needed to treat patients with anxiety disorders versus patients with clinical depression. The reasons for this dosing difference are not clear.

In contrast to SSRIs, there is an ascending dose-antidepressant response curve with venlafaxine.177 This antidepressant at its starting dosage of 75 mg/day produces approximately the same number of responders as do the SSRIs, but the percentage of responders increase with higher doses consistent with its apparent dual mechanism of antidepressant action (ie, serotonin and then norepinephrine uptake inhibition). Consistent with its pharmacology, higher doses of venlafaxine also cause a higher incidence of serotonin- and norepinephrine-mediated adverse effects, including the potential to increase blood pressure as previously discussed.

There are no published fixed-dose studies which empirically establish that doses of mirtazapine higher than 15 mg/day are more effective or better tolerated. There are anecdotal and theoretical reasons to suspect that higher doses will decrease the problem of sedation (ie, offsetting arousal effects as a result of alpha-2-adrenergic blockade at higher concentrations [Figure 6.4]). However, in the case of bupropion and nefazodone, the optimal dose in terms of efficacy and tolerability is quite variable and requires empiric dose titration.103

Most of the antidepressants can be given once a day. The exceptions are bupropion (including the sustained-release formulation) and nefazodone which require at least twice-a-day dosing (Table 7.11). While once-a-day dosing is desirable, the available literature suggests that problems with compliance do not become an issue until the required dosing frequency is 3 times a day or more.