Chapter 12. Tricyclic and Tetracyclic Drugs

Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

1 of 33

10/05/2009 15:56

Print Close Window

12. Craig Nelson: Chapter 12. Tricyclic and Tetracyclic Drugs, in The American Psychiatric Publishing Textbook of

Psychopharmacology, 4th Edition. Edited by Alan F. Schatzberg, Charles B. Nemeroff. Copyright ©2009 American

Psychiatric Publishing, Inc. DOI: 10.1176/appi.books.9781585623860.409505. Printed 5/10/2009 from

www.psychiatryonline.com

Textbook of Psychopharmacology >

Chapter 12. Tricyclic and Tetracyclic Drugs

TRICYCLIC AND TETRACYCLIC DRUGS: INTRODUCTION

The tricyclic antidepressant agents hold an important place in the history of treatments for

depression. They were the first class of antidepressant compounds to be widely used in depression

and remained the first-line treatment for more than 30 years. The observation of their activity led

to theories of drug action involving norepinephrine and serotonin. Indeed, this

“psychopharmacological bridge” suggested that alterations of these neurotransmitters might cause

depression (Bunney and Davis 1965; Prange 1965; Schildkraut 1965). The tricyclics were

extensively studied, and through this study the field developed several principles for the

management of depressive illness. For example, in addition to understanding the need for adequate

dose and duration during acute treatment, the importance of continuation treatment was described.

The adverse events associated with these agents required that psychiatrists become familiar with a

variety of syndromes, such as anticholinergic delirium, delayed cardiac conduction, precipitation of

acute glaucoma, and orthostatic hypotension. The observation that tricyclic plasma concentrations

varied widely stimulated interest in understanding the metabolism of tricyclics. The field was

introduced to the concepts of polymorphisms in the mephenytoin and debrisoquine pathways (later

relabeled the cytochrome P450 [CYP] 2C19 and 2D6 pathways), which in part explained the widely

varying blood levels. Knowledge of the widely varying drug concentrations raised questions about

the relationships of clinical activity and drug concentrations. Although effects of antipsychotics and

a few other drugs on tricyclic plasma levels had been described in the 1970s, it was the observation

of the effect of fluoxetine on desipramine coupled with the aggressive marketing of the selective

serotonin reuptake inhibitors (SSRIs) that greatly expanded our awareness of and knowledge about

drug interactions. Finally, our knowledge of how these drugs worked became the basis for the

discovery of new drugs such as the SSRIs.

HISTORY AND DISCOVERY

Although iminodibenzyl had been synthesized in 1889 (Byck 1975), it was not until after 1948 that

derivatives of this compound were investigated for their potential usefulness in human subjects.

Ironically, the properties that were of interest—the antihistaminic and sedative actions—would later

be viewed as the “side effects” of these compounds. In the 1950s, investigation of one of these

compounds led to the serendipitous observation that chlorpromazine had “antipsychotic” effects

(Delay and Deniker 1952). The tricyclic compound imipramine was closely related to

chlorpromazine, differing only in the substitution of an ethylene linkage for sulfur. In 1957, Roland

Kuhn, a Swiss psychiatrist, investigated the clinical effects of imipramine in human subjects in part

to determine if its sedative properties would be useful (Kuhn 1958, 1970). He found that

imipramine was not useful for calming agitated patients; however, he observed that it did appear to

ameliorate symptoms in depressed patients. As with lithium and chlorpromazine earlier, the

discovery of the psychotropic effects of the tricyclics was a chance observation.

After imipramine was introduced, several other antidepressant compounds were developed and

marketed. These compounds had a basic tricyclic structure and also shared many of the secondary

effects for which the tricyclics came to be known. Later, other heterocyclic compounds, such as

maprotiline, were also marketed. These agents had somewhat similar structures and secondary

effects.

STRUCTURE–ACTIVITY RELATIONSPrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

2 of 33

10/05/2009 15:56

Tricyclic and tetracyclic compounds are categorized on the basis of their chemical structure (Figure

12–1). The tricyclics have a central three-ring structure, hence the name. The tertiary-amine

tricyclics, such as amitriptyline and imipramine, have two methyl groups at the end of the side

chain. These compounds can be demethylated to secondary amines, such as desipramine and

nortriptyline. The tetracyclic compound maprotiline (Ludiomil) has a four-ring central structure.

Five tertiary amines have been marketed in the United States—amitriptyline (Elavil), clomipramine

(Anafranil), doxepin (Sinequan), imipramine (Tofranil), and trimipramine (Surmontil). The three

secondary-amine compounds are desipramine (Norpramin, Pertofrane), nortriptyline (Aventyl,

Pamelor), and protriptyline (Vivactil). All of these compounds, in addition to amoxapine (Asendin)

and maprotiline (Ludiomil), have been approved for use in major depression with the exception of

clomipramine (Anafranil), which in the United States is approved for use only in

obsessive-compulsive disorder (OCD).

FIGURE 12–1. Drugs marketed in the United States as tricyclics (1–9) and a tetracyclic (10).

Source. Reprinted from Potter WZ, Manji HK, Rudorfer MV: “Tricyclics and Tetracyclics,” in The American

Psychiatric Press Textbook of Psychopharmacology, Second Edition. Edited by Schatzberg AF, Nemeroff CB.

Washington, DC, American Psychiatric Press, 1998, p. 200. Copyright 1998, American Psychiatric Press, Inc.

Used with permission.

Although the tricyclics are named for their central three-ring structure, the nature of the side chain

appears more important for their function. The tertiary tricyclics—amitriptyline, imipramine, andPrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

3 of 33

10/05/2009 15:56

clomipramine—are more potent in blocking the serotonin transporter. The secondary tricyclics are

much more potent in blocking the norepinephrine transporter (Table 12–1) (Bolden-Watson and

Richelson 1993; Tatsumi et al. 1997).

TABLE 12–1. Affinity of tricyclics and tetracyclics for the neurotransmitter transporters and

specific receptors (expressed as equilibrium dissociation constants)

Potency uptake blockade Receptor binding affinity

Drug 5-HT NE DA

1

2

H1

M1

5-HT1A

5-HT2

Amitriptyline 4.3 35 3,250 27 940 1.1 18 450 18

Amoxapine 58.0 16 4,310 50 2,600 25 1,000 220 1.0

Clomipramine 0.28 38 2,190 38 3,200 31 37 7,000 27

Desipramine 17.6 0.83 3,190 130 7,200 110 198 6,400 350

Doxepin 68.0 29.5 12,100 24 1,100 0.24 80 276 27

Imipramine 1.4 37 8,500 90 3,200 11 90 5,800 150

Maprotiline 5,800.0 11.1 1,000 90 9,400 2 570 12,000 120

Nortriptyline 18.0 4.37 1,140 60 2,500 10 150 294 41

Protriptyline 19.6 1.41 2,100 130 6,600 25 25 3,800 67

Trimipramine 149.0 2,450 3,780 24 680 0.27 58 8,400 32

Reference

Pentolamine

15

Yohimbine

1.6

D-Chlorpheniramine

15

Atropine

2.4

Serotonin

0.72

Ketanserin

2.5

Note. Affinity and potency = equilibrium dissociation constants in molarity. 1 = 1-adrenergic; 2 =

2-adrenergic; DA = dopamine; 5 HT = serotonin; 5-HT1A = serotonin1A; 5-HT2 = serotonin2; H1 = histamine1;

M1 = muscarinic1; NE = norepinephrine.

Source. Uptake potency data adapted from Tatsumi et al. 1997. Receptor affinity data adapted from Richelson

and Nelson 1984.

The nature of the central three-ring structure produces other differences. For example,

amitriptyline and nortriptyline share a similar dibenzocycloheptadiene structure, although their side

chains are identical to those of imipramine and desipramine, respectively. Similar to the

imipramine–desipramine pair, amitriptyline and nortriptyline have similar differences in potency in

blocking the serotonin and norepinephrine transporters; however, compared with imipramine,

amitriptyline is more anticholinergic and antihistaminic, has greater 1-adrenergic receptor

blockade, and is somewhat more potent in blocking the serotonin transporter (see Table 12–1)

(Cusack et al. 1994; Richelson and Nelson 1984; Tatsumi et al. 1997).

Of the tricyclics, clomipramine is the most potent in blocking the serotonin transporter, although its

metabolite desmethylclomipramine is a potent norepinephrine reuptake inhibitor (Hall and Ogren

1981). Doxepin is structurally most similar to amitriptyline but is more similar in functional potency

to imipramine (Pinder et al. 1977). Doxepin is the most antihistaminic of these compounds.

Protriptyline is structurally most similar to nortriptyline but is more slowly cleared (has a longer

elimination half-life) and is given in doses about one-third those of nortriptyline (Moody et al.

1977).Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

4 of 33

10/05/2009 15:56

The structure of amoxapine differs from the structures of the other tricyclics. With a central

three-ring structure and a side chain that differ from those of the tricyclics, amoxapine is

structurally more similar to the neuroleptic loxapine, from which it is derived. Similar to the

secondary tricyclics, it is a potent norepinephrine reuptake inhibitor. Unlike all of the other

compounds in this group, amoxapine, and particularly its metabolite 7-hydroxyamoxapine, blocks

postsynaptic dopamine receptors (Coupet et al. 1979). As a result, it is the only compound in the

group that has antipsychotic activity in addition to antidepressant effects.

Maprotiline also differs from the others in this group. Although maprotiline is referred to as a

heterocyclic or tetracyclic, these terms do not explain its inclusion in the group. The side chain,

however, is identical to that in desipramine, nortriptyline, and protriptyline. As would be predicted

from this similarity, maprotiline is most potent in blocking the norepinephrine transporter (Randrup

and Braestrup 1977).

As discussed below, it is important to remember that in humans, the tertiary tricyclic compounds

are demethylated to secondary tricyclic compounds, all of which are potent norepinephrine

reuptake inhibitors. As a result, when these tertiary compounds are administered to human

subjects, the effects of both the metabolite and the parent have to be taken into account.

PHARMACOLOGICAL PROFILE

Reuptake Blockade

Early in the history of the tricyclic and tetracyclic antidepressants, the ability of these compounds

to block the transporter site for norepinephrine was described (Axelrod et al. 1961) (see Table

12–1). The tertiary amines have greater affinity for the serotonin transporter, whereas the

secondary amines are relatively more potent at the norepinephrine transporter. During the

administration of amitriptyline, imipramine, or clomipramine, these tertiary amines are

demethylated to secondary amines; thus, both serotonergic and noradrenergic effects occur. In

addition, because dopamine is inactivated by norepinephrine transporters in the frontal cortex

(Bymaster et al. 2002), norepinephrine reuptake inhibitors would be expected to increase

dopamine concentrations in that region.

Receptor Sensitivity Changes

The initial reuptake blockade described above is followed by a sequence of events (Blier et al. 1987;

Charney et al. 1991; Tremblay and Blier 2006). The tertiary tricyclic compounds inhibit the uptake

of serotonin, and serotonin levels rise. As the result of inhibitory feedback from the presynaptic

somatodendritic serotonin1A (5-HT1A) autoreceptor, the firing rate of the presynaptic serotonin

neuron falls, and concentrations of 5-hydroxyindoleacetic acid (5-HIAA), the major metabolite of

serotonin, decline rapidly. During a 10- to 14-day period, the presynaptic autoreceptor is

desensitized, and at this point, the tonic firing rate returns to its pretreatment rate. With both a

normal firing rate and reuptake blockade, serotonin transmission is enhanced.

The tricyclic agents also sensitize or upregulate postsynaptic 5-HT1A receptors (de Montigny and

Aghajanian 1978). These changes further enhance the effects of serotonin. The changes in pre- and

postsynaptic receptor sensitivity occur over a 2-week period. Consequently, the timing of these

changes is more consistent with the onset of early antidepressant response than with the initial

uptake blockade (Charney et al. 1991).

The tricyclics also have actions at the 5-HT2 receptor. Depression is associated with an increase in

postsynaptic 5-HT2 receptor density (Arora and Meltzer 1989; Cheetham et al. 1988; Stanley and

Mann 1983; Yates et al. 1990). A variety of antidepressants, including the tricyclics, downregulate

the 5-HT2 receptors (Goodwin et al. 1984; Peroutka and Snyder 1980). The 5-HT2 receptors

mediate excitatory effects, whereas the 5HT1A receptors generally have inhibitory effects; thus,

these two systems act in opposition. In preclinical experiments, when the 5-HT2 receptor was

blocked by an antagonist, the effects of serotonin were enhanced (Lakoski and Aghajanian 1985;

Marek et al. 2003). Some of the tricyclics—particularly doxepin, amitriptyline, and amoxapine—havePrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

5 of 33

10/05/2009 15:56

5-HT2 antagonist properties relatively comparable to their reuptake potency (see Table 12–1)

(Tatsumi et al. 1997). These properties should enhance their serotonergic effects.

The sequence of events with chronic dosing in the noradrenergic system is more complicated

(Tremblay and Blier 2006). As in the serotonergic system, reuptake inhibition results in a rapid

decline in norepinephrine turnover, as reflected by a fall in concentrations of

3-methoxy-4-hydroxyphenylglycol (MHPG), a metabolite of norepinephrine, and attenuation of the

firing rate of the noradrenergic neuron. This effect appears to be mediated by the presynaptic

somatodendritic 2-adrenergic autoreceptor, which provides inhibitory feedback to the presynaptic

neuron. In contrast to the serotonergic system, the firing rate of noradrenergic neurons remains

inhibited with chronic treatment (Szabo and Blier 2001), suggesting that somatodendritic 2

receptors do not desensitize. Norepinephrine concentrations do increase at postsynaptic sites such

as the hippocampus and frontal cortex. This may indicate desensitization of terminal 2

autoreceptors.

With chronic treatment, the postsynaptic -adrenergic receptor is downregulated or decreased in

density. In fact, downregulation of the -adrenergic receptor was once proposed as a mechanism of

tricyclic action (Sulser et al. 1978; Vetulani and Sulser 1975). Current evidence suggests that

-adrenergic receptor downregulation is more likely a compensatory change. Overall, chronic

administration of a noradrenergic reuptake inhibitor appears to override the downregulation of the

postsynaptic -receptor, resulting in enhanced noradrenergic transmission. This effect is

manifested as enhanced formation of the second messenger cyclic adenosine monophosphate

(cAMP) (Duman et al. 1997) and is reflected clinically by a persistent increase in heart rate (Roose

et al. 1998; Rosenstein and Nelson 1991).

Alternatively, some of the actions of noradrenergic reuptake inhibitors may be mediated by

postsynaptic 1 and 2 receptors, which do not appear to be downregulated during chronic

treatment. The net effect of norepinephrine reuptake inhibitors on noradrenergic transmission

during chronic treatment is further complicated by regional differences in the adrenergic receptors.

Nevertheless, the antidepressant effects of norepinephrine reuptake inhibitors do appear to be

mediated by norepinephrine, because inhibition of catecholamine synthesis with

-methyl-p-tyrosine (AMPT) results in relapse of depressive symptoms (H. L. Miller et al. 1996).

Secondary Effects

The tricyclic and tetracyclic compounds also have a variety of other actions mediated by other

receptors (Cusack et al. 1994; Richelson and Nelson 1984) (see Table 12–1). For example, these

compounds block muscarinic receptors, producing anticholinergic effects. Although these

anticholinergic effects have generally been thought to mediate adverse effects, a recent

double-blind, randomized crossover study in 19 subjects with major depression found that the

anticholinergic drug scopolamine had a beneficial effect on depressive and anxious symptoms

(Furey and Drevets 2006). The tricyclics also block histamine1 (H1) receptors and 1- and

2-adrenergic receptors, resulting in a variety of other effects (as discussed in the next section).

Tricyclics act on fast sodium channels, which explains their adverse cardiac effects. Although the

actions of tricyclics on sodium channels have traditionally been considered problematic for patients

with depression, these same actions may contribute to the beneficial effects of tricyclics on pain

(Priest and Kaczorowski 2007). The potency of secondary effects of the tricyclics and tetracyclics

varies considerably. Amitriptyline is the most anticholinergic not only of the agents in this group

but also of all antidepressants. Desipramine is the least anticholinergic among the tricyclics.

Doxepin is the most potent H1 antagonist in this group, although another antidepressant,

mirtazapine, is even more potent. The consequences of these secondary effects are discussed

below.

PHARMACOKINETICS AND DISPOSITION

Absorption

Absorption of the tricyclic and tetracyclic drugs occurs in the small intestine and is rapid andPrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

6 of 33

10/05/2009 15:56

reasonably complete. Peak levels are reached within 2–8 hours following ingestion. Exceptions

include protriptyline (peak levels reached between 6 and 12 hours after ingestion) and maprotiline

(peak levels not reached until 8 hours or longer). Although peak levels may have implications for

side effects, which can occur quickly, the timing of peak levels is relatively unimportant with

respect to efficacy because the antidepressant action of these drugs occurs over several weeks.

Volume of Distribution

The tricyclic and tetracyclic compounds are basic lipophilic amines and are concentrated in a variety

of tissues throughout the body. As a result, they have a high volume of distribution. For example,

concentrations of these drugs in cardiac tissue exceed concentrations in plasma.

Plasma Protein Binding

The tricyclic and tetracyclic compounds are extensively bound to plasma proteins (e.g., 90% or

greater) because of their lipid solubility. Exceptions are the hydroxy metabolites, which have lower

plasma protein binding than the parent compounds.

First-Pass Metabolism

Following absorption, the tricyclics are taken up in the circulation but pass first through the liver,

and metabolism of the drug begins—the so-called first-pass effect. As a result, the amount of the

compound that enters the systemic circulation is reduced.

Hepatic Metabolism

Hepatic metabolism is the principal method of clearance for the tricyclic and tetracyclic compounds.

Only a small portion of drug is eliminated by the kidney. Rates of hepatic metabolism vary widely

from person to person, resulting in dramatic differences in steady-state plasma concentrations.

Elimination half-lives for most of the tricyclic and tetracyclic compounds average about 24 hours or

longer; thus, the drugs can be given once a day (Table 12–2). Amoxapine has a shorter half-life

than the other tricyclics and is an exception.

TABLE 12–2. Dosage, clearance, and apparent therapeutic plasma concentrations of tricyclics and

tetracyclics

Plasma Therapeutic

Drug Half-life

(hours)

Clearance

(L/hour)

Dosage range

(mg/day)

Plasma level

(ng/mL)

Tertiary tricyclics

Amitriptyline 5–45 20–70 150–300

Clomipramine 15–60 20–120 150–300

>150a

Doxepin 10–25 40–60 150–300

Imipramine 5–30 30–100 150–300

>200a

Trimipramine 15–40 40–105

Secondary

tricyclics

Desipramine 10–30 80–170 75–300 >125

Nortriptyline 20–55 15–80 50–150 50–150

Protriptyline 55–200 5–25 15–60

Tetracyclics

Amoxapine 5–10 225–275 150–300

Maprotiline 25–50 15–35 100–225Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

7 of 33

10/05/2009 15:56

aTotal concentration of the parent compound and the desmethyl metabolite.

Source. Adapted from Nelson JC: “Tricyclic and Tetracyclic Drugs,” in Comprehensive Textbook of

Psychiatry/VII, 7th Edition. Edited by Kaplan HI, Sadock BJ. Baltimore, MD, Lippincott Williams & Wilkins,

2000, p. 2494. Copyright 2000, Lippincott Williams & Wilkins. Used with permission.

Hepatic metabolism of the tricyclics and tetracyclics occurs along two principal metabolic pathways.

Demethylation of the side chain converts the tertiary amines to secondary amines—for example,

amitriptyline is converted to nortriptyline—and the characteristics of the compound are altered. The

tertiary amines are relatively more serotonergic, whereas the demethylated amines are relatively

more noradrenergic. The other pathway in hepatic metabolism is hydroxylation of the ring

structure. Hydroxylation results in the formation of hydroxy metabolites. In some cases, the levels

of the metabolite are substantial. The concentration of 10-hydroxynortriptyline usually exceeds that

of the parent compound (Bertilsson et al. 1979). Usually 2-hydroxydesipramine is present at levels

approximately 40%–50% of those present in the parent compound, but these ratios are quite

variable, depending on the rate of hydroxylation (Bock et al. 1983; Potter et al. 1979). Thus, in

extensive metabolizers, the ratio of hydroxy metabolite to parent compound can be quite high, but

total drug levels are low. Hydroxyimipramine and hydroxyamitriptyline are present at very low

concentrations and are clinically unimportant. The hydroxy metabolites are then conjugated and

excreted. The conjugated metabolites are not active.

Hydroxynortriptyline and hydroxydesipramine both block the norepinephrine transporter

(Bertilsson et al. 1979; Potter et al. 1979). Both have been shown to have antidepressant activity

(Nelson et al. 1988b; Nordin et al. 1987). The potency of hydroxydesipramine is comparable to that

of the parent compound in terms of norepinephrine reuptake blockade. There are two isomers of

hydroxynortriptyline, E- and Z-10-hydroxynortriptyline. E-10-hydroxynortriptyline is present at

levels four times higher than those of the Z isomer and is about 50% as potent as nortriptyline in

blocking norepinephrine uptake. The clinical significance of high levels of less potent

hydroxynortriptyline is not entirely clear. In particular, it is not clear whether high levels of less

potent hydroxynortriptyline might interfere with the action of nortriptyline—a question of interest

because such an effect might explain the therapeutic window described for this drug. Both

hydroxynortriptyline and hydroxydesipramine are less anticholinergic than their parent compounds.

The hydroxy metabolites may have other effects. Early studies suggested that hydroxynortriptyline

concentrations were disproportionately associated with cardiac conduction abnormalities

(Schneider et al. 1988; Young et al. 1985), but later studies indicated that the E enantiomer of

10-hydroxynortriptyline was less cardiotoxic (Pollock et al. 1992).

The principal metabolic pathway for amoxapine is hydroxylation, during which 7-hydroxyamoxapine

and 8-hydroxyamoxapine are produced (Coupet et al. 1979). These compounds differ:

7-hydroxyamoxapine has high-potency neuroleptic properties but a short half-life;

8-hydroxyamoxapine is metabolized more slowly and appears to contribute to the drug’s

antidepressant action.

In recent years, identification of the specific isoenzyme pathways involved in the metabolism of a

variety of drugs, including the tricyclics, has been the focus of intensive study. The CYP2D6

pathway appears responsible for hydroxylation of desipramine and nortriptyline (Brosen et al.

1991). In fact, desipramine has been considered the prototypic substrate for CYP2D6 because it has

no other major metabolic pathways. Demethylation of the tertiary-amine compounds appears to

involve a number of CYP isoenzymes, including 1A2, 3A4, and 2C19. These hepatic isoenzymes are

under the control of specific genes, and the gene loci have been identified for several of these

isoenzymes, including CYP2D6. Approximately 5%–10% of Caucasians are homozygous for the

recessive autosomal 2D6 trait, resulting in deficient hydroxylation of desipramine and nortriptyline

(Brosen et al. 1985; Evans et al. 1980). These individuals are termed poor metabolizers, while

those with adequate 2D6 enzyme are referred to as extensive metabolizers. Approximately 20% of

individuals of Asian descent have a genetic polymorphism resulting in deficient CYP2C19

metabolism. This pathway is involved in the metabolism of the tertiary tricyclic compounds.Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

8 of 33

10/05/2009 15:56

The variability in plasma concentrations that results from these metabolic differences is substantial.

For example, in a sample of 83 inpatients who were given a fixed dose of 2.5 mg/kg of

desipramine, we observed steady-state plasma concentrations ranging from 20 ng/mL to 934

ng/mL (Nelson 1984). Even among extensive metabolizers, there can be variability in the rates of

metabolism, resulting in the term ultrarapid metabolizers. Various methods have been used to

phenotype the individuals who are slow or fast metabolizers. For example, formation of the

debrisoquine metabolite in the urine has been used to characterize the metabolic rate of CYP2D6

(Brosen et al. 1991; Evans et al. 1980). Recent work in this area has shifted to genotyping the

involved isoenzymes. In clinical practice, blood levels of the compounds themselves are more often

used as a crude index of the rate of metabolism.

As noted above, desipramine has often been used as a substrate for 2D6 because 2D6 is the only

major metabolic pathway for this compound. While desipramine may be useful for examination of

2D6 inhibition, it may overestimate the magnitude of drug interactive effects for those agents that

have multiple pathways.

Steady-State Concentrations

Steady state is an important pharmacological concept for clinicians to understand if drug

monitoring is employed. Steady state is that point, on a fixed dose, at which plasma concentrations

of the drug reach a plateau. Steady state is achieved after five half-lives. At this point, the

concentration of the drug should be 97% of the maximal concentration achieved for that dose. In

fact, after three half-lives, the drug will have achieved about 87% of the steady-state

concentration. If blood level monitoring is employed, a sample is drawn before the next dose is

given, usually in the morning after the patient’s level has reached a steady state. Steady-state drug

concentrations should remain relatively stable as long as the dose is constant, the patient is

compliant, and no interactive drugs are added.

The day-to-day biological variability of drug concentrations at steady state is not frequently

described. In inpatients at steady state, the coefficient of variation (SD/mean) of desipramine is

approximately 10%–15% (J. C. Nelson, unpublished data, 1985). In outpatients, this variability

may increase. This means that if the average plasma concentration is 150 ng/mL, two-thirds of

samples obtained will be ±10%, or between 135 ng/mL and 165 ng/mL. Research studies of drug

concentrations usually employ an average of two or three plasma samples to reduce the effect of

this variability. If only one sample is drawn, the clinician needs to remember that even if the

laboratory error is low, there will be moderate biological variability. Single blood levels are better

viewed as estimates than as precise measures.

When the drug concentration is measured, the total of both the free and bound drug is reported.

Few laboratories are prepared to measure free levels, yet drug concentrations in the cerebrospinal

fluid are proportional to the free levels. The free concentration is dependent on dose and hepatic

clearance but is not affected by plasma protein binding (Greenblatt et al. 1998). The latter is often

misunderstood. Factors that affect plasma proteins—malnutrition, inflammation—may lead to

changes in the bound fraction, but the absolute free concentration is unaffected. If another drug

affects binding, the absolute free concentration remains unaffected. In these instances, the free

fraction may change because the bound portion declines, not because there is a change in the free

concentration.

Linear Kinetics

Most of the tricyclics have linear kinetics; that is, concentration increases in proportion to dose

within the therapeutic range. There are exceptions. Desipramine, for example, appears to have

nonlinear kinetics in the usual dose range (Nelson and Jatlow 1987). Rapid metabolizers of

desipramine are most likely to display nonlinear changes during the time when the dose is

increasing. In these patients, a disproportionate rise in the drug concentration occurs when the

dose is increased. In cases of overdose, nonlinear changes are more likely to occur, and the

clinician cannot assume that usual rates of drug elimination will be maintained.Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

9 of 33

10/05/2009 15:56

Effects of Aging

Many changes in the pharmacodynamics and pharmacokinetics of drug treatment occur with aging,

yet some may be relatively unimportant (Greenblatt et al. 1998). The ratio of fat to lean body mass

increases, and cardiac output and hepatic blood flow decrease. There may be further changes

associated with medical illness. But the clinical importance of these changes is usually relatively

minor because of the dramatic variability of hepatic metabolism. Age-related changes in

metabolism vary with the isoenzymes involved. The activity of the CYP3A4 pathway does slow with

age (von Moltke et al. 1995). Most studies of the tertiary amines, such as imipramine, suggest that

concentrations of these drugs are increased somewhat in older individuals (Abernathy et al. 1985;

Benetello et al. 1990; Furlanut and Benetello 1990). Alternatively, most studies of nortriptyline

(Bertilsson 1979; Katz et al. 1989; Smith et al. 1980; Young et al. 1984; Ziegler and Biggs 1977)

and desipramine (Abernathy et al. 1985; Nelson et al. 1985, 1995) indicate that ratios of blood level

to dosage of these drugs are relatively unaffected by aging, suggesting that the 2D6 isoenzyme is

not similarly affected. In addition, the relationship of nortriptyline and desipramine plasma

concentrations to therapeutic effects appears to be relatively similar in younger and older adults

(Katz et al. 1989; Nelson et al. 1985, 1995; Young et al. 1988). Renal clearance of the hydroxy

metabolites does decrease with age (Nelson et al. 1988a; Young et al. 1984). As a result,

concentrations of hydroxynortriptyline may be substantially elevated in older patients.

In children, the clearance of tricyclic compounds is increased. Half-lives of imipramine are shorter

and ratios of desmethylimipramine to imipramine are higher, consistent with more rapid

metabolism (Geller 1991; Rapoport and Potter 1981). Alternatively, a study of desipramine in

children found that the clearance of both desipramine and hydroxydesipramine was increased so

that hydroxy metabolite–parent compound ratios were not elevated (Wilens et al. 1992).

Relationship of Plasma Concentration to Clinical Action

Plasma Concentration and Response

Marked interindividual variability of tricyclic plasma concentrations was described by Hammer and

Sjöqvist in 1967. This finding suggested that drug level monitoring might ensure that therapeutic

blood levels are achieved and might help to avoid toxic levels. In carefully selected inpatients with

endogenous or melancholic major depression, treatment with adequate levels of imipramine or

desipramine resulted in robust response rates of about 85% (Glassman et al. 1977; Nelson et al.

1982). For several years, the relationship of tricyclic blood levels to response and the utility of

monitoring blood levels were the focus of considerable attention and debate.

A task force of the American Psychiatric Association (1985) that reviewed these studies concluded

that relationships between plasma level and response had been demonstrated for imipramine,

desipramine, and nortriptyline (see Table 12–2). For imipramine, drug levels above 200 ng/mL

were more effective than lower levels (Glassman et al. 1977; Reisby et al. 1977). For desipramine,

levels above 125 ng/mL were more effective (Nelson et al. 1982). For both desipramine and

imipramine, blood levels in excess of 300 ng/mL were more likely to be associated with serious

side effects. Effective plasma concentrations were also established for nortriptyline, but the

relationship appeared to be curvilinear. For this drug, plasma levels between 50 ng/mL and 150

ng/mL were more effective than lower or higher levels (Åsberg et al. 1971; Kragh-Sørenson et al.

1973, 1976).

For amitriptyline, it has been more difficult to establish a therapeutic relationship between plasma

levels and response (American Psychiatric Association 1985). In part, this difficulty may be related

to the fact that during amitriptyline administration, three active compounds are present

(amitriptyline, nortriptyline, and hydroxynortriptyline), and it is unclear if their effects are additive

or if there is a more complicated relationship (Breyer-Pfaff et al. 1982). During amitriptyline

administration, responders usually have total amitriptyline and nortriptyline levels in the

neighborhood of 150–250 ng/mL (Kupfer et al. 1977), but there is not good agreement between

studies. For clomipramine, blood levels of 150–300 ng/mL (total of clomipramine andPrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

10 of 33

10/05/2009 15:56

desmethylclomipramine) have been suggested for antidepressant effectiveness. Higher levels are

usually employed in the treatment of OCD. The data relating blood levels and response are limited

for the other tricyclic and tetracyclic compounds.

The therapeutic utility of blood level monitoring has been the subject of controversy. Blood

level–response relationships have been demonstrated in melancholic inpatients. But similar

relationships have proven difficult to demonstrate in depressed outpatients. In outpatients,

drug–placebo differences are often small, and the effect of drug treatment is harder to detect.

Depressed outpatients may be more heterogeneous and include individuals who are not responsive

to any drug treatment. Finally, many studies are not designed to detect blood level–response

relationships. Fixed dosing is required, and the plasma concentrations achieved must fall above and

below the suspected threshold. If all patients achieve adequate drug concentrations, no

relationship with response will be found. It is logical to conclude that blood level relationships

determined in severely depressed inpatients might be used as a guide for treatment of outpatients,

but this assumption has not been empirically validated.

Plasma Concentration and Toxicity

The alternative question is whether blood level monitoring might help to avoid toxicity. A variety of

data support this view. The risk of delirium is substantially increased at amitriptyline plasma

concentrations above 450 ng/mL and is moderately increased at concentrations above 300 ng/mL

(Livingston et al. 1983; Preskorn and Simpson 1982). But amitriptyline is the most anticholinergic

tricyclic and is most likely to produce delirium. The risk of first-degree atrioventricular block is also

increased with plasma concentrations of imipramine greater than 350 ng/mL (Preskorn and Irwin

1982). The risk of seizures also increases at higher doses and, presumably, higher blood levels,

although a clear plasma-level threshold for seizures has not been demonstrated. Following

overdose, tricyclic blood levels greater than 1,000 ng/mL can be achieved, and the risks of

delirium, stupor, cardiac abnormalities, and seizures are all substantially increased (Preskorn and

Irwin 1982; Rudorfer and Young 1980; Spiker et al. 1975).

The value of blood level monitoring to avoid serious adverse effects has been hard to demonstrate

because rates of serious toxicity are low so that large samples are required to demonstrate any

increase in risk at higher blood levels. For some adverse reactions (e.g., delirium), early warning

signs may prompt dose reduction. Alternatively, there may be no warning for seizures or cardiac

arrhythmia, and blood level monitoring might be most useful for reducing the risk of those adverse

events.

If blood level monitoring is undertaken, the clinician needs to remember that the patient needs to

be at steady state, the blood sample should be drawn before the next dose (a trough level), and the

sample should be sent to the laboratory promptly. For a quantitative estimate the laboratory will

usually employ high-performance liquid chromatography (HPLC). In a competent laboratory, the

coefficient of variation for an HPLC assay is usually less than 10%. This assay is relatively specific;

however, other drugs can interfere. Because there are many modifications of the HPLC technique,

the interfering drugs will vary by site. Under the best of circumstances, there will still be biological

variability of the compound (discussed earlier in subsection “Steady-State Concentrations”). Add to

this occasional missed doses or laboratory problems, and there will be considerable

sample-to-sample variability. For these reasons, the clinician should not view the concentration

reported as a precise measure. Yet, because concentrations vary across such a wide range, it may

be very helpful to know if the level is low (e.g., 25–75 ng/mL), moderate (e.g., 100–300 ng/mL), or

high (e.g., 300–1,000 ng/mL).

Prospective Dosing Techniques

Conventional dosing requires administration of a given dose for a long enough period of time to

determine whether that dose is effective. Sometimes two or three trials are needed to determine

the effective dose. The possibility of rapid dosage adjustment using plasma levels was suggested

when several investigators demonstrated a relationship between an initial timed blood sample andPrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

11 of 33

10/05/2009 15:56

the final steady-state level (Alexanderson 1972; Brunswick et al. 1979; Cooper and Simpson 1978;

Potter et al. 1980). This method appeared most applicable for nortriptyline, which has linear

kinetics, but was also useful for desipramine, because the targeted level was within a broad range.

Clinical studies using blood levels to adjust dose were reported for amitriptyline (Dawling et al.

1984; Madakasira and Khazanie 1985), but sedation and anticholinergic side effects limited the rate

at which the dose could be adjusted. Alternatively, a clinical study using rapid dosing of

desipramine found that treatment could be initiated at full dose once the dose needed to reach a

therapeutic level was determined from a 24-hour blood level following a test dose (Nelson et al.

1987). Sixteen of the 18 patients who completed treatment had plasma levels within the targeted

range, and side effects appeared to be no greater than those experienced with more gradual

dosing.

The practical application of these methods was limited by laboratory issues. The laboratory

performing the assay had to be prepared to determine drug concentrations accurately at very low

levels (below the therapeutic range) and had to be able to report results quickly. Most labs were

not prepared to do either. A more practical and clinically feasible method is to start the drug at a

low or moderate fixed dose, obtain a blood sample after 5–7 days on that dose, and then make

further adjustments based on that level.

There are exceptions. Elderly depressed patients often require gradual dosing in order to assess

tolerance. In panic patients, lower starting doses are employed to avoid exacerbation of panic

attacks.

MECHANISM OF ACTION

Early biochemical theories of depression were in large part based on the knowledge of drug action.

The observation that the tricyclic agents increased the availability of norepinephrine and serotonin

suggested that depression resulted from a deficit in these neurotransmitters (Bunney and Davis

1965; Prange 1965; Schildkraut 1965). This work stimulated interest in the role of these

neurotransmitters in the etiology of depression, and several abnormalities were identified. Yet, it

remains unclear which, if any, of these abnormalities play a central role in causing depression or

are responsible for the vulnerability to becoming depressed.

Recent challenge studies in depressed patients do confirm that the actions of antidepressant drugs

are mediated by serotonin and norepinephrine. For example, administration of a tryptophan-free

diet rapidly depletes serotonin and, in depressed patients who have been successfully treated,

causes relapse (Delgado et al. 1990). In addition, tryptophan depletion caused relapse in patients

who were treated with serotonergic agents, whereas those who were treated with norepinephrine

reuptake inhibitors were relatively unaffected. Alternatively, administration of AMPT, which

interrupts the synthesis of catecholamines, caused relapse in patients who were being successfully

treated with noradrenergic agents but not those receiving serotonergic drugs (Delgado et al. 1993).

Tryptophan depletion in untreated depressed patients, however, had no effect on the patients’

depression. These studies provide supporting evidence that serotonin and norepinephrine mediate

antidepressant effects, but they do not necessarily imply that alterations in these neurotransmitter

systems are central to the pathophysiology of depression.

The synaptic effects of tricyclic and tetracyclic agents on norepinephrine and serotonin transporters

and receptors were described in detail earlier (see section “Pharmacological Profile” earlier in this

chapter).

The early theories of depression that focused on depletion of norepinephrine or serotonin

suggested that it might be possible to identify “serotonergic” and “noradrenergic” depressions and

that such identification would help the clinician select the appropriate type of antidepressant

(Beckmann and Goodwin 1975; Maas et al. 1972). A number of studies investigated the predictive

value of urinary MHPG, a metabolite of norepinephrine, but a definite predictive link with a

noradrenergic antidepressant was not established. Some of these studies were, in part, hampered

by the use of agents such as amitriptyline and imipramine, which are not very selective. However,Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

12 of 33

10/05/2009 15:56

even those studies that examined the ability of MHPG to predict response to more selective agents,

such as zimelidine, fluoxetine, and desipramine, failed to demonstrate clear predictive utility

(Bowden et al. 1993; Potter 1984). The data, taken together, suggest that urinary MHPG is not a

clinically useful predictor. Nevertheless, these studies do not rule out the possibility that there are

depressions in which serotonin or norepinephrine plays a relatively more prominent role.

More recently, research into the mechanism of action of the tricyclics and other antidepressant

drugs has shifted to include consideration of factors affecting postsynaptic signal transduction

(Manji et al. 1995). These factors include coupling of G proteins to the adrenergic receptor or to

adenylyl cyclase and the activity of membrane phospholipases and protein kinases. Other novel

targets, including glucocorticoid receptors (Barden 1996), neurotrophic factors (Duman et al.

1997), and gene expression (Lesch and Manji 1992; Nibuya et al. 1996; Schwaninger et al. 1995),

have been explored.

INDICATIONS AND EFFICACY

Major Depression

The efficacy of the tricyclic and tetracyclic compounds in major depression is well established. The

evidence for their effectiveness has been reviewed previously (Agency for Health Care Policy and

Research 1993; Davis and Glassman 1989). Imipramine is the most extensively studied tricyclic

antidepressant, in part because for many years it was the standard agent against which other new

drugs were compared. In 30 of 44 placebo-controlled studies, imipramine was more effective than

placebo. If data from these studies are combined, 65% of 1,334 patients completing treatment with

imipramine were substantially improved, whereas 30% of those on placebo improved.

Intention-to-treat response rates for placebo-controlled studies of imipramine in outpatients were

51% for imipramine and 30% for placebo (Agency for Health Care Policy and Research 1993). In

most comparison studies, the other tricyclic and tetracyclic antidepressants have been found to be

comparable to imipramine in efficacy.

The tricyclic compounds are also effective when used for maintenance treatment. Early studies

demonstrated that maintenance treatment with a tricyclic would reduce the relapse rate associated

with placebo by about 50% (Davis 1976). These studies, however, usually employed low doses.

Subsequently, the Pittsburgh group found that imipramine, at full dose, effectively maintained

nearly 80% of the depressed patients for a 3-year period compared with 10% of those on placebo

(Frank et al. 1990). In this study, maintenance psychotherapy had an intermediate effect, with

about 30% of the patients remaining well. Although this seminal study demonstrated the

impressive value of maintenance treatment with full-dose imipramine, the magnitude of the

findings may reflect characteristics of the sample treated. The sample comprised patients with

recurrent depression who might have been expected to do poorly on placebo. In addition, the

patients selected for the study had a history of symptom-free periods between prior episodes,

suggesting that these patients might be more likely (than patients with a history of residual

symptoms) to have a complete response to treatment. In practice, clinicians may encounter

patients with chronic depression, patients with residual symptoms, or patients with comorbid

medical and psychiatric disorders. For such patients, drug treatment may be more effective than

placebo, but the actual number of patients whose depression remains in remission may be lower.

The U.S. Food and Drug Administration (FDA) has approved all of the tricyclic and tetracyclic

compounds discussed in this chapter for the treatment of depression with the exception of

clomipramine. In Europe, clomipramine is also used for depression; in fact, it is regarded by many

as the most potent antidepressant.

Melancholia or Severe Depression

The efficacy of the tricyclic compounds appears to vary in different subtypes of depression. Four

decades ago, when Kuhn studied imipramine, the prevailing view was that it was essential to

establish efficacy of an antidepressant in patients with endogenous depression or at least in those

with severe depression (Kuhn 1970). The early studies of imipramine and the other tricyclicPrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

13 of 33

10/05/2009 15:56

compounds were frequently conducted in hospitalized patients with severe or endogenous

depression, and in these patients the tricyclics were found to be effective. In fact, these agents may

be especially effective in this group. Two studies of imipramine and desipramine found rates of

response of about 85% in severely depressed hospitalized patients who did not have a refractory

history, did not have prominent personality disorder, received an adequate plasma concentration of

the drug, and completed treatment (Glassman et al. 1977; Nelson et al. 1982).

When the SSRIs were introduced, it was suggested that they might be less effective than the

tricyclic antidepressants in treating severe or melancholic depression. In a large meta-analysis of

more than 100 studies comparing tricyclic antidepressants and SSRIs, Anderson (2000) found that

in general, these agents had comparable efficacy. When individual agents and patient

characteristics were considered, the only tricyclic agent that appeared to be more effective than the

SSRIs was amitriptyline, and the only patient characteristic was inpatient status. In a separate

meta-analysis of 25 inpatient studies (Anderson 1998), the advantage of the tricyclics appeared

limited to those with dual action, namely amitriptyline and clomipramine. Two of the most

frequently cited studies in this regard were the two Danish University Antidepressant Group (1986,

1990) studies that found clomipramine to be more effective than paroxetine or citalopram in

severely depressed inpatients. Recent clinical trials of antidepressants were usually conducted in

outpatients with depression of moderate severity. In outpatients, the designation of melancholia

does not appear to predict an advantage for tricyclic antidepressants versus SSRIs (Anderson and

Tomenson 1994; Montgomery 1989). Although the question of the best treatment for severely

melancholic inpatients lingers, the expense of these studies, and the shift in practice to treatment

of depression in outpatient settings, has substantially decreased interest in this question.

Anxious Depression

Anxious depression is not recognized in DSM-IV-TR (American Psychiatric Association 2000) as a

subtype of depression; nevertheless, it has been frequently studied. Three of the tricyclic and

tetracyclic compounds—doxepin, amoxapine, and maprotiline—have received FDA approval for use

in patients with depression and symptoms of anxiety. For many years, clinical lore suggested that

amitriptyline was most effective for anxious depression. Direct comparison studies, however, have

found little indication that one of these compounds is better than another for treatment of anxious

depression. Depressed patients who are anxious may respond less well than less anxious patients.

This has been observed with amitriptyline (Kupfer and Spiker 1981), imipramine (Roose et al.

1986), and desipramine (Nelson et al. 1994). Yet these drugs are still more effective than placebo

in anxious depressed patients, and it is not established that other classes of antidepressants are

more effective in these patients.

Atypical Depression

A series of studies by the Columbia University Group examined the efficacy of imipramine in

depressed patients with atypical features (Liebowitz et al. 1984, 1988). These depressed patients

had reactive mood and reversed vegetative symptoms, severe fatigue, or rejection sensitivity.

Imipramine was more effective than placebo but significantly less effective than the monoamine

oxidase inhibitor (MAOI) phenelzine. Other investigators have reported the value of switching from

a tricyclic to an MAOI in tricyclic-refractory depressed patients, especially those with atypical

features (McGrath et al. 1987; Thase et al. 1992). In fact, the validity and utility of the atypical

subtype of depression were in large part supported by this observed difference. However, this

subtype has not been shown to be preferentially responsive to SSRIs (Fava et al. 1997), nor has

any second-generation antidepressant been shown to be superior to any other in treating atypical

depression.

Psychotic Depression

In 1975, Glassman et al. observed that imipramine was less effective in patients with major

depression who had delusions. Later, Chan et al. (1987), in reviewing several studies addressing

this question and involving more than 1,000 patients, found that antidepressants—usuallyPrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

14 of 33

10/05/2009 15:56

tricyclics—given alone were effective in approximately two-thirds of the nonpsychotic patients but

only about one-third of those with psychotic features. Although the definition of psychosis has

undergone several changes, currently it is defined in DSM-IV-TR as major depression with delusions

or hallucinations. Several open studies reviewed elsewhere (Nelson 1987) and one prospective

study (Spiker et al. 1985) found that the tricyclics, when combined with an antipsychotic, are

effective in psychotic depression.

Anton and Burch (1990) suggested that because of its antipsychotic effects, amoxapine might be

effective for psychotic depression. In a double-blind study, these researchers demonstrated that

amoxapine was comparable in efficacy to the combination of perphenazine and amitriptyline in

treating psychotic depression (Anton and Burch 1990).

Bipolar Depression

Thirty years ago, it was suggested that the MAOI antidepressants might be more effective than the

tricyclics in treating bipolar depression (Himmelhoch et al. 1972). Later, Himmelhoch et al. (1991)

demonstrated in a double-blind study that tranylcypromine was more effective than imipramine for

bipolar depression. In addition, tricyclics are more likely than other agents to induce mania (Weir

and Goodwin 1987). As a result, the tricyclics are not recommended for monotherapy of bipolar

depression.

Chronic Major Depression and Dysthymia

Imipramine appears to be effective in treating chronic depression and dysthymia and to be

relatively comparable to sertraline in efficacy (Keller et al. 1998; Kocsis et al. 1988; Thase et al.

1996). Imipramine and desipramine have both been studied in controlled trials and have been

found to be more effective than placebo both for acute treatment and for maintenance treatment

(N. L. Miller et al. 2001).

Late-Life Depression

Gerson et al. (1988) reviewed the studies of tricyclic antidepressants reported prior to 1986. They

found 13 placebo-controlled trials but noted several problems, such as lack of diagnostic criteria,

inclusion of younger patients, and dosing issues. Although tricyclics were effective, overall drug and

placebo response rates in these older patients appeared to be lower than rates in nonelderly

patients (Agency for Health Care Policy and Research 1993). Katz et al. (1990) performed one of

the first placebo-controlled trials of nortriptyline in the treatment of patients older than 80 years

living in a residential care facility. Nortriptyline was more effective than placebo. The doses

employed and levels achieved were similar to those in younger subjects. This study remains the

only study to date showing an advantage for an antidepressant over placebo in depressed patients

older than 75 years.

Depression in Children

In children and adolescents, the tricyclic antidepressants have not demonstrated superiority over

placebo (Ryan 1992).

Obsessive-Compulsive Disorder

Unlike depression, which responds to a variety of antidepressant agents, OCD appears to require

treatment with a serotonergic agent. Clomipramine, the most serotonergic of the tricyclics, is

approved by the FDA for use in OCD, and its efficacy in this disorder is well established (Greist et al.

1995). Studies comparing its effectiveness with noradrenergic agents such as desipramine found

that clomipramine was substantially superior (Leonard et al. 1989). Although the SSRIs are

effective in treating OCD, there is a suggestion that clomipramine may be superior (Greist et al.

1995). Whether this putative superiority is due to the dual mechanism of clomipramine or to other

factors is unclear.

Panic DisorderPrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

15 of 33

10/05/2009 15:56

None of the tricyclic or tetracyclic drugs are approved for use in panic disorder. Yet imipramine was

the first drug described for use in this disorder (Klein 1964). In fact, observation of the effects of

imipramine helped to establish the diagnostic utility of panic disorder. The efficacy of both tertiary

and secondary tricyclics has been demonstrated in controlled trials (Jobson et al. 1978; Munjack et

1. 1988; Zitrin et al. 1980). In treating this disorder, the drug is initiated at a low dose to avoid

exacerbation of panic symptoms.

Attention-Deficit/Hyperactivity Disorder

The efficacy of the stimulant drugs in treating attention-deficit/hyperactivity disorder (ADHD) is

well established. The tricyclics, especially desipramine, also appear to be of value. In one study,

desipramine, given at doses greater than 4 mg/kg for 3–4 weeks, was effective in two-thirds of the

children, whereas placebo was effective in only 10% (Biederman et al. 1989). Desipramine was

also found to be more effective than placebo in adults with ADHD (Wilens et al. 1996). One of the

advantages of desipramine is its low potential for abuse. Unfortunately, five cases of sudden death

were reported in the early 1990s in children being treated with desipramine (Riddle et al. 1991,

1993). All were under the age of 12 years. As a result, desipramine is now contraindicated in

children younger than 12 years (discussed in greater detail below; see section “Side Effects and

Toxicology”). Given that tricyclics as a group share the same adverse cardiac effects, there is

reason to be concerned that other tricyclics might also have safety issues in young children.

Pain Syndromes

The tricyclics and maprotiline have been widely used in various chronic pain syndromes. In a

review of the literature, O’Malley et al. (1999) identified 56 controlled studies involving tricyclic

antidepressant therapy for various pain syndromes, including headache (21 studies), fibromyalgia

(18 studies), functional gastrointestinal syndromes (11 studies), idiopathic pain (8 studies), and

tinnitus (2 studies), and Salerno et al. (2002) identified 7 more placebo-controlled trials of

tricyclics or maprotiline used for chronic back pain. These agents were quite effective; in fact, the

mean effect size (0.87) and the drug–placebo difference in response rates (32%) observed in pain

syndromes are more robust than those usually observed in placebo-controlled studies in

depression. In studies in which depression was also assessed, improvement in pain appeared to be

independent of improvement in depression. Thus, the analgesic effects of these compounds were

not simply the result of their antidepressant effects.

The mechanism of these agents’ analgesic effects appears to differ from that of their

antidepressant effects. The antinociceptive actions of the antidepressants result from actions on

descending norepinephrine and serotonin pathways in the spinal cord (Yoshimura and Furue 2006).

In animals, norepinephrine reuptake inhibitors and combined norepinephrine–serotonin reuptake

inhibitors appear to be more potent than SSRIs (Mochizucki 2004). In humans, there is some

evidence that the combined-action agents amitriptyline and clomipramine are more effective than

the SSRI fluoxetine (Max et al. 1992) or the norepinephrine-selective agents maprotiline (Eberhard

et al. 1988) and nortriptyline (Panerai et al. 1990). In humans, antidepressant dosing and timing of

effects for pain differ from those observed in depression. For example, usual dosages of

amitriptyline required for pain management ( 75 mg/day) are lower than those required to treat

depression (15–300 mg/day), and response occurs more quickly, usually within the first 1 or 2

weeks.

Other Indications

Imipramine has been used for treatment of nocturnal enuresis in children with FDA approval, and

controlled trials indicate that it is clearly effective (Rapoport et al. 1980). The dose of imipramine is

usually 25–50 mg at bedtime. Amitriptyline and nortriptyline also appear to be useful, although

they are not approved for use in this disorder. The mechanism of action is unclear but may in part

be anticholinergic. It is not clear, however, that the risk of cardiac problems would be substantially

less with tricyclics other than desipramine in children younger than 12 years, although the low

doses required may reduce this risk.Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

16 of 33

10/05/2009 15:56

Tricyclic antidepressant drugs have been extensively studied in patients with schizophrenia.

However, in the absence of a major depressive syndrome, these agents appear to be of limited

value (Siris et al. 1978).

SIDE EFFECTS AND TOXICOLOGY

The delineation of side effects during the treatment of depressed patients is complicated because

depression itself is accompanied by a variety of somatic symptoms. For example, headache,

constipation, and drowsiness—symptoms usually considered as “side effects”—have been observed

in more than 50% of untreated inpatients with major depression if these symptoms were each

directly assessed (Nelson et al. 1984). During treatment, patients may be quick to label these

somatic symptoms as side effects even if the symptoms were preexisting. Another manifestation of

this issue is the rate of spontaneously reported “side effects” on placebo in clinical trials. One of

the best examples is headache. Clinical trial data for recently marketed antidepressants indicate

that the rate of headaches on placebo in depressed outpatients ranges from 17% to 24%

(Physicians’ Desk Reference 2002). For fluoxetine, sertraline, paroxetine, and bupropion, the rate

for drug was only 1%–2% higher than that for placebo. For venlafaxine XR and citalopram, the rate

of headaches was higher for placebo than for drug. A strong argument can be made that headache

is usually a symptom of depression. Of course, these mean values conceal the possibility that a

symptom may worsen or emerge during treatment in some patients and improve with treatment in

others. In groups of patients, however, the strongest predictor of overall somatic symptom severity

is the severity of the depression at the time of assessment (Nelson et al. 1984), and the best

intervention may be more aggressive treatment.

Another general factor contributing to side effects is the patient’s vulnerability. For example, one of

the best predictors of orthostatic hypotension during treatment is the presence of orthostatic

hypotension prior to treatment (Glassman et al. 1979). Seizures are most likely in a patient with a

history of seizures (Rosenstein et al. 1993). Cardiac conduction problems are most likely to occur in

patients with preexisting conduction delay (Roose et al. 1987a).

The final manifestation of somatic symptoms during treatment is the net result of the interaction of

direct effects of the medication on specific organs, the indirect effects of the medication on

depression and its associated somatic symptoms, and the patient’s vulnerability to certain

symptoms. The attribution of cause—that is, whether a physical symptom is a “side effect” of a drug

or a symptom of depression—involves a judgment about whether the symptom is new or has

worsened during drug treatment.

Antidepressant drugs do, of course, have direct effects on a variety of organs and can produce

adverse effects. The in vitro potency or affinity of antidepressant compounds for various receptor

sites (see Table 12–1) is one method for comparing the likelihood that various agents will produce

specific side effects. A related issue is how the in vitro potency of a secondary effect relates to the

potency of the primary action of the drug. If the secondary effect is more potent, it will occur at

concentrations below the therapeutic level of the drug. An example for the tricyclics is orthostatic

hypotension, which often manifests at plasma concentrations below the usual antidepressant

threshold. Alternatively, in patients without preexisting medical illness, the proarrhythmic and

proconvulsant effects of the tricyclic antidepressants are uncommon at therapeutic concentrations

but become more frequent at levels encountered in overdose.

Central Nervous System Effects

The principal action of the tricyclic and tetracyclic agents in the central nervous system is to

alleviate depression. In particular, they reduce the symptoms of depression rather than simply

elevating mood. Nondepressed subjects given imipramine may feel sleepy, quieter, light-headed,

clumsy, and tired. These effects are generally unpleasant (DiMascio et al. 1964).

The anticholinergic and antihistaminic effects of the tricyclics and tetracyclics can produce

confusion or delirium. The incidence of delirium is dose-dependent and increases at blood levels

above 300 ng/mL. One study reported that 67% of patients with blood levels above 450 ng/mLPrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

17 of 33

10/05/2009 15:56

developed delirium when receiving the tertiary amines, particularly amitriptyline (Livingston et al.

1983; Preskorn and Simpson 1982). The clinician should be alert to the possibility of delirium in a

patient whose depression is worsening during treatment. This can be especially problematic in

patients with psychotic depression. Patients with concurrent dementia are particularly vulnerable

to the development of delirium, and the more anticholinergic tricyclics should be avoided in these

patients. Intramuscular or intravenous physostigmine can be used to reverse or reduce the

symptoms of delirium. Although physostigmine may represent a useful diagnostic test, its short

duration of action makes the continued use of this agent difficult.

Seizures can occur with all of the tricyclic and tetracyclic agents and are dosage- and blood

level–related (Rosenstein et al. 1993). For clomipramine, the risk for seizures is reported to be

0.5% at dosages up to 250 mg/day. At dosages above 250 mg/day, the seizure risk increases to

1.67% (new drug application data on file with the FDA). For maprotiline, the overall risk of seizures

is reported to be 0.4%, but, again, this risk increases at dosages above the maximum

recommended dose of 225 mg/day (Dessain et al. 1986). The seizure risk for some of the older

compounds was not as well established at the time of marketing. A retrospective meta-analysis of

imipramine found an estimated seizure rate of 1 per 1,000 patients receiving less than 200 mg/day

(Peck et al. 1983). At dosages above 200 mg/day, the rate was 0.6%. Another large review (Jick et

1. 1983) found similar dose-dependent rates for amitriptyline and doxepin. Rates of 1%–4% have

been reported at doses between 250 mg/day and 450 mg/day, but the samples in these studies

were often small, and the confidence intervals for these rates were large. Consistent with a

dose-dependent effect, the risk of seizures is substantially increased following overdose (Spiker et

1. 1975). Seizure rates for the secondary amines have not been well described. Because the risk of

convulsions is clearly increased in patients with predisposing factors such as a prior history of

seizures, brain injury, or presence of neuroleptics; because the rates are low (e.g., 1/100); and

because sample size in drug trials is often on the order of 200 patients, inclusion of a few patients

who have a significant vulnerability to seizures can have a marked effect on the rate of seizures

reported. The mechanism by which tricyclics produce seizures is not well understood. It has been

suggested that antidepressant drugs induce convulsions by acting at the -aminobutyric acid

(GABA) receptor chloride–ionophore complex, where they inhibit chloride conductance (Escorihuela

et al. 1989; Malatynska et al. 1988).

A fine, rapid tremor can occur with use of tricyclic agents. Because this tremor is dose dependent,

tends to occur at higher levels, and is not a typical depressive symptom, development of a tremor

may be a clinical indicator of an elevated blood level (Nelson et al. 1984). Dose reduction will often

lead to improvement in the tremor.

Because the 7-hydroxy metabolite of amoxapine has neuroleptic properties, administration of

amoxapine carries the potential risk of neuroleptic malignant syndrome, which has been reported

(Lesaca 1987; Madakasira 1989; Taylor and Schwartz 1988; Washington et al. 1989), and tardive

dyskinesia. The occurrence of these adverse events is rare; however, the seriousness of the risk

and the availability of many alternatives suggest that use of amoxapine should be reserved for

patients whose clinical condition warrants the use of an agent with antipsychotic properties.

Anticholinergic Effects

The tricyclics block muscarinic receptors and can cause a variety of anticholinergic side effects,

such as dry mouth, constipation, blurred vision, and urinary hesitancy. These effects can precipitate

an ocular crisis in patients with narrow-angle glaucoma. The tricyclic and tetracyclic compounds

vary substantially in their muscarinic potency (see Table 12–1). Amitriptyline is the most potent,

followed by clomipramine. Of the tricyclics, desipramine is the least anticholinergic. Amoxapine and

maprotiline also have minimal anticholinergic effects. Anticholinergic effects can contribute to

tachycardia, but tachycardia also occurs as a result of stimulation of -adrenergic receptors in the

heart. Thus, tachycardia regularly occurs in patients receiving desipramine, which is minimally

anticholinergic (Rosenstein and Nelson 1991).

Although anticholinergic effects may be annoying, they are usually not serious. They can, however,Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

18 of 33

10/05/2009 15:56

become severe. An ocular crisis in patients with narrow-angle glaucoma is an acute condition

associated with severe pain. Urinary retention can be associated with stretch injuries to the

bladder. Constipation can progress to severe obstipation. (Paralytic ileus has been described but is

rare.) In these conditions, medication must be discontinued and appropriate supportive measures

instituted. Elderly patients are at greatest risk for severe adverse consequences. The frequency of

severe anticholinergic adverse reactions is increased by concomitant neuroleptic administration.

Use of nortriptyline or desipramine, either of which is less anticholinergic, can help to reduce the

likelihood of these problems.

Anticholinergic effects may benefit from other interventions. Bethanechol (Urecholine) at a dosage

of 25 mg three or four times a day may be helpful in patients with urinary hesitancy. The regular

use of stool softeners helps to manage constipation. Patients with narrow-angle glaucoma who are

receiving pilocarpine eyedrops regularly can be treated with a tricyclic, as can those who have had

an iridectomy. Tricyclic agents do not affect patients with chronic open-angle glaucoma.

Antihistaminic Effects

Several of the tricyclic compounds and maprotiline have clinically significant antihistaminic effects.

Doxepin is the most potent H1 receptor blocker in this group. It is more potent than the commonly

administered antihistamine diphenhydramine. More recently, however, it has been surpassed by

mirtazapine and olanzapine, which are even more potent antihistamines. Central H1 receptor

blockade can contribute to sedation and delirium and also appears to be related to the increased

appetite and associated weight gain that patients may develop with chronic treatment. Because of

their sedating effects, the tricyclic antidepressants, especially amitriptyline, have been used as

hypnotics. Given their cardiac effects and the frequency of lethal overdose, this practice should be

discouraged.

Cardiovascular Effects

Orthostatic hypotension is one of the most common reasons for discontinuation of tricyclic

antidepressant treatment (Glassman et al. 1979). It can occur with all of the tricyclics but appears

to be less pronounced with nortriptyline (Roose et al. 1981; Thayssen et al. 1981). The

1-adrenergic blockade associated with the tricyclics contributes to orthostatic hypotension;

however, it is the postural reflex that is primarily affected. Resting supine blood pressure may be

unaffected or can even be elevated (Walsh et al. 1992). Orthostatic hypotension is most likely to

occur or is most severe in patients who have preexisting orthostatic hypotension (Glassman et al.

1979). It is also aggravated by concurrent antihypertensive medications, especially

volume-depleting diuretic agents. The elderly are more likely to have preexisting hypotension and

are also more vulnerable to the consequences of orthostatic hypotension, such as falls and hip

fractures.

Often, orthostatic hypotension occurs at low blood levels, so that dosage reduction is not a helpful

management strategy. Gradual dose adjustment may allow accommodation to the subjective

experience of light-headedness, but the actual orthostatic blood pressure changes do not

accommodate within a reasonable period of time (e.g., 4 weeks) (Roose et al. 1998). Thus, unless

the plasma level is elevated and the dose can be reduced, patients who experience serious

symptomatic orthostatic hypotension may not be treatable with a tricyclic antidepressant.

Fludrocortisone (Florinef) has been used to raise blood pressure, but in this author’s experience it

is not very effective. If patients are receiving antihypertensives, it may be possible and helpful to

reduce the dose of these agents.

Desipramine has been reported to raise supine blood pressure in younger patients, although it is

not clear this effect is limited to that age group (Walsh et al. 1992). This effect may be similar to

that reported for venlafaxine.

Tachycardia occurs with all the tricyclics, not just the more anticholinergic agents. Both supine and

postural pulse changes can occur, and the standing pulse can be markedly elevated. A relatively

recent study of nortriptyline, dosed to a therapeutic plasma concentration, found a mean pulse risePrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

19 of 33

10/05/2009 15:56

of 11% (8 beats per minute) (Roose et al. 1998). Patients do not accommodate to the pulse rise,

which can persist for months. Tachycardia is more prominent in younger patients, who appear more

sensitive to sympathomimetic effects, and is one of the most common reasons for drug

discontinuation in adolescents. A persistent pulse rise in older patients, however, increases cardiac

work and may be clinically significant in patients with ischemic heart disease.

The effect of tricyclic antidepressants on cardiac conduction has been a subject of great interest.

Cardiac arrhythmia is the principal cause of death following overdose (Biggs et al. 1977; Pimentel

and Trommer 1994; Spiker et al. 1975). As a result of this observation, for many years there was

great concern about the use of tricyclic antidepressants in patients with and without heart disease.

The effect of these agents has now been well described. Apparently, through inhibition of

Na+/K+-ATPase, the tricyclics stabilize electrically excitable membranes and delay conduction,

particularly His ventricular conduction. Consequently, the tricyclics have type I antiarrhythmic

qualities or quinidine-like effects.

At therapeutic blood levels, the tricyclics can have beneficial effects on ventricular excitability. In

patients with preexisting conduction delay, however, the tricyclic antidepressants can further delay

conduction and cause heart block (Glassman and Bigger 1981; Roose et al. 1987b). A pretreatment

QTc interval of 450 milliseconds or greater indicates that conduction is already delayed, that a

tricyclic may aggravate this condition, and that the patient is not a candidate for tricyclic

antidepressant treatment. High drug plasma levels further increase the risk of cardiac toxicity. For

example, first-degree atrioventricular heart block is increased with imipramine plasma

concentrations above 350 ng/mL (Preskorn and Irwin 1982).

The tricyclic antidepressants do not reduce cardiac contractility or cardiac output (Hartling et al.

1987; Roose et al. 1987a). Studies using radionuclide angiography indicate no adverse effect of

imipramine or doxepin on cardiac output, even in patients with diminished left ventricular ejection

fractions. But orthostatic hypotension was common in these studies and could be severe in these

patients.

Glassman et al. (1993), noting that the type I antiarrhythmic drugs given following myocardial

infarction actually increased the risk of sudden death, suggested that the tricyclics may pose

similar risks. The risk of sudden death is also increased when heart rate variability is reduced, and

the tricyclics reduce heart rate variability (Roose et al. 1998).

As mentioned earlier (see subsection “Attention-Deficit/Hyperactivity Disorder”), sudden death has

been reported in five children under the age of 12 years who were receiving desipramine (Riddle et

1. 1991, 1993). It was suggested that the immature conduction system in some children might

render them more vulnerable to the cardiac effects of desipramine. Subsequently, a study was

conducted in 71 children with 24-hour cardiac monitoring (Biederman et al. 1993). No cardiac

abnormalities were observed. Wilens et al. (1992) examined the possibility that

hydroxydesipramine might reach unusually high levels in children and adolescents, but such levels

were not found. A study of electrocardiographic parameters in that sample failed to show a

relationship between those parameters and concentrations of desipramine or hydroxydesipramine

(Wilens et al. 1993). Although these studies failed to reveal a mechanism for the sudden deaths

reported, they do suggest that these events are not predictable, that they are not dose-dependent

cardiac effects, and that usual blood level or electrocardiogram monitoring is not likely to identify

those at risk.

To summarize the clinical implications of these cardiac effects, the clinician may wish to consider

the following. In adults without cardiac disease, orthostatic hypotension may occur with tricyclic

use, but conduction problems are not likely. In patients with preexisting conduction delay, the

tricyclics may cause heart block. In patients with ischemic heart disease, continued use of tricyclics

will increase cardiac work and reduce heart rate variability, possibly increasing the risk of sudden

death. Children younger than 12 years also appear vulnerable to the risk of sudden death during

tricyclic administration, possibly because of cardiac conduction effects or reduced heart rate

variability. Cardiac arrhythmia is the most common cause of death with tricyclic overdose. ThesePrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

20 of 33

10/05/2009 15:56

cardiac safety issues, coupled with the recently reported safety of the SSRI sertraline when

administered for depression following myocardial infarction (Glassman et al. 2002), indicate that

the tricyclics are relatively contraindicated in patients with ischemic heart disease and that their

use should be reserved for patients whose illnesses are refractory to other treatments.

Hepatic Effects

Acute hepatitis has been associated with administration of imipramine (Horst et al. 1980; Moskovitz

et al. 1982; Weaver et al. 1977) and desipramine (Powell et al. 1968; Price and Nelson 1983). Mild

increases of liver enzymes (less than three times normal) are not uncommon and usually can be

monitored safely over a period of days or weeks without apparent harmful consequences. These

changes in liver enzymes do not appear to be related to drug concentrations (Price et al. 1984).

Acute hepatitis is relatively uncommon but can occur. The etiology is not well established but in

some cases appears to be a hypersensitivity reaction. It is characterized by very high enzyme levels

(e.g., aspartate aminotransferase [AST] levels >800), which develop within days. The enzyme

pattern can be either hepatocellular or cholestatic. Enzyme changes may precede clinical

symptoms, especially in the hepatocellular form. If a random blood test indicates mildly elevated

liver enzymes, enzyme levels can be followed for a few days. Because of the rapid rise in liver

enzyme levels in acute hepatitis, that condition will become evident quickly and will be easily

distinguished from mild, persistent enzyme level elevations.

Acute hepatitis is a dangerous and potentially fatal condition. The antidepressant must be

discontinued and should not be introduced again because the next reaction may be more severe.

Unfortunately, it is not uncommon for the patient to be receiving several medications, so that the

offending agent may be hard to identify. The risk of severe drug-induced hepatitis is not well

established. This author has observed four cases associated with desipramine in the course of

treating approximately 500 patients.

Other Side Effects

Increased sweating can occur with the tricyclic compounds and occasionally can be marked. The

mechanism for this symptom is unclear but may be associated with noradrenergic effects. Another

side effect for which the mechanism is unclear is carbohydrate craving. This effect, when coupled

with antihistaminic effects, can lead to significant weight gain. One report of outpatients treated

with amitriptyline found an average gain of 7 kg during a 6-month period (Berken et al. 1984).

Weight gain appears to be greater with the tertiary compounds (Fernstrom et al. 1986) and is less

common with nortriptyline, desipramine, and protriptyline. Sexual dysfunction has been described

with the tricyclics but generally is less common with this group than during treatment with the

SSRIs. This side effect appears to be associated with the more serotonergic compounds and does

occur with clomipramine. Tricyclic antidepressants can cause allergic skin rashes, which are

sometimes associated with photosensitivity reactions. Various blood dyscrasias also have been

reported; fortunately, these are very rare.

Overdose

Because antidepressants are used for depressed patients who are at risk for overdose, the lethality

of antidepressant drugs in overdose is of great concern. A tricyclic overdose of 10 times the total

daily dose can be fatal (Gram 1990; Rudorfer and Robins 1982). Death most commonly occurs as a

result of cardiac arrhythmia. However, seizures and central nervous system depression can occur.

Although the use of tricyclics for the treatment of depression has declined, amitriptyline remains

widely used for other disorders, such as pain.

The total number of deaths associated with amitriptyline is comparable to that for all other

tricyclics and tetracyclics combined. Overdoses often include multiple drugs. In 2006, the American

Association of Poison Control Centers (Bronstein et al. 2007) began reporting deaths based on

single substance ingestions, which gives a more accurate estimate of lethality. There were 2,730

ingestions of amitriptyline, with 6 deaths, and 1,938 ingestions of all other tricyclic and tetracyclic

antidepressants, with 6 deaths. By comparison, there were 19,598 ingestions of various SSRIs, withPrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

21 of 33

10/05/2009 15:56

3 deaths. The mortality rate for the cyclic antidepressants was 257 per 100,000, while the rate for

the SSRIs was 15.3 per 100,000, a 17-fold difference. All of the tricyclic and tetracyclic compounds

are dangerous in overdose. Desipramine appears to have a particularly high fatality rate.

Amoxapine has been reported to produce high rates of seizure in overdose. But the differences

among these drugs are relatively minor in comparison with the improved safety of the

second-generation antidepressant agents.

Teratogenicity

Although it would be ideal to discontinue all drugs during pregnancy, the patient and physician are

faced with a dilemma. The risk of relapse is a serious concern in patients with recurrent depression

because the risk may be increased during pregnancy or the postpartum period. This risk is

particularly high for patients with a prior history of depression during or following pregnancy. The

long history of tricyclic use without observation of birth defects argues for the safety of these

agents. Of course, the patient must be informed of the possible risks and benefits of taking the drug

and of discontinuing treatment before making a decision.

If tricyclics are continued during pregnancy, dosage adjustment may be required because of

metabolic changes related to the pregnancy (Altshuler and Hendrick 1996). Drug withdrawal

following delivery can occur in the infant and is characterized by tachypnea, cyanosis, irritability,

and poor sucking reflex. The drugs in this class should be discontinued 1 week prior to delivery if

possible. The tricyclics are excreted in breast milk at concentrations similar to those in plasma. The

actual quantity delivered, however, is very small, so that drug levels in the infant are usually

undetectable (Rudorfer and Potter 1997).

DRUG–DRUG INTERACTIONS

Pharmacodynamic Interactions

Pharmacodynamic interactions are those in which the action of one drug affects the action of the

other. More commonly, the effects of the two drugs are additive and result in an adverse event.

Perhaps the best example is the interaction of the tricyclics with MAOI drugs. The most dangerous

sequence is to give a large dose of a tricyclic to a patient who is already taking an MAOI. This can

result in a sudden increase in catecholamines and a potentially fatal hypertensive reaction. These

two compounds have been used together to treat patients with refractory depression (Goldberg and

Thornton 1978; Schuckit et al. 1971). Treatment is begun with lower doses, and either the two

compounds are started together or the tricyclic is started first. Once begun, coadministration may

actually reduce the risk of tyramine reactions (Pare et al. 1985); however, because the protective

effect is variable and unpredictable, the usual MAOI diet is maintained.

Perhaps the most common pharmacodynamic interaction is when two psychotropic drugs are added

together, resulting in increased sedation. This interaction might occur when tricyclics are combined

with antipsychotic agents or with benzodiazepines. Other pharmacodynamic interactions can occur.

By blocking the transporters, the tricyclics block the uptake and thus interfere with the action of

guanethidine. Desipramine and the other tricyclics reduce the effect of clonidine.

Quinidine is an example of a drug with a potential dynamic and kinetic interaction with tricyclics.

Because the tricyclics have quinidine-like effects, the effects of tricyclics and quinidine on cardiac

conduction are potentially additive. In addition, quinidine is a potent CYP2D6 isoenzyme inhibitor

that can raise tricyclic levels, further adding to the problem.

Pharmacokinetic Interactions

Recently, pharmacokinetic interactions have received considerable attention. One type of

pharmacokinetic interaction is enzyme inhibition. A number of drugs can block the metabolic

pathways of the tricyclics, resulting in higher and potentially toxic levels. Desipramine has been of

particular interest because its metabolism is fairly simple, occurring via the CYP2D6 isoenzyme.

Because there are no major alternative pathways, inhibition of CYP2D6 can result in very high

desipramine plasma levels, and toxicity can occur (Preskorn et al. 1990). A number of drugs inhibitPrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

22 of 33

10/05/2009 15:56

2D6. Quinidine, mentioned above, is a very potent 2D6 inhibitor. Other drugs commonly used in

psychiatry that inhibit CYP2D6 include the SSRIs fluoxetine and paroxetine, duloxetine, bupropion,

and some antipsychotics. Fluoxetine and paroxetine at usual doses raise desipramine levels, on

average, three- to fourfold in extensive metabolizers (Preskorn et al. 1994). In slow metabolizers,

enzyme inhibitors have less of an effect, because the patients are already deficient in the enzyme

and the drug level is already high. In ultrarapid metabolizers, fluoxetine and paroxetine may cause

a greater increase in desipramine levels, but these patients are likely to have very low initial

desipramine levels. Sertraline 50 mg/day increases desipramine levels, on average, about

30%–40%, which is not a clinically meaningful difference (Preskorn et al. 1994). At higher doses,

there is proportionally greater inhibition, but the increase is still substantially less than the

300%–400% increase that occurs with fluoxetine or paroxetine. The magnitude of the effect of

bupropion on CYP2D6 has not been reported but appears to be clinically significant. Venlafaxine,

nefazodone, mirtazapine, and citalopram appear to have minimal effects on 2D6. 2D6 inhibitors

would be expected to block nortriptyline metabolism, but the magnitude of this interaction has not

been well studied.

Antipsychotic agents such as chlorpromazine and perphenazine also inhibit 2D6 (Gram et al. 1974;

Nelson and Jatlow 1980). At usual doses, perphenazine raises desipramine levels, on average,

twofold, but this effect varies with dose and with the neuroleptic employed. Haloperidol can also

inhibit the CYP2D6 pathway, but in this author’s experience, this effect is not likely to be clinically

meaningful at low dosages (e.g., <10 mg/day).

Because the tertiary tricyclics are metabolized by several pathways (CYP1A2, 3A4, 2C19), a

selective inhibitor of one pathway would be likely to have less of an effect on these compounds.

Drug interactions with the tertiary amines do occur but appear to be less robust. Methylphenidate

appears to inhibit demethylation of imipramine to desipramine. At this point, numerous drug

interactions have been described, although many are of doubtful clinical significance (for

comprehensive reviews, see Nemeroff et al. 1996; Pollock 1997).

The other type of pharmacokinetic drug interaction is enzyme induction. The result of this

interaction may render the drug acted upon ineffective. Unlike enzyme inhibition, which occurs

quickly, enzyme induction requires synthesis of new enzyme. As a result, the full effect of an

enzyme inducer may take 2–3 weeks to develop. If the inducer is discontinued, the effect takes 2–3

weeks to dissipate. Barbiturates and carbamazepine are potent inducers of CYP3A4. Phenytoin also

can induce this enzyme, but its effects on the tricyclics appear to be less dramatic. Although

CYP2D6 is a noninducible isoenzyme, phenobarbital reduces the availability of desipramine

substantially. Apparently when CYP3A4 is induced, it becomes an important metabolic pathway for

desipramine and the other tricyclics. In this author’s experience, it can be difficult to attain an

effective blood level of desipramine in the presence of a barbiturate.

Nicotine induces the CYP1A2 pathway and may lower concentrations of the tertiary tricyclics, but

the secondary tricyclics (e.g., desipramine, nortriptyline) appear to be less affected.

Alcohol has a complicated interaction with the tricyclics. Acute ingestion of alcohol can reduce

first-pass metabolism, resulting in higher tricyclic levels. Because tricyclic overdose is often

associated with alcohol ingestion, this is an important interaction, resulting in higher tricyclic

levels. Alternatively, chronic use of alcohol appears to induce hepatic isoenzymes and may lower

tricyclic levels (Shoaf and Linnoila 1991).

The tricyclics themselves produce some enzyme inhibition, but few clinically significant interactions

have been described. The tertiary tricyclics compete with warfarin for some metabolic enzymes

(e.g., CYP1A2) and may raise warfarin levels.

CONCLUSION

The tricyclic drugs were the mainstay of treatment for depression for three decades. Although the

second-generation antidepressants appear to be better tolerated, no new agent has been shown to

be more effective than the tricyclics, and if anything, there has been concern that the new agentsPrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

23 of 33

10/05/2009 15:56

may be less effective. The tricyclics were “dirty” drugs; that is, they had multiple actions. Although

their side effects have been emphasized, these multiple actions may have contributed to their

efficacy. Not only does amitriptyline block uptake of 5-HT, but its metabolite blocks uptake of

norepinephrine, and in addition, amitriptyline is a 5-HT2 antagonist. The principal drawback of this

class of agents is the risk of serious cardiac adverse effects. They can aggravate arrhythmia in

patients with preexisting conduction delay. They also may increase the risk of sudden death in

children and in patients with ischemic heart disease. Moreover, a week’s supply of medication taken

in overdose could be fatal. As noted, even though the use of the tricyclics has diminished,

amitriptyline remains a common cause of death by overdose in the United States. Because of these

adverse effects, it is unlikely that there will be a resurgence of interest in the tricyclics.

Nevertheless, the efficacy of these agents across a range of disorders, including pain, suggests the

possible advantage of antidepressant drugs with multiple actions.

REFERENCES

Abernathy DR, Greenblatt DJ, Shader RI: Imipramine and desipramine disposition in the elderly. J

Pharmacol Exp Ther 232:183–188, 1985

Agency for Health Care Policy and Research: Clinical Practice Guideline: Depression in Primary Care:

Treatment of Major Depression, Vol 2. Rockville, MD, U.S. Government Printing Office, 1993

Alexanderson B: Pharmacokinetics of nortriptyline in man after single and multiple oral doses: the

predictability of steady-state plasma concentrations from single-dose plasma-level data. Eur J Clin

Pharmacol 4:82–91, 1972 [PubMed]

Altshuler LL, Hendrick VC: Pregnancy and psychotropic medication: changes in blood levels. J Clin

Psychopharmacol 16:78–80, 1996 [PubMed]

American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th

Edition, Text Revision. Washington, DC, American Psychiatric Association, 2000

American Psychiatric Association Task Force: Tricyclic antidepressants—blood level measurements

and clinical outcome: an APA Task Force report. Task Force on the Use of Laboratory Tests in

Psychiatry. Am J Psychiatry 142:155–162, 1985

Anderson I: SSRIs versus tricyclics antidepressants in depressed inpatients: a meta-analysis of

efficacy and tolerability. Depress Anxiety 7 (suppl 1):11–17, 1998

Anderson I: Selective serotonin reuptake inhibitors versus tricyclics antidepressant: a

meta-analysis of efficacy and tolerability. J Affect Disord 58:19–36, 2000 [PubMed]

Anderson I, Tomenson B: A meta-analysis of the efficacy of selective serotonin reuptake inhibitors

compared to tricyclic antidepressants in depression (abstract). Neuropsychopharmacology 10

(suppl):106, 1994

Anton RF, Burch EA: Amoxapine versus amitriptyline combined with perphenazine in the treatment

of psychotic depression. Am J Psychiatry 147:1203–1208, 1990 [PubMed]

Arora RC, Meltzer HY: Serotonergic measures in the brains of suicide victims: 5-HT2 binding sites in

frontal cortex of suicide victims and control subjects. Am J Psychiatry 146:730–736, 1989

[PubMed]

Åsberg M, Cronholm B, Sjöqvist F, et al: Relationship between plasma level and therapeutic effect of

nortriptyline. BMJ 3:331–334, 1971

Axelrod J, Whitby LG, Herting G: Effect of psychotropic drugs on the uptake of H3-norepinephrine

by tissues. Science 133:383–384, 1961 [PubMed]

Barden N: Modulation of glucocorticoid receptor gene expression by antidepressant drugs.

Pharmacopsychiatry 29:12–22, 1996 [PubMed]

Beckmann H, Goodwin FK: Antidepressant response to tricyclics and urinary MHPG in unipolarPrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

24 of 33

10/05/2009 15:56

patients. Arch Gen Psychiatry 32:17–22, 1975 [PubMed]

Benetello P, Furlanut M, Zara G, et al: Imipramine pharmacokinetics in depressed geriatric patients.

Int J Clin Pharmacol Res 10:191–195, 1990 [PubMed]

Berken GN, Weinstein DO, Stern WC: Weight gain: a side effect of tricyclic antidepressants. J Affect

Disord 7:133–138, 1984 [PubMed]

Bertilsson L, Mellstrom B, Sjöqvist P: Pronounced inhibition of noradrenaline uptake by 10-hydroxy

metabolites of nortriptyline. Life Sci 25:1285–1292, 1979 [PubMed]

Biederman J, Baldessarini RJ, Wright V, et al: a double-blind placebo-controlled study of

desipramine in the treatment of ADD, I: efficacy. J Am Acad Child Adolesc Psychiatry 32:805–813,

1989

Biederman J, Baldessarini RJ, Goldblatt A, et al: A naturalistic study of 24-hour electrocardiographic

recording and echocardiographic findings in children and adolescents treated with desipramine. J

Am Acad Child Adolesc Psychiatry 32:805–813, 1993 [PubMed]

Biggs JT, Spiker DG, Petit JM, et al: Tricyclic antidepressant overdose. JAMA 238:135–138, 1977

[PubMed]

Blier P, de Montigny C, Chaput Y: Modification of the serotonin system by antidepressant

treatments: implications for the therapeutic response in major depression. J Clin Psychopharmacol

7:24S–35S, 1987

Bock J, Nelson JC, Gray S, et al: Desipramine hydroxylation: Variability and effect of antipsychotic

drugs. Clin Pharmacol Ther 33:190–197, 1983

Bolden-Watson C, Richelson E: Blockade of newly developed antidepressants of biogenic amine

uptake into rat brain synaptosomes. Life Sci 52:1023–1029, 1993 [PubMed]

Bowden CL, Schatzberg AF, Rosenbaum A, et al: Fluoxetine and desipramine in major depressive

disorder. J Clin Psychopharmacol 13:305–310, 1993 [PubMed]

Breyer-Pfaff U, Gaertner HJ, Kreuter F, et al: Antidepressive effect and pharmacokinetics of

amitriptyline with consideration of unbound drug and 10-hydroxynortriptyline plasma levels.

Psychopharmacology (Berl) 76:240–244, 1982 [PubMed]

Bronstein AC, Spyker DA, Cantilena LR, et al: 2006 Annual Report of the American Association of

Poison Control Centers’ National Poison Data System (NPDS). Clin Toxicol 45:815–917, 2007

[PubMed]

Brosen K, Otton SV, Gram LF: Sparteine oxidation polymorphism in Denmark. Acta Pharmacol

Toxicol 57:357–360, 1985 [PubMed]

Brosen K, Zeugin T, Myer UA: Role of P450IID6, the target of the sparteine/debrisoquin oxidation

polymorphism, in the metabolism of imipramine. Clin Pharmacol Ther 49:609–617, 1991 [PubMed]

Brunswick DJ, Amsterdam JD, Mendels J, et al: Prediction of steady-state imipramine and

desmethylimipramine plasma concentrations from single dose data. Clin Pharmacol Ther

25:605–610, 1979 [PubMed]

Bunney WE, Davis JM: Norepinephrine in depressive reactions: a review. Arch Gen Psychiatry

13:483–494, 1965 [PubMed]

Byck R: Drugs and the treatment of psychiatric disorders, in The Pharmacological Basis of

Therapeutics, 5th Edition. Edited by Gilman AG, Goodman IS. New York, Macmillan, 1975, pp

152–200

Bymaster FP, Katner JS, Nelson DL, et al: Atomoxetine increases extracellular levels of

norepinephrine and dopamine in prefrontal cortex of rat: a potential mechanism for efficacy in

attention deficit/hyperactivity disorder. Neuropsychopharmacology 27:699–711, 2002 [PubMed]Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

25 of 33

10/05/2009 15:56

Chan CH, Janicak PG, Davis JM: Response of psychotic and nonpsychotic depressed patients to

tricyclic antidepressants. J Clin Psychiatry 48:197–200, 1987 [PubMed]

Charney DS, Delgado PL, Price LH, et al: The receptor sensitivity hypothesis of antidepressant

action: a review of antidepressant effects on serotonin function, in The Role of Serotonin in

Psychiatric Disorders. Edited by Brown SL, van Praag HM. New York, Brunner/Mazel, 1991, pp

29–56

Cheetham SC, Crompton MR, Katona CLE, et al: Brain 5-HT2 receptor binding in suicide victims.

Brain Res 443:272–280, 1988 [PubMed]

Cooper TB, Simpson GM: Prediction of individual dosage of nortriptyline. Am J Psychiatry

135:333–335, 1978 [PubMed]

Coupet I, Rauh CE, Szucs-Myers VA, et al: 2-Chloro-11(piperazinyl)[b,f][1,4]oxazepine

(amoxapine), an antidepressant with antipsychotics properties—a possible role for

7-hydroxyamoxapine. Biochem Pharmacol 28:2514–2515, 1979 [PubMed]

Cusack B, Nelson A, Richelson E: Binding of antidepressants to human brain receptors: focus on

newer generation compounds. Psychopharmacology 114:559–565, 1994 [PubMed]

Danish University Antidepressant Group: Citalopram: clinical effect profile in comparison with

clomipramine. A controlled multicenter study. Psychopharmacology (Berl) 90:131–138, 1986

Danish University Antidepressant Group: Paroxetine: a selective serotonin reuptake inhibitor

showing better tolerance, but weaker antidepressant effect than clomipramine in a controlled

multicenter study. J Affect Disord 18:289–299, 1990

Davis JM: Overview: maintenance therapy in psychiatry, II: affective disorders. Am J Psychiatry

133:1–13, 1976 [PubMed]

Davis JM, Glassman AH: Antidepressant drugs, in Comprehensive Textbook of Psychiatry. Edited by

Kaplan HI, Sadock BJ. Baltimore, MD, Williams & Wilkins, 1989, pp 1627–1655

Dawling S, Ford S, Rangedara DC, et al: Amitriptyline dosage prediction in elderly patients from

plasma concentration at 24 hours after a single 100 mg dose. Clin Pharmacokinet 9:261–266, 1984

[PubMed]

Delay J, Deniker P: Trente-huit cas de psychoses traitees par la cure prolongee et continue de 4560

1. Le Congres des Al. et Neurol. de Langue Fr. in Compte rendu du Congres. Masson et Cie, Paris,

1952

Delgado PL, Charney DS, Price LH, et al: Serotonin function and the mechanism of antidepressant

action. Reversal of antidepressant-induced remission by rapid depletion of plasma tryptophan. Arch

Gen Psychiatry 47:411–418, 1990 [PubMed]

Delgado PL, Miller HL, Salomon RM, et al: Monoamines and the mechanism of antidepressant action:

effects of catecholamine depletion on mood of patients treated with antidepressants.

Psychopharmacol Bull 29:389–396, 1993 [PubMed]

de Montigny C, Aghajanian GK: Tricyclic antidepressants: long-term treatment increases

responsivity of rat forebrain neurons to serotonin. Science 202:1303–1306, 1978

Dessain EC, Schatzberg AF, Woods BT, et al: Maprotiline treatment in depression: a perspective on

seizures. Arch Gen Psychiatry 43:86–90, 1986 [PubMed]

DiMascio A, Heninger G, Klerman GL: Psychopharmacology of imipramine and desipramine: a

comparative study of their effects in normal males. Psychopharmacologia 5:361–371, 1964

[PubMed]

Duman RS, Heninger GR, Nestler EJ: A molecular and cellular theory of depression. Arch Gen

Psychiatry 54:597–606, 1997 [PubMed]Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

26 of 33

10/05/2009 15:56

Eberhard G, von Knorring L, Nilsson HL, et al: A double-blind randomized study of clomipramine

versus maprotiline in patients with idiopathic pain syndromes. Neuropsychobiology 19:25–34, 1988

[PubMed]

Escorihuela RM, Boix F, Corda MG, et al: Chronic but not acute antidepressant treatment increases

pentetrazol-induced convulsions in mice. J Pharm Pharmacol 41:143–144, 1989 [PubMed]

Evans DAP, Mahgoub A, Sloan TP, et al: A family and population study of the genetic polymorphism

of debrisoquin oxidation in a white British population. J Med Genet 17:102–105, 1980 [PubMed]

Fava M, Uebelacker LA, Alpert JE, et al: Major depression subtypes and treatment response. Biol

Psychiatry 42:568–576, 1997 [PubMed]

Fernstrom MD, Krowinski RL, Kupfer DJ: Chronic imipramine treatment and weight gain. Psychiatr

Res 17:269–273, 1986 [PubMed]

Frank E, Kupfer DJ, Perel JM, et al: Three-year outcomes for maintenance therapies in recurrent

depression. Arch Gen Psychiatry 47:1093–1099, 1990 [PubMed]

Furey ML, Drevets WC: Antidepressant efficacy of the antimuscarinic drug scopolamine: a

randomized, placebo-controlled clinical trial. Arch Gen Psychiatry 63:1121–1129, 2006 [PubMed]

Furlanut M, Benetello P: The pharmacokinetics of tricyclic antidepressant drugs in the elderly.

Pharmacol Res 22:15–25, 1990 [PubMed]

Geller B: Psychopharmacology of children and adolescents: pharmacokinetics and relationships of

plasma/serum levels to response. Psychopharmacol Bull 27:401–409, 1991 [PubMed]

Gerson SC, Plotkin DA, Jarvik LF: Antidepressant drug studies, 1964 to 1986: empirical evidence for

aging patients. J Clin Psychopharmacol 8:311–322, 1988 [PubMed]

Glassman AH, Bigger JT: Cardiovascular effects of therapeutic doses of tricyclic antidepressants.

Arch Gen Psychiatry 38:815–820, 1981 [PubMed]

Glassman AH, Kantor S, Shostak M: Depression, delusions, and drug response. Am J Psychiatry

132:716–719, 1975 [PubMed]

Glassman AH, Perel JM, Shostak M, et al: Clinical implications of imipramine plasma levels for

depressive illness. Arch Gen Psychiatry 34:197–204, 1977 [PubMed]

Glassman AH, Bigger JT Jr, Giardina EV, et al: Clinical characteristics of imipramine induced

orthostatic hypotension. Lancet 1(8114):468–472, 1979 [PubMed]

Glassman AH, Roose SP, Bigger JT Jr: The safety of tricyclic antidepressants in cardiac patients:

risk/benefit reconsidered. JAMA 269:2673–2675, 1993 [PubMed]

Glassman AH, O’Connor CM, Califf RM, et al: Sertraline treatment of major depression in patients

with acute MI or unstable angina. JAMA 288:701–709, 2002 [PubMed]

Goldberg RS, Thornton WE: Combined tricyclic-MAOI therapy for refractory depression: a review,

with guidelines for appropriate usage. J Clin Pharmacol 18:143–147, 1978 [PubMed]

Goodwin GM, Green AR, Johnson P: 5-HT2 receptor characteristics in frontal cortex and 5-HT2

receptor mediated head-twitch behavior following antidepressant treatment to mice. Br J

Pharmacol 83:235–242, 1984 [PubMed]

Gram LF: Inadequate dosing and Pharmacokinetic variability as confounding factors in assessment

of efficacy of antidepressants. Clin Neuropharmacol 13 (suppl):S35–S44, 1990

Gram LF, Overo KF, Kirk L: Influence of neuroleptics and benzodiazepines on metabolism of tricyclic

antidepressants in man. Am J Psychiatry 131:863–866, 1974 [PubMed]

Greenblatt DJ, Moltke LL, Shader RI: Pharmacokinetics of psychotropic drugs, in Geriatric

Psychopharmacology. Edited by Nelson JC. New York, Marcel Dekker, 1998, pp 27–41Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

27 of 33

10/05/2009 15:56

Greist JH, Jefferson JW, Kobak KA, et al: Efficacy and tolerability of serotonin transport inhibitors in

obsessive-compulsive disorder: a meta-analysis. Arch Gen Psychiatry 52:53–60, 1995 [PubMed]

Hall H, Ogren SO: Effects of antidepressant drugs on different receptors in the brain. Eur Pharmacol

70:393–407, 1981 [PubMed]

Hammer W, Sjöqvist F: Plasma levels of monomethylated tricyclic antidepressants during treatment

with imipramine-like compounds. Life Sci 6:1895–1903, 1967 [PubMed]

Hartling OJ, Marving J, Knudsen P, et al: The effect of the tricyclic antidepressant drug,

nortriptyline on left ventricular ejection fraction and left ventricular volumes. Psychopharmacology

91:381–383, 1987 [PubMed]

Himmelhoch JM, Detre T, Kupfer DJ, et al: Treatment of previously intractable depressions with

tranylcypromine and lithium. J Nerv Ment Dis 155:216–220, 1972 [PubMed]

Himmelhoch JM, Thase ME, Mallinger AG, et al: Tranylcypromine versus imipramine in anergic

bipolar depression. Am J Psychiatry 148:910–916, 1991 [PubMed]

Horst DA, Grace ND, Le Compte PM: Prolonged cholestasis and progressive hepatic fibrosis

following imipramine therapy. Gastroenterology 79:550–554, 1980 [PubMed]

Jick H, Dinan BJ, Hunter JR, et al: Tricyclic antidepressants and convulsions. J Clin Psychopharmacol

3:182–185, 1983 [PubMed]

Jobson K, Linnoila M, Gillam J, et al: Successful treatment of severe anxiety attacks with tricyclic

antidepressants: a possible mechanism of action. Am J Psychiatry 135:863–864,1978 [PubMed]

Katz IR, Simpson GM, Jethanandani V, et al: Steady state pharmacokinetics of nortriptyline in the

frail elderly. Neuropsychopharmacology 2:229–236, 1989 [PubMed]

Katz IR, Simpson GM, Curlik SM, et al: Pharmacologic treatment of major depression for elderly

patients in residential care settings. J Clin Psychiatry 51 (7, suppl):41–47, 1990

Keller MB, Gelenberg AJ, Hirschfeld RMA, et al: The treatment of chronic depression: a double-blind,

randomized trial of sertraline and imipramine. J Clin Psychiatry 59(pt 2):598–607, 1998

Klein DF: Delineation of two drug responsive anxiety syndromes. Psychopharmacologia 5:397–408,

1964 [PubMed]

Kocsis JH, Frances AJ, Voss CB, et al: Imipramine treatment for chronic depression. Arch Gen

Psychiatry 45:253–257, 1988 [PubMed]

Kragh-Sørenson P, Åsberg M, Eggert-Hansen C: Plasma nortriptyline levels in endogenous

depression. Lancet 1(7795):113–115, 1973 [PubMed]

Kragh-Sørenson P, Hansen CE, Baastrup PC, et al: Self-inhibiting action of nortriptyline

antidepressant effect at high plasma levels: a randomized double-blind study controlled by plasma

concentrations in patients with endogenous depression. Psychopharmacologia 45:305–312, 1976

[PubMed]

Kuhn R: The treatment of depressive states with G22355 (imipramine hydrochloride). Am J

Psychiatry 115:459–464, 1958 [PubMed]

Kuhn R: The imipramine story, in Discoveries in Biological Psychiatry. Edited by Ayd FJ, Blackwell B.

Philadelphia, PA, JB Lippincott, 1970, pp 205–217

Kupfer D, Hanin I, Spiker DG, et al: Amitriptyline plasma levels and clinical response in primary

depression. Clin Pharmacol Ther 22:904–911, 1977 [PubMed]

Kupfer DJ, Spiker DG: Refractory depression: prediction of nonresponse by clinical indicators. J Clin

Psychiatry 42:307–312, 1981 [PubMed]

Lakoski JM, Aghajanian GK: Effects of ketanserin on neuronal responses to serotonin in thePrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

28 of 33

10/05/2009 15:56

prefrontal cortex, lateral geniculate and dorsal raphe nucleus. Neuropharmacology 24:265–273,

1985 [PubMed]

Leonard HL, Swedo S, Rapoport JL, et al: Treatment of obsessive compulsive disorder in children

and adolescents with clomipramine and desipramine: a double-blind crossover comparison. Arch

Gen Psychiatry 46:1088–1092, 1989 [PubMed]

Lesaca T: Amoxapine and neuroleptic malignant syndrome. Am J Psychiatry 144:1514, 1987

[PubMed]

Lesch KP, Manji HK: Signal-transducing G proteins and antidepressant drugs: evidence for

modulation of alpha subunit gene expression in rat brain. Biol Psychiatry 32:549–579, 1992

[PubMed]

Liebowitz MR, Quitkin FM, Stewart JW, et al: Phenelzine vs imipramine in atypical depression. Arch

Gen Psychiatry 41:669–677, 1984 [PubMed]

Liebowitz MR, Quitkin FM, Stewart JW, et al: Antidepressant specificity in atypical depression. Arch

Gen Psychiatry 45:129–137, 1988 [PubMed]

Livingston RL, Zucker DK, Isenberg K, et al: Tricyclic antidepressants and delirium. J Clin Psychiatry

44:173–176, 1983 [PubMed]

Maas JW, Fawcett JA, Dekirmenjian H: Catecholamine metabolism, depressive illness, and drug

response. Arch Gen Psychiatry 26:353–363, 1972

Madakasira S: Amoxapine-induced neuroleptic malignant syndrome. DICP 23:50–55, 1989

[PubMed]

Madakasira S, Khazanie PG: Reliability of Amitriptyline dose prediction based on single-dose plasma

levels. Clin Pharmacol Ther 37:145–149, 1985 [PubMed]

Malatynska E, Knapp RJ, Ikeda M, et al: Antidepressants and seizure-interactions at the

GABA-receptor chloride-ionophore complex. Life Sci 43:303–307,1988 [PubMed]

Manji HK, Potter WZ, Lenox RH: Signal transduction pathways: molecular targets for lithium’s

actions. Arch Gen Psychiatry 52:531–543, 1995 [PubMed]

Marek GJ, Carpenter LL, McDougle CJ, et al: Synergistic action of 5-HT2A antagonists and selective

serotonin reuptake inhibitors in neuropsychiatric disorders Neuropsychopharmacology 28:402–412,

2003 [PubMed]

Max MB, Lynch SA, Muir J, et al: Effects of desipramine, amitriptyline, and fluoxetine on pain in

diabetic neuropathy. N Engl J Med 326:1250–1256, 1992 [PubMed]

McGrath PJ, Stewart JW, Harrison W, et al: Treatment of tricyclic refractory depression with a

monoamine oxidase inhibitor antidepressant. Psychopharmacol Bull 23:169–172, 1987 [PubMed]

Miller HL, Delgado PL, Salomon RM, et al: Clinical and biochemical effects of catecholamine

depletion on antidepressant-induced remission of depression. Arch Gen Psychiatry 53:117–128,

1996 [PubMed]

Miller NL, Kocsis JH, Leon AC, et al: Maintenance desipramine for dysthymia: a placebo-controlled

study. J Affect Disord 64:231–237, 2001 [PubMed]

Mochizucki D: Serotonin and noradrenaline reuptake inhibitors in animal models of pain. Hum

Psychopharmacol 19 (suppl 1): S15–S19, 2004

Montgomery SA: The efficacy of fluoxetine as an antidepressant in the short and long term. Int Clin

Psychopharmacol 4 (suppl):113–119, 1989

Moody JP, Whyte SF, MacDonald AJ, et al: Pharmacokinetic aspects of protriptyline plasma levels.

Eur J Clin Pharmacol 11:51–56, 1977 [PubMed]Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

29 of 33

10/05/2009 15:56

Moskovitz R, DeVane CL, Harris R, et al: Toxic hepatitis and single daily dosage imipramine therapy.

J Clin Psychiatry 43:165–166, 1982 [PubMed]

Munjack DJ, Usigli R, Zulueta A, et al: Nortriptyline in the treatment of panic disorder and

agoraphobia with panic attacks. J Clin Psychopharmacol 8:204–207, 1988 [PubMed]

Nelson JC: Use of desipramine in depressed inpatients. J Clin Psychiatry 45:10–15, 1984 [PubMed]

Nelson JC: The use of antipsychotic drugs in the treatment of depression, in Treating Resistant

Depression. Edited by Zohar J, Belmaker RH. New York, PMA Publishing, 1987, pp 131–146

Nelson JC, Jatlow P: Neuroleptic effect on desipramine steady state plasma concentrations. Am J

Psychiatry 137:1232–1234, 1980 [PubMed]

Nelson JC, Jatlow PI: Nonlinear desipramine kinetics: prevalence and importance. Clin Pharm Ther

41:666–670, 1987 [PubMed]

Nelson JC, Jatlow P, Quinlan DM, et al: Desipramine plasma concentrations and antidepressant

response. Arch Gen Psychiatry 39:1419–1422, 1982 [PubMed]

Nelson JC, Jatlow PI, Quinlan DM: Subjective complaints during desipramine treatment: relative

importance of plasma drug concentrations and the severity of depression. Arch Gen Psychiatry

41:55–59, 1984 [PubMed]

Nelson JC, Jatlow P, Mazure C: Desipramine plasma levels and response in elderly melancholics. J

Clin Psychopharmacol 5:217–220, 1985 [PubMed]

Nelson JC, Jatlow PI, Mazure C: Rapid desipramine dose adjustment using 24-hour levels. J Clin

Psychopharmacol 7:72–77, 1987 [PubMed]

Nelson JC, Mazure C, Attilasoy E, et al: Hydroxy-desipramine in the elderly. J Clin Psychopharmacol

8:428–433, 1988a

Nelson JC, Mazure C, Jatlow PI: Antidepressant activity of 2-hydroxy-desipramine. Clin Pharmacol

Ther 44:283–288, 1988b

Nelson JC, Mazure CM, Jatlow PI: Characteristics of desipramine refractory depression. J Clin

Psychiatry 55:12–19, 1994 [PubMed]

Nelson JC, Mazure CM, Jatlow PI: Desipramine treatment of major depression in patients over 75

years of age. J Clin Psychopharmacol 15:99–105, 1995 [PubMed]

Nemeroff CB, DeVane CL, Pollock BG: Newer antidepressants and the cytochrome P450 system. Am

J Psychiatry 153:311–320, 1996 [Full Text] [PubMed]

Nibuya M, Nestler EJ, Duman RS: Chronic antidepressant administration increases the expression of

cAMP response element-binding protein (CREB) in rat hippocampus. J Neurosci 16:2365–2372,

1996 [PubMed]

Nordin C, Bertilsson L, Siwers B: Clinical and biochemical effects during treatment of depression

with nortriptyline—the role of 10-hydroxynortriptyline. Clin Pharmacol Ther 42:10–19, 1987

[PubMed]

O’Malley PG, Jackson JL, Santoro J, et al: Antidepressant therapy for unexplained symptoms and

symptom syndromes. J Fam Pract 48:980–990, 1999 [PubMed]

Panerai AE, Monza G, Movilia P, et al: A randomized, within-patient, cross-over, placebo-controlled

trial on the efficacy and tolerability of the tricyclic antidepressants chlorimipramine and

nortriptyline in central pain. Acta Neurol Scand 82:34–38, 1990 [PubMed]

Pare CMB, Mousawi MA, Sandler M, et al: Attempts to attenuate the “cheese effect”: combined drug

therapy in depressive illness. J Affect Disord 9:137–141, 1985 [PubMed]

Peck AW, Stern WC, Watkinson C: Incidence of seizures during treatment with tricyclicPrint: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

30 of 33

10/05/2009 15:56

antidepressant drugs and bupropion. J Clin Psychiatry 44 (5, pt 2):197–201, 1983

Peroutka SJ, Snyder SH: Long term antidepressant treatment decreases spiroperidol-labeled

serotonin receptor binding. Science 210:88–90, 1980 [PubMed]

Physicians’ Desk Reference. Montvale, NJ, Medical Economics Company, 2002

Pimentel L, Trommer L: Cyclic antidepressant overdoses: a review. Emerg Med Clin North Am

12:533, 1994 [PubMed]

Pinder RM, Brogden RN, Speight TM, et al: Doxepin up-to-date: a review of its pharmacological

properties and therapeutic efficacy with particular reference to depression. Drugs 13:161–218,

1977 [PubMed]

Pollock BG: Drug interactions, in Geriatric Psychopharmacology. Edited by Nelson JC. New York,

Marcel Dekker, 1997, pp 43–60

Pollock BG, Everett G, Perel JM: Comparative cardiotoxicty of nortriptyline and its isomeric

10-hydroxymetabolites. Neuropsychopharmacology 6:1–10, 1992 [PubMed]

Potter WZ: Psychotherapeutic drugs and biogenic amines: current concepts and therapeutic

indications. Drugs 28:127–143, 1984 [PubMed]

Potter WZ, Calil NM, Manian AA, et al: Hydroxylated metabolites of tricyclic antidepressants:

preclinical assessment of activity. Biol Psychiatry 14:601–613, 1979 [PubMed]

Potter WZ, Zavadil AP, Kopin IJ, et al: Single-dose kinetics predict steady-state concentrations of

imipramine and desipramine. Arch Gen Psychiatry 36:314–320, 1980

Powell WJ, Koch-Weser J, Williams RA: Lethal hepatic necrosis after therapy with imipramine and

desipramine. JAMA 206:1791–1792, 1968

Prange AI: The pharmacology and biochemistry of depression. Dis Nerv Syst 25:217–221, 1965

Preskorn SH, Irwin HA: Toxicity of tricyclic antidepressants: kinetics, mechanism, intervention: a

review. J Clin Psychiatry 43:151–156, 1982 [PubMed]

Preskorn SH, Simpson S: Tricyclic antidepressant–induced delirium and plasma drug concentration.

Am J Psychiatry 139:822–823, 1982 [PubMed]

Preskorn SH, Beber JH, Faul JC, et al: Serious adverse effects of combining fluoxetine and tricyclic

antidepressants. Am J Psychiatry 147:532, 1990 [PubMed]

Preskorn SH, Alderman J, Chung M, et al: Pharmacokinetics of desipramine co-administered with

sertraline or fluoxetine. J Clin Psychopharmacol 14:90–98, 1994 [PubMed]

Price LH, Nelson JC: Desipramine associated hepatitis. J Clin Psychopharmacology 3:243–246, 1983

[PubMed]

Price LH, Nelson JC, Jatlow P: Effects of desipramine on clinical liver function tests. Am J Psychiatry

414:798–800, 1984

Priest BT, Kaczorowski GJ: Blocking sodium channels to treat neuropathic pain. Expert Opin Ther

Targets 11:291–306, 2007 [PubMed]

Randrup A, Braestrup C: Uptake inhibition of biogenic amines by newer antidepressant drugs:

relevance to the dopamine hypothesis of depression. Psychopharmacology (Berl) 53:309–314, 1977

[PubMed]

Rapoport J, Potter WZ: Tricyclic antidepressants: use in pediatric psychopharmacology, in

Pharmacokinetics: Youth and Age. Edited by Raskin A, Robinson D. Amsterdam, The Netherlands,

Elsevier, 1981, pp 105–123

Rapoport JL, Mikkelson EJ, Zavadil AP, et al: Childhood enuresis, II: psychopathology, tricyclic

concentrations in plasma and anti-enuretic effect. Arch Gen Psychiatry 37:1146–1152, 1980Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

31 of 33

10/05/2009 15:56

[PubMed]

Reisby N, Gram LF, Bech P, et al: Imipramine: clinical effects and pharmacokinetic variability.

Psychopharmacology 54:263–272, 1977 [PubMed]

Richelson E, Nelson A: Antagonism by antidepressants of neurotransmitter receptors of normal

human brain in vitro. J Pharmacol Exp Ther 230:94–102, 1984 [PubMed]

Riddle MA, Nelson JC, Kleinman CS, et al: Sudden death in children receiving Norpramin: a review of

three reported cases and commentary. J Am Acad Child Adolesc Psychiatry 30:104–108, 1991

[PubMed]

Riddle MA, Geller B, Ryan N: Another sudden death in a child treated with desipramine. J Am Acad

Child Adolesc Psychiatry 32:792–797, 1993 [PubMed]

Roose SP, Glassman AH, Siris SG, et al: Comparison of imipramine and nortriptyline-induced

orthostatic hypotension: a meaningful difference. J Clin Psychopharmacol 1:316–319, 1981

[PubMed]

Roose SP, Glassman AH, Walsh BT, et al: Tricyclic nonresponders, phenomenology and treatment.

Am J Psychiatry 143:345–348, 1986 [PubMed]

Roose SP, Glassman AH, Giardina EG, et al: Cardiovascular effects of imipramine and bupropion in

depressed patients with congestive heart failure. J Clin Psychopharmacol 7:247–251, 1987a

Roose SP, Glassman AH, Giardina EG, et al: Tricyclic antidepressants in depressed patients with

cardiac conduction disease. Arch Gen Psychiatry 44:273–275, 1987b

Roose S, Laghrissi-Thode F, Kennedy JS, et al: A comparison of paroxetine and nortriptyline in

depressed patients with ischemic heart disease. JAMA 279:287–291, 1998 [PubMed]

Rosenstein DL, Nelson JC, Jacobs JC: Seizures associated with antidepressants: a review. J Clin

Psychiatry 54:289–299, 1993 [PubMed]

Rosenstein DL, Nelson JC: Heart rate during desipramine treatment as an indicator of

beta1-adrenergic function. Society of Biological Psychiatry Scientific Program. Biol Psychiatry

29:132A, 1991

Rudorfer MV, Potter WZ: The role of metabolites of antidepressant in the treatment of depression.

CNS Drugs 7:273–312, 1997

Rudorfer MV, Robins E: Amitriptyline overdose: clinical effects of tricyclic antidepressant plasma

levels. J Clin Psychiatry 43:457–460, 1982 [PubMed]

Rudorfer MV, Young RC: Desipramine: cardiovascular effects and plasma levels. Am J Psychiatry

137:984–986, 1980 [PubMed]

Ryan ND: The pharmacologic treatment of child and adolescent depression. Psychiatric Clin North

Am 15:29–40, 1992 [PubMed]

Salerno SM, Browning R, Jackson JL: The effect of antidepressant treatment on chronic back pain.

Arch Intern Med 162:19–24, 2002 [PubMed]

Schildkraut JJ: The catecholamine hypothesis of affective disorders: a review of supporting

evidence. Am J Psychiatry 122:509–522, 1965 [PubMed]

Schneider LS, Cooper TB, Severson JA, et al: Electrocardiographic changes with nortriptyline and

10-hydroxynortriptyline in elderly depressed outpatients. J Clin Psychopharmacol 8:402–408, 1988

[PubMed]

Schuckit M, Robins E, Feighner J: Tricyclic antidepressants and monoamine oxidase inhibitors:

combination therapy in the treatment of depression. Arch Gen Psychiatry 24:509–514, 1971

[PubMed]Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

32 of 33

10/05/2009 15:56

Schwaninger M, Schofl C, Blume R, et al: Inhibition by antidepressant drugs of cyclic AMP response

element-binding protein/cyclic AMP response element-directed gene transcription. Mol Pharmacol

47:1112–1118, 1995 [PubMed]

Shoaf SE, Linnoila M: Interaction of ethanol and smoking on the pharmacokinetics and

pharmacodynamics of psychotropic medications. Psychopharmacol Bull 27: 577–594, 1991

[PubMed]

Siris SG, van Kammen DP, Docherty JP: Use of antidepressant drugs in schizophrenia. Arch Gen

Psychiatry 35:1368–1377, 1978 [PubMed]

Smith RC, Reed K, Leelavathi DE, et al: Pharmacokinetics and the effects of nortriptyline in geriatric

depressed patients. Psychopharmacol Bull 16:54–57, 1980 [PubMed]

Spiker DG, Weiss JC, Chang S, et al: Tricyclic antidepressant overdose: clinical presentation and

plasma levels. Clin Pharmacol Ther 18:539–546, 1975 [PubMed]

Spiker DG, Weiss JC, Dealy RS, et al: The pharmacologic treatment of delusional depression. Am J

Psychiatry 142:430–436, 1985 [PubMed]

Stanley M, Mann JJ: Increased serotonin-2 binding sites in frontal cortex of suicide victims. Lancet

1(8318):214–216, 1983 [PubMed]

Sulser F, Vetulani J, Mobley PL: Mode of action of antidepressant drugs. Biochem Pharmacol

27:257–261, 1978 [PubMed]

Szabo ST, Blier P: Effect of the selective noradrenergic reuptake inhibitor reboxetine on the firing

activity of noradrenaline and serotonin neurons. Eur J Neurosci 13:2077–2087, 2001 [PubMed]

Tatsumi M, Groshan K, Blakely RD, et al: Pharmacological profile of antidepressants and related

compounds at human monoamine transporters. Eur J Pharmacol 340:249, 1997 [PubMed]

Taylor NE, Schwartz HI: Neuroleptic malignant syndrome following amoxapine overdose. J Nerv

Ment Dis 176:249–251, 1988 [PubMed]

Thase ME, Malinger AG, McKnight D, et al: Treatment of imipramine-resistant recurrent depression,

IV: a double-blind crossover study of tranylcypromine for anergic bipolar depression. Am J

Psychiatry 149:195–198, 1992 [PubMed]

Thase ME, Fava M, Halbreich U, et al: A placebo-controlled, randomized clinical trial comparing

sertraline and imipramine for the treatment of dysthymia. Arch Gen Psychiatry 53:777, 1996

[PubMed]

Thayssen P, Bjerre M, Kragh-Sorenson P, et al: Cardiovascular effects of imipramine and

nortriptyline in elderly patients. Psychopharmacology 74:360–364, 1981 [PubMed]

Tremblay P, Blier P: Catecholaminergic strategies for the treatment of major depression. Curr Drug

Targets 7:149–158, 2006 [PubMed]

Vetulani J, Sulser F: Action of various antidepressant treatments reduces reactivity of

noradrenergic cyclic AMP generating systems in limbic forebrain. Nature 257:395–496, 1975

von Moltke LL, Greenblatt DJ, Harmatz JS, et al: Psychotropic drug metabolism in old age: principles

and problems of assessment, in Psychopharmacology: The Fourth Generation of Progress. Edited by

Bloom FE, Kupfer DJ. New York, Raven, 1995, pp 1461–1469

Walsh T, Hadigan CM, Wong LM: Increased pulse and blood pressure associated with desipramine

treatment of bulimia nervosa. J Clin Psychopharmacol 12:163–168, 1992 [PubMed]

Washington C, Haines KA, Tam CW: Amoxapine-induced neuroleptic malignant syndrome. DICP

23:713, 1989 [PubMed]

Weaver GA, Pavinoc D, Davis JS: Hepatic sensitivity to imipramine. Dig Dis 22:551–553, 1977

[PubMed]Print: Chapter 12. Tricyclic and Tetracyclic Drugs http://www.psychiatryonline.com/popup.aspx?aID=409509&print=yes…

33 of 33

10/05/2009 15:56

Weir TA, Goodwin FK: Can antidepressants cause mania and worsen the course of affective illness?

Am J Psychiatry 144:1403–1411, 1987

Wilens TE, Biederman J, Baldessarini RJ, et al: Developmental changes in serum concentrations of

desipramine and 2-hydroxydesipramine during treatment with desipramine. J Am Acad Child

Adolesc Psychiatry 31:691–698, 1992 [PubMed]

Wilens TE, Biederman J, Baldessarini RJ, et al: Electrocardiographic effects of desipramine and

2-hydroxydesipramine in children, adolescents, and adults treated with desipramine. J Am Acad

Child Adolesc Psychiatry 32:789–804, 1993

Wilens TE, Biederman J, Prince J, et al: Six week, double-blind, placebo-controlled study of

desipramine for adult attention deficit hyperactivity disorder. Am J Psychiatry 153:1147–1153,

1996 [Full Text] [PubMed]

Yates M, Leake A, Candy JM, et al: 5-HT2 receptor changes in major depression. Biol Psychiatry

27:489–496, 1990 [PubMed]

Yoshimura M, Furue H: Mechanisms for the anti-nociceptive actions of the descending

noradrenergic and serotonergic systems in the spinal cord. J Pharmacol Sci 101:107–117, 2006

[PubMed]

Young RC, Alexopoulos GS, Shamoian CA, et al: Plasma 10-hydroxy-nortriptyline in elderly

depressed patients. Clin Pharmacol Ther 35:540–544, 1984 [PubMed]

Young RC, Alexopoulos GS, Shamoian CA, et al: Plasma 10-hydroxynortriptyline and ECG changes in

elderly depressed patients. Am J Psychiatry 142:866–868, 1985 [PubMed]

Young RC, Alexopoulos GS, Shindledecker R, et al: Plasma 10-hydroxynortriptyline and therapeutic

response in geriatric depression. Neuropsychopharmacology 1:213–215, 1988 [PubMed]

Ziegler VE, Biggs JT: Tricyclic plasma levels: effect of age, race, sex, and smoking. JAMA

238:2167–2169, 1977 [PubMed]

Zitrin CM, Klein DF, Woerner MG: Treatment of agoraphobia with group exposure in vivo and

imipramine. Arch Gen Psychiatry 37:63–72, 1980 [PubMed]

Copyright © 2009 American Psychiatric Publishing, Inc. All Rights Reserved.