sign in sign up
<<
Home , Antidepressant

John P. Feighner

Mechanism of Action of

John P. Feighner, M.D.

The of is a field that has evolved rapidly in just under 5 decades. Early antidepressant medications— (TCAs) and inhibi- © Copyrighttors (MAOIs)—were discovered2000 throughPhysicians astute clinical observations.Postgraduate These first-generation Press, medica- Inc. tions were effective because they enhanced or noradrenergic mechanisms or both. Unfor- α tunately, the TCAs also blocked histaminic, cholinergic, and 1- receptor sites, and this action brought about unwanted such as , dry mouth, , drowsiness, and . MAOIs can interact with to cause potentially lethal hypertension and present potentially dangerous interactions with a number of medications and over-the-counter . The newest generation of antidepressants, including the single-receptor selective in- hibitors (SSRIs) and multiple-receptor antidepressants , , , trazo- done, and , target one or more specific receptor sites without, in most cases, activat- ing unwanted sites such as and acetylcholine. This paper discusses the new antidepressants, particularly with regard to , and looks at future developments in the treatment of depression. (J Clin Psychiatry 1999;60[suppl 4]:4Ð11) One personal copy may be printed

n just under 5 decades, the psychopharmacology of de- noted the role of in depression in the mid- I pression has evolved from medications discovered 1950s, and Schildkraut2 first proposed the catecholamine through serendipitous clinical observations to drugs de- hypothesis in 1965. The hypothesis suggested that at least signed to target a single receptor site to ones that target some depression could be traced to lowered norepinephrine multiple sites that have been found to be associated with levels at important sites located in the depression. The first antidepressants were identified brain; higher-than-normal levels were suggested to cause through a process comprising serendipitous observations manic behavior. Clinical observations were consistent with followed by minor structural modifications to produce the catecholamine hypothesis. Medications with or structural analogues (e.g., secondary and tertiary amine depressive effects in humans— and tetrabena- tricyclic antidepressants). Most of these agents acted at zine—depleted catecholamines in the brain. , several sites in the brain, each of which might or might not a , released norepinephrine into the brain. Mono- affect depression. In most cases, the mechanism of action amine oxidase inhibitors (MAOIs) increased catecholamine was unknown. The next generation of antidepressants (the levels in the brain; the TCA inhibited the cel- selective serotonin reuptake inhibitors, or SSRIs) was ra- lular uptake and inactivation of serotonin and norepineph- tionally designed to act on one specific , rine. Soon, Lapin and Oxenkrug,3 among others, proposed i.e., serotonin. The newest generation of psychotropic that abnormally low levels of serotonin at certain brain re- medications (e.g., venlafaxine, mirtazapine) was tailored ceptor sites were related to depression; abnormally high to act at several specific receptor sites that are implicated levels were related to . Working separately, Kuhn4 in in depression. Table 11 shows the presence and extent of Switzerland and Lehmann and Kline5 in the United States antidepressant action on each receptor site. This paper re- reported 2 classes of drugs—TCAs and MAOIs—effective counts the brief history of antidepressant medications, in treating depression. These drugs affect both norepineph- with particular emphasis on mechanism of action. rine and serotonin receptors in the brain. The history of the psychopharmacology of depression TCAs interact with a number of other receptor sites, in- is only several decades old. Seymour S. Kety and others cluding histamine, acetylcholine, and epinephrine, and also have substantial atropine-like effects, including dry mouth, dizziness, blurred vision, constipation, sedation, and orthostatic , which can cause falls. Car- From John P. Feighner, M.D., Inc., San Diego, Calif. diovascular effects are common, often restricting treat- Presented at the symposium “Assessing Antidepressant Efficacy: A Reexamination,” held January 20–21, 1998, ment at sufficient dosages. Death from TCA overdose gen- Phoenix, Ariz., which was supported by an unrestricted erally has a cardiovascular etiology secondary to the educational grant from Wyeth-Ayerst Laboratories. Reprint requests to: John P. Feighner, M.D., 5375 Mira -like effect. In overdose, TCAs slow intraven- Sorrento Pl., Suite 210, San Diego, CA 92121. tricular conduction, which causes complete heart block or

4 J Clin Psychiatry 1999;60 (suppl 4) Mechanism of Action of Antidepressant Medications

10 Table 1. Effects of Single-Receptor and Multiple-Receptor drugs. Gardner et al. reviewed literature on dietary re- Newer Antidepressants on CNS * strictions during MAOI treatment and also conducted Antidepressant Serotonin Norepinephrine tyramine assays. They proposed that dietary restrictions be Single-receptor limited and asserted that MAOIs are effective alternatives ++++ 0 0 when patients do not respond to TCAs, particularly for ++++ 0 0 , , and bipolar depression. ++++ 0 0/+ Multiple-receptor Bupropion + + ++a,b SINGLE-RECEPTOR ANTIDEPRESSANTS Mirtazapine +++c ++a 0 d Nefazodone +++ +0 Selective Serotonin Reuptake Inhibitors © Copyright ++d 002000 Physicians Postgraduate Press, Inc. Venlafaxine ++++ ++ 0/+ On the basis of the hypothesis that alterations in recep- *Data from reference 1. Symbols: ++++ = high, +++ = moderate, tor sensitivity may play a role in both the mechanism of ++ = low, + = very low, 0 = none. a action of antidepressant drugs and the pathophysiology of a2 presynaptic antagonist. bAcutely, increases; chronically, stabilizes. the depressive disorders, researchers in Scandinavia began c 5-HT2 and 5-HT3 antagonist. to search for an agent that would act only at serotonin re- d5-HT antagonist. 2 ceptors. was the first such compound to be clinically tested and made available.11 However, because of significant neurotoxicity and immunogenicity, the com- ventricular reentry . TCAs can also be pound was withdrawn from the worldwide market. None- seizurogenic in overdose, and patients who experience sei- theless, clinical trials and clinical experience demon- zures may have sometimes life-threatening broadening of strated that this was an effective antidepressant. Even the QRS duration or hypotension.6 OneDeToledo personal et al.7 copysug- maythough be printedzimelidine was withdrawn from the worldwide gested that in combination with valproic market, the result of this methodical research led to a new acid can lead to increased levels of clomipramine and sub- generation of antidepressants, the SSRIs, which have sequent in patients predisposed to these events. since become the most widely utilized class of antidepres- The increased burden of the TCAs can lead to sants on an international basis. These agents, which in- reduced tolerability in patients, especially when the drugs clude fluoxetine, paroxetine, sertraline, , and are administered over the long term. Brodkin et al.8 looked , combine the effectiveness of their older coun- at 35 adults with DSM-IV diagnoses of pervasive develop- terparts with a much-improved side effect profile. mental disorders treated with clomipramine. They found SSRIs are now prescribed widely in this country and in 13 patients with clinically significant adverse effects, in- many others. Pincus et al.12 tracked the prescriptions writ- cluding 3 patients with seizures. Despite these drawbacks, ten for psychotropic medications in the United States from TCAs were still widely prescribed through the 1970s and 1988 to 1994. Of the patients who were treated for depres- 1980s. Olfson and Klerman,9 analyzing data from 3 Na- sion, 60% were prescribed antidepressants—34% SSRIs tional Ambulatory Medical Care Surveys, found that, in and 26% non-SSRIs and the older generic antidepressants. the 3 years studied (1980, 1985, and 1989), TCAs were the The SSRI class of drugs has a wide range of clinical appli- type of antidepressant most often prescribed by office- cation in the full spectrum of depressive disorders and in based psychiatrists, and they continue to be used com- many other psychiatric disorders, including but not limited monly today. Many European psychiatrists still believe to obsessive-compulsive disorder,13 ,14 so- that the TCAs, and particularly the tertiary amines—such cial ,15,16 eating disorders,17,18 and premenstrual dys- as clomipramine and , which have substantial phoric disorders.19 SSRIs are well tolerated, with safety in effects on both norepinephrine and serotonin—are the overdose and low seizurogenicity. These properties have most potent antidepressants. been confirmed by acute and long-term maintenance stud- MAOIs, which also interact with several receptor sites, ies. Even though SSRIs cost more than the older generic have their own set of problems. Because MAOIs can inter- antidepressants, pharmacoeconomic studies have consis- act with tyramine, causing potentially lethal hypertension, tently demonstrated their indirect and direct cost- patients taking them must adhere to a diet that restricts or effectiveness in terms of use of medical facilities and re- eliminates tyramine-containing foods. This is less critical gained productivity.20 for reversible MAOIs (RIMAs) such as . However, some investigators have suggested that the MAOIs also present interaction problems with a number SSRIs tend to lose efficacy over time. In a review of the of drugs, including other MAOIs, TCAs, and SSRIs. Other literature on clinical trials of antidepressants, Byrne and problematic side effects of the MAOIs include hypoten- Rothschild21 found that depressive symptoms returned sion, weight gain, and . Nevertheless, during maintenance treatment in 9% to 57% of patients, there are a number of patients with depressive disorders most of whom were treated with SSRIs. Fava et al.22 re- who respond better to MAOIs than to any other class of ported that 26 among 77 patients taking a 20 mg/day main-

J Clin Psychiatry 1999;60 (suppl 4) 5 John P. Feighner tenance dose of fluoxetine under double-blind conditions subjects (MADRS: 76% vs. 47%, p = .024; HAM-D: 76% lost efficacy. vs. 41%, p = .006; CGI-Improvement: 76% vs. 47%) for the first 4 weeks of the study. Fewer venlafaxine-treated Selective Norepinephrine Reuptake Inhibitors patients (18%) than fluoxetine-treated patients (37%) with- A number of projects to develop a specific norepineph- drew from the study for any reason. The researchers con- rine have failed, mostly because of car- cluded that venlafaxine was more effective than fluoxetine diovascular side effects. (still being tested in in the treatment of severely depressed hospitalized pa- the United States but in use in Great Britain) appears to be tients. effective and well tolerated.23 Guelfi et al.26 performed a randomized, double-blind, parallel-group study of 93 patients hospitalized with severe MULTIPLE-RECEPTOR© Copyright ANTIDEPRESSANTS 2000 Physiciansdepression Postgraduate who were randomly Press, assigned toInc. treatment with venlafaxine (N = 46) or (N = 47) for 4 weeks. Ef- Venlafaxine and Venlafaxine XR ficacy was measured with the HAM-D, the MADRS, and With the evolution of antidepressants that target spe- the CGI. At day 4, the MADRS score had decreased a cific single receptors have come antidepressant medica- mean ± SD of 3.3 ± 4.6 points in the venlafaxine-treated tions designed to interact with more than one receptor site. patients and 1.0 ± 4.9 points in the placebo-treated patients Venlafaxine, mirtazapine, trazodone, and nefazodone, for (p = .026). At week 1, the HAM-D score had decreased example, affect both serotonin and norepinephrine re- 5.7 ± 6.2 in the venlafaxine group and 3.1 ± 6.8 in the pla- ceptors. Serotonin-norepinephrine reuptake inhibitors cebo group (p = .043). Response, defined as a 50% de- (SNRIs) are a new class of antidepressant, of which crease in MADRS scores, occurred at a rate of 65% for the venlafaxine is currently the only member. This new class venlafaxine group and 28% for the placebo group at end- acts on both serotonin and norepinephrine,One personal but it does copy not maypoint be (p printed≤ .001). All of these results acheived statistical interact with histaminic and cholinergic-adrenergic recep- signifigance for venlafaxine versus placebo and support tors and thus avoids troublesome adverse events such as early onset of action. dry mouth, hypotension, and sedation. Venlafaxine also Delayed onset of action common to all antidepressant interacts very weakly with dopamine receptors, an action medications has prompted the hypothesis that antide- that may have clinical applicability at very high doses. It pressant action is mediated by neuroadaptive changes re- can be titrated to relatively high doses to enhance response sulting from repeated administration that results in down- when clinically indicated. There is a dose response with regulation of the β-adrenergic receptor. Unlike other this , and at higher doses, the adrenergic effects of the antidepressants, venlafaxine has an acute onset of down- drug are increased. regulation of β-adrenergic receptors, which suggests a pos- A number of studies indicate the potential therapeutic sible, as yet unproven, mechanism for early onset of ac- superiority of venlafaxine over SSRIs. Dierick et al.24 con- tion. On the basis of the results of their study, Guelfi et al. ducted an 8-week double-blind comparison of outpatients concluded that venlafaxine constitutes a rapid and effec- with major depression taking fluoxetine (N = 151) or ven- tive treatment for hospitalized patients with major depres- lafaxine (N = 153). Seventy-two percent of patients taking sion and melancholia. venlafaxine or 60% of patients taking fluoxetine regis- Derivan et al.27 evaluated data from 2 randomized, tered a meaningful clinical response, defined as at least a double-blind, placebo-controlled studies to measure the 50% decrease from baseline to endpoint (6 weeks) in the onset of action for venlafaxine in a retrospective study and total Hamilton Rating Scale for Depression (HAM-D) found that venlafaxine-treated patients showed response scores. by day 7. Each study employed the rapid titration of doses Clerc et al.25 conducted a double-blind randomized trial of venlafaxine to a mean daily dose of 200 mg/day within 7 of venlafaxine and fluoxetine in severely depressed inpa- to 8 days. Data were assessed using 3 statistical method- tients. In this 6-week study, venlafaxine-treated patients ologies—traditional analysis of depression scale scores, received 200 mg/day and fluoxetine-treated patients re- pattern analysis based on timing and persistence of re- ceived 40 mg/day. Efficacy was rated using the Clinical sponse, and survival analysis of sustained response. Tradi- Global Impressions scale (CGI), the Montgomery-Asberg tional assessment observed the onset of antidepressant Depression Rating Scale (MADRS), and the HAM-D. The activity (defined as the time of the first statistically signifi- researchers found differences favoring venlafaxine at cant differences between treatment groups): all statistically most time points. The CGI scores followed a similar pat- significant differences between treatment groups occurring tern, with significant differences seen at week 4. The within the first 2 weeks were noted. The pattern analysis MADRS and HAM-D improvements of 50% or greater was developed by Quitkin and colleagues28,29 to recognize from baseline to endpoint were considered meaningful distinctions between true or specific responses and placebo clinical response. Venlafaxine-treated subjects responded or nonspecific responses. In this case, a CGI-Improvement at a significantly higher rate than did the fluoxetine-treated score of 1 or 2 indicated improvement, with distinctions

6 J Clin Psychiatry 1999;60 (suppl 4) Mechanism of Action of Antidepressant Medications made between early (within 2 weeks) and delayed (after 2 mg/day of venlafaxine or 161 patients taking 40 mg/day weeks) responses. The survival analysis of sustained re- of fluoxetine in an 8-week, double-blind comparison. sponse regarded a patient as a sustained responder if a re- Mahapatra and Hackett32 conducted a randomized, sponse (as defined above) was measured for at least 2 con- double-blind study comparing venlafaxine with dothiepin, secutive weeks and then continued for the balance of a TCA used widely in the United Kingdom. Seventeen scheduled treatment. Pattern analysis revealed that, in percent of patients taking dothiepin experienced standing study 1, more than half the venlafaxine-treated patients systolic pressure changes, as opposed to 2% of ven- demonstrated early persistent responses as opposed to lafaxine-treated patients. Electrocardiogram (ECG) inter- 15% of the group receiving placebo, a statistically signifi- val changes with either clinical significance or potentially cant (p < .0001) difference. In study 2, more than 30% of clinical significance were found more often in the venlafaxine-treated© Copyright patients and 4% 2000 of patients Physicians receiving dothiepin-treated Postgraduate than in the Press, venlafaxine-treated Inc. group. placebo demonstrated early persistent responses to treat- This author,33 in a survey of MEDLINE literature covering ment. Survival analysis of sustained response, in study 1, 15 years, found clinically significant increases in blood disclosed that sustained responses were measured for 15% pressure (increase in diastolic blood pressure of ≥ 15 mm of venlafaxine-treated patients, as opposed to 1% of pa- Hg and to ≥ 105 mg Hg from baseline) in 5.5% of patients tients receiving placebo. By day 15, 25% of patients taking more than 200 mg/day of venlafaxine. There was a treated with venlafaxine and 4% of patients receiving pla- lower incidence of clinically significant blood pressure cebo had demonstrated a sustained response. In study 2, increases in patients taking less than 200 mg/day of venla- 4% of venlafaxine-treated patients, as opposed to 1% of faxine. patients receiving placebo, registered a sustained re- Augustin et al.,34 in a survey of venlafaxine clinical tri- sponse; 18% of venlafaxine-treated patients versus 2% of als, also found the frequency of hypertension in patients to placebo-treated patients had achievedOne sustained personal responses copy maybe bedose printed dependent. Hypertension (defined as treatment- by day 15. emergent diastolic blood pressure > 90 mm Hg and dia- Increased blood pressure is problematic in a small per- stolic blood pressure > 10 mm Hg above baseline for 3 centage (2%Ð4%) of venlafaxine-treated patients taking consecutive visits) occurred at a rate of 1.1% in patients doses under 200 mg/day. When dosage is increased to taking 75 mg/day (a rate equal to that in patients taking 200 mg and above, the incidence is somewhat higher placebo). In patients taking 225 mg/day, the rate of hyper- (6%Ð8%), and blood pressure monitoring is required by tension was 2.2%; in patients taking 375 mg/day, it was the Food and Drug Administration. On a clinical basis, the 4.5%. The authors noted the limited number of venlafax- same monitoring is required of patients treated with TCAs ine-treated patients reporting sustained hypertension. and MAOIs. Benkert et al.30 reported that increases in su- Drug-drug interactions affect the potential usefulness pine diastolic blood pressure of potentially clinical signifi- of any . Most psychotropic drugs, including cance (of at least 15 mm Hg and to at least 105 mm Hg) antidepressants, are metabolized by cytochrome P450 occurred in 15% of patients taking venlafaxine and in 18% (CYP) isoenzymes. Most antidepressants competitively of patients taking placebo, with no significant differences inhibit the isoenzymes CYP1A2, CYP2D6, CYP3A3/4 between groups. (Doses of venlafaxine were rapidly esca- (the designation of 2 very similar isoenzymes, lated to 375 mg/day and decreased to 150 mg/day after 10 CYP3A3 and CYP3A4), CYP2C8/9, and CYP2C19, days.) Guelfi et al.26 reported that 3 patients taking 150 to among others. CYP2D6 plays a role in the of 375 mg/day of venlafaxine and 2 treated with a placebo , , clomipramine, , recorded blood pressure increases. Postural hypotension some β-blockers, paroxetine, fluoxetine, and phenothia- occurred in both treatment groups, although twice as often zine . Several SSRIs inhibit CYP2D6. among venlafaxine-treated patients as among those taking CYP3A3/4 metabolizes sertraline, among other agents. placebo. Drug-drug interactions can occur when inhibition of the However, other researchers have failed to find signifi- CYP system causes additive or synergistic drug effects. cant increases in blood pressure in studies of venlafaxine. For example, when a lipophilic β-blocker and a CYP2D6- Gründer et al.31 compared the effects on blood pressure of inhibitor SSRI are used together, the resulting increased patients taking venlafaxine with those of patients taking concentrations of the β-blocker may cause bradycardia. imipramine. Patients taking venlafaxine had their doses Ereshefsky35 has commented that venlafaxine has a favor- rapidly titrated to 375 mg/day, then reduced to 150 mg/day able drug interaction profile, noting that it does not sub- at day 28. Both the group treated with imipramine and the stantially inhibit the activity of isoenzymes CYP2C9, group treated with venlafaxine sustained systolic blood CYP2D6, CYP1A2, or CYP3A3/4. pressure reductions of 5% (imipramine: 4 mm Hg; Because venlafaxine tends to have fewer problems re- venlafaxine: 7 mm Hg) after 14 days. Dierick et al.24 re- lated to drug interactions than older antidepressants, it is ported no clinically significant changes in heart rate or ideal for administration to geriatric patients taking mul- blood pressure in either 153 patients treated with 75 to 150 tiple medications for comorbid medical disorders. Khan et

J Clin Psychiatry 1999;60 (suppl 4) 7 John P. Feighner al.36 conducted an open-label clinical study evaluating pa- Mirtazapine tient acceptance and safety of venlafaxine in depressed ge- Mirtazapine, a noradrenergic and specific serotonergic riatric patients. Fifty-eight patients were recruited for the antidepressant, acts on both serotonin and norepinephrine study, and 24 completed the full 12 months of the study. through a mechanism different from reuptake or enzyme Most of those who failed to complete the study withdrew inhibition. SSRIs apparently work through a single mech- because of treatment-emergent study events; no particular anism of action. Specifically, they inhibit the neuronal up- adverse event occurred more often than another. The most take pump for 5-HT. The subsequent increase in 5-HT common adverse events were , nausea, dry availability, as well as its duration of action, seems to be mouth, and sweating. Only 1 patient had an adverse event responsible for both the beneficial and adverse effects judged definitely© Copyright drug related. Two 2000 patients presented Physicians with commonly Postgraduate associated with Press,SSRI therapy. Inc. The down- elevated blood pressure that was deemed probably drug re- regulation of postsynaptic 5-HT2A receptors and presynap- lated. Adverse conditions in these 3 patients resolved with tic 5-HT1D receptors appears to be responsible for the anti- no medical consequences after they had withdrawn from action of this class of drugs and the adverse the study. events frequently accompanying SSRI use, such as gastro- The availability of venlafaxine extended-release (XR) intestinal disturbance and restlessness. formulation substantially enhances tolerability and ease of Mirtazapine blocks the serotonin receptor subtypes 37 α administration. Rudolph and Derivan reported a double- 5-HT2A, 5-HT2C, and 5-HT3, as well as 2-adrenergic re- blind, placebo-controlled trial investigating the safety and ceptors.39,40 This blockade results in an increase of norad- efficacy of extended-release venlafaxine and fluoxetine in renergic activity and specific serotonergic activity. This a population of 301 patients with major depressive disor- mechanism of action means that many of the side effects der. Patients initially received 75 mg/day of venlafaxine common to TCAs—dry mouth, drowsiness, and constipa- XR, 20 mg/day of fluoxetine, or placebo.One personalVenlafaxine copy XR maytion—and be printed SSRIs—gastrointestinal distress, , and doses could be raised to 150 mg/day at 2 weeks and to 225 sexual dysfunction—will be minimized. Mirtazapine does mg/day at 4 weeks; fluoxetine could be raised to 40 have a substantial antihistaminic effect, however, which mg/day and to 60 mg/day at the same intervals. Seventy- can cause sedation—which can be used therapeutically— one percent of venlafaxine XRÐtreated patients achieved a increased appetite, and weight gain. These side effects oc- CGI global improvement score of 1 or 2 (very much im- cur more often with mirtazapine-treated patients than with proved or much improved), while 62% of fluoxetine- placebo-treated groups. Mirtazapine appears safe in over- treated patients and 52% of placebo-treated patients regis- dose, and a half-life of 20 to 40 hours makes it suitable for tered the same scores in these categories. Thirty-seven once-daily bedtime dosing. The metabolism of this medi- percent of the venlafaxine XRÐtreated patients, 22% of the cation does not depend principally on the isoenzymes fluoxetine-treated patients, and 18% of the placebo-treated CYP2D6 or CYP2C19. In addition, in vitro studies indi- patients achieved full remission (HAM-D total score ≤ 7). cate that mirtazapine does not inhibit the isoenzymes In each of these groups, the difference between rating scale CYP1A2, CYP2D6, or CYP3A. scores of patients treated with venlafaxine XR and those of placebo-treated patients was statistically significant. This Bupropion study provides evidence that full remission may occur The mechanism of action of bupropion is not well un- more frequently in venlafaxine-treated patients than in derstood: it increases whole body turnover of norepineph- those treated with fluoxetine. rine, and to a lesser extent blockades the reuptake of dopa- Double-blind trials have recently demonstrated that ven- mine. Bupropion does not inhibit monoamine oxidase or lafaxine XR is both clinically and statistically superior to interact with histaminic or cholinergic receptors or α placebo as an medication. In an 8-week, double- 1-adrenoceptors, a fact that enhances tolerability. The blind study,38 377 outpatients who met DSM-IV criteria for drug is seizurogenic at levels slightly higher than those of generalized disorder (GAD) but not for major de- typically used TCAs, but the sustained-release formula- pressive disorder were randomly assigned to take 75 tion reduces the risk of seizures. Although clinical trials mg/day, 150 mg/day, or 225 mg/day of venlafaxine XR or have indicated the efficacy of bupropion in seriously de- placebo. The primary outcome measures were the Hamil- pressed inpatient and outpatient populations, bupropion is ton Rating Scale for Anxiety (HAM-A) total score, the not generally used as a broad-spectrum antidepressant.41 HAM-A psychic anxiety factor score, and the CGI scale. In Hurt et al.42 conducted a double-blind, placebo- all cases, the group taking 225 mg/day of venlafaxine XR controlled trial testing the efficacy of 100, 150, and 300 registered statistically significant improvements over the mg/day of bupropion in in 615 patients. group taking placebo. On the basis of these data, the authors After 7 weeks, all groups receiving bupropion improved suggest that venlafaxine XR is safe and effective in the more than the placebo-treated group, with 44 percent of treatment of GAD and may provide an important alterna- the group receiving 300 mg/day of bupropion responding; tive to anxiolytic medications that are currently available. at 1 year 23% of this group reported a sustained response.

8 J Clin Psychiatry 1999;60 (suppl 4) Mechanism of Action of Antidepressant Medications

Bupropion, under the trade name Zyban, is approved by LINKAGE OF SEROTONIN AND the United States Food and Drug Administration for smok- NOREPINEPHRINE IN DEPRESSION ing cessation and is now utilized in primary care . Vetulani and Sulser,47 among others,48 have demon- Trazodone and Nefazodone strated a link between the interaction of serotonin and nor- Trazodone and its later analogue, nefazodone, act as re- epinephrine. Norepinephrine and serotonin receptors inter- uptake inhibitors for both serotonin and norepinephrine act anatomically and pharmacologically. Compounds that α and interact with 1-adrenoceptors. The most potent action affect only serotonin have modulatory effects on both do- is blockade of 5-HT2 postsynaptic receptors. A modest an- pamine and norepinephrine. There is progressive evidence tidepressant, trazodone is used mostly for its and that a single amine system dysfunction is unlikely to be the anxiolytic© effects, Copyright particularly since 2000 the advent Physicians of SSRIs. pathophysiologic Postgraduate mechanism Press,of depression. Inc. Both seroton- One side effect of trazodone is . ergic agents and noradrenergic agents down-regulate the Nefazodone is a more potent antidepressant than trazo- β-adrenoceptor. There is a specific pharmacodynamic link done, and it is much less likely than trazodone to cause between serotonin and norepinephrine effects at the priapism. Common side effects include sedation, impaired G-protein level. In addition, there are recep- concentration, and lethargy. The lethargy can negate the tors in cell bodies containing norepinephrine and serotonin. therapeutic effect of the drug and may make long-term This fact demonstrates a link with glucocorticoid receptors maintenance problematic. On the other hand, lack of in both serotonin and norepinephrine and may be the link antihistaminic and anticholinergic activity improves toler- between stress-induced or stress-aggravated affective ability and safety. Nefazodone lacks quinidine-like episodes. activity with safety in overdose, has a low rate of Communication between serotonergic and noradrener- seizurogenicity, and low rates of sexualOne dysfunction, personal espe- copy maygic bereceptors printed ameliorated by activation of protein kinase cially when compared with SSRIs, venlafaxine, TCAs, has been demonstrated repeatedly.49,50 In theory, then, and MAOIs. drugs that have both noradrenergic and serotonergic ef- Like trazodone, nefazodone does not interact with his- fects, such as venlafaxine, could potentiate the cascade of taminergic or cholinergic receptors.43 It has somewhat less events that leads to downstream changes that may relate α affinity for the 1-adrenergic receptor than many other an- more directly to the core pathophysiology of depression. tidepressants. This receptor is believed to be responsible This could enhance the speed and quality of clinical re- for the sedative side effects of trazodone and may also be sponse and may account for a more robust and earlier onset related to the occurrence of priapism. Nefazodone has a of action demonstrated specifically by venlafaxine, which α 27 lower affinity for 1-adrenergic receptors when compared has a demonstrated dual mechanism of action. with trazodone and amitriptyline.43Ð45 Looking at the re- Cleare et al.51 used d-, a specific 5-HT re- sults of clinical trials examining the effects of nefazodone leasing agent lacking the effects of in a combined population of 593 depressed patients, dl-fenfluramine, as a serotonergic neuroendocrine chal- Augustin et al.34 reported no dose-dependent incidents of lenge in subjects with unipolar major depression. After cor- abnormal ejaculation or orgasm. Overall, 0.2% of the tisol and responses to 30 mg of d-fenfluramine population experienced this adverse event. Blocking were measured, the patients were randomly assigned to

5-HT2A as well as inhibiting the reuptake of 5-HT, nefazo- treatment with either a specific norepinephrine reuptake in- done appears to possess a dual mechanism of action on hibitor, a TCA, or placebo. The patients treated with the the serotonergic system. The agent produces a dose- norepinephrine reuptake inhibitor demonstrated a substan- dependent reuptake inhibition of 5-HT. Nefazodone also tial increase in 5-HTÐmediated cortisol responses whether modestly inhibits the reuptake of norepinephrine. This in- or not the depression responded to treatment. The authors hibition is less apparent after the chronic administration of concluded that antidepressants selectively modifying nor- nefazodone than after acute administration. Nefazodone adrenergic function affect 5-HT function, measured by neu- appears to be safe in overdose. roendocrine testing, as well. A recent double-blind, placebo-controlled inpatient In a placebo-controlled procedure, Mann et al.52 com- study conducted by Feighner et al.46 evaluated nefazodone pared regional brain glucose metabolism after the in the treatment of 120 inpatients with marked-to-severe serotonin-releasing drug dl-fenfluramine had been admin- major depressive disorder. Fifty percent of nefazodone- istered to 6 healthy inpatients and 6 drug-free inpatients treated patients responded, compared with 29% of with moderately severe major depression. They reported placebo-treated patients. In this study, nefazodone was su- that, upon evidence of positron emission tomography (PET) perior to placebo in the treatment of major depression, scans, healthy inpatients evinced several statistically sig- with significant (p < .001) clinical benefits based on nificant (p < .01) areas of both increased and decreased me- HAM-D and MADRS scores noted as early as week 1 of tabolism. Depressed patients did not exhibit these metabolic treatment. changes. The authors assert that the results of this study

J Clin Psychiatry 1999;60 (suppl 4) 9 John P. Feighner provide visual evidence corroborating the role of impaired 3. Lapin IP, Oxenkrug GF. Intensification of the central serotonergic pro- cesses as a possible determinant of the thymoleptic effect. Lancet 1969;1: serotonergic transmission in depression. 132Ð136 53 Geracioti et al. found no substantial differences 4. Kuhn R. The treatment of depressive states with G22355 (imipramine hy- in concentrations of the neurotransmitters , drochloride). Am J Psychiatry 1958;115:459Ð464 5. Lehmann HE, Kline NS. Clinical discoveries with antidepressant drugs. In: 5-hydroxyindoleacetic acid (5-HIAA), norepinephrine, or Parnham MJ, Bruinvels J, eds. Discoveries in , vol 1. 3-methyoxy-4-hydroxyphenylglycol (MHPG) in the cere- Amsterdam, the Netherlands: Elsevier; 1983:209Ð221 brospinal fluid (CSF) of 10 patients with major depression 6. Taboulet P, Michard F, Muszynski J, et al. Cardiovascular repercussions of seizures during cyclic antidepressant poisoning. Clin Toxicol 1995;33: and 10 healthy controls. They did observe, however, a 205Ð211 negative linear relationship between mean concentrations 7. DeToledo JC, Haddad H, Ramsay RE. Status epilepticus associated with the combination of valproic acid and clomipramine. Ther Drug Monit of 5-HIAA and norepinephrine in the CSF of the healthy © Copyright 2000 Physicians 1996;19:71Ð73Postgraduate Press, Inc. volunteers that was absent in the CSF of the depressed 8. Brodkin ES, McDougle CJ, Cohen DJ, et al. Clomipramine in adults with patients. They report that these findings support the hy- pervasive developmental disorders: a prospective open-label investigation. pothesis that a disturbance in the relationship between ser- J Child Adolesc Psychopharmacol 1997;7:109Ð121 9. Olfson M, Klerman GL. Trends in the prescription of antidepressants by otonergic and noradrenergic systems can be found in de- office-based psychiatrists. Am J Psychiatry 1993;150:571Ð577 pressive illness when no simple 5-HT or norepinephrine 10. Gardner DM, Shulman KI, Walker SE, et al. The making of a user friendly deficit or surplus exists. MAOI diet. J Clin Psychiatry 1996;57:99Ð104 11. Montgomery SA, McAuley R, Rani SJ, et al. A double-blind comparison of zimelidine and amitriptyline in endogenous depression. Acta Psychiatr FUTURE RESEARCH Scand 1981;63(suppl 290):314Ð327 12. Pincus HA, Tanielian TL, Marcus SC. Prescribing trends in psychotropic medications: primary care, psychiatry, and other medical specialities. Both serotonergic and noradrenergic agents down- JAMA 1998;279:526Ð531 regulate the β-adrenoceptors. There is a specific pharma- 13. Leonard HL. New developments in the treatment of obsessive-compulsive codynamic link between 5-HT and norepinephrineOne personal effects copy may disorder.be printed J Clin Psychiatry 1997;58(suppl 14):39Ð45 14. Oehrberg S, Cristiansen PE, Behnke K, et al. Paroxetine in the treatment of at the G-protein level. There are glucocorticoid receptors panic disorder: a randomised, double-blind, placebo-controlled study. Br J in cell bodies that contain both serotonin and norepineph- Psychiatry 1995;167:374Ð379 rine receptors, indicating a likely link between glucocorti- 15. Katzelnick DJ, Kobak KA, Greist JH, et al. Sertraline for social phobia: a double-blind, placebo-controlled crossover study. Am J Psychiatry 1995; coid receptors in both of these systems and perhaps a link 152:1368Ð1371 between stress and induced or aggravated affective epi- 16. Davidson JRT. Pharmacotherapy of disorder. J Clin Psychia- sodes. A cross-talk between 5-HT and norepinephrine re- try 1998;59:47Ð53 17. Marcus MD, Wing RR, Ewing L, et al. A double-blind, placebo-controlled ceptors expedited by activation of a protein kinase has trial of fluoxetine plus behavior modification in the treatment of obese been repeatedly demonstrated. In theory, antidepres- binge-eaters and nonÐbinge-eaters. Am J Psychiatry 1990;147:876Ð881 sants—such as venlafaxine—that affect both 5-HT and 18. Hudson JI, McElroy SL, Raymond NC, et al. Fluvoxamine treatment of : a multicenter, placebo-controlled trial. In: New Re- norepinephrine receptors could create reactions that po- search Program and Abstracts of the 147th Annual Meeting of the Ameri- tentiate events leading to downstream changes relative to can Psychiatric Association; May 26, 1994; Philadelphia, Pa. Abstract the core pathophysiology of depression. These changes NR620:218 19. Steiner M, Steinberg S, Stewart D, et al. Fluoxetine in the treatment of pre- could improve the rapidity and quality of clinical re- menstrual dysphoria. Canadian Fluoxetine/Premenstrual Dysphoria Col- sponse. laborative Study. N Engl J Med 1995;332:1529Ð1534 The overview of the history of antidepressant medica- 20. Jönsson B, Bebbington PE. What price depresssion? the cost of depression and the cost-effectiveness of pharmacological treatment. Br J Psychiatry tion has gone from the early discoveries of TCAs and 1994;164:665Ð673 MAOIs to drugs designed to target specific receptor sites 21. Byrne SE, Rothschild AJ. Loss of antidepressant efficacy during mainte- in the brain. Increasing knowledge enables us not only to nance therapy: possible mechanisms and treatments. J Clin Psychiatry 1998;59:279Ð288 create more effective antidepressants rationally but also to 22. Fava M, Rappe SM, Pava JA, et al. in patients on long-term fluox- understand the limitations of existing drugs. etine treatment: response to increased fluoxetine dose. J Clin Psychiatry 1995;56:52Ð55 Drug names: amitriptyline (Elavil and others), bupropion (Wellbutrin, 23. Mucci M. Reboxetine: a review of antidepressant tolerability. J Psycho- Zyban), citalopram (Celexa), clomipramine (Anafranil), desipramine pharmacol 1997;11:S33ÐS37 (Norpramin and others), fenfluramine (Pondimin), fluoxetine (Prozac), 24. Dierick M, Ravizza L, Realini R, et al. A double-blind comparison of ven- fluvoxamine (Luvox), imipramine (Tofranil and others), mirtazapine lafaxine and fluoxetine for treatment of major depression in outpatients. (Remeron), nefazodone (Serzone), nortriptyline (Pamelor and others), Prog Neuropsychopharmacol Biol Psychiatry 1995;30:57Ð71 paroxetine (Paxil), reserpine (Serpasi and others), risperidone 25. Clerc GE, Ruimy P, Verdeau-Pailles J, on behalf of the Venlafaxine French (Risperdal), sertraline (Zoloft), trazodone (Desyrel and others), valproic Inpatient Study Group. A double-blind comparison of venlafaxine and flu- acid (Depakene and others), venlafaxine (Effexor). oxetine in patients hospitalized for major depression and melancholia. Int Clin Psychopharmacol 1994;9:139Ð143 26. Guelfi JD, White C, Hackett D, et al. Effectiveness of venlafaxine in pa- REFERENCES tients hospitalized for major depression and melancholia. J Clin Psychiatry 1995;56:450Ð458 1. Schatzberg AF, DeBattista C, Ereshefsky L, et al. Current psychotropic 27. Derivan A, Entsuah AR, Kikta D. Venlafaxine: measuring the onset of anti- dosing and monitoring guidelines. Prim Psychiatry 1997:35Ð61 depressant action. Psychopharmacol Bull 1995;31:439Ð447 2. Schildkraut JJ. The catecholamine hypothesis of affective disorders: a 28. Quitkin FM, Rabkin JK, Ross D, et al. Identification of true drug response review of supporting evidence. Am J Psychiatry 1965;122:509Ð522 to antidepressants. Arch Gen Psychiatry 1984;41:782Ð786

10 J Clin Psychiatry 1999;60 (suppl 4) Mechanism of Action of Antidepressant Medications

29. Quitkin FM, Rabkin JD, Markowitz JM, et al. Use of pattern analysis to of nefazodone, a novel antidepressant drug. J Clin Psychiatry 1995; identify true drug response. Arch Gen Psychiatry 1987;44:259Ð264 56(suppl 6):3Ð11 30. Benkert O, Gründer G, Wetzel H, et al. A randomized, double-blind com- 44. Davis R, Whittington R, Bryson HM. Nefazodone: a review of its pharma- parison of a rapidly escalating dose of venlafaxine and imipramine in inpa- cology and clinical efficacy in the management of major depression. Drugs tients with major depression and melancholia. J Psychiatr Res 1998; 1997;53:608Ð636 30:441Ð451 45. Stoudemire A. Expanding psychopharmacologic treatment options for the 31. Gründer G, Wetzel H, Schlöβer R, et al. Subchronic antidepressant treat- depressed medical patient. Psychosomatics 1995;36:S19ÐS26 ment with venlafaxine or imipramine and effects on blood pressure and 46. Feighner J, Targum SD, Bennett ME, et al. A double-blind, placebo- heart rate: assessment by automatic 24-hour monitoring. Pharmaco- controlled trial of nefazodone in the treatment of patients hospitalized for psychiatry 1996;29:72Ð78 major depression. J Clin Psychiatry 1998;59:246Ð253 32. Mahapatra SN, Hackett D. A randomised, double-blind, parallel-group 47. Vetulani J, Sulser F. Action of various antidepressant treatments reduces comparison of venlafaxine and dothiepin in geriatric patients with major reactivity of noradrenergic cyclic AMP-generating system in limbic fore- depression. Int J Clin Pract 1997;51:209Ð213 brain [letter]. Nature 1975;257:495Ð496 33. Feighner© JP. Cardiovascular Copyright safety in depressed 2000 patients: Physicians focus on venla- 48. TaylorPostgraduate SS, Buechler JA, Yonemoto Press, W. cAMP-dependent Inc. protein kinase: faxine. J Clin Psychiatry 1995;56:574Ð579 framework for a diverse family of regulatory enzymes. Annu Rev Biochem 34. Augustin BG, Cold JA, Jann MW. Venlafaxine and nefazodone, two 1990;59:971Ð1005 pharmacologically distinct antidepressants. Pharmacotherapy 1997;17: 49. Nibuya M, Nestler EJ, Duman RS. Chronic antidepressant administration 511Ð530 increases the expression of cAMP response element binding protein 35. Ereshefsky L. Drug-drug interactions involving antidepressants: focus on (CREB) in rat hippocampus. J Neurosci 1996;16:2365Ð2372 venlafaxine. J Clin Psychopharmacol 1996;16(suppl 2):37SÐ53S 50. Popoli M, Vocaturo C, Perez J, et al. Presynaptic Ca2+/calmodulin-depen- 36. Khan A, Rudolph R, Baumel B, et al. Venlafaxine in depressed geriatric dent protein kinase II: autophosphorylation and activity increase in the hip- outpatients: an open-label study. Psychopharmacol Bull 1995;31:753Ð758 pocampus after long-term blockade of serotonin. Mol Pharmacol 1995; 37. Rudolph RL, Derivan AT, for the Venlafaxine XR 211 Study Group. 48:623Ð629 Double-blind, placebo-controlled study of once-daily venlafaxine XR and 51. Cleare AJ, Murray RM, O’Keane V. Do noradrenergic reuptake inhibitors fluoxetine in depressed outpatients. In: Scientific Abstracts of the 36th An- affect serotonergic function in depression? Psychopharmacology 1997; nual Meeting of the American College of ; De- 134:406Ð410 cember 8Ð12, 1997; Kamuela, Hawaii. page 173 52. Mann JJ, Malone KM, Diehl DJ, et al. Demonstration in vivo of reduced 38. Haskins T, Rudolph R, Pallay A, et al, for the Venlafaxine XR 210 Study serotonin responsivity in the brain of untreated depressed patients. Am J Group. Double-blind, placebo-controlled studyOne of once personal daily venlafaxine copy may Psychiatrybe printed 1996;153:174Ð182 XR in outpatients with generalized (GAD). In: Scientific 53. Geracioti T Jr, Loosen PT, Ekhator NN, et al. Uncoupling of serotonergic Abstracts of the 36th Annual Meeting of the American College of Neuro- and noradrenergic systems in depression: preliminary evidence from con- psychopharmacology; December 8Ð12, 1997; Kamuela, Hawaii. page 173 tinuous cerebrospinal fluid sampling. Depress Anxiety 1997;6:89Ð94 39. Preskorn SH. Selection of an antidepressant: mirtazapine. J Clin Psychiatry 1997;58(suppl 6):3Ð8 40. Stimmel GL, Depheide JA, Stahl SM. Mirtazapine: an antidepressant with DISCLOSURE OF OFF-LABEL USAGE noradrenergic and specific serotonergic effects. Pharmacotherapy 1997;17: 10Ð21 41. Ascher JA, Cole JO, Colin J-N, et al. Bupropion: a review of its mechanism The author of this article has determined that, to the of antidepressant activity. J Clin Psychiatry 1995;56:395Ð401 best of his clinical estimation, no investigational or off-la- 42. Hurt RD, Sachs DP, Glover ED, et al. A comparison of sustained-release bel information about pharmaceutical agents has been pre- bupropion and placebo for smoking cessation. N Engl J Med 1997;337: 1195Ð1202 sented that is outside Food and Drug AdministrationÐ 43. Taylor DP, Carter RB, Eison AS, et al. Pharmacology and neurochemistry approved labeling.

J Clin Psychiatry 1999;60 (suppl 4) 11