The American Psychiatric Association (APA) has updated its Privacy Policy and Terms of Use, including with new information specifically addressed to individuals in the European Economic Area. As described in the Privacy Policy and Terms of Use, this website utilizes cookies, including for the purpose of offering an optimal online experience and services tailored to your preferences.

Please read the entire Privacy Policy and Terms of Use. By closing this message, browsing this website, continuing the navigation, or otherwise continuing to use the APA's websites, you confirm that you understand and accept the terms of the Privacy Policy and Terms of Use, including the utilization of cookies.

×
Published Online:

Abstract

Objective: Few controlled studies have addressed the issue of which antidepressant medications should be recommended for outpatients who have not responded to multiple treatment trials. This study compared the efficacy of switching to mirtazapine to that of switching to a tricyclic antidepressant (nortriptyline) following two prospective, consecutive, unsuccessful medication treatments for nonpsychotic major depressive disorder. Method: Following lack of remission or an inability to tolerate an initial trial of citalopram for up to 12 weeks (first step) and a second trial with either monotherapy involving another antidepressant or augmentation of citalopram with bupropion or buspirone (second step), adult outpatients (N=235) with nonpsychotic major depressive disorder were randomly assigned to 14 weeks of treatment with mirtazapine (up to 60 mg/day) (N=114) or nortriptyline (up to 200 mg/day) (N=121). The primary outcome, symptom remission, was defined a priori as a total exit score of ≤7 on the 17-item Hamilton Rating Scale for Depression. The 16-item Quick Inventory of Depressive Symptomatology–Self-Report (QIDS-SR 16 ), obtained at treatment visits, provided secondary outcomes of remission (score ≤5 at exit) and response (≥50% reduction in score from baseline). Results: For mirtazapine, remission rates were 12.3% and 8.0% per the Hamilton and QIDS-SR 16 scores, respectively. For nortriptyline, remission rates were 19.8% and 12.4%, respectively. QIDS-SR 16 response rates were 13.4% for mirtazapine and 16.5% for nortriptyline. Neither response nor remission rates statistically differed by treatment, nor did these two treatments differ in tolerability or adverse events. Conclusions: Switching to a third antidepressant monotherapy regimen after two consecutive unsuccessful antidepressant trials resulted in low remission rates (<20%) among patients with major depressive disorder.

Major depressive disorder is associated with substantial morbidity, mortality, family burden, and health care costs (1) . In double-blind efficacy trials typically conducted with outpatients who have uncomplicated, nonchronic, nonrefractory major depressive disorder, initial treatment with antidepressant medications appears to lead to remission in only 35%–47% of patients (2 , 3) . In a recent report, among representative patients in primary and psychiatric care, roughly 30% reached remission after 8–12 weeks of therapy with citalopram (4) . Consequently, a large majority of patients will warrant subsequent treatment regimens in order to achieve remission.

Most second or third treatment steps following initial antidepressant failure have been evaluated in uncontrolled trials, typically in psychiatric outpatients with minimal general medical comorbidity recruited as volunteers and treated in research clinics (5) , which hampers the generalizability of findings to clinical practice. There is a clear paucity of data on whether switching antidepressants for the third time is a clinically useful strategy for depressed patients who have not had adequate responses to two prior antidepressant treatments. In a recent uncontrolled efficacy trial, switching to a third antidepressant led to remission rates of 36% to 50%, similar to the rates of 48% to 50% observed with the first and second antidepressant trials within the same study (6) . These results suggest that the strategy of switching antidepressants following two consecutive antidepressant treatment failures has robust efficacy. However, an uncontrolled effectiveness trial with 115 patients found only a 21% remission rate for the third antidepressant trial (7) . Only one trial has included participants who had not received adequate benefit from two initial antidepressant treatments who were then randomly assigned to at least two potentially effective third-step treatments (venlafaxine or paroxetine). An overall remission rate of 27% after 4 weeks of treatment was found (8) . In that study, however, nonresponse to the first antidepressant trial was only ascertained historically, and the second treatment had to have been prescribed by the investigator at an effective dose only 4 weeks or more before the first day of the study, or 2 weeks or more if a safety problem had caused discontinuation.

Two open trials have suggested the potential usefulness of switching to the atypical antidepressant mirtazapine (9) or to the tricyclic antidepressant nortriptyline (10) for patients with major depression whose index episode did not respond adequately to at least one previous antidepressant trial.

This report focuses on the outcomes of participants who were randomly assigned to one of two third-step antidepressant switch strategies (mirtazapine or nortriptyline) as part of the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial (11 , 12) . STAR*D is the first trial to evaluate the relative effectiveness of switching to these two antidepressants for primary and psychiatric care patients with major depressive disorder who did not adequately benefit from an initial prospective trial with a selective serotonin reuptake inhibitor (SSRI) and a subsequent antidepressant medication. The equipoise stratified randomized design allowed patients at specific steps to select the treatment strategies they considered acceptable and from which their random assignment would be determined (13) . This study assessed the effectiveness of switching to mirtazapine compared with switching to nortriptyline for treatment of major depressive disorder following two consecutive unsuccessful antidepressant regimens.

Method

Participants

The institutional review boards at the STAR*D National Coordinating Center, the Data Coordinating Center, all regional centers, and relevant clinical sites as well as the NIMH Data Safety and Monitoring Board (Bethesda, Md.) approved and monitored the protocol. All participants provided written informed consent at enrollment into the initial (Level 1) treatment with citalopram and at enrollment into each subsequent treatment step, including the third-step (Level 3) antidepressant switch strategies reported herein.

Outpatients with a primary diagnosis of nonpsychotic major depressive disorder, per DSM-IV and confirmed by a checklist completed by the clinical research coordinators, were enrolled between July 2001 and April 2004 at 18 primary and 23 psychiatric practice settings serving both public and private sector patients. Advertising for participants was proscribed. Broad inclusion and minimal exclusion criteria were used to maximize the generalizability of findings (4 , 11 , 12) . More specifically, excluded patients were those with bipolar or psychotic disorders, primary diagnoses of obsessive-compulsive or eating disorders, general medical conditions contraindicating the use of protocol medications in the first two treatment steps, substance dependence (only if it required inpatient detoxification), a clear history of nonresponse or inability to tolerate any protocol treatment during the first two treatment steps for the current major depressive episode, and those who were pregnant or breastfeeding (4 , 11 , 12) .

Eligible participants for second-step treatment (Level 2) had either not achieved remission or were not able to tolerate treatment with citalopram (up to 12 weeks). Nonremission following citalopram was defined as a score of >5 on the clinician-rated, 16-item Quick Inventory of Depressive Symptomatology (QIDS-C 16 ) (1417) obtained at the last citalopram treatment visit.

Participants in the second step (Level 2) of STAR*D agreed to random assignment to one of four alternate monotherapy options (bupropion [sustained release], cognitive therapy, sertraline, or venlafaxine [extended release]) or one of three citalopram augmentation options (bupropion [sustained release], buspirone, or cognitive therapy) (13) . Thus, Level 2 of STAR*D evaluated the comparative effectiveness of different medication switch strategies (18) and augmentation options (19) . Those achieving remission in Level 2 entered a 12-month naturalistic follow-up phase. Those achieving response but not remission (QIDS-C 16 score of 6–8) at this or subsequent levels were encouraged to enter the next level, but they could opt to enter the naturalistic follow-up. Those with an unsatisfactory response entered Level 3. Participants with an unsatisfactory response to both citalopram (Level 1) and to cognitive therapy during Level 2, whether cognitive therapy was used as a switch strategy or an augmentation option, entered Level 2A, which compared the effectiveness of two medication switch strategies (bupropion [sustained release] or venlafaxine [extended release]). Level 2A ensured that all participants who entered Level 3 had an unsatisfactory response to two different antidepressants. Those with remission or a satisfactory response to Level 2A entered the 12-month follow-up, whereas those with an unsatisfactory response entered Level 3. Therefore, eligible participants for third-step treatment entered Level 3 if they had not achieved remission or were unable to tolerate Level 2 or Level 2A treatments. Study participants were not required to meet major depressive disorder criteria at the time of entry into Level 3, as long as they had met major depressive disorder criteria at entry into Level 1 and had not adequately responded or been able to tolerate previous levels.

Level 3 compared the relative effectiveness of two antidepressant switch strategies (mirtazapine or nortriptyline) and two augmentation options (lithium or T 3 thyroid hormone) using the equipoise stratified randomized design (12) . This report presents the main outcomes that compared the two Level 3 switch strategies (mirtazapine versus nortriptyline) to which participants were randomly assigned in a 1:1 ratio, stratified by acceptability to subjects and regional center.

Protocol Treatment

To mimic clinical practice, enhance safety, and ensure vigorous dosing, participants and treating clinicians were not masked to either treatment assignment or dose. A clinical treatment manual (www.star-d.org) recommended starting doses and dose changes for each medication guided by both symptom ratings (per the QIDS-C 16 [ 15 , 16 ]) and side effect ratings (per the Frequency, Intensity, and Burden of Side Effects Rating [12] ) obtained at each treatment visit. Furthermore, didactic instruction, clinical research coordinator support, and a centralized monitoring system (20) with feedback constituted intense efforts to assure timely dose increases when inadequate symptom reduction occurred in the context of acceptable side effects. Clinical management aimed to achieve symptom remission (operationally defined for clinic personnel as a QIDS-C 16 score ≤5 at exit from the treatment). The protocol recommended medication clinic visits at weeks 0, 2, 4, 6, 9, and 12, but visit schedules were flexible (e.g., the week 2 visit could occur acceptably within 6 days of week 2). Extra visits could be scheduled if needed, and for a participant who showed a response or remission only at week 12, two additional visits were permitted to determine whether that status was sustained.

At participant entry into this Level 3 switch trial, citalopram, along with the Level 2 augmenting agents bupropion and buspirone, were discontinued without tapering at the initial Level 3 treatment visit, as were the Level 2 and 2A monotherapy regimens (bupropion, sertraline, and venlafaxine). Either mirtazapine or nortriptyline was begun without a washout period. Recommended mirtazapine doses were 15 mg/day for the first 7 days, 30 mg/day by day 8, 45 mg/day by day 28, and, if necessary, 60 mg/day by day 42. Recommended nortriptyline doses were 25 mg/day for 3 days, 50 mg/day for 4 days, and then 75 mg/day by day 8, 100 mg/day by day 28, and, if necessary, 150 mg/day by day 42. These dosing recommendations were flexible, based on clinical judgment informed by the side effect and the QIDS-C 16 ratings obtained at each treatment visit, and guided by the clinical manual (www.star-d.org). Clinicians were allowed (but not required) to use measurements of nortriptyline blood levels to guide dosing decisions. During the course of the study, nortriptyline blood levels were obtained in 33.9% of the participants; the mean nortriptyline blood level was 91.66 ng/ml (SD=86.03).

Concomitant Treatments

Stimulant, anticonvulsant, antipsychotic, mood stabilizing, nonprotocol antidepressant medications, and potential antidepressant augmenting agents (e.g., buspirone) were proscribed. Otherwise, any concomitant medication was allowed to manage concurrent general medical conditions or protocol antidepressant side effects (e.g., sexual dysfunction), as were anxiolytics (except alprazolam) and sedative hypnotics (including trazodone, ≤200 mg h.s., for sleep).

Measures

Clinical and demographic features were defined at baseline study entry before Level 1 treatment (4) . Baseline measures included the Cumulative Illness Rating Scale (21) to assess general medical conditions, the Psychiatric Diagnostic Screening Questionnaire (22 , 23) to assess comorbid psychiatric disorders, and the 30-item Inventory of Depressive Symptomatology–Clinician-Rated (24) completed by both the clinical research coordinator and the research outcomes assessor to assess depression severity and selected symptom features. Assessments of overall functioning (Short-Form Health Survey [25] , Work Productivity and Activity Impairment Questionnaire [26] , Work and Social Adjustment Scale [27] ) and satisfaction (Quality of Life Enjoyment and Satisfaction Questionnaire [28] ) were collected by an automated interactive voice response telephone system (2931) .

The primary outcome, symptom remission, was defined as a total score ≤7 on the 17-item Hamilton Rating Scale for Depression obtained via telephone-based structured interviews (English or Spanish) conducted by independent, treatment-masked research outcomes assessors within 5 days of entry and exit from this study. The Hamilton depression scale items of psychomotor retardation and agitation were evaluated by the research outcomes assessors on the basis of study participant self-report. The intraclass correlation coefficient for Hamilton depression scale administered via telephone by research outcomes assessors and in person by the clinical research coordinators was 0.66 (N=3,876). In addition, research outcomes assessors participated in numerous training sessions throughout the conduct of the study to prevent rater drift and to ensure a consistent approach to symptom assessment. Secondary outcomes included QIDS-SR 6 scores and side effect ratings obtained at each treatment visit. For QIDS-SR 16 scores, remission was defined as a total score ≤5 at Level 3 exit, while response was a reduction of ≥50% from the Level 3 baseline score.

Statistical Methods

Student’s t tests, Wilcoxon tests, and chi-square tests were used to compare the baseline clinical and demographic features, treatment features, side effects, and serious adverse event rates across treatments and for the entire cohort.

All analyses were conducted using all participants randomly assigned to each treatment group (13) . Remission was defined as Hamilton depression scale total score ≤7 (per masked rater assessment) and QIDS-SR 16 score ≤5 (obtained at each treatment visit). The remission threshold for the self-reported depressive symptom inventory score was established using item response theory analysis and was chosen because it corresponds to a score of 7 or less on the Hamilton depression scale (15) . Logistic regression models were used to compare the remission and response rates after adjusting for the effect of regional center, the two acceptability strata, and baseline clinical and demographic factors that were not balanced across treatment groups (history of prior suicide attempt was the only such factor identified). Times to first remission (QIDS-SR 16 score ≤5) and first response (≥50% reduction from baseline) were defined as the first observed point using clinic visit data. Log-rank tests were used to compare the cumulative proportion with remission and response across the two treatment groups. Additional exploratory logistic regression analyses were conducted to determine if there was a differential effect of treatment among those who were unable to tolerate their prior medication treatment.

When outcome Hamilton depression scale scores were missing, participants were assumed to not have achieved remission (as defined in the original proposal [12] ). Sensitivity analyses were conducted to determine if the method of addressing the missing data had an impact on the results of the study. An additional method of addressing the missing data, using an imputed value generated from an item response theory analysis of the relationship between the Hamilton depression scale and the QIDS-SR 16 score, was used in the analysis of remission based on the Hamilton depression scale.

RESULTS

As previously reported (4) , the overall Hamilton depression scale remission rate in Level 1 was 27.5% (N=790 of 2,876). At Level 2, remission rates, based on the Hamilton depression scale, did not differ significantly between the three medication switch strategies: 21.3% (N=51 of 239) with sustained-release bupropion, 17.6% (N=42 of 238) with sertraline, and 24.8% (N=62 of 250) with extended-release venlafaxine (18). Level 2 Hamilton depression scale remission rates were also similar across drug augmentation strategies: 29.7% (N=83 of 279) achieved remission with citalopram plus sustained-release bupropion, and 30.2% (N=86 of 285) achieved remission with citalopram plus buspirone (19) . Level 2A Hamilton depression scale remission rates did not differ significantly between the two medications: 13.3% (N=2 of 15) with sustained-release bupropion and 6.3% (N=1 of 16) with extended-release venlafaxine.

Patient Disposition

Figure 1 shows the generation of the Level 3 treatment groups by acceptability strata. Of 250 participants accepting random assignment to at least the two medication switch options, 88.4% (N=221) had accepted only these two switch strategies, while 11.6% (N=29) had accepted the two switch strategies and the two augmentation options (i.e., lithium or thyroid hormone).

Figure 1. The STAR*D Level 3 Treatment Groups

Demographic and Clinical Characteristics

Demographic characteristics of the patients by Level 3 medication are presented in Table 1 . Table 2 reveals that trial participants (N=235) often had recurrent (69.6%), early onset (first major depressive episode prior to age 18) (28.4%), or chronic (index episode ≥2 years) (28.9%) depressive episodes with low functioning. One or more concurrent axis I disorders, per the Psychiatric Diagnostic Screening Questionnaire, were present in 66.4% of the participants. Of the 235 enrollees, 69.7% did not have an adequate response or were unable to tolerate one of the three medication switch strategies at Level 2 (bupropion, sertraline, or venlafaxine) or Level 2A (bupropion, venlafaxine). The remaining participants did not have an adequate response or were unable to tolerate the two Level 2 augmentation options (citalopram plus bupropion or citalopram plus buspirone). Among the participants in the present study, 49.5% were unable to tolerate their Level 2 treatment, defined as exiting Level 2 prior to 4 weeks for any reason, or exiting at or after 4 weeks because of intolerable side effects. The two medication groups were similar, except that more participants in the mirtazapine group (24.6%) had previously attempted suicide than those in the nortriptyline group (12.4%).

Treatment Features

The course of treatment with each medication in this trial (Level 3) was similar for both medications and was adequately dosed for substantial periods of time ( Table 3 ). The Level 3 12-week completion rates were 33.3% for mirtazapine-treated patients and 30.6% for nortriptyline-treated patients.

Symptomatic Outcomes

Remission rates, per the Hamilton depression scale, did not differ by treatment. They were 12.3% (N=14 of 114) for mirtazapine and 19.8% (N=24 of 121) for nortriptyline ( Table 4 ). As measured by QIDS-SR 16 scores, remission (8.0% versus 12.4%) and response (13.4% versus 16.5%) rates also did not differ by treatment nor did percent reductions in QIDS-SR 16 score (baseline to exit). Even when stratified by ability to tolerate prior level medication, the treatment outcomes were not significantly different.

The treatments also did not differ in either time to remission (log rank χ 2 =0.18, p=0.68) ( Figure 2 ) or time to response (log rank χ 2 =0.23, p=0.60) ( Figure 3 ) as determined by the QIDS-SR 16 score. For those who did achieve remission, the mean time to remission was 5.7 weeks (SD=4.0) for mirtazapine and 6.3 weeks (SD=4.4) for nortriptyline. Similarly, for those who did achieve response, the mean time to response was 6.9 weeks (SD=4.0) for mirtazapine and 6.3 weeks (SD=4.1) for nortriptyline.

Figure 2. Time to Remission a for Outpatients With Nonpsychotic Major Depressive Disorder Randomly Assigned to Mirtazapine or Nortriptyline After Two Consecutive Failed Medication Treatments

a Score ≤5 on the 16-item Quick Inventory of Depressive-Symptomatology–Self Rated.

Figure 3. Time to Response a for Outpatients With Nonpsychotic Major Depressive Disorder Randomly Assigned to Mirtazapine or Nortriptyline After Two Consecutive Failed Medication Treatments

a ≥50% improvement from Level 3 baseline score on the 16-item Quick Inventory of Depressive-Symptomatology–Self Rated.

Tolerability and Adverse Events

Side effects and serious adverse events were similar for the two treatment groups ( Table 5 ). The two antidepressant treatments did not differ in overall side effect ratings or in the proportion of participants with any psychiatric serious adverse events. No participants completed suicide during this trial; four participants were hospitalized for suicidal ideation or attempt, all in the mirtazapine group. The exit rate in the first 2 weeks of the trial were 11.5% for mirtazapine and 7.4% for nortriptyline (p=0.29).

DISCUSSION

We found no statistical difference in remission rates for participants treated with mirtazapine or nortriptyline as a third-step medication strategy after inadequate response or intolerance to two previous medication treatments for depression. Remission rates (as assessed by the Hamilton depression scale and QIDS-SR 16 ) were 12.3% and 8.0%, respectively, for mirtazapine and 19.8% and 12.4% for nortriptyline. In addition, the two antidepressants did not differ in tolerability or adverse events. No treatment group differences could be identified for response rates (per QIDS-SR 16 scores), time to response or remission, serious adverse events, or side effects. A relatively small proportion of study participants continued their Level 3 treatment longer than 12 weeks, as one would expect in an effectiveness trial with “real world” populations.

The modest remission rates (less than 20%) achieved in this trial were not likely due to either medication underdosing or to inadequate treatment durations for either medication, given the vigorous dosing schedule and diligent implementation of the protocol. As described earlier, a sensitivity analysis was conducted to evaluate the methods used to address the missing Hamilton depression scale data. Both the multiple imputation approach, and the use of values imputed from the QIDS-C 16 scores at exit based on item response theory (15) revealed remarkably similar findings, indicating that the analyses were not sensitive to the missing data methodology. In addition, no differential treatment effects were found after we controlled for the imbalance across the two groups of those unable to tolerate their Level 2/2A treatment. Secondary analyses using Hamilton depression scale remission rates did not identify a differential treatment effect among those who were and were not intolerant to Level 2/2A treatment.

From a pharmacological standpoint, serotonin reuptake inhibition was the primary pharmacological action of citalopram, the Level 1 treatment, and most of the Level 2/2A treatments (i.e., those involving either a switch to sertraline or citalopram augmentation). Two of the Level 2 switch strategies did not selectively affect the serotonergic neurotransmission system: sustained-release bupropion is a norepinephrine/dopamine reuptake inhibitor, and extended-release venlafaxine is a serotonin/norepinephrine reuptake inhibitor. It is noteworthy that the two Level 3 switch strategies involved agents with pharmacological actions different from those of previous levels. Mirtazapine, an antagonist of presynaptic alpha 2-adrenergic autoreceptors and heteroreceptors on both the norepinephrine and serotonin neurons, is also a potent antagonist of postsynaptic serotonin 5-HT 2 and 5-HT 3 receptors (32) and is an inhibitor of the release of corticotropin-releasing hormone (CRH) from CRH-containing neurons (33) . On the other hand, nortriptyline, a secondary-amine tricyclic antidepressant, strongly inhibits the norepinephrine transporter (34) . It also significantly antagonizes serotonin 5-HT 2 receptors (in particular, 5-HT 2C receptors [35] ) and modestly inhibits GABA transporters (36) , with only mild inhibition of the serotonin transporter (34) . The outcomes observed with these two Level 3 switch agents, despite their distinctive pharmacological actions compared with the primary reuptake inhibition of Level 1 and 2 medications, suggest that there is no clear advantage in switching to either one of these two treatments for outpatients who have not responded to multiple treatment trials and that the use of successive monotherapies results in only modest remission rates, even when the antidepressants have pharmacological profiles that clearly differ from those of the previous agents. While some depression treatment guidelines suggest the utility of three sequential monotherapies (37) , these results suggest that this approach yields rather limited results.

This study indicates that fewer than one in five depressed patients may achieve remission upon switching to another antidepressant medication after two unsuccessful medication treatments. Contrary to previous efficacy trials in major depressive disorder with 27%–50% remission rates upon switching to other antidepressants in patients who had not benefited adequately from two antidepressant trials (6 , 8) , this effectiveness study suggests that switching antidepressants, as a third-step treatment for such patients, provides rather modest chances of remission. Rates reported in previous open-label switch efficacy studies following two antidepressant treatments often involved the recruitment of subjects with nonchronic depression and minimal concurrent general medical comorbidity and whose treatment resistance was, in most cases, not prospectively determined. Our findings, in contrast, are generalizable to most adult outpatients with nonpsychotic major depressive disorder treated in real-world, primary or specialty care settings, although perhaps enriched with uninsured (40.5%) and unemployed (48.5%) populations.

Study limitations include the lack of a placebo control condition and nonmasked treatment delivery, although assessors of the primary outcome (Hamilton depression scale) were masked to treatment, and the QIDS-SR 16 score and the Hamilton depression scale ratings were in agreement. While a placebo control condition could have helped to determine whether improvement was due to spontaneous improvement or to nonspecific aspects of treatment, such a control is not required to discern whether these two treatments differed. Further, switching to a placebo after two consecutive failed treatment trials would have raised insurmountable human participant concerns and likely would have limited generalizability if many participants refused random assignment. A blinded placebo control condition could also have led to less vigorous dosing, given the high prevalence of multiple general medical conditions in our participants.

One might raise the issue that the failure to achieve remission in the prior level despite significant improvement could have yielded relatively high rates of remission in Level 3 with only modest reductions in depressive symptoms. However, since patients had the option of choosing to be randomly assigned only to augmentation strategies at Level 3, one may assume that those study participants who had improved significantly in the previous level would be less likely to accept switching treatments as an option. This is also suggested by the fairly significant level of depression severity at the time of entry into Level 3 (mean Hamilton depression scale score of 19.2) for those who agreed to be randomly assigned to a medication switch. Another limitation relates to the fact that participants were immediately switched to nortriptyline or mirtazapine. It is possible that discontinuation-emergent adverse events may have occurred and may have limited some participants’ ability to tolerate such a switch. On the other hand, there was no significant difference in exit rates between the two treatments during the first 2 weeks of the study. In addition, a previous open trial demonstrated the feasibility and safety of an abrupt switch from SSRIs to mirtazapine, without significant immediate dropouts (9) .

One might argue that a relative limitation of our study is its overall size, and that our study was not powered to detect a difference of 10% or more in remission rates, assuming the lowest remission rate was 12.3%, since a 10% difference in an adequately powered study would have required the random assignment of over 480 patients. However, our study was the largest ever (N=235) among investigations of depressed patients having completed two consecutive failed antidepressant trials. In addition, a difference of at least 15% between the two treatments was identified as the minimum clinically meaningful difference. Therefore, with this cohort, there was adequate power (at least 80%) to detect a 15% difference, if the lowest remission rate was 12.3%, and we can conclude that, based on our assumptions, neither treatment was significantly more effective than the other.

Finally, one may argue that defining remission as the primary outcome of the study may fail to distinguish between those who achieve remission with very small percentage changes in depressive symptoms (as in the case of those partial responders who entered Level 3) and those who achieve it with greater than 50% reductions in overall symptoms. On the other hand, the 8.0% and 12.3% remission rates with mirtazapine and nortriptyline per QIDS-SR 16 scores are comparable to the 7.1% and 10.9% mean percentage reductions in QIDS-SR 16 scores observed with these two treatments, suggesting that results based on remission are comparable to more traditional approaches that focus on changes in symptom severity.

CONCLUSION

This is the first report of a third-step medication switch among depressed participants who had not adequately benefited from two prior medication treatment trials. It is important to note that this trial included self-referred outpatients with major depressive disorder in real-world practices, with substantial and representative levels of chronicity and axis I comorbidity. Only 10%–20% of patients achieved remission upon switching to another antidepressant medication, with no clinically significant differences in efficacy or tolerability between the two medications. Contrary to previous efficacy trials in major depressive disorder, study results suggest that switching antidepressants as a third-step treatment after two consecutive antidepressant medication trials have failed provides only a modest chance of producing remission in major depressive disorder.

Received Oct. 12, 2005; revision received Jan. 24 and March 13, 2006; accepted March 14, 2006. From the Department of Psychiatry, University of Texas Southwestern Medical Center, Dallas; the Epidemiology Data Center, University of Pittsburgh Graduate School of Public Health, Pittsburgh; the New York State Psychiatric Institute and Columbia College of Physicians and Surgeons, New York; the Depression Clinical and Research Program, Massachusetts General Hospital, Boston; the Department of Psychiatry, University of Pittsburgh School of Medicine, Pittsburgh; and the National Institute of Mental Health, Bethesda, Md. Address correspondence and reprint requests to Dr. Fava, Depression Clinical and Research Program, Massachusetts General Hospital, Bulfinch 351, 55 Fruit St., Boston, MA 02114; [email protected] (e-mail).Supported by NIMH funding under contract N01MH90003 to UT Southwestern Medical Center at Dallas (A.J. Rush, principal investigator).The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government. The authors appreciate the support of Bristol-Myers Squibb Company, Forest Pharmaceuticals Inc., GlaxoSmithKline, King Pharmaceuticals, Organon Inc., Pfizer Inc, and Wyeth-Ayerst Laboratories for providing medications at no cost for this trial. The authors would also like to acknowledge the editorial support of Jon Kilner, M.S., M.A., and the secretarial support of Fast Word Information Processing Inc. (Dallas).Additional information on the STAR*D trial is included in a data supplement that accompanies the online version of this article.

CME Disclosure Maurizio Fava, M.D., has received research support from Abbott Laboratories; Lichtwer Pharma GmbH; Lorex Pharmaceuticals; as well as honoraria from EPIX Pharmaceuticals; Bayer AG; Compellis; Cypress Pharmaceuticals, Janssen Pharmaceutica; Knoll Pharmaceutical Company; Lundbeck; Dov Pharmaceuticals; Biovail Pharmaceuticals, Inc; BrainCells, Inc; Cypress Pharmaceuticals; Fabre-Kramer Pharmaceuticals, Inc; Grunenthal GmBH; MedAvante, Inc; Sepracor; and Somerset Pharmaceuticals. In addition, Dr Fava has received both research support and honoraria from Aspect Medical Systems; Astra-Zeneca; Bristol-Myers Squibb Company; Cephalon; Eli Lilly & Co; Forest Pharmaceuticals, Inc; GlaxoSmithKline; Johnson & Johnson Pharmaceuticals; Novartis; Organon, Inc; Pamlab; Pharmavite; Pfizer, Inc; Roche; Sanofi/Synthelabo; Solvay Pharmaceuticals, Inc; and Wyeth-Ayerst Laboratories. A. John Rush, M.D., has provided scientific consultation to or served on Advisory Boards for Advanced Neuronetic Systems, Inc., Bristol-Myers Squibb Company, Cyberonics Inc., Eli Lilly & Company, Forest Pharmaceuticals Inc., GlaxoSmithKline, Health Technology Systems, Merck Co. Inc., Neuronetics, Ono Pharmaceuticals, Organon Pharmaceuticals Inc., Pfizer Inc., the Urban Institute, and Wyeth-Ayerst Laboratories Inc. He has received royalties from Guilford Press and Health Technology Systems and research/grant support from the Robert Wood Johnson Foundation, the National Institute of Mental Health, and the Stanley Foundation. He has been on speaker bureaus for Cyberonics, Inc., Eli Lilly & Company, Forest Pharmaceuticals Inc., GlaxoSmithKline, and Merck Co. Inc., and owns stock in Pfizer Inc. Andrew A. Nierenberg, M.D., has provided scientific consultation for Eli Lilly & Company, Genaissance, GlaxoSmithKline, Innapharma, Sepracor, and Shire, and has received research support from Bristol-Myers Squibb Company, Cederroth, Cyberonics, Inc., Forest Pharmaceuticals Inc., Janssen Pharmaceutica, Lichtwer Pharma, and Pfizer Inc. Dr. Nierenberg has received both research support and honoraria from Eli Lilly & Company, GlaxoSmithKline, and Wyeth-Ayerst Laboratories. Jonathan Alpert, M.D., Ph.D., has received research support from Aspect Medical Systems, Eli Lilly & Co., Forest Pharmaceuticals, and Pfizer Inc.; has received research support and honoraria from Organon Inc., and has received research support from and served as a consultant for Pharmavite LLC and Pamlab LLC. Patrick J. McGrath, M.D., has had research support from Pfizer Inc. and Organon, Inc., and has been a consultant to GlaxoSmithKline and Somerset Pharmaceuticals. Michael E. Thase, M.D., has provided scientific consultation to AstraZeneca, Bristol-Myers Squibb Company, Cephalon, Cyberonics, Inc., Eli Lilly & Company, Forest Pharmaceuticals, Inc., GlaxoSmithKline, Janssen Pharmaceutica, Novartis, Organon, Inc., Pfizer Inc., and Wyeth-Ayerst Laboratories. Dr. Thase has been on the speakers bureaus for AstraZeneca, Eli Lilly & Company, GlaxoSmithKline, Organon, Inc., and Wyeth-Ayerst Laboratories. Diane Warden, Ph.D., M.B.A., owns stocks of Pfizer Inc. and previously owned stock of Bristol-Myers Squibb Company. Madhukar H. Trivedi, M.D., has provided scientific consultation or served on Advisory Boards for Bristol-Myers Squibb Company, Cyberonics Inc., Eli Lilly & Company, Forest Pharmaceuticals, Johnson & Johnson, Pfizer, Inc., Sepracor, and Wyeth-Ayerst Laboratories. He has been on speaker bureaus for Bristol-Myers Squibb Company, Cyberonics, Inc., Eli Lilly & Company, Forest Pharmaceuticals, and Wyeth-Ayerst Laboratories. He has received research support from Bristol-Myers Squibb Company, Cephalon, Inc., Corcept Therapeutics, Inc., Eli Lilly & Company, GlaxoSmithKline, Janssen Pharmaceutica, National Institute of Mental Health, National Alliance for Research in Schizophrenia and Depression, Pfizer Inc., Predix Pharmaceuticals, and Wyeth-Ayerst Laboratories. George Niederehe, Ph.D., Stephen R. Wisniewski, Ph.D., Melanie Biggs, Ph.D., James F. Luther, M.A., and Louise Ritz, M.B.A., have no conflicts to report. APA policy requires disclosure by CME authors of unapproved or investigational use of products discussed in CME programs. Off-label use of medications by individual physicians is permitted and common. Decisions about off-label use can be guided by scientific literature and clinical experience.

References

1. Cassano P, Fava M: Depression and public health: an overview. J Psychosom Res 2002; 53:849–857Google Scholar

2. Thase ME, Entsuah AR, Rudolph RL: Remission rates during treatment with venlafaxine or selective serotonin reuptake inhibitors. Br J Psychiatry 2001 178:234–241Google Scholar

3. Thase ME, Haight BR, Richard N, Rockett CB, Mitton M, Modell JG, Vanmeter S, Harriett AE, Wang Y: Remission rates following antidepressant therapy with bupropion or selective serotonin reuptake inhibitors: a meta-analysis of original data from 7 randomized controlled trials. J Clin Psychiatry 2005; 66:974–981Google Scholar

4. Trivedi MH, Rush AJ, Wisniewski SR, Nierenberg AA, Warden D, Ritz L, Norquist G, Howland RH, Lebowitz B, McGrath PJ, Shores-Wilson K, Biggs MM, Balasubramani GK, Fava M, and STAR*D Study Team: Evaluation of outcomes with citalopram for depression using measurement-based care in STAR*D: implications for clinical practice. Am J Psychiatry 2006; 163:28–40Google Scholar

5. Fava M: Management of nonresponse and intolerance: switching strategies. J Clin Psychiatry 2000; 6(suppl 2):10–12Google Scholar

6. Quitkin FM, McGrath PJ, Stewart JW, Deliyannides D, Taylor BP, Davies CA, Klein DF: Remission rates with 3 consecutive antidepressant trials: effectiveness for depressed outpatients. J Clin Psychiatry 2005; 66:670–676Google Scholar

7. Petersen T, Papakostas GI, Posternak MA, Kant A, Guyker WM, Iosifescu DV, Yeung AS, Nierenberg AA, Fava M: Empirical testing of two models for staging antidepressant treatment resistance. J Clin Psychopharmacol 2005; 25:336–341Google Scholar

8. Poirier MF, Boyer P: Venlafaxine and paroxetine in treatment-resistant depression. double-blind, randomised comparison. Br J Psychiatry 1999; 175:12–16Google Scholar

9. Fava M, Dunner DL, Greist JH, Preskorn SH, Trivedi MH, Zajecka J, Cohen M: Efficacy and safety of mirtazapine in major depressive disorder patients after SSRI treatment failure: an open-label trial. J Clin Psychiatry 2001; 62:413–420Google Scholar

10. Nierenberg AA, Papakostas GI, Petersen T, Kelly KE, Iacoviello BM, Worthington JJ, Tedlow J, Alpert JE, Fava M: Nortriptyline for treatment-resistant depression. J Clin Psychiatry 2003; 64:35–39Google Scholar

11. Fava M, Rush AJ, Trivedi MH, Nierenberg AA, Thase ME, Sackeim HA, Quitkin FM, Wisniewski S, Lavori PW, Rosenbaum JF, Kupfer DJ: Background and rationale for the sequenced treatment alternatives to relieve depression (STAR*D) study. Psychiatr Clin North Am 2003; 26:457–494Google Scholar

12. Rush AJ, Fava M, Wisniewski SR, Lavori PW, Trivedi MH, Sackeim HA, Thase ME, Nierenberg AA, Quitkin FM, Kashner TM, Kupfer DJ, Rosenbaum JF, Alpert J, Stewart JW, McGrath PJ, Biggs MM, Shores-Wilson K, Lebowitz BD, Ritz L, Niederehe G: Sequenced treatment alternatives to relieve depression (STAR*D): rationale and design. Control Clin Trials 2004; 25:119–142Google Scholar

13. Lavori PW, Rush AJ, Wisniewski SR, Alpert J, Fava M, Kupfer DJ, Nierenberg A, Quitkin FM, Sackeim HA, Thase ME, Trivedi M: Strengthening clinical effectiveness trials: equipoise-stratified randomization. Biol Psychiatry 2001; 50:792–801Google Scholar

14. Rush AJ, Carmody TJ, Reimitz PE: The Inventory of Depressive Symptomatology (IDS): clinician (IDS-C) and self-report (IDS-SR) ratings of depressive symptoms. Int J Methods Psychiatr Res 2000; 9:45–59Google Scholar

15. Rush AJ, Trivedi MH, Ibrahim HM, Carmody TJ, Arnow B, Klein DN, Markowitz JC, Ninan PT, Kornstein S, Manber R, Thase ME, Kocsis JH, Keller MB: The 16-item Quick Inventory of Depressive Symptomatology (QIDS), clinician rating (QIDS-C), and self-report (QIDS-SR): a psychometric evaluation in patients with chronic major depression. Biol Psychiatry 2003; 54:573–583Google Scholar

16. Trivedi MH, Rush AJ, Ibrahim HM, Carmody TJ, Biggs MM, Suppes T, Crismon ML, Shores-Wilson K, Toprac MG, Dennehy EB, Witte B, Kashner TM: The Inventory of Depressive Symptomatology, clinician rating (IDS-C) and self-report (IDS-SR), and the Quick Inventory of Depressive Symptomatology, clinician rating (QIDS-C) and self-report (QIDS-SR) in public sector patients with mood disorders: a psychometric evaluation. Psychol Med 2004; 34:73–82Google Scholar

17. Rush AJ, Bernstein IH, Trivedi MH, Carmody TJ, Wisniewski S, Mundt JC, Shores-Wilson K, Biggs MM, Woo A, Nierenberg AA, Fava M: An evaluation of the Quick Inventory of Depressive Symptomatology and the Hamilton Rating Scale for Depression: A Sequenced Treatment Alternatives to Relieve Depression trial report. Biol Psychiatry 2006; 59:493–501Google Scholar

18. Rush AJ, Trivedi MH, Wisniewski SR, Stewart J, Nierenberg AA, Thase ME, Ritz L, Biggs MM, Warden D, Luther JF, Shores-Wilson K, Niederehe G, Fava M, for the STAR*D Trial Team: Bupropion-SR, sertraline, or venlafaxine-XR after failure of SSRIs for depression. N Engl J Med 2006; 354:1231–1242Google Scholar

19. Trivedi MH, Fava M, Wisniewski SR, Thase ME, Quitkin F, Warden D, Ritz L, Nierenberg A, Niederehe G, Biggs M, Luther JF, Shores-Wilson K, Rush AJ, for the STAR*D Study Team: A comparison of citalopram augmentation with bupropion-SR and buspirone following SSRI failure for depressed outpatients: A STAR*D report. N Engl J Med (in press)Google Scholar

20. Wisniewski SR, Eng H, Meloro L, Gatt R, Ritz L, Stegman D, Trivedi M, Biggs MM, Friedman E, Shores-Wilson K, Warden D, Bartolowits D, Martin JP, Rush AJ: Web-based communications and management of a multi-center clinical trial: the sequenced treatment alternatives to relieve depression (STAR*D) project. Clin Trials 2004; 1:387–398Google Scholar

21. Linn BS, Linn MW, Gurel L: Cumulative Illness Rating Scale. J Am Geriatr Soc 1968; 16:622–626Google Scholar

22. Zimmerman M, Mattia JI: A self-report scale to help make psychiatric diagnoses: the Psychiatric Diagnostic Screening Questionnaire. Arch Gen Psychiatry 2001; 58:787–794Google Scholar

23. Rush AJ, Zimmerman M, Wisniewski SR, Fava M, Hollon SD, Warden D, Biggs MM, Shores-Wilson K, Shelton RC, Luther JF, Thomas B, Trivedi MH: Comorbid psychiatric disorders in depressed outpatients: demographic and clinical features. J Affect Disord 2005; 87:43–55Google Scholar

24. Rush AJ, Gullion CM, Basco MR, Jarrett RB, Trivedi MH: The Inventory of Depressive Symptomatology (IDS): psychometric properties. Psychol Med 1996; 26:477–486Google Scholar

25. Ware J Jr, Kosinski M, Keller SD: A 12-item short-form health survey: construction of scales and preliminary tests of reliability and validity. Med Care 1996; 34:220–233Google Scholar

26. Reilly MC, Zbrozek AS, Dukes EM: The validity and reproducibility of a work productivity and activity impairment instrument. Pharmacoeconomics 1993; 4:353–365Google Scholar

27. Mundt JC, Marks IM, Shear MK, Greist JH: The Work and Social Adjustment Scale: a simple measure of impairment in functioning. Br J Psychiatry 2002; 180:461–464Google Scholar

28. Endicott J, Nee J, Harrison W, Blumenthal R: Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q): a new measure. Psychopharmacol Bull 1993; 29:321–326Google Scholar

29. Trivedi MH, Rush AJ, Wisniewski SR, Warden D, McKinney W, Downing M, Berman SR, Farabaugh A, Luther J, Nierenberg AA, Lis J, Sackeim H: Factors associated with health-related quality of life among outpatients with major depressive disorder: a STAR*D report. J Clin Psychiatry 2006; 67:185–195Google Scholar

30. Mundt JC: Interactive voice response systems in clinical research and treatment. Psychiatr Serv 1997; 48:611–612Google Scholar

31. Kobak KA, Greist JH, Jefferson JW, Mundt JC, Katzelnick DJ: Computerized assessment of depression and anxiety over the telephone using interactive voice response. MD Comput 1999; 16:63–68Google Scholar

32. Stimmel GL, Dopheide JA, Stahl SM: Mirtazapine: an antidepressant with noradrenergic and specific serotonergic effects. Pharmacotherapy 1997; 17:10–21Google Scholar

33. Fabricio AS, Tringali G, Pozzoli G, Navarra P: Mirtazapine acutely inhibits basal and K(+)-stimulated release of corticotropin-releasing hormone from the rat hypothalamus via a non-genomic mechanism. Psychopharmacology (Berl) 2005; 178:78–82Google Scholar

34. Frazer A: Serotonergic and noradrenergic reuptake inhibitors: prediction of clinical effects from in vitro potencies. J Clin Psychiatry 2001; 62(suppl 12):16–23Google Scholar

35. Akiyoshi J, Isogawa K, Yamada K, Nagayama H, Fujii I: Effects of antidepressants on intracellular Ca2+ mobilization in CHO cells transfected with the human 5-HT2C receptors. Biol Psychiatry 1996; 39:1000–1008Google Scholar

36. Nakashita M, Sasaki K, Sakai N, Saito N: Effects of tricyclic and tetracyclic antidepressants on the three subtypes of GABA transporter. Neurosci Res 29(1):87–91, 1997 Sep.Google Scholar

37. Crismon ML, Trivedi M, Pigott TA, Rush AJ, Hirschfeld RM, Kahn DA, DeBattista C, Nelson JC, Nierenberg AA, Sackeim HA, Thase ME: The Texas Medication Algorithm Project: report of the Texas Consensus Conference Panel on Medication Treatment of Major Depressive Disorder. J Clin Psychiatry 1999; 60:142–156Google Scholar