Chronic Depression: Medication

ORIGINAL ARTICLE Chronic Depression Medication (Nefazodone) or Psychotherapy (CBASP) Is Effective When the Other Is Not

Alan F. Schatzberg, MD; A. John Rush, MD; Bruce A. Arnow, PhD; Phillip L. C. Banks, MS; Janice A. Blalock, PhD; Frances E. Borian, RN; Robert Howland, MD; Daniel N. Klein, PhD; James H. Kocsis, MD; Susan G. Kornstein, MD; Rachel Manber, PhD; John C. Markowitz, MD; Ivan Miller, PhD; Philip T. Ninan, MD; Barbara O. Rothbaum, PhD; Michael E. Thase, MD; Madhukar H. Trivedi, MD; Martin B. Keller, MD

Context: Although various strategies are available to man- Main Outcome Measures: The 24-item Hamilton Rat- age nonresponders to an initial treatment for depres- ing Scale for Depression, administered by raters blinded sion, no controlled trials address the utility of switching to treatment, the Clinician Global Impressions–Severity from an antidepressant medication to psychotherapy or scale, and the 30-item Inventory for Depressive Symp- vice versa. tomatology–Self-Report.

Objective: To compare the responses of chronically de- Results: Analysis of the intent-to-treat sample revealed pressed nonresponders to 12 weeks of treatment with that both the switch from nefazodone to CBASP and the either nefazodone or cognitive behavioral analysis sys- switch from from CBASP to nefazodone resulted in clini- tem of psychotherapy (CBASP) who were crossed over cally and statistically significant improvements in symp- to the alternate treatment (nefazodone, n=79; CBASP, toms. Neither the rates of response nor the rates of re- n=61). mission were significantly different when the groups of completers were compared. However, the switch to CBASP Design: Crossover trial. following nefazodone therapy was associated with significantly less attrition due to adverse events, which may Setting: Twelve academic outpatient psychiatric centers. explain the higher intent-to-treat response rate among Patients: There were 140 outpatients with chronic ma- those crossed over to CBASP (57% vs 42%). jor depressive disorder; 92 (65.7%) were female, 126 (90.0%) were white, and the mean age was 43.1 years. Conclusions: Among chronically depressed individu- Thirty participants dropped out of the study prema- als, CBASP appears to be efficacious for nonresponders turely, 22 in the nefazodone group and 8 in the CBASP to nefazodone, and nefazodone appears to be effective group. for CBASP nonresponders. A switch from an antidepressant medication to psychotherapy or vice versa appears Interventions: Treatment lasted 12 weeks. The dos- to be useful for nonresponders to the initial treatment. age of nefazodone was 100 to 600 mg/d; CBASP was pro- vided twice weekly during weeks 1 through 4 and weekly thereafter. Arch Gen Psychiatry. 2005;62:513-520

SUBSTANTIAL PROPORTION nicians as to the preferable next step.3-5 Im- of patients treated for de- portantly, many depressed patients ini- pression do not respond to tially receive psychotherapy alone. the initial trial of either an Surprisingly few studies6 have evaluated the antidepressant medica- role of medication following nonresponse tion or depression-targeted psycho- to psychotherapy.7 None has evaluated the A1 therapy. This so-called stage I antidepres- efficacy of psychotherapy following non- sant resistance2,3 is one of several degrees response to medication. Surprisingly few of resistance defined by prior treatment his- controlled switch studies are available even tory. Numerous treatment options are avail- for medication-to-medication switches.7,8 able in such situations, including switch- This report provides the first prospective ing medication type or class, augmenting controlled trial evidence, using blinded rater or combining medications, and switching methodology, to examine switching from Author Affiliations are listed at to or augmenting with psychotherapy. nefazodone therapy to psychotherapy (spe- the end of this article. There are few controlled data to guide cli- cifically, cognitive behavioral analysis sys-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, MAY 2005 WWW.ARCHGENPSYCHIATRY.COM 513

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/29/2021 minimally adequate trial of nefazodone (ie, at least 4 weeks at Ն300 Assessed for Eligibility mg/d) were not eligible. Patients could not have been treated with (N = 1035) benzodiazepines within 2 weeks, fluoxetine or monoamine oxi- dase inhibitors within 4 weeks, electroconvulsive therapy within Randomized Not Randomized (n = 354) 6 months, or depot neuroleptics within 6 months. No psycho- (n = 681) • Did Not Meet Inclusion therapy outside the study protocol was permitted. Criteria (n = 235) • Withdrew Consent (n = 47) INITIAL ACUTE-PHASE TREATMENT TRIAL • Other Reasons (n = 72) Participants were recruited via advertisements in newspapers, Nefazodone CBASP Combination radio, and printed flyers as well as through physician refer- (n = 226) (n = 228) (n = 227) rals. After providing written informed consent, subjects un- derwent a complete medical history interview, physical exami- Nonresponders Nonresponders nation, electrocardiogram, and laboratory screening test battery (n = 73) (n = 83) to confirm medical eligibility. No nonprotocol psychotropic medications were allowed during the study, including anxio- Declined Entered CBASP Entered Declined lytics and sedative hypnotics. Patients agreed to randomiza- Crossover (Intention-to- Nefazodone Crossover tion to CBASP, nefazodone, or the combination for this 12- (n = 12) Treat Sample) (Intention-to- (n = 4) week acute-phase trial (Figure 1). (n = 61) Treat Sample) (n = 79) Patients randomized to nefazodone monotherapy or combined nefazodone therapy and CBASP received open-label treatment with oral nefazodone. Those patients receiving nefaz- Dropped Completers Completers Dropped odone, either with or without CBASP, received an initial daily Out (n = 53) (n = 57) Out (n = 8) (n = 22) dose of 200 mg (100 mg twice a day), which was increased to 300 mg/d during the second week. Thereafter, when indicated, the dose was increased in increments of 100 mg/d to a Figure 1. Randomization and disposition. maximum of 600 mg/d. To remain in the study, patients were required to take a minimum of 300 mg/d by week 3. Patients tem of psychotherapy [CBASP])9 or switching from CBASP noncompliant with prescribed dosing were dropped from the to nefazodone therapy following completion with nonre- study. At the end of the initial acute-phase trial, the overall fi- sponse to a 12-week trial of the alternative treatment. nal average daily dosage of nefazodone was 461.0 mg/d (SD, 143.3 mg/d). Psychopharmacology visits were conducted according to a manual13 for clinical management (eg, review of METHODS symptoms, adverse events, illnesses, and concomitant medications) and were limited to 15 to 20 minutes. SUBJECTS Cognitive behavioral analysis system of psychotherapy, which was developed to address the specific challenges of treating in- This protocol was part of a multiphase collaborative research pro- dividuals with chronic depression, incorporates both cogni- gram studying chronic depressive disorders (see Keller et al10 for tive behavioral and interpersonal features.9 Treatment was con- a full description of study design, rationale, and methods). The ducted according to a manual14 specifying sessions twice weekly institutional review boards at all 12 sites approved the study. during weeks 1 through 4 and weekly thereafter until week 12. Briefly, outpatients 18 to 75 years of age who were in a current Sessions could be held twice weekly during weeks 5 through 8 major depressive episode were eligible to enroll at one of 12 cen- if the patient was having trouble mastering situational analy- ters if they met the DSM-IV criteria for chronic major depres- sis, the core procedure in CBASP.9 Patients who attended fewer sive disorder (ie, current major depressive episode Ն2 years in than 13 sessions or missed 3 consecutive sessions were dropped duration), “double depression” (ie, a current major depressive from the study. The overall mean number of total CBASP ses- episode superimposed on antecedent dysthymic disorder), or re- sions for the intent-to-treat acute-phase sample was 16 (SD, 5.4). current major depressive disorder with incomplete interepi- Psychotherapists had either (1) an MD or PhD degree along sode recovery. Additionally, patients had to have a score of 20 with at least 2 years of post-training experience or (2) an MSW or higher on the 24-item Hamilton Rating Scale for Depression degree along with at least 5 years of post-degree experience. (HRS-D24) both at screening and following a 2-week drug-free All psychotherapists attended a 2-day training workshop and period at baseline prior to treatment initiation. met the criteria for mastery of CBASP treatment procedures on The Structured Clinical Interview for DSM-IV Axis I Disor- 2 pilot training cases, as assessed by videotape-based adher- ders11 and an abbreviated version of the Structured Clinical In- ence rating, before they were allowed to see study patients. Ad- terview for DSM-IV Personality Disorders12 were used to estab- herence monitoring to the CBASP protocol took place through- lish the intake diagnoses. Patients with organic mental syndromes; out the study. All study CBASP sessions were videotaped. The bipolar disorder or cyclothymia; schizophrenia or other psy- CBASP supervisors at each site conducted weekly adherence chotic disorders; obsessive-compulsive disorder; or schizotypal, monitoring by videotape review using a standardized adher- antisocial, or severe borderline personality disorder were ex- ence rating scale. Site supervisor adherence was similarly moni- cluded, as were those with principal current diagnoses of panic, tored by a CBASP coordinator. generalized anxiety, or posttraumatic stress disorders. Patients were not eligible if they had abused alcohol or drugs within the last 6 months or suffered from bulimia or anorexia nervosa within 1 ASSESSMENT OF RESPONSE year of intake. Patients considered to be at immediate risk of sui- cide, to have medical contraindications to antidepressant therapy, Vital signs and adverse events (volunteered or observed) were or to have significant, unstable general medical disorders were assessed at each visit. The primary outcome measure was the 15-17 also excluded. Patients who had not responded previously to a HRS-D24, which was completed once weekly through week

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, MAY 2005 WWW.ARCHGENPSYCHIATRY.COM 514

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/29/2021 4 and biweekly thereafter. Secondary outcomes included the STATISTICAL ANALYSES Clinical Global Impressions–Severity scale (CGI-S),18 the 30- item Inventory for Depressive Symptomatology–Self-Report The analyses performed included the analysis of baseline and (IDS-SR ),19,20 and the Hamilton Rating Scale for Anxiety Psy- 30 demographic characteristics to assess comparability of the two chic Anxiety factor score (HAM-A ). All clinical ratings were PA crossover-phase treatment groups as well as comparisons of ef- completed by an independent evaluator without knowledge of ficacy parameters and incidence of adverse events. Time to re- the treatment received. sponse and likelihood of response (Ն50% reduction in HRS- A remission was defined as a final HRS-D score of 8 or lower 24 D score) were tested with Kaplan-Meier survival analysis. The at both weeks 10 and 12. A response without remission was 24 time to a reduction of 50% or greater in HRS-D score was de- declared if there was a reduction of 50% or greater in the total 24 fined as the number of days from randomization to the cross- HRS-D score and there was a total score of 15 or lower at both 24 over treatment until the first postrandomization observation weeks 10 and 12 but remission was not achieved. In addition, of a decrease of 50% or greater in HRS-D total score. If a a single overall response/remission rate was calculated that com- 24 decrease of 50% or greater was not observed, the time was right- bined those in the remission group with those who demon- censored using the visit date of the last usable postrandomiza- strated response without remission. All other patients were con- tion HRS-D score from the entire 24-week period. Kaplan- sidered nonresponders. 24 Meier plots for the survival distribution for time to a reduction of 50% or greater in HRS-D24 score were generated and comparisons were made using the log-rank test statistic (Figure 2). ACUTE-PHASE OUTCOME To further explore response, we performed categorical analyses (eg, response and remission rates, attrition rates, and rates A total of 454 patients were randomized either to CBASP mono- of adverse events) using the Fisher exact probability test, therapy (n=228) or to nefazodone monotherapy (n=226). Of McNemar test (for within-subject comparisons), or Cochran- these, 173 in the CBASP group (75.9%) and 167 in the nefaz- Mantel-Haenszel ␹2 test (stratified by study site) as appropri- odone group (73.9%) completed the 12-week acute-phase trial. ate. Changes in continuously distributed efficacy variables (HRS- Among completers, 33.5% (56/167) of those in the nefaz- D24, CGI-S, IDS-SR30, and HAM-APA) were analyzed using analysis odone group and 28.3% (49/173) of those in the CBASP group of covariance at the endpoint visit with the crossover-phase base- had response without remission. These response rates did not ␹2 line value as the covariate and fixed effects for treatment and significantly differ from one another ( 1=1.08, P=.30). Remis- site. These analyses were performed separately for the com- sion rates for completers were 21.6% (36/167) for nefazodone pleter sample and the intent-to-treat sample. For the intent- and 23.7% (41/173) for CBASP; again, these remission rates were ␹2 to-treat sample, response status was determined based on the not significantly different ( 1=0.22, P=.64). last available observation and the endpoint symptom severity Among the patients randomized at study baseline to either used in the analyses on continuous variables. An exploratory CBASP monotherapy or nefazodone monotherapy, 216 in the analysis examined the correlation between the percentage change CBASP group and 220 in the nefazodone group constituted the in HRS-D score from entry to exit in the acute phase and the intent-to-treat sample. For this sample, response and remis- 24 percentage change in HRS-D24 score from entry to exit in the sion rates were determined based on the last available obser- crossover phase (overall and by treatment group). vation. The response rates without remission were 18.6% A piecewise mixed-effects random regression model was used (41/220) in the nefazodone group and 14.4% (31/216) in the to assess whether there were differences between nefazodone CBASP group. These response rates were not significantly dif- ␹2 and CBASP in the rate (linear slope) of improvement in symp- ferent ( 1=1.45, P=.23). Remission rates were 29.1% (64/ toms, with change in HRS-D scores from baseline to week 4 220) for nefazodone therapy and 33.3% (72/216) for CBASP; 24 ␹2 as one variable and from week 4 to week 12 (or the last visit) these remission rates were not significantly different ( 1=0.91, as a second variable. The model included a random intercept P=.34). and a random slope. We hypothesized that patients receiving Altogether, 156 nonresponders randomized to nefazodone nefazodone in the crossover phase (ie, patients who were alone or CBASP alone completed the acute-phase trial (73 who switched from CBASP to nefazodone) would exhibit a more rapid started in the nefazodone group and 83 who started in the CBASP rate of improvement in symptoms during the first 4 weeks than group). The 73 nonresponders in the nefazodone group re- patients who were switched from nefazodone to CBASP, just ceived a mean final daily dosage of 491.1 mg/d (SD, 125.4 mg/d; as observed in the acute phase. This analysis examined the lin- range, 100-600 mg/d). Responders to nefazodone received a non- ear slopes between weeks 0 to 4 and weeks 4 to 12 in the cross- significantly higher final daily dosage (520.1 mg/d; F1,163=2.73, over phase. P=.10). The 83 nonresponders in the CBASP group attended All statistical tests used 2-tailed probability values with un- a total of 17.7 therapy sessions (SD, 1.9; range, 13-22). Re- adjusted significance levels of PՅ.05. sponders to CBASP attended a slightly greater number of sessions than nonresponders (18.2; F1,171=3.75, P=.054). RESULTS SWITCH PROTOCOL Of the 156 monotherapy nonresponders who com- There was no washout period for the switch (crossover) phase. pleted the initial acute-phase trial, 140 (89.7%) con- Nonresponders to nefazodone simply stopped taking the medi- sented to begin the alternate treatment. There were no cation and began CBASP treatment if they met the criteria to significant differences in sociodemographic and clinical enter the crossover phase. Conversely, CBASP nonresponders measures between those who agreed and those who de- stopped psychotherapy and began nefazodone treatment. All those who agreed to this crossover phase signed a new written clined to participate in the crossover trial. Twelve (16%) informed consent. Visits, assessments, medication dosing, CBASP of the 73 patients initially treated with nefazodone de- procedures, visit and session frequency, and outcome defini- clined the crossover treatment, while 4 (5%) of the 83 tions during the new (crossover) treatment were identical to patients who initially received CBASP declined to pro- ␹2 the methods used in the initial acute-phase trial. ceed with crossover to nefazodone ( 1=5.7, P=.02).

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, MAY 2005 WWW.ARCHGENPSYCHIATRY.COM 515

0.9

CBASP Followed by Nefazodone 0.8 Nefazodone Followed by CBASP

0.7

0.6

0.5

0.4 Survival Distribution

0.3

0.2

0.1

0 20406080100 120 140 160 180 200 Time to 50% Reduction in HRS-D24 Score, d

Figure 2. Kaplan-Meier plot of the time to a 50% reduction in the HRS-D24 (24-item Hamilton Rating Scale for Depression) score from the baseline of the acute phase. CBASP indicates cognitive behavioral analysis system of psychotherapy.

SAMPLE CHARACTERISTICS ␹2=5.03, P=.03). There was no significant difference in rates of remission or response without remission be- A total of 61 patients were switched from nefazodone to tween the two groups. In the completer sample, there were CBASP, and 79 CBASP nonresponders were switched to no significant between-group differences in the rates of nefazodone. With the exception of the percentage of pa- response without remission, remission, or overall re- tients with a lifetime history of anxiety disorders, base- sponse. line characteristics did not differ significantly between Table 3 presents the observed values for HRS-D24, the two groups (Table 1). The mean dosage of nefaz- CGI-S, IDS-SR30, and HAM-APA at the baseline of the cross- odone at the endpoint of the crossover phase was 419.0 over phase and at the endpoint. The change from base- mg/d (SD, 154.7 mg/d). The mean number of psycho- line to endpoint and the adjusted means of the change therapy sessions attended during the crossover phase was in each measure are also presented. The results of the 16.5 (SD, 4.4). Patients receiving nefazodone were sched- analysis of covariance for all 4 parameters indicated that uled to see a pharmacotherapist at each of 9 scheduled patients in both groups improved over time. In the intent- study visits during the crossover phase. The mean num- to-treat sample, the group switched from nefazodone to ber of study visits attended during crossover for those CBASP had a significantly greater improvement in HRS- in the nefazodone condition was 7.8 (SD, 1.8, n=61). D24 and HAM-APA scores than the group switched from CBASP to nefazodone. However, there were no signifi- ATTRITION AND TOLERABILITY cant differences in outcome on these 4 measures when completers were compared. During the crossover phase, 30 patients dropped out, 22 A piecewise random regression analysis comparing the (28%) in the nefazodone group and 8 (13%) in the CBASP rates of symptomatic improvement (linear slope) based ␹2 group ( 1=4.61, P=.03). This difference was probably on the HRS-D24 score (Figure 3) revealed no signifi- caused by the fact that more patients receiving nefaz- cant difference between the two groups during the first odone dropped out because of adverse events (18% vs 4 weeks of the crossover phase (F1,138=0.16, P=.69). How- ␹2 3%; 1=6.79, P=.009). ever, a significantly greater rate of symptomatic improvement was evident for weeks 4 to 12 among the patients RESPONSE AND REMISSION who were switched from nefazodone to CBASP (F1,138=10.38, P=.002). Response and remission rates are presented in Table 2 for both the intent-to-treat and completer samples. In the intent-to-treat sample, overall response rates were sig- COMMENT nificantly higher for patients who crossed over to CBASP from nefazodone (57% [35/61]) than for patients who Results of this controlled, rater-blinded switch study in- crossed over to nefazodone from CBASP (42% [33/79]; dicate that a switch strategy for nonresponders provides

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, MAY 2005 WWW.ARCHGENPSYCHIATRY.COM 516

Total Nefazodone → CBASP CBASP → Nefazodone Variable (n = 140) (n = 61) (n = 79) Statistics*

2 Female, No. (%) 92 (65.7) 44 (72) 48 (61) ␹1 = 2.70, P = .10

Age, y† 43.1 ± 11 42.3 ± 12.0 43.7 ± 10.4 F1,127 = 0.56, P = .46 2 White, No. (%) 126 (90.0) 55 (90) 71 (90) ␹1 = 0.03, P = .87 2 Marital status, No. (%)‡ ␹3 = 2.12, P = .55 Currently married or cohabiting 63 (45.0) 29 (48) 34 (43) Single 40 (28.6) 14 (23) 26 (33) Widowed 3 (2.1) 2 (3) 1 (1) Divorced/separated 34 (24.3) 16 (26) 18 (23) 2 Employment status, No. (%)‡ ␹4 = 2.15, P = .71 Work for pay 96 (68.6) 42 (69) 54 (68) Housewife 17 (12.1) 8 (13) 9 (11) Student 4 (2.9) 3 (5) 1 (1) Retired 4 (2.9) 1 (2) 3 (4) Unemployed 18 (12.9) 6 (10) 12 (15) Data missing 1 (0.7) 1 (2) 0 (0) 2 Depression diagnosis, No. (%)‡ ␹2 = 2.19, P = .34 Chronic major depression 56 (40.0) 27 (44) 29 (37) Double depression 53 (37.9) 24 (39) 29 (37) Recurrent depression, incomplete interepisode recovery 31 (22.1) 10 (16) 21 (27)

Age at onset of initial episode of MDD, y† 26.4 ± 12.4 25.7 ± 12.0 26.9 ± 12.8 F1,114 = 2.32, P = .13

Duration of current MDD, y† 8.0 ± 9.28 7.5 ± 9.1 8.3 ± 9.45 F1,127 = 0.19, P = .66

Age at onset of dysthymia, y† 21.6 ± 14.87 20.8 ± 13.1 22.3 ± 16.4 F1,41 = 0.46, P = .50

Duration of current dysthymia, y† 20.2 ± 15.59 20.6 ± 16.26 19.8 ± 15.3 F1,36 = 0.66, P = .42 2 History of alcohol abuse/dependence, No. (%) 43 (30.7) 17 (28) 27 (34) ␹1 = 2.42, P = .12 2 Lifetime comorbid anxiety disorder, No. (%) 41 (29.3) 24 (39) 17 (22) ␹1 = 3.84, P = .05 2 Previous treatment with psychotherapy, No. (%) 94 (67.1) 40 (66) 54 (68) ␹1 = 0.45, P = .50 2 Previous treatment with antidepressant, No. (%) 95 (67.9) 41 (67) 54 (68) ␹1 = 0.00, P = .98

Abbreviations: CBASP, cognitive behavioral analysis system of psychotherapy; MDD, major depressive disorder. *For categorical variables, the ␹2 values were determined using the Cochran-Mantel-Haenszel test stratified by site. Analysis of variance F-test values were determined for continuous variables. †Values are mean ± SD. ‡Percentages may not add to 100% because of rounding.

Table 2. Response* and Remission† Rates at the End of the Crossover Phase

No. (%)‡

Nefazodone → CBASP CBASP → Nefazodone ␹2§ P Value Completers (n = 110) No. of subjects 53 57 Response without remission 21 (40) 14 (25) 2.73 .10 Remission 13 (25) 16 (28) 0.01 .92 Overall response/remission 34 (64) 30 (53) 2.86 .09 Intent-to-Treat Sample (n = 140) No. of subjects 61 79 Response without remission 13 (21) 12 (15) 0.95 .33 Remission 22 (36) 21 (27) 2.63 .11 Overall response/remission 35 (57) 33 (42) 5.03 .03

Abbreviations: CBASP, cognitive behavioral analysis system of psychotherapy; HRS-D24, 24-item Hamilton Rating Scale for Depression. *Response was defined as a reduction of 50% or greater in the HRS-D24 score and a total score of 15 or lower. †Remission was defined as a final HRS-D24 score of 8 or lower. ‡Percentages may not add to overall values because of rounding. §The ␹2 values were determined using the Cochran-Mantel-Haenszel test; df = 1 for all ␹2 values.

substantial benefit for chronically depressed patients. rate of approximately 50%. As in the initial trial,10 CBASP Treatment with nefazodone following nonresponse to was associated with significantly less attrition due to ad- CBASP and treatment with CBASP following nonre- verse events than nefazodone, which is in keeping with sponse to nefazodone were equally effective for patients the fact that patients were not blind to their treatment who completed 12 weeks of treatment, with a response and the fact that psychotherapy has few adverse events.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, MAY 2005 WWW.ARCHGENPSYCHIATRY.COM 517

Nefazodone → CBASP CBASP → Nefazodone

Outcome Adjusted Adjusted Measure Baseline Endpoint Change Change† Baseline Endpoint Change Change† Statistics‡ Intention-to-Treat Sample

HRS-D24 24.1 (6.5) 13.4 (8.4) −10.7 (9.1) −11.1 (1.1) 23.2 (6.5) 15.4 (9.1) −7.8 (10.2) −8.1 (1.0) F1,126 = 4.12, P = .045

CGI-S 4.2 (0.8) 3.1 (1.1) −0.9 (1.0) −0.9 (0.2) 4.0 (0.6) 3.0 (1.3) −1.1 (1.4) −1.1 (0.1) F1,98 = 0.27, P = .60

IDS-SR30 33.9 (10.2) 20.4 (12.3) −13.7 (11.8) −14.6 (1.6) 33.4 (9.6) 22.2 (13.8) −11.2 (14.0) −11.3 (1.4) F1,124 = 2.34, P = .13

HAM-APA 1.5 (0.6) 0.8 (0.6) −0.6 (0.6) −0.7 (0.1) 1.5 (0.5) 1.0 (0.7) −0.4 (0.8) −0.4 (0.1) F1,125 = 4.21, P = .04 Completers

HRS-D24 24.1 (6.6) 11.9 (7.7) −12.2 (8.3) −12.6 (1.1) 23.6 (6.5) 13.3 (8.3) −10.2 (10.2) −10.2 (1.1) F1,96 = 2.26, P = .14

CGI-S 4.3 (0.9) 2.9 (1.1) −1.2 (1.0) −1.1 (0.2) 4.1 (0.6) 2.6 (1.3) −1.4 (1.4) −1.4 (0.2) F1,70 = 1.25, P = .27

IDS-SR30 34.0 (10.7) 18.5 (11.8) −15.2 (11.0) −15.9 (1.6) 34.2 (9.7) 19.3 (12.6) −15.0 (13.5) −14.5 (1.6) F1,95 = 0.36, P = .55

HAM-APA 1.4 (0.6) 0.7 (0.6) −0.7 (0.6) −0.6 (0.1) 1.5 (0.6) 0.9 (0.6) −0.6 (0.8) −0.6 (0.1) F1,96 = 1.51, P = .22

Abbreviations: CBASP, cognitive behavioral analysis system of psychotherapy; CGI-S, Clinical Global Impressions–Severity scale; HAM-APA, Hamilton Rating Scale for Anxiety Psychic Anxiety factor subscale; HRS-D24, 24-item Hamilton Rating Scale for Depression; IDS-SR30, 30-item Inventory for Depressive Symptomatology–Self-Report. *Values are mean (SD). †Least squares adjusted means. ‡P values are comparing adjusted mean change from baseline to end point.

rable in the nefazodone groups whether it was the first 30 CBASP Followed by Nefazodone or second treatment received, whereas a significantly Nefazodone Followed by CBASP smaller proportion of patients dropped out while receiv- 25 ing CBASP when it was the second treatment in the se- quence. It appears that the experience of an unsuccess- 20 ful initial trial of pharmacotherapy may have strengthened

Score the motivation of patients to participate in an adequate 24 15 trial of psychotherapy, whereas nonresponse to CBASP

HRS-D did not improve patients’ ability to tolerate unpleasant 10 adverse effects of medication.

5 Interestingly, the overall response and remission rates observed in the intent-to-treat crossover sample were simi- 10 0 lar to those observed in the acute phase of this study as Baseline Week 1 Week 2 Week 3 Week 4 Week 6 Week 8 Week 10 Week 12 well as to those observed by our collaborative research Visit group8 in another sample of chronically depressed patients who were crossed over to sertraline or imipra- Figure 3. Mean HRS-D24 (24-item Hamilton Rating Scale for Depression) scores in the crossover phase. CBASP indicates cognitive behavioral analysis mine following nonresponse to one of these two antide- system of psychotherapy. pressant medications. These results underscore how important it is that clinicians not be discouraged— Although there was a significant difference in the over- though their chronically depressed patients may be—by all response rate between the two crossover groups in the initial treatment resistance. intent-to-treat sample (57% response for those who Few data are available to guide the clinician about the crossed over to CBASP vs 42% for those who crossed over best way to treat depressed patients who have not re- to nefazodone), this was likely due to the higher overall sponded to an initial adequate course of treatment. The attrition rate among those receiving nefazodone than present study provides support for a switch from medi- among those receiving CBASP. A slower titration of nefaz- cation to psychotherapy and vice versa. Other strategies odone might have resulted in a lower number of pa- include switching within treatment (eg, switch to an- tients dropping out of the crossover trial because of ad- other antidepressant medication) or augmentation either verse events, potentially increasing the response and with another treatment of the same modality or with a remission rates. treatment of a different modality (ie, combined treat- The finding of differential attrition was unexpected. ments). Compared with medication augmentation strat- In the acute-phase trial, although the reasons for prema- egies for nonresponders to an antidepressant medica- ture termination were different in the CBASP and nefaz- tion, switching antidepressant classes has the advantages odone groups, overall dropout rates were almost iden- of simplicity and parsimony (eg, lower cost and lower tical. Specifically, greater attrition due to adverse events risk of drug-drug interactions). However, further stud- in the nefazodone group was essentially matched by at- ies are required to test the relative short- and long-term trition due to dissatisfaction with treatment in the CBASP effectiveness of the different strategies as well as their rela- group. Across the sequential trials, attrition was compa- tive acceptability. In the present study, a larger propor-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, MAY 2005 WWW.ARCHGENPSYCHIATRY.COM 518

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/29/2021 tion of participants were willing to cross over to medi- randomization of nonresponders to alternate treat- cation after nonresponse to psychotherapy than the ments, interpretation of these comparative results re- reverse. mains tentative. On the other hand, these results have Although significant improvement was observed in ap- high applicability to routine practice, where treatment proximately 50% of patients who completed the cross- would never be chosen by randomization. over treatment, only about 1 in 4 patients (28% [16/57] This is the first prospective, comparative random- for nefazodone and 25% [13/53] for CBASP) achieved full ized study of switching depressed outpatients from medi- remission; that is, about half of the responders still had cation to psychotherapy and vice versa in the context of significant residual symptoms at the end of the cross- nonresponse to but completion of the first treatment. over phase. Responders without remission have less ro- Given the magnitude of the problem of nonresponse to bust social recoveries than those who achieve full remis- antidepressant medication, randomized controlled stud- sion21,22 and are at higher risk for subsequent relapse.23 ies of switch strategies as well as augmentation strate- Chronically depressed patients might benefit from the gies are desperately needed. The only other large con- combination of both CBASP and nefazodone or from trolled crossover trial found that roughly 50% of longer courses of treatment. Consistent with this, roughly imipramine nonresponders responded to sertraline and 40% of chronically depressed patients who had a re- vice versa; this study was conducted among chronically sponse but not a remission after 12 weeks of acute- depressed outpatients who completed 12 weeks of acute phase treatment with either imipramine or sertraline con- treatment with either of these two medications.8 verted to full remission during continuation therapy with It is essential to know more about the range of effi- the same agent.24 cacy of other second-step treatments. For example, we Several methodological limitations affect interpreta- do not know whether nonresponders to either nefaz- tion of these results. First, in the absence of a placebo odone or CBASP, given a different medication as a switch condition, it is difficult to determine whether the effects strategy, would have achieved equivalent or even greater that are observed are due to treatment. We did not use a success. The recently initiated Sequenced Treatment Al- placebo-controlled design in the crossover phase be- ternatives to Relieve Depression (STAR*D) (www.star-d cause of concerns about withholding active treatment from .org) randomized controlled trial is comparing 4 differ- chronically depressed patients who had already not re- ent switch treatments (sustained-release bupropion, sponded to initial treatment with nefazodone or CBASP. cognitive therapy, sertraline, or extended-release venla- Absence of a placebo control group is less critical in the faxine) in outpatients who do not respond to a selective study of populations in which low placebo response rates serotonin reuptake inhibitor (citalopram) as the first treat- have been documented. Such populations include those ment. The STAR*D trial will determine whether a switch with treatment-resistant depression (placebo response within the class of selective serotonin reuptake inhibi- rates of 10%-20%)25-27 and those with a chronic depres- tors, an out-of-class switch, a switch to a dual-action agent, sive illness (double depression; placebo response rate, or a switch to cognitive therapy is most effective. 13%).28 The findings of the present study provide the first evi- Second, switching treatments only for nonre- dence from a controlled trial of the value of switching to sponders who were able to complete the acute trial lim- an antidepressant medication or to psychotherapy after its the generalizability of findings to such acute study com- the failure of the alternative treatment in an initial trial. pleters. If we had switched treatments at the point where The evidence specifically pertains to nefazodone and nonresponse was evident rather than at a specific pre- CBASP. The general principle may hold for other anti- determined time point, we might have observed higher depressant medications, as suggested by the findings of (or lower) response or remission rates during the cross- Thase et al8 with imipramine and sertraline. It remains over trial. Similarly, we cannot determine how well the to be determined whether other forms of depressed- switch strategy would have worked for those who de- focused psychotherapies would yield similar benefits to cided to drop out prior to completing the acute phase of CBASP. It would be of interest, for example, to deter- treatment. Third, although patients with a number of co- mine whether therapies with a more interpersonal or psy- morbid conditions were allowed to participate in this chodynamic focus are useful when a cognitive behav- study, the exclusion of patients with greater levels of co- ioral strategy is not. morbidity further limits generalizability of our findings. For patients with chronic depression, the present re- We do not know whether such excluded patients would sults provide a strong basis for switching to CBASP after have responded preferentially to one or the other cross- a medication does not produce a response and, con- over treatment. versely, for switching to medication after patients do not Fourth, nonresponders in the two groups who en- respond to an adequate trial of psychotherapy. tered the crossover phase may not have been compa- rable since they were not rerandomized at initiation of Submitted for Publication: November 17, 2003; final re- the switch. However, our confidence in the validity of vision received October 8, 2004; accepted November 16, comparing these groups is strengthened by the absence 2004. of significant differences in sociodemographic and clini- Author Affiliations: Department of Psychiatry and Be- cal characteristics in the main trial10 and in the cross- havioral Sciences, Stanford University School of Medi- over trial. In fact, minimal improvement during the ini- cine, Stanford, Calif (Drs Schatzberg, Arnow, and Man- tial treatment trial was predictive of success (albeit weakly) ber); Department of Psychiatry, University of Texas with both crossover strategies. Nevertheless, without re- Southwestern Medical Center at Dallas (Drs Rush and

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, MAY 2005 WWW.ARCHGENPSYCHIATRY.COM 519

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/29/2021 Trivedi); i3STATPROBE Inc, Dublin, Ohio (Mr Banks); 9. McCullough JP Jr. Treatment for Chronic Depression: Cognitive Behavioral Analy- Behavioral Science Department, University of Texas MD sis System of Psychotherapy (CBASP). New York, NY: Guilford Press; 2000. 10. Keller MB, McCullough JP, Klein DN, Arnow B, Dunner DL, Gelenberg AJ, Marko- Anderson Cancer Center, Houston (Dr Blalock); Bristol- witz JC, Nemeroff CB, Russell JM, Thase ME, Trivedi MH, Zajecka J. A compari- Myers Squibb Co, Plainsboro, NJ (Ms Borian); Depart- son of nefazodone, the cognitive behavioral-analysis system of psychotherapy, ment of Psychiatry, University of Pittsburgh School of and their combination for the treatment of chronic depression. N Engl J Med. Medicine (Drs Howland and Thase), Pittsburgh, Pa; De- 2000;342:1462-1470. partment of Psychology, State University of New York 11. First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Interview for DSM-IV Axis I Disorders–Patient Edition (SCID-I/P, Version 2.0). New York, NY: at Stony Brook (Dr Klein); Department of Psychiatry, Cor- Biometrics Research Department, New York State Psychiatric Institute; 1995. nell University Medical College, New York, NY (Drs Koc- 12. First MB, Gibbon M, Spitzer RL, Williams JBW, Benjamin LS. Structured Clinical sis and Markowitz); Department of Psychiatry, Virginia Interview for DSM-IV Axis II Disorders (SCID-II). Washington, DC: American Psy- Commonwealth University, Richmond (Dr Kornstein); chiatric Press; 1997. Department of Psychiatry and Human Behavior, Brown 13. Fawcett J, Epstein P, Fiester SJ, Elkin I, Autry JH. Clinical management–imipramine/ placebo administration manual: NIMH Treatment of Depression Collaborative Re- Medical School, Providence, RI (Drs Miller and Keller); search Program. Psychopharmacol Bull. 1987;23:309-324. and Department of Psychiatry, Emory University School 14. McCullough JP. Therapist Manual for Cognitive Behavioral Analysis System of of Medicine, Atlanta, Ga (Drs Ninan and Rothbaum). Psychotherapy. Richmond: Virginia Commonwealth University; 1995. Correspondence: Alan F. Schatzberg, MD, Department 15. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960; of Psychiatry and Behavioral Sciences, Stanford Univer- 23:56-62. 16. Hamilton M. Development of a rating scale for primary depressive illness. Br J sity School of Medicine, 401 Quarry Rd, Stanford, CA Soc Clin Psychol. 1967;6:278-296. 94305-5717 ([email protected]). 17. Miller IW, Bishop S, Norman WH, Maddever H. The modified Hamilton Rating Scale Funding/Support: This multicenter project was sup- for Depression: reliability and validity. Psychiatry Res. 1985;14:131-142. ported in part by a series of grants from Bristol-Myers 18. Guy W. ECDEU Assessment Manual for Psychopharmacology. Rockville, Md: Psy- Squibb Co, New York, NY. chopharmacology Research Branch, National Institutes of Health; 1976. Publi- cation ADM 76-338. 19. Rush AJ, Giles DE, Schlesser MA, Fulton CL, Weissenburger JE, Burns CT. REFERENCES The Inventory for Depressive Symptomatology (IDS): preliminary findings. Psy- chiatry Res. 1986;18:65-87. 1. Depression Guideline Panel. Depression in Primary Care: Treatment of Major 20. Rush AJ, Gullion CM, Basco MR, Jarrett RB, Trivedi MH. The Inventory of De- Depression. Vol 2. Rockville, Md: Agency for Health Care Policy and Research, pressive Symptomatology (IDS): psychometric properties. Psychol Med. 1996; Public Health Service, Dept of Health and Human Services; 1993. Clinical Prac- 26:477-486. tice Guideline, No. 5. AHCPR Publication 93-0551. 21. Miller IW, Keitner GI, Schatzberg AF, Klein DN, Thase ME, Rush AJ, Markowitz 2. Thase ME, Rush AJ. Treatment-resistant depression. In: Bloom FE, Kupfer DJ, JC, Schlager DS, Kornstein SG, Davis SM, Harrison WM, Keller MB. The treat- eds. Psychopharmacology: The Fourth Generation of Progress. New York, NY: ment of chronic depression, part 3: psychosocial functioning before and after Raven Press; 1995:1081-1097. treatment with sertraline and imipramine. J Clin Psychiatry. 1998;59:608-619. 3. Thase ME, Rush AJ. When at first you don’t succeed: sequential strategies for 22. Judd LL, Akiskal HS, Maser JD, Zeller PJ, Endicott J, Coryell W, Paulus MP, Ku- antidepressant nonresponders. J Clin Psychiatry. 1997;58:23-29. novac JL, Leon AC, Mueller TI, Rice JA, Keller MB. A prospective 12-year study 4. Rush AJ, Kupfer DJ. Strategies and tactics in the treatment of depression. In: of subsyndromal and syndromal depressive symptoms in unipolar major de- Gabbard GO, ed. Treatment of Psychiatric Disorders. 4th ed. Washington, DC: pressive disorders. Arch Gen Psychiatry. 1998;55:694-700. American Psychiatric Association; 2001:1417-1439. 23. Paykel ES, Scott J, Teasdale JD, Garland A, Moore R, Jenaway A, Cornwall PL, 5. Fava M, Rush AJ, Trivedi MH, Nierenberg AA, Thase ME, Sackeim HA, Quitkin Hayhurst H, Abbott R, Pope M. Prevention of relapse in residual depression by FM, Wisniewski S, Lavori PW, Rosenbaum JF, Kupfer DJ; for the STAR*DIn- cognitive therapy. Arch Gen Psychiatry. 1999;56:829-835. vestigators Group. Background and rationale for the Sequenced Treatment Al- 24. Koran LM, Gelenberg AJ, Kornstein SG, Howland RH, Friedman RA, DeBattista ternatives to Relieve Depression (STAR*D) study. Psychiatr Clin North Am. 2003; C, Klein D, Kocsis JH, Schatzberg AF, Thase ME, Rush AJ, Hirschfeld RM, La- 26:457-494. Vange LM, Keller MB. Sertraline versus imipramine to prevent relapse in chronic 6. Stewart JW, Mercier MA, Agnosti V, Guardino M, Quitkin FM. Imipramine is ef- depression. J Affect Disord. 2001;65:27-36. fective after unsuccessful cognitive therapy: sequential use of cognitive therapy 25. Joffe RT, Singer W, Levitt AJ, MacDonald C. A placebo-controlled comparison and imipramine in depressed outpatients. J Clin Psychopharmacol. 1993;13: of lithium and triiodothyronine augmentation of tricyclic antidepressants in uni- 114-119. polar refractory depression. Arch Gen Psychiatry. 1993;50:387-393. 7. Rush AJ, Thase ME. Psychotherapies for depressive disorders: a review. In: Maj 26. McAskill R, Mir S, Taylor D. Pindolol augmentation of antidepressant therapy M, Sartorius N, eds. Depressive Disorders. Chichester, England: John Wiley & Sons [published correction appears in Br J Psychiatry. 1998;173:443]. Br J Psychiatry. Ltd; 1999:161-206. WPA Series: Evidence and Experience in Psychiatry; vol 1. 1998;173:203-208. 8. Thase ME, Rush AJ, Howland RH, Kornstein SG, Kocsis JH, Gelenberg AJ, 27. Landen M, Goran B, Agren H, Fahlen T. A randomized, double-blind, placebo- Schatzberg AF, Koran LM, Keller MB, Russell JM, Hirschfeld RM, LaVange LM, controlled trial of buspirone in combination with an SSRI in patients with treat- Klein DN, Fawcett J, Harrison W. Double-blind switch study of imipramine or ment refractory depression. J Clin Psychiatry. 1998;59:664-668. sertraline treatment of antidepressant-resistant chronic depression. Arch Gen 28. Kocsis JH, Frances AJ, Voss C, Mann JJ, Mason BJ, Sweeney J. Imipramine treat- Psychiatry. 2002;59:233-239. ment for chronic depression. Arch Gen Psychiatry. 1988;45:253-257.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 62, MAY 2005 WWW.ARCHGENPSYCHIATRY.COM 520