Sertraline is more effective than imipramine in the treatment of non-melancholic depression:

Progress in Neuro-Psychopharmacology & Biological Psychiatry 27 (2003) 493 – 500 www.elsevier.com/locate/pnpbp

Sertraline is more effective than imipramine in the treatment of non-melancholic depression: Results from a multicentre, randomized study Enrique Bacaa, Manuel Gonza´lez de Cha´vezb, Mauro Garcı´a-Toroc, Francisco Pe´rez-Arnaud, Alberto Porras-Chavarinoe,* a Department of Psychiatry, Hospital Puerta de Hierro, Madrid, Spain Department of Psychiatry, Hospital Gregorio Maran˜o´n, Madrid, Spain c Hospital Psiquia´trico, Palma de Mallorca, Spain d C.S.M. Prat de Llobregat, Barcelona, Spain e Medical Unit, Pfizer S.A., Parque Empresarial La Moraleja, Avda. Europa 20 B, 28108 Alcobendas Madrid, Spain b

Accepted 22 January 2003

Abstract The acute treatment efficacy, tolerability, and effects on health-related quality of life of sertraline (50 – 200 mg/day) versus imipramine (75 – 225 mg/day) were compared in outpatients with non-melancholic depression. The study employed an open-label, parallel-group design. One hundred and sixteen patients were randomized to receive sertraline and 123 to receive imipramine for 8 weeks. In the intent-to-treat (ITT), last-observation-carried-forward (LOCF) analysis, sertraline produced statistically significantly greater improvements in depressive (21-item Hamilton Depression Rating Scale [HAM-D21] scores of 24.9 and 24.4 were reduced to 10.3 and 13.1 at endpoint, P < .005) and anxiety symptoms (Hamilton Anxiety Rating Scale [HAM-A] scores of 21.8 and 21.9 were reduced to 9.5 and 13.9, P < .01), as well as in response (69.0% versus 53.7% at endpoint, P=.016) and remission rates (51.3% versus 38.0% at endpoint, P=.041) from week 4 onwards compared with imipramine. The proportion of patients who were ‘very much improved’ or ‘much improved’ (Clinical Global Impressions Scale of Improvement [CGI-I] score of 1 or 2) was significantly higher at endpoint in the sertraline group (76.1%) than in the imipramine group (62.8%) ( P=.028). At week 8, patients in both treatment groups showed clear improvements in quality of life, although nonstatistically significant differences were evident in the quality of life of sertraline- versus imipramine-treated patients. Sertraline was significantly superior in tolerability with less discontinuations due to adverse events (10.3%) compared with the imipramine group (24.4%) ( P=.004). It was concluded that sertraline is more effective than imipramine in the acute treatment of depressive and anxiety symptoms in patients with nonmelancholic depression. D 2003 Elsevier Science Inc. All rights reserved. Keywords: Clinical trial; Depressive disorder; Dysthymic disorder; Imipramine; Sertraline

1. Introduction Many studies have shown that the selective serotonin reuptake inhibitors (SSRIs) are an effective treatment for depression (Amsterdam, 1998; Kernick, 1997; Reimherr et al., 1990). Studies also suggest that SSRIs are better tol-

Abbreviations: BQOL, Batelle Quality of Life Questionnaire; CGI-I, Clinical Global Impressions Scale of Improvement; CGI-S, Clinical Global Impressions Scale of Severity; HAM-A, Hamilton Anxiety Rating Scale; HAM-D21, 21-item Hamilton Depression Rating Scale; ITT, Intention-totreat; LOCF, last-observation-carried-forward; SSRIs, selective serotonin reuptake inhibitors; TCAs, Tricyclic antidepressants. * Corresponding author. Tel.: +34-91-4909825; fax: +34-91-4909720. E-mail address: [email protected] (A. Porras-Chavarino).

erated and have higher patient compliance with treatment when compared to tricyclic antidepressants (TCAs) (Deakin, 1996; Kasper et al., 1992; Moller, 1997). The results of a recent meta-analysis by the Cochrane Group (Geddes et al., 2002) found an effect of size (for the difference between SSRIs and TCAs) of 0.044 (95% CI, 0.020– 0.107), suggesting no statistically or clinically significant efficacy difference between the two drug classes. This meta-analysis also specifically examined the efficacy of SSRIs versus imipramine. Based on data from 23 available trials, the between-class effect of size was 0.040 (95% CI, = 0.126 –0.046), again suggesting no difference in efficacy. Sertraline was the only SSRI that did not contribute data to the SSRI versus imipramine metaanalysis.

0278-5846/03/$ – see front matter D 2003 Elsevier Science Inc. All rights reserved. doi:10.1016/S0278-5846(03)00038-1

494

E. Baca et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 27 (2003) 493–500

Recent meta-analyses (Anderson, 2000) have also evaluated the comparative tolerability of SSRIs and TCAs. The weight of available evidence from short-term efficacy studies suggests that SSRIs are notably better tolerated, with rates of treatment discontinuations that are in the range of 12 –27% lower compared to TCAs, both overall (relative risk 0.88; 95% CI, 0.83 – 0.93) and due to side effects (relative risk, 0.73; 95% CI, 0.67 – 0.80) (Anderson, 2000). Classic randomized, placebo-controlled clinical trials provide valid and reliable information about the pharmacologic efficacy of a medication or class of medications. Because of the methodological rigor of their design though, such trials have limited generalizability to actual clinical practice. Recently, there has been a growing interest in ‘‘effectiveness’’ trials that are less methodologically rigorous, and therefore permit the evaluation of the effectiveness and patient acceptance of an antidepressant treatment in a way that is more readily generalizable to clinical practice. The large NIMH-funded Star * D program is an example of such a trial that is ongoing (Rush et al., 2002). The purpose of the current study was to evaluate the comparative effectiveness and patient acceptance of sertraline versus imipramine in outpatients diagnosed with nonmelancholic depression. Random assignment to treatment was employed, but in all other regards, study design and study procedures were similar to clinical practice (e.g., openlabel treatment, titration based on tolerability and clinician judgment).

2. Methods 2.1. Patient population Patients eligible for participation in the study were outpatients of both genders, aged  18 years, not suffering from significant concomitant medical diseases and with a DSM-III-R (American Psychiatric Association, 1987) diagnosis of major depression (single or recurrent) with or without dysthymia. Patients were required to have a score of at least 18 at baseline on the 21-item Hamilton Depression Rating Scale (HAM-D21) (Hamilton, 1960, 1967). Patients were excluded if they fulfilled the criteria for meeting DSM-III-R (American Psychiatric Association, 1987) and ICD-10 (World Health Organization, 1992) depression of the melancholic type, or if their HAM-D21 score had decreased by more than 50% between screening and baseline assessments, or if they had previously failed to respond to antidepressant treatment or had a history of psychoses. Women were excluded if pregnant or if not using a reliable method of contraception throughout the study. 2.2. Drug administration Patients not receiving psychoactive treatment started study treatment without washout period. Patients currently

taking psychoactive medication entered a washout period whose duration corresponded to five half-lives of their previous medication, or 1 week, whichever was longer. Patients were then randomly assigned to start on an oral dose of 50 mg sertraline (hydrochloride) or 75 mg imipramine (hydrochloride), once daily. Depending on clinical response and tolerability, the dose of either drug could be increased in 50 mg/day increments to sertraline 100, 150, or 200 mg/day, or imipramine 125, 175, or 225 mg/day for imipramine. Doses were titrated up to the next level at the investigator’s discretion, after a minimum of 2 weeks at the previous level. Once patients had reached a clinically effective dose, they were maintained on that dose throughout the study unless side effects necessitated a reduction in dose. 2.3. Study design The study had a multicentre, randomized, open-label, parallel-group design. Subjects were randomized in a 1:1 ratio to treatment with flexible doses of sertraline or imipramine for 8 weeks. The study protocol was approved by the relevant ethics committees, and was conducted in accordance with the Declaration of Helsinki (Hongkong revision, 1989). All subjects gave written informed consent to participate. 2.4. Assessments The following physician-rated instruments were used in the study at baseline and assessments at 2, 4, 6, and 8 weeks: HAM-D21 (Hamilton, 1960, 1967), the Hamilton Anxiety Rating Scale (HAM-A) (Hamilton, 1967), and the Clinical Global Impressions Scales (CGI) of Severity (CGI-S) and Improvement (CGI-I) (Guy, 1976). In addition, patients completed the Batelle Quality of Life Questionnaire (BQOL) (Revicki et al., 1992) at baseline and week 8 of the study. Safety was assessed by recording all adverse events during the study and by haematology, biochemistry, and urinalysis on samples taken at baseline and at the final study visit. In addition, the patient and investigator made global assessments of tolerability at the final visit according to a four-point scale (‘excellent’, ‘good’, ‘satisfactory,’ and ‘poor’). 2.5. Statistical methods Efficacy variables were analysed using the last-observation-carried-forward (LOCF) analysis performed on the intent-to-treat (ITT) group (all randomized patients for whom any post-randomization efficacy assessment was available: 113 sertraline, 121 imipramine). All patients who had taken any study medication and for whom any safety information was available were included in the safety analysis. The primary efficacy measures were the proportions of responders to treatment and patients who were in remission

E. Baca et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 27 (2003) 493–500

according to the HAM-D criterion. Response to treatment was defined as a reduction in HAM-D21 score of  50% from baseline (Frank et al., 1991), and remission as a score of  8 on the HAM-D21 scale (Thase et al., 2001). Note that the consensus definition of remission is a HAM-D total score < 7 based on the 17-item scale. Because a 21-item HAM-D scale was used in the current study, the HAM-D remission cutoff score has been raised by one point to 8. The proportions of responders according to the CGI criterion of ‘very much improved’ (CGI-I = 1) were also calculated. The proportion of responders or those in remission was analysed by the chisquare test. Repeated measures ANOVA was used to examine group differences on HAM-D21, HAM-A, and BQOL scores. The required sample size was determined as follows. Based on a review of the literature, a response rate to imipramine of about 60% could be expected. A response rate to sertraline in a clinical sample was estimated at 75%, based on a similar large open-label study (Lydiard et al., 1999). To detect a difference between those response rates with a=.05 and b = 0.2 (i.e., 80% power), at least 149 patients in each treatment group would be required. Tolerability data on the safety population (116 sertraline, 123 imipramine) were analysed using descriptive statistics. The proportions of patients suffering adverse events or discontinuing therapy were also calculated on the safety population. All statistical analyses were done using SAS. All tests were two sided, and statistical significance was taken at the 5% level.

3. Results 3.1. Subject characteristics A total of 239 patients (i.e., the safety population) from 12 centres in Spain entered the study, 116 in the sertraline group and 123 in the imipramine group. The study therefore had slightly lower statistical power than had been planned in the protocol, which specified that each group should have had at least 149 patients. The two groups were well matched demographically and in severity and type of depression (Table 1). 3.2. Dosages In the ITT population, the mean doses ( ± S.D.) at the end of the study were 89.1 ± 44.5 mg/day for patients randomized to sertraline and 101.6 ± 39.1 mg/day for those randomized to imipramine. 3.3. Discontinuations In the safety population groups, more sertraline-treated patients (93:116; 80.2%) than imipramine-treated patients (78:123; 63.4%) completed the 8-week study period. Stat-

495

Table 1 Baseline demographic characteristics of study subjects (ITT analysis)

Number of patients Gender Male Female Age (years) Mean (range)

Sertraline

Imipramine

113

121

23 (20%) 90 (80%)

27 (22%) 94 (78%)

39 (18 – 72)

41 (18 – 72)

DSM-III-R diagnosis at baseline [n (%)] Major depression 53 (47%) Dysthymia 42 (37%) Double depression 18 (16%)

60 (50%) 40 (33%) 21 (17%)

Rating scale scores at baseline Mean HAM-D21 ( ± S.D.) Mean HAM-A ( ± S.D.) Mean CGI-S (range)

24.9 ( ± 4.8) 21.8 ( ± 6.1) 4.4 (2 – 6)

24.4 ( ± 4.5) 21.9 ( ± 6.8) 4.5 (3 – 6)

Marital status Unmarried Married Separated Widowed

36 (32%) 64 (57%) 9 (8%) 4 (4%)

34 (28%) 71 (59%) 10 (8%) 6 (5%)

Employment status Employed Unemployed Homemaker Student Retired

44 (39%) 20 (18%) 35 (31%) 7 (6%) 7 (6%)

55 (45%) 16 (13%) 41 (34%) 4 (3%) 4 (3%)

istically, significantly more patients in the imipramine group (30; 24.4%) than in the sertraline group (12; 10.3%) discontinued treatment due to adverse events ( P=.004). Patients could discontinue due to more than one adverse event. The most commonly reported adverse events leading to discontinuation were anxiety (6.0%), insomnia (1.7%), sweating (1.7%), vomiting (1.7%), and dizziness (1.7%) in the sertraline group and anxiety (8.9%), dry mouth (5.7%), tremor (5.7%), constipation (3.2%), vomiting (3.2%), and dizziness (3.2%) in the imipramine group. 3.4. Efficacy Although both treatments resulted in a substantial reduction in HAM-D21 scores from baseline, the magnitude of reduction with sertraline was significantly greater than with imipramine at all assessment times from week 4 onwards ( P < .005, repeated measures ANOVA, multiple comparisons using Bonferroni’s aproach; ITT LOCF analysis) (Fig. 1). Using the HAM-D criterion, the response rates were significantly higher in the sertraline group (69.0% at endpoint) than in the imipramine group (53.7% at endpoint) at all assessment times from week 4 onwards ( P < .05, chisquare test; ITT LOCF analysis). Similarly, remission rates were significantly higher in the sertraline group (51.3% at endpoint) than in the imipramine group (38.0% at endpoint)

496

E. Baca et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 27 (2003) 493–500

Fig. 1. Mean absolute values of HAM-D21 score during treatment (repeated measures ANOVA; ITT LOCF analysis).

at all assessment times from week 4 onwards ( P < .05, chisquare test; ITT LOCF analysis) (Fig. 2a and b). Mean HAM-A scores decreased substantially compared to baseline during treatment with both drugs (Fig. 3).

Fig. 2. Response (a) and remission (b) rates. Response is defined as a reduction from baseline in HAM-D21 score  50%. Remission is defined as a score of  8 on the HAM-D21 scale (chi-square test; ITT LOCF analysis).

Fig. 3. Mean absolute values of HAM-A score during treatment (repeated measures ANOVA; ITT LOCF analysis).

However, the reduction was greater with sertraline than with imipramine from week 2 onwards ( P < .01, repeated measures ANOVA, Bonferroni corrected; ITT LOCF analysis). The proportion of patients who achieved a HAM-A score of  8 was significantly greater in the sertraline group than the imipramine group from week 2 onwards ( P < .05, chi-square test; ITT LOCF analysis). Similarly, the proportion of patients with a  50% reduction from baseline in HAM-A score was significantly greater in the sertraline group from week 4 onwards ( P < .05, chi-square test; ITT LOCF analysis). Superiority of sertraline over imipramine was further demonstrated on the CGI-I and CGI-S scales ( P < .05, chisquare test; ITT LOCF analysis). The proportion of patients who were ‘very much improved’ or ‘much improved’ (CGII = 1 or 2) was significantly higher at endpoint in the sertraline group (76.1%) than in the imipramine group (62.8%) ( P=.028, chi-square test; ITT LOCF). In addition, a statistically significantly higher proportion of sertralinetreated patients than imipramine-treated patients was classified as treatment responders using the stringent CGI criterion (CGI-I = 1, ‘very much improved’) at all assessments from week 4 onwards (Fig. 4a). The results on the CGI-S scale were similar, using the same stringent criterion (CGI-S = 1), with statistically significant superiority of sertraline over imipramine at weeks 6 and 8 ( P < .05, chi-square test; ITT LOCF analysis) (Fig. 4b). The proportion of patients who achieved a CGI-S score of 1 or 2 was also significantly greater for sertraline than imipramine at weeks 6 and 8 ( P < .05, chi-square test; ITT LOCF analysis) At endpoint, the proportion of patients achieving a CGI-S score of 1 or 2 was 63.7% for sertraline versus 47.1% for imipramine ( P=.022, chi-square test; ITT LOCF analysis).

E. Baca et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 27 (2003) 493–500

497

Table 2 Mean change from baseline in HAM-D21 score in subgroups (ITT LOCF analysis) Subgroup

Sertraline n

Fig. 4. Proportion of patients who achieved scores of 1 (very much improved) on the CGI-I scale (a) and scores of 1 (normal, not ill) on the CGI-S scale (b) (chi-square test; ITT LOCF analysis).

Patients in both treatment groups showed clear improvements from baseline in quality of life assessed by the BQOL ( P < .05, repeated measures ANOVA, Bonferroni corrected for multiple comparisons). At endpoint, there were numerical advantages for sertraline over imipramine on all items, though none achieved statistical significance ( P>.05, repeated measures ANOVA, Bonferroni corrected for multiple comparisons). 3.5. Subgroup analyses The changes in HAM-D21 score were analysed in various subgroups of patients randomized to treatment with either drug (Table 2). The subgroup analyses were regarded as exploratory and were analysed using descriptive statistics. Compared to imipramine, patients treated with sertraline showed clinically meaningful efficacy advantages in several important subgroups. Endpoint improvement in HAM-D scores was 3.5 – 5 points greater on sertraline in women and young adults, in the anxious subtype of depression, and in severe depression and ‘‘double depression’’ (major depression superimposed on dysthymia). 3.6. Tolerability The results showed that sertraline was better tolerated than imipramine, with clear differences between treatments

Imipramine

Mean

S.D.

n

Mean

S.D.

Gender Men Women

23 90

12.6 15.1

7.4 8.3

27 94

11.1 11.3

8.6 8.8

Age (years) 18 – 44 45 – 64  65

74 34 5

15.3 13.3 12.2

8.0 7.9 11.6

76 40 5

11.7 10.3 12.2

8.5 8.7 13.0

Diagnosis Major depression Dysthymia Both

53 42 18

15.4 13.4 14.9

8.8 7.5 7.9

60 40 21

11.5 11.7 9.7

8.8 8.9 8.3

Baseline HAM-D21 score < 24 53  24 60

11.5 17.4

7.3 7.9

57 64

8.5 13.7

7.5 9.0

Baseline HAM-A score < 22 67  22 46

13.0 16.9

7.6 8.5

72 49

10.8 11.9

8.2 9.5

in the percentages of patients reporting adverse events at any time throughout the study. There were no serious adverse events during the study, defined as being potentially life threatening or that require hospitalisation. Furthermore, no adverse events were rated as severe in the sertraline group; in contrast, there were four adverse events rated as severe in the imipramine group. The most frequent adverse events (experienced by 10% or more of patients in either treatment group and regardless of severity) are shown in Table 3. The differences between sertraline- and imipramine-treated patients were particularly pronounced for anticholinergic effects, such as dry mouth (16.4% and 60.2%, respectively) and constipation (2.6% and 29.3%, respectively). Serotonergic adverse events, such as nausea, occurred more frequently in the sertraline (15.5%) than in the imipramine group (6.5%). Given the significant improvement in anxiety observed on both drugs

Table 3 Adverse events with a frequency of at least 10% in either group

Anxiety Constipation Diarrhoea Dizziness Dry mouth Gastritis Insomnia Nausea Sweating Tremor

Sertraline (%)

Imipramine (%)

11.2 2.6 12.1 9.5 16.4 12.9 8.6 15.5 6.9 6.9

14.6 29.3 0.8 16.3 60.2 6.5 10.6 6.5 19.5 20.3

498

E. Baca et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 27 (2003) 493–500

(Fig. 3), it may appear puzzling that anxiety was also recorded as an adverse event. This is a common finding in treatment studies of affective and anxiety disorders, where any transient worsening in anxiety symptomatology, regardless of causality, is recorded as an adverse event. During the study, no clinically significant laboratory abnormalities were considered to be related to treatment.

4. Discussion This study demonstrated consistent advantages of sertraline in treatment for depression and associated anxiety symptoms and tolerability compared with imipramine in the treatment of outpatients with non-melancholic depression. These advantages were present consistently on all efficacy and tolerability variables used in this study. 4.1. Efficacy in depression The superior efficacy of sertraline was evident from week 4 onward, and included significantly greater improvement in depressive symptoms as measured by the HAM-D (Fig. 1), as well as significantly higher endpoint response rates (Fig. 2). These findings differ from the results of a previously reported meta-analysis (Geddes et al., 2002) that found no efficacy difference for imipramine compared to other SSRIs. This meta-analysis did not include sertraline data. The emerging gold standard of antidepressant efficacy is remission, most frequently defined as an endpoint HAMD17 score  7. When a 21-item HAM-D is used, as in the current study, remission criteria  8 have been reported and appear to yield rates that are within one to two points of the HAM-D21 criteria (Thase et al., 2001). In the current study, remission rates were significantly higher on sertraline compared to imipramine (51% versus 38%). This result for sertraline is equivalent to the highest remission rates ever reported in comparable large outpatient depression studies. The highest remission rates have been reported by venlafaxine in combined data from four SSRI comparator studies (not placebo controlled, and not involving sertraline). In this combined analysis, 6 – 8 weeks of venlafaxine treatment resulted in remission rates of 48% (using HAM-D17  17 criteria). In contrast, the SSRI group achieved a remission rate of less than 40%. 4.2. Efficacy in depression subgroups The clinical presentation of depression, especially in the primary care setting, frequently is complicated by syndromic or subsyndromic levels of anxiety. More than 70% of patients suffering from major depression in primary care also suffer from a current anxiety disorder (Sartorius et al., 1996; Kessler et al., 1996; Olfson et al., 1997). The efficacy of an antidepressant is significantly enhanced, therefore, if it has additional anxiolytic efficacy. Sertraline has demonstra-

ted significant efficacy across a broad spectrum of anxiety disorders, including panic disorder (Pohl et al., 1998; Pollack et al., 1998; Londborg et al., 1998), obsessive – compulsive disorder (Greist et al., 1995), posttraumatic stress disorder (Davidson et al., 2001; Brady et al., 2000), social anxiety disorder (Van Ameringen et al., 2001), and generalized anxiety disorder (Brawman-Mintzer et al., 2002). In the current study, patients with the anxious depression subtype, defined by a baseline HAM-A score  22, showed a clinically meaningful improvement in endpoint HAM-D total score that was five points greater than imipramine. Patients also showed parallel improvement in anxiety symptomatology. Treatment with sertraline resulted in significantly greater reduction in HAM-A scores from week 2 onward (Fig. 3). In addition to efficacy benefits in anxious depression, sertraline also showed a numerical superiority over imipramine in two other clinically important depression subgroups (Table 2), patients with severe depression (baseline HAM-D21  24) and in patients suffering from double depression (major depression superimposed on chronic dysthymia). 4.3. Tolerability One of the notable findings of the current trial was the significantly greater tolerability of sertraline compared to imipramine. The side effect burden was markedly higher on imipramine (Table 3), largely contributed, as expected, by anticholinergic adverse events. In addition, the total discontinuation rate (imipramine 35.0% versus sertraline 17.7%), as well as the rate of discontinuation due to adverse events (imipramine 24.4% versus sertraline 10.3%) was more than double on imipramine compared to sertraline. These results are consistent with attrition data reported from meta-analyses of SSRIs versus TCAs (Anderson, 2000) and are especially striking because of the lower doses of imipramine used in the current study (mean dose, 102 mg/day). These differences in tolerability and treatment acceptance have significant clinical implications given that depression requires long-term treatment. Further evidence for the better tolerability of sertraline compared with imipramine is provided by the positive ratings for overall tolerability, which patients and investigators recorded for sertraline. More than half the patients and investigators rated imipramine treatment as having ‘poor’ or ‘bad’ tolerability, whereas sertraline therapy was rated ‘excellent’ or ‘good’ by more than 70% of both patients and investigators. 4.4. Quality of life Although patients in both treatment groups showed clear improvements in quality of life and sertraline demonstrated advantages over imipramine in all nine dimensions of the quality-of-life assessment, statistically significant differen-

E. Baca et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 27 (2003) 493–500

ces between both treatments were not found. It is notable that quality of life was assessed only at baseline and week 8 and, consequently, that only those subjects who completed 8 weeks of treatment were included in the analysis. As the dropout rate due to adverse events was significantly higher in the imipramine group, it is possible that the potential superiority of sertraline over imipramine on quality of life was underestimated in this study. 4.5. Limitations of the study It is important to note the limitations of the current study, some of which are the consequence of conducting a clinical effectiveness trial rather than a classic doubleblind, placebo-controlled efficacy study. These include lack of a placebo control to provide assay sensitivity for the trial (Temple and Ellenberg, 2000) and lack of blinding methodology. The latter may be of less importance than it seems at first since both study treatments are known to be effective antidepressants. Still, this is clearly a study limitation whose full effect is uncertain. Finally, an important limitation is that the mean dose of imipramine (102 mg) was lower than may have been necessary to optimize efficacy. This means that no inferences concerning the comparative pharmacologic efficacy of sertraline versus imipramine can be made. All that can be said is that from a practical clinical standpoint, sertraline was a significantly more effective and well-tolerated treatment, and that optimal dosing of imipramine appeared to be limited by lack of patient acceptance of its higher side effect burden.

5. Conclusions This study has shown that sertraline is a more effective and better tolerated treatment for depressive and anxiety symptoms than imipramine in outpatients with non-melancholic depression.

Acknowledgements Authors thank to additional investigators and collaborating sites: S. Oliveros (Hospital Puerta de Hierro, Madrid); B. Franco (C.S.M. Colmenar Viejo, Madrid); A. Rivera, B. Penasa, I. Lo´pez (C.S.M. Salamanca, Madrid); E. Capdevila (Centro de Psiquiatrı´a y Psicoterapia, Madrid); F. Teba (C.S.M. Prat de Llobregat, Barcelona); C. Gasto´, J. Blanch, L. Pintor (Hospital Clinic, Barcelona); J. Pe´rez de los Cobos, J. Fa´bregas, L. de Angel (Centro de Psicoterapia, Barcelona); R. Noguera, L. Jorda´ (Hospital Psiquia´trico, Gerona) A. Gonza´lez (H. Psiquia´trico, Palma de Mallorca). This study was supported by a research grant from Pfizer Spain.

499

References American Psychiatric Association, 1987. Diagnostic and Statistical Manual of Mental Disorders. Revised, 3rd ed. American Psychiatric Press, Washington, DC. Amsterdam, J., 1998. Selective serotonin reuptake inhibitor efficacy in severe and melancholic depression. J. Psychopharmacol. 12, S99 – S111 (Suppl.). Anderson, I.M., 2000. Selective serotonin reuptake inhibitors versus tricyclic antidepressants: a meta-analysis of efficacy and tolerability. J. Affect. Disord. 58, 19 – 36. Brady, K., Pearlstein, T., Asnis, G.M., Baker, D., Rothbaum, B., Sikes, C.R., Farfel, G.M., 2000. Efficacy and safety of sertraline treatment of posttraumatic stress disorder: a randomized controlled trial. JAMA 283, 1837 – 1844. Brawman-Mintzer, O., Rickels, K., Rynn, M., Knapp, R., Weihs, K., Shrivastava, R.K., Tilker, H., Patterson, W., Sambunaris, A., Rapaport, M., 2002. A placebo-controlled, double-blind, flexible dosage study to evaluate the efficacy and tolerability of sertraline in the treatment of DSMIV generalized anxiety disorder. Eur. Neuropsychopharmacol. 12 (Suppl. 3), S365. Davidson, J.R., Rothbaum, B.O., van der Kolk, B.A., Sikes, C.R., Farfel, G.M., 2001. Multicenter, double-blind comparison of sertraline and placebo in the treatment of posttraumatic stress disorder. Arch. Gen. Psychiatry 58, 485 – 492. Deakin, J.F.W., 1996. Does selectivity matter? Int. Clin. Psychopharmacol. 11 (Suppl. 1), 13 – 17. Frank, E., Prien, R.F., Jarrett, R.B., Keller, M.B., Kupfer, D.J., Lavori, P.W., Rush, A.J., Weissman, M.M., 1991. Conceptualization and rationale for consensus definitions of terms in major depressive disorder. Remission, recovery, relapse, and recurrence. Arch. Gen. Psychiatry 48, 851 – 855. Geddes, J.R., Freemantle, N., Mason, J., Eccles, M.P., Boynton, J., 2002. SSRIs versus other antidepressants for depressive disorder. Cochrane Database Syst. Rev. 2 (CD001851). Greist, J.H., Jefferson, J.W., Kobak, K.A., Chouinard, G., DuBoff, E., Halaris, A., Kim, S.W., Koran, L., Liebowtiz, M.R., Lydiard, B., 1995. A 1 year double-blind placebo-controlled fixed dose study of sertraline in the treatment of obsessive – compulsive disorder. Int. Clin. Psychopharmacol. 10, 57 – 65. Guy, W., 1976. Clinical global impressions. In: Guy, W. (Ed.), ECDEU Assessment Manual for Psychopharmacology. NIMH, Rockville, pp. 217 – 222. Hamilton, M., 1960. A rating scale for depression. J. Neurol. Neurosurg. Psychiatry 23, 56 – 62. Hamilton, M., 1967. Development of a rating scale for primary depressive illness. Br. J. Soc. Clin. Psychol. 6, 278 – 296. Kasper, S., Fuger, J., Moller, H.J., 1992. Comparative efficacy of antidepressants. Drugs 43 (Suppl. 2), 11 – 22. Kernick, D.P., 1997. Which antidepressant? A commentary from general practice on evidence- based medicine and health economics. Br. J. Gen. Pract. 47, 95 – 98. Kessler, R.C., Nelson, C., McGonagle, K.A., Liu, J., Swartz, M., Blazer, D.G., 1996. Comorbidity of DSM-III-R major depressive disorder in the general population: results from the US National Comorbidity Survey. Br. J. Psychiatry 168 (Suppl. 30), 17 – 30. Londborg, P.D., Wolkow, R., Smith, W.T., DuBoff, E., England, D., Ferguson, J., Rosenthal, M., Weise, C., 1998. Sertraline in the treatment of panic disorder. A multi-site, double-blind, placebo-controlled, fixeddose investigation. Br. J. Psychiatry 173, 54 – 60. Lydiard, R.B., Perera, P., Batzar, E., Clary, C.M., 1999. From the bench to the trench: a comparison of sertraline treatment of major depression in clinical and research patient samples. Prim. Care Companion 1, 154 – 162. Moller, H.J., 1997. Pharmacotherapy of depressed patients. Increasing the therapeutic arsenal with a new generation of antidepressive drugs. Fortschr. Med. 115, 30 – 36. Olfson, M., Fireman, B., Weissman, M.M., Leon, A.C., Sheehan, D.V., Kathol, R.G., Hoven, C., Farber, L., 1997. Mental disorders and dis-

500

E. Baca et al. / Progress in Neuro-Psychopharmacology & Biological Psychiatry 27 (2003) 493–500

ability among patients in a primary care group practice. Am. J. Psychiatry 154, 1734 – 1740. Pohl, R.B., Wolkow, R.M., Clary, C.M., 1998. Sertraline in the treatment of panic disorder: a double-blind multicenter trial. Am. J. Psychiatry 155, 1189 – 1195. Pollack, M.H., Otto, M.W., Worthington, J.J., Manfro, G.G., Wolkow, R., 1998. Sertraline in the treatment of panic disorder: a flexible-dose multicenter trial. Arch. Gen. Psychiatry 55, 1010 – 1016. Reimherr, F.W., Chouinard, G., Cohn, C.K., Cole, J.O., Itil, T.M., Lapierre, Y.D., Masco, H.L., Mendels, J., 1990. Antidepressant efficacy of sertraline: a double-blind, placebo- and amitriptyline-controlled, multicenter comparison study in outpatients with major depression. J. Clin. Psychiatry 51 (Suppl. B), 18 – 27. Revicki, D.A., Turner, R., Brown, R., Martindale, J.J., 1992. Reliability and validity of a health-related quality of life battery for evaluating outpatient antidepressant treatment. Qual. Life Res. 1, 257 – 266. Rush, A.J., Trivedi, M.H., Fava, M., Nierenberg, A., Rosenbaum, J.F., Thase, M., Howland, R., Kupfer, D.J., Wisniewski, S., Quitkin, F., Sackheim, H., McGrath, P., 2002. Sequenced treatment alternatives to

relieve depression (Star * D). Web site accessed on November 5, 2002: http://www.edc.gsph.pitt.edu/stard/. Sartorius, N., Ustun, T.B., Lecrubier, Y., Wittchen, H.U., 1996. Depression comorbid with anxiety: results from the WHO study on psychological disorders in primary health care. Br. J. Psychiatry Suppl. 30, 38 – 43. Temple, R., Ellenberg, S.S., 2000. Placebo-controlled trials and active-control trials in the evaluation of new treatments: part 1. Ethical and scientific issues. Ann. Intern. Med. 133, 455 – 463. Thase, M.E., Entsuah, A.R., Rudolph, R.L., 2001. Remission rates during treatment with venlafaxine or selective serotonin reuptake inhibitors. Br. J. Psychiatry 178, 234 – 241. Van Ameringen, M.A., Lane, R.M., Walker, J.R., Bowen, R.C., Chokka, P.R., Goldner, E.M., Johnston, D.G., Lavallee, Y.J., Nandy, S., Pecknold, J.C., Hadrava, V., Swinson, R.P., 2001. Sertraline treatment of generalized social phobia: a 20-week, double-blind, placebo-controlled study. Am. J. Psychiatry 158, 275 – 281. World Health Organization, 1992. The ICD-10 Classification of Mental and Behavioural Disorders. Diagnostic Criteria for Research. World Health Organization, Geneva.