Randomized, Placebo-Controlled trial of nefazodone maintenance treatment in preventing recurrence in chronic depression
Introduction
Epidemiologic studies suggest that approximately 17% of individuals will have an episode of major depressive disorder (MDD) in their lifetime (Kessler et al 1994). The extensive impairments in occupational and social functioning Hays et al 1995, Wells et al 1992 due to MDD have led to its identification as the most disabling illness in the developed countries by the World Health Organization (Murray and Lopez 1997). Chronic depressions account for much of the disability associated with MDD. Compared with nonchronic depression, chronic depression is associated with even greater impairments in psychosocial and work functioning Hays et al 1995, Miller et al 1998, Wells et al 1992, higher health care utilization (Howland 1993), and more frequent suicide attempts and hospitalization (Klein et al 1998). At any given time, at least 3% of the U.S. population has some form of chronic depression Kessler et al 1994, Weissman et al 1988.
Because of the chronic and recurrent nature of MDD, psychopharmacologic treatment has evolved to include an acute phase that aims to achieve a clinically significant reduction of symptoms, a continuation phase (4–9 months) to prevent a relapse of the index episode, and, in some cases, a maintenance phase (1 year or more) designed to prevent new episodes (Kupfer 1991). Continuation treatment is recommended for all cases of MDD (Depression Guideline Panel 1993), based on data showing a 25% relapse rate within 2 months of discontinuation of acute antidepressant treatment (Maj et al 1992) and numerous controlled continuation treatment trials demonstrating substantially lower relapse rates for active drug compared with placebo Coppen et al 1978, Doogan and Caillard 1992, Feiger et al 1999, Mindham et al 1973, Montgomery and Dunbar 1993, Montgomery et al 1988, Thase et al 2001.
Maintenance treatment has been recommended for individuals with a high likelihood of recurrence of MDD, including those with chronic forms of MDD (Depression Guideline Panel 1993). Several studies have demonstrated prophylactic efficacy of maintenance antidepressant therapy for patients with a history of previous episodes of MDD Frank et al 1990, Hochstrasser et al 2001, Montgomery and Dunbar 1993, Montgomery et al 1988, Prien et al 1984, Versiani et al 1999 and with dysthymia (Kocsis et al 1996), but only one study (Keller et al 1998) has demonstrated maintenance phase prophylactic efficacy of an antidepressant for patients with chronic forms of MDD.
Psychotherapy has also been examined for the prevention of relapse and recurrence of MDD. Short-term (acute phase only) cognitive therapy (Beck et al 1979) has reduced relapse relative to discontinued medication Blackburn et al 1986, Evans et al 1992, Kovacs et al 1981, and continuation cognitive therapy has reduced relapse relative to short-term cognitive therapy without continuation treatment. For patients with residual depressive symptoms after pharmacotherapy, cognitive therapy has been found to reduce recurrence rates relative to maintenance medication only (Paykel et al 1999) and treatment-as-usual (Teasdale et al 2000). The role of acute or continuation psychotherapy (with or without medication) in preventing recurrences during maintenance treatment has not been examined for patients with chronic forms of MDD.
We recently reported (Keller et al 2000) the acute-phase, 12-week outcomes of a comparison of nefazodone, cognitive behavioral analysis system of psychotherapy (CBASP; McCullough 2000), and their combination in the treatment of chronic forms of MDD. Within the intent-to-treat sample, combined nefazodone and psychotherapy resulted in significantly greater acute-phase treatment response rates (73%) when compared with nefazodone alone (48%) and psychotherapy alone (48%). The gains made during the acute phase were generally maintained and consolidated during the continuation phase (Kocsis et al, unpublished data).
This article reports on results of the maintenance phase of this study, specifically a placebo-controlled examination of nefazodone as a maintenance treatment for patients with chronic forms of MDD who had achieved a clinical response over the course of acute and continuation treatment with either nefazodone or the combination of nefazodone and psychotherapy. Our primary hypothesis was that nefazodone would be superior to placebo in preventing recurrences during maintenance treatment. Secondarily, we hypothesized that prior combined treatment (during acute and continuation phases) would be associated with a lower recurrence rate during maintenance treatment, compared with prior treatment with nefazodone alone.
In conducting this trial, we encountered a methodological problem that may increasingly be an issue in long-term treatment studies for MDD. The growing awareness of the general public of treatment options for MDD, and the success of these treatments, may make it increasingly difficult to retain patients assigned to placebo in a prevention study. Particularly for patients with chronic or recurrent depression, sustained improvement in depressive symptoms provides a strong motivation for preferring to remain on active treatment. Patients randomly assigned to placebo may be inclined to discontinue treatment if there is any evidence (e.g., lack of medication side effects or re-emergence of one or two salient depressive symptoms) that they are no longer on active, previously successful treatment.
To address this issue, we applied a “competing risk” analytic strategy David and Moeschberger 1978, Gooley et al 1999, Prentice et al 1978 that has been applied to the study of other medical disorders in which, for example, risk for other causes of death “competes” with risk for death from the target illness under study. Recent examples employing such competing risk analysis include studies of the following: durability of aortic pericardial bioprosthesis (Banbury et al 2001); risk factors for sudden infant death syndrome (Pollack and Frohna 2001); incidence rates of opportunistic infections in acquired immune deficiency syndrome patients (Yan et al 2000); the probability of dying from prostate cancer given tumor histology and age of diagnosis (Albertsen et al 1998); and incidence of Alzheimer’s disease in the Framingham study (Beiser et al 2000). In the current context, the competing risk approach allows the risk for recurrence to be examined adjusting for the “competing risk” of study discontinuation for reasons other than recurrence.
Section snippets
Methods and materials
The study rationale and methods, detailed elsewhere (Keller et al 2000), are briefly summarized here.
Patient characteristics
Patient disposition leading up to and including the maintenance phase is given in Figure 1. Two hundred sixty-nine patients who responded to either nefazodone alone or combined treatment during the acute phase were subsequently enrolled in the continuation phase. Of these, 165 maintained their response and elected to participate in the maintenance phase. Safety analyses were conducted on these 165 patients. The sample evaluated for efficacy was 160 patients (n = 76 in nefazodone; n = 84 in
Discussion
The results of this multicenter, placebo-controlled maintenance trial suggest that nefazodone may protect against the recurrence of MDD in patients with chronic depression. At the end of the 1-year maintenance phase, the conditional probability of recurrence was 30.3% for nefazodone-treated patients compared with 47.5% for placebo-treated patients. This finding, however, was achieved in the context of a competing risks data analytic approach that focused on recurrences of new episodes of MDD by
Acknowledgements
This work was supported by grants from Bristol-Myers Squibb, Inc.
References (56)
- et al.
Age and value size effect on the long-term durability of the Carpentier-Edwards aortic pericardial bioprosthesis
Ann Thorac Surg
(2001) - et al.
A two-year naturalistic follow-up of depressed patients treated with cognitive therapy, pharmacotherapy and a combination of both
J Affect Disord
(1986) - et al.
Global mortality, disability, and the contribution of risk factorsGlobal Burden of Disease Study
Lancet
(1997) - et al.
A competing risk model of sudden infant death syndrome incidence in two US birth cohorts
J Pediatrics
(2001) - et al.
Selective serotonin reuptake inhibitor discontinuation syndromeA randomized clinical trial
Biol Psychiatry
(1998) - et al.
Competing risk adjustment reduces overestimation of opportunistic infection rates in AIDS
J Clin Epidemiol
(2000) - et al.
Competing risk analysis of men aged 55 to 74 years at diagnosis managed conservatively for clinically localized prostate cancer
JAMA
(1998) - et al.
Computing estimates of incidence, including lifetime riskAlzheimer’s disease in the Framingham Study. The Practical Incidence Estimators (PIE) macro
Stat Med
(2000) - et al.
Controlled acute and follow-up trial of cognitive therapy and pharmacotherapy in out-patients with recurrent depression
Br J Psychiatry
(1997) - et al.
Continuation therapy with amitriptyline in depression
Br J Psychiatry
(1978)