Funded by the National Institute of Mental Health, the STAR*D project is one of the largest depression treatment studies ever conducted, with more than 4,000 participants. Results from the second phase of the study will be published over the next year. In this issue PT readers will find a preliminary review of data drawn from the first 1,500 enrollees.
One in five outpatients seeking care for their major depression experiences episodes lasting two years or more, and chronic depressive episodes are associated with socioeconomic disadvantage and racial/ethnic minority status (Gilmer et al., 2005). These and other findings are emerging from recent reports on the first 1,500 enrollees in one of the largest depression treatment studies ever conducted, the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) project.
A seven-year, $34.8 million project, STAR*D is funded by the National Institute of Mental Health under a contract with the University of Texas (UT) Southwestern Medical Center at Dallas (the national coordinating center for the project). The study involves 4,041 patients (18 to 75 years of age) diagnosed with nonpsychotic major depression as determined by having a score of greater than or equal to 14 (moderate severity) on the 17-item Hamilton Depression Rating Scale (HAM-D-17).
At the American Psychiatric Association's annual meeting last year, project director, A. John Rush Jr., MD, professor and vice chairman for research in psychiatry at UT Southwestern, said the project's objective is to identify "the preferred treatments in treatment-resistant depression."
"We are looking for the next best step treatments for depressions not responding satisfactorily," he explained, adding that remission rather than response is the goal (Rush, 2005).
Now nearing completion, the study involved four possible treatment levels lasting up to 12 weeks, with an option to go to 14 weeks if symptom ratings were close to remission, Rush said. Level 1 entailed an adequate trial (in terms of dose and duration) with citalopram (Celexa), a selective serotonin reuptake inhibitor. Patients who did not achieve remission or could not tolerate citalopram were eligible to participate in a series of randomizations with options.
In the Equipoise Stratified Randomized Design (ESRD), "We allowed patients to choose whether they would like to receive augmentation or switch [treatments]," Rush said.
Options at Level 2 were bupropion SR (Wellbutrin SR), sertraline (Zoloft), venlafaxine XR (Effexor XR) or cognitive therapy (CT) as monotherapies, or CT, bupropion SR or buspirone (BuSpar) as augmenting agents.
Those patients without remission after two adequate trials with antidepressants could proceed to Level 3. Options at Level 3 were mirtazapine (Remeron) or nortriptyline (Aventyl, Pamelor) as monotherapies, or augmentation with lithium (Eskalith, Lithobid) or a thyroid hormone (T3).
If no remission was achieved after Levels 1, 2 and 3, patients could participate in Level 4, which meant switching to tranylcypromine (Parnate), a non-hydrazine reversible monoamine oxidase inhibitor, or to a combination of venlafaxine XR and mirtazapine. Patients remitting at any level proceeded into a 12-month naturalistic follow-up phase.
As of press time, Rush told Psychiatric Times that enrollment is complete for STAR*D, but there are 10 patients still in follow-up. He expects the findings on the first treatment strategy (Level 1) with citalopram to be published this month in the American Journal of Psychiatry.
Secondary goals of the project as outlined by Rush included identifying predictors and baseline moderators of response or remission; creating an infrastructure for ancillary studies, such as defining the effect of alcohol and drug abuse on treatment outcomes for major depressive disorder (MDD); and simplification of outcome measures for use in subsequent trials.
The STAR*D is remarkable for its scope, coordination and "real world" approach, said William S. Gilmer, MD, the first author of a preliminary STAR*D report on chronic depression (Gilmer et al., 2005).
The STAR*D infrastructure included the national coordinating center in Dallas, a data coordinating center in Pittsburgh and 14 regional centers across the United States, each of which supervised two to four clinical sites providing primary or specialty care. Nearly half of the clinical sites (18 of 41) were primary care settings. Clinicians at all the sites used identical assessment and therapeutic procedures as defined by the STAR*D protocol. In addition to the HAM-D-17, some other baseline assessments were the 16-item Quick Inventory of Depressive Symptomatology-Clinician Rating (QIDS-C16), 30-item Inventory of Depressive Symptomatology-Clinician Rating (IDS-C-30), 14-item Cumulative Illness Rating Scale (CIRS), the Work and Social Adjustment Scale (WSAS) and the Quality of Life Enjoyment and Satisfaction Questionnaire (QLESQ).
Of the 4,041 enrollees in the study, two-thirds were female, Rush said, and there was a good representation of racial and ethnic minorities. Among the participants, 17% were African-American; 2%, Asian; 73%, Caucasian; and 8%, other. Twelve percent of the enrollees classified themselves as Hispanic (University of Texas Southwestern Medical Center at Dallas, 2004).
"One of the very important things about this study and what makes it so unique is that these are real patients being given care in the community. They are not receiving treatment in tertiary care clinics," said Gilmer, associate professor of psychiatry and behavioral science at Northwestern University's Fineberg School of Medicine.
Gilmer, who is clinical director of the Asher Center for the Study and Treatment of Depressive Disorders, explained that he is involved in several studies evaluating investigational treatments, and "typically those studies developed or supported by the pharmaceutical industry have multiple exclusions. They often exclude anybody who has any comorbid psychiatric illness or who has any significant medical illness."
Exclusion criteria in STAR*D, Gilmer added, was kept to a minimum to ensure that the findings generalized to clinical practice. For example, many of the STAR*D patients have medical illnesses, such as cancer or cardiac disease, or psychiatric comorbidities.
At APA, Rush noted that patients enrolled in STAR*D "turned out to be a much more severe chronic and recurrent sample than we had anticipated." The average HAM-D-17 score was 20, and more than 80% of the participants had chronic or recurrent depression.
Patients also had clear views as to whether they wanted to be switched to a different antidepressant or take an augmenting drug and whether they wanted to receive or exclude CT.
While the total enrollment in STAR*D was more than 4,000, the study investigators decided to issue preliminary reports on the baseline characteristics of the first 1,500 enrolled patients. For instance, they identified characteristics associated with chronicity, recurrence or the presence of various symptom complexes in depression.
"The preliminary analyses were looking for basic associations, with the intent that we would use the analyses from the first 1,500 patients to develop more complex analyses for the remaining 2,541," Gilmer said.
Most of the baseline papers have been published or are in press, according to Gilmer. The second wave of papers on the subsequent 2,541 patients will be published over the next year or year and a half.
When asked about the recently published article that he co-authored on factors associated with chronic depressive episodes, Gilmer responded:
"The most important findings are that chronic depressions (those depressions that last for at least two years) are very common in clinical practice both in primary care and psychiatric settings; that chronic depressions are associated with a number of factors suggesting socioeconomic disadvantage; and that there is a tremendous amount of morbidity associated with chronic depressions, in particular, lower quality of life, lower levels of work and social functioning, greater medical comorbidity and greater disease burden. The challenge is, of course, to determine which comes first--chronic depressions or greater socioeconomic and medical burden. The arrows of influence probably go in both directions."
Another disturbing finding, he said, was "people with chronic depressions are much more likely to have made previous suicide attempts."
One preliminary finding Gilmer described as provocative was the higher rate of chronicity in patients enrolled in primary care settings as compared to those enrolled in psychiatric settings.
"This finding may have significant public health consequences, and needs further investigation," he said.
Speculating on why a higher rate of chronicity is found among patients enrolled in primary care settings, Gilmer suggested that the presence of prominent medical illness may obscure the recognition of depression in primary care settings or lead to a lesser prioritization of depression treatment.
Another possibility, he said, may relate to socioeconomic factors. Patients who are unemployed, who have low incomes and who rely on public insurance may not have access to or be able to afford psychiatric treatment and are consequently seen in primary care clinics.
A third possibility is that patients fear being stigmatized as having a mental illness by seeking care from psychiatrists or other mental health professionals, so they go instead to primary care clinics.
As mentioned earlier, there were associations between minority race and ethnicity and chronicity. In the preliminary study, Gilmer and colleagues reported that blacks were more likely than whites to report chronic depressions (odds ratio=1.585) and Hispanics were more likely than non-Hispanics to be chronically depressed (odds ratio=1.617). The finding of higher chronicity among blacks and Hispanics, Gilmer speculated, may relate to other factors, such as unemployment or lack of insurance.
Looking at psychiatric comorbidities and chronic depression, Gilmer said the research team found that generalized anxiety disorder, as identified by the Psychiatric Diagnostic Screening Questionnaire (PDSQ), correlated with chronicity, but not panic disorder nor alcohol and substance abuse.
A surprising finding for Gilmer and his fellow investigators was that the chronic depressions looked no different from nonchronic depressions in terms of symptom profiles.
"They had the same levels of severity on rating scales, and they appeared to have similar depressive symptoms that made up their clinical pictures," he said.
The key difference was that chronic depressions, as compared to nonchronic depressions, were associated with greater levels of disability and lesser levels of quality of life.
The associations identified in the preliminary sample of 1,500 patients, Gilmer said, will be more fully tested in future analyses of the larger sample.
Researchers also will be evaluating how specific patient characteristics identified in the preliminary analyses may influence treatment response.
"So, for example, does greater medical burden influence whether you get well or not? Does being socioeconomically disadvantaged make you more or less likely to respond to treatment? Are there racial or ethnic differences in response?" Gilmer said.
In his own clinical practice, Gilmer said he looks for factors that might be contributing to patients' chronic depressions.
"When evaluating a patient who has been experiencing chronic depression, I first assess what factors might be contributing to that person having a more chronic course. Do they have other chronic medical conditions that are undertreated or have they simply been inadequately treated for their current depression? It is not unusual to see patients with chronic depression who have been on inadequate doses of an antidepressant for several months or even years," he said.
The findings from the STAR*D analyses will be very important for clinicians, Gilmer said, since they will provide guidance on the types of interventions necessary for patients with chronic depressions, "especially in this age of managed care and limited resources."
"It is important to identify those patients who are at high risk for chronicity and provide them with optimized treatment and, when possible, case management that can address some of the socioeconomic factors and other life adversities that put them at risk," he said.
When asked about evaluating new treatments not yet available at the time of STAR*D's initiation, Gilmer responded:
"We have this research infrastructure in place in real world settings. We hope to be able to continue this type of study in the future, looking at newer treatments as they evolve."
Other Preliminary Findings
In recent months, several preliminary analyses from the STAR*D have been published. Some of the findings are highlighted below.
Gender differences. Women had an earlier onset of their first major depressive episode (mean=24.3 years) than men (mean of 26.5 years), and they commonly reported concurrent symptoms consistent with anxiety disorders, somatoform disorder and bulimia as well as atypical symptoms. Symptoms consistent with both alcohol and drug use/dependence more commonly were found in depressed men than women (Marcus et al., 2005).
Age differences. When compared with younger or middle-age patients, those 51 years of age and older were less likely to report hypersomnia, more likely to endorse mid-nocturnal insomnia, less likely to report irritability, less likely to report weight increase, less likely to report negative views of the self and future and slightly more likely to report reduced sexual interest. When compared to younger patients (18 to 35 years of age), older and midlife (36 to 50 years of age) patients were more likely to report gastrointestinal symptoms (e.g., diarrhea or constipation) and to experience psychomotor slowing (Husain et al., 2005).
Recurrent versus single-episode depression. As compared to patients with single-episode depression, patients with recurrent depression were older, had an earlier age of onset and were more likely to have a positive family history of depression. They also reported higher levels of depressive symptoms, and a greater number and higher total scores on comorbid medical conditions than first-episode patients (Hollon et al., 2005).
Issues of onset. A preliminary analysis looking at factors differentiating early and late onset in major depression found that major depressive disorder that begins before age 18 is a particularly severe and chronic condition with a distinct set of demographic and clinical correlates, including female gender, more episodes, more suicidality, greater symptom severity and atypical symptom features. Additionally, early-onset major depression was associated with significant psychosocial consequences, specifically lower educational attainment and marriage rates (Zisook et al., 2004).
Comorbid conditions. Concurrent Axis I conditions (most often anxiety disorder) are very common with major depressive disorder. Depressed outpatients with more concurrent Axis I conditions had earlier ages at first onset of major depressive disorder, longer histories of MDD, greater depressive symptom severity and more general medical comorbidity, as well as poorer physical and mental function, health perceptions and life satisfaction (Rush et al., 2005).
Gilmer WS, Trivedi MH, Rush AJ et al. (2005), Factors associated with chronic depressive episodes: a preliminary report from the STAR-D project. Acta Psychiatr Scand 112(6):425-433.
Hollon SD, Shelton RC, Wisniewski S et al. (2005), Presenting characteristics of depressed outpatients as a function of recurrence: preliminary findings from the STAR*D clinical trial. J Psychiatr Res Oct. 20 [Epub ahead of print].
Husain MM, Rush AJ, Sackeim HA et al. (2005), Age-related characteristics of depression: a preliminary STAR*D report. Am J Geriatr Psychiatry 13(10):852-860.
Marcus SM, Young EA, Kerber KB et al. (2005), Gender differences in depression: findings from the STAR*D study. J Affect Disord 87(2-3):141-150.
Rush AJ (2005), An update on the STAR*D trial. Symposium 9C. Presented at the 158th annual meeting of the American Psychiatric Association. Atlanta; May 23.
Rush AJ, Zimmerman M, Wisniewski SR et al. (2005), Comorbid psychiatric disorders in depressed outpatients: demographic and clinical features. J Affect Disord 87(1):43-55.
University of Texas Southwestern Medical Center at Dallas (2004), STAR*D Participant News 3(3):1.
Zisook S, Rush AJ, Albala A et al. (2004), Factors that differentiate early vs. later onset of major depression disorder. Psychiatry Res 129(2):127-140.