STAR^*D Findings May Guide Treatment Recommendations

One-third of patients with major depression who take an adequate dose of an SSRI antidepressant for a sufficient period can expect to achieve full remission of their depression, a new government report says. An additional 10 percent to 15 percent of patients can expect significant improvement, short of remission.

The findings are from the first phase of data from the largest medication treatment trial for major depression undertaken to date. Known as STAR^*D (Sequenced Treatment Alternatives to Relieve Depression), the study included 2,876 patients with major depressive disorder. The $35 million, six-year protocol involving researchers and clinicians at 23 psychiatric and 18 primary care treatment sites across the country is being funded by the National Institute of Mental Health. The results of the first phase of the four-phase trial were published in the January American Journal of Psychiatry.

STAR^*D was designed to help determine the most effective treatment strategies for major depressive disorder by focusing on“ real-world” patients with depression. The four phases of the trial are geared toward shedding light on how best to treat individuals who do not respond to the first antidepressant tried.

“STAR^*D is important because it represents the largest group of moderately to severely depressed patients studied, most of whom had either chronic or recurrent major depression, and the majority of whom also had concurrent other general medical or psychiatric disorders,” said A. John Rush, M.D., vice chair of research in psychiatry at the University of Texas Southwestern Medical Center and STAR^*D's principal investigator, in a press release. “What is also important is that nearly half of these patients significantly benefited from and showed a response to the medications, with 1 in 3 achieving a symptom-free state after the first round of medication.”

“In the field of depression treatment, the burning question is, `What are the best approaches for treatment of depression, and, more importantly, what are the next subsequent steps to take if the first step doesn't lead to full remission?” added Madhukar Trivedi, M.D., in a press release. Trivedi is a professor of psychiatry also at the University of Texas Southwestern Medical Center and the lead author of last month's report.

What sets STAR^*D apart from other studies of depression treatment, Trivedi continued, is that “we entered into the community of practitioners who were not doing research studies themselves, but were routinely treating patients with depression—both in primary care and specialty care.”

Rather than evaluating the effects of standardized treatment protocols on individuals who volunteer for clinical studies—such as those conducted for Food and Drug Administration approvals of medications—the patients in STAR^*D “were already being treated for depression,” and many had multiple other medical illnesses, as well as psychiatric comorbidities. “The results from this study can thus easily be transferred to routine clinical practice in both primary and specialty care.”

Citalopram Used in Round One

During phase one of the study, all of the nearly 3,000 patients enrolled were given the highly selective serotonin reuptake inhibitor citalopram (Celexa) for up to 14 weeks.

All patients were screened at entry into the study with the Psychiatric Diagnostic Screening Questionnaire, as well as either the 30-item Inventory of Depressive Symptomatology (IDS) or the Hamilton Rating Scale for Depression (HamD) to estimate the severity of depression.

The primary research outcome was measured by HamD score through the 14 weeks of the study. The secondary outcomes for phase one were based on scores from the Quick Inventory of Depressive Symptoms–Clinician (QIDS-C) or– Self-Report (QIDS-SR), collected at baseline and then at each treatment visit. The study protocol recommended visits at weeks 2, 3, 6, 9, and 12, with an optional week 14 visit if needed.

The study used a “measurement-based care” treatment manual to guide clinical decision making throughout the study. Citalopram was started at 20 mg a day and increased to 40 mg a day by week 4 and then to 60 mg a day by week 6. Dosing adjustments were based on length of time the patient had been on a dose, symptom response, and side effects. Clinicians had flexibility to initiate lower doses or titrate doses upward more slowly when clinically appropriate.

Citalopram was selected “as a representative SSRI given the absence of discontinuation symptoms, demonstrated safety in elderly and medically fragile patients, once-a-day dosing, few dose-adjustment steps, and favorable drug-drug interaction profile,” the researchers noted in their report.

The stated goal of citalopram treatment was to achieve symptom remission (defined as a HamD score less than or equal to 7 or a QIDS-C or QIDS-SR score less than or equal to 5). Patients were classified as a responder if they achieved a reduction of at least 50 percent of their baseline QIDS-SR score.

One-Third Reach Remission

About 62 percent of patients enrolled in the study came from a psychiatric setting. Subjects who were members of minority groups made up 24 percent of the total patient population. Depressive symptoms were on average moderate to severe (HamD scores greater than 21), and more than 75 percent of the study population met DSM-IV criteria for recurrent or chronic depression. The average length of illness was 15.5 years, and the average patient had 3.3 comorbid medical conditions.

The average dose of citalopram was 41.8 mg a day and did not differ between those who achieved remission and those who did not. Doses in primary care settings (40.6 mg/day) were not significantly different from doses in psychiatric care settings (42.5 mg/day). Patients took citalopram for 10 weeks on average; however, those who achieved remission tended to take the medication longer (12.0 weeks) than those who did not (9.3 weeks), although the difference was not statistically significant.

Overall, 27.5 percent of patients met HamD criteria for remission (score of less than 7) compared with 32.9 percent who met QIDS-SR criteria for remission (score of less than 5). Remission rates did not significantly differ between primary care settings (26.6 percent by HamD; 32.5 percent by QIDS-SR) and psychiatric care settings (28 percent by HamD; 33.1 percent by QIDS-SR).

The a priori defined response rate overall was 47 percent, defined as a 50 percent reduction in symptoms, and again it did not significantly differ in primary care (46 percent) and psychiatric care settings (48 percent). The mean time to response was 5.7 weeks (5.7 primary care; 5.6 psychiatric care), and the mean time to remission was 6.7 weeks (six weeks primary care; seven weeks psychiatric care).

“There does seem to be a group of social, demographic, and clinical factors that may predict a better outcome when treating depression,” Trivedi noted. Higher remission rates were found in Caucasians, women, better-educated and higher-income patients, individuals with private insurance, those with fewer medical and psychiatric comorbidities, better pretreatment physical and mental function, and greater satisfaction with life. No differences were found, however, in family history of depressive disorder, history of attempted suicide, current age, age at onset of illness, or duration of current episode.

To Be Continued...

Patients who did not achieve remission by the end of phase one of STAR^*D were encouraged to enter the next phase of the study. In phase two, patients were offered seven treatment alternatives, including switching to different medications or trying psychotherapy, or adding another medication or psychotherapy to existing citalopram therapy. Additional treatments were offered to patients who were still not symptom free in phases three and four of the study. No placebos were used in any phase of STAR^*D.

“One of the reasons we likely saw such positive results in this trial is because of the diligent management of patients,” said Rush.“ The approach, called measurement-based care, utilized routine measurement of symptoms and side effects to guide treatment and dosing and can be easily implemented in busy primary care and psychiatric practices.”

“This study raises the bar for studies of real-life outcomes of treatment for depression by focusing on recovery rather than mere improvement of symptoms,” commented Robert Freedman, M.D., editor in chief of the American Journal of Psychiatry. “While we are pleased about the recovery of so many people,” Freedman said in a journal press release,“ the question of how the majority who do not achieve recovery should be treated is of obvious concern. For now, the data offer new treatment guidelines for clinicians and realistic estimates of the likelihood of successful outcomes for their patients.”

Darrel Regier, M.D., M.P.H., executive director of the American Psychiatric Institute for Education and Research and director of APA's Division of Research, added, “This study should prompt both primary care physicians and psychiatrists to monitor their patients' response to treatment with brief assessment instruments. Currently available instruments such as the QIDS [used in STAR^*D] or the PHQ-9 [nine-question Patient Health Questionnaire] may be used to correctly identify and modify treatment for the two-thirds of patients who don't achieve remission with initial treatment.”

NIMH Director Thomas Insel, M.D., said in a press release, “The real goal of STAR^*D is how best to help the 70 percent of patients for whom treatment with a representative SSRI is not enough for remission. As the results of subsequent levels of the trial are revealed in the coming months, we will learn the effectiveness of other treatment options.”

“Evaluation of Outcomes With Citalopram for Depression Using Measurement-Based Care in STAR^*D: Implications for Clinical Practice” is posted at<http://ajp.psychiatryonline.org/cgi/content/full/163/1/28>. More information on STAR^*D methods and protocols is posted at<www.clinicaltrials.gov/ct/show/NCT00021528?order=1> and<www.edc.gsph.pitt.edu/stard/>▪

Am J Psychiatry 2005 163 28