1. Discuss the possible reasons for failure to achieve an acceptable response in the majority of patients treated with antidepressants.

2. Describe the algorithms and results at the 4 treatment levels of the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study.

3. Recognize those patients in whom remission of depression is less likely.

4. Prescribe the adequate duration of acute treatment necessary to achieve remission.

5. Discuss the overall conclusions of the STAR*D study.

Sidney Zisook, MD
Professor of Psychiatry, University of California, San Diego, School of Medicine, Director, Outpatient Psychiatric Services, and Director, Department of Psychiatry Residency Training Program, San Diego Veterans Affairs Medical Center, La Jolla, CA

Introduction: majority of patients treated with antidepressants do not have acceptable response; reasons— nonadherence; inability to tolerate side effects of medication enough to achieve therapeutic dose; inability to achieve and maintain remission; multiple strategies used in clinical practice; few Food and Drug Administration (FDA)–approved treatments for patients who fail first antidepressant medication

Systematic approach to treatment-resistant depression (7 Ds): definition—“good enough” response (patients should have no symptoms, good function, and feel good about themselves); goals of treatment complete remission, recovery from episode, and protection against further episodes; diagnostic issues—consider other psychiatric conditions possibly missed (eg, comorbid anxiety disorder), atypical features, psychotic features, general medical condition that interferes with patient’s response (eg, thyroid disease), and medications (eg, polypharmacy, benzodiazepines); dose—whether adequate; tendency to underdose; duration—adequate trial should be ≥8 wk, with 3 to 4 of those weeks on highest dose tolerated, until robust response achieved; different approaches—maximizing other strategies that may help (eg, exercise, nutrition, light, psychotherapy); one study suggested that patients with chronic severe depression who have history of childhood sexual abuse or significant trauma do not respond as well to medications alone, compared to similar patients without this history; drugs; determination—possibly most important

Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study: goal—empirically define preferred treatments for treatment-resistant depression; National Institute of Mental Health (NIMH)–funded study; had 14 regional centers nationwide; each center had 2 to 3 sites; inclusion criteria—diagnosis of major depression; treatment-seeking patients; age range 18 to 75 yr; Hamilton Rating Scale for Depression (HAM-D) score ≥14; antidepressant medications appropriate; essentially no exclusions (ie, as long as patient did not require in- house detoxification, included in study); baseline clinical features—75% of patients had recurrent depression; 15% had ≥10 episodes of major depression in lifetime; >50% had positive family history of depression; 18% had attempted suicide; ≈25% had chronic depression at time of entrance to study; 53% had significant anxious features; distribution of age at onset—mean age for first episode of depression ≈25 yr, but more than one-third had first episode in childhood or adolescence (≈13% had first episode at >45 yr of age); first episode can start at childhood or adolescence; those with earlier-onset depression have more chronic, more severe, more recurrent, and more comorbid course; also tend to be more suicidal and more challenging to treat; design of study—after consent obtained, all participants placed on selective serotonin reuptake inhibitor (SSRI), citalopram (Celexa); patient started on 20 mg and by 4 wk, if no robust response seen, dose increased to 40 mg; if still no good response after 4 wk, increased to 60 mg; treated for ≥14 wk; goal remission; if patient in remission after treatment with citalopram, followed for 1 yr; if response inadequate or remission not achieved by end of 14 wk, patient taken to level 2, then level 3, and finally level 4; 4 levels with 14 wk of treatment and different algorithm for each level

Level 1: treatment with citalopram for 14 wk; 30% to 45% of patients had remission (same rates seen in efficacy studies in which participants had nonchronic depression, few comorbidities, and no suicidality); ≈66% of participants either dropped out, had too many side effects, or did not remit; of those who remitted, 53% remitted within first 6 wk (23% remitted at ≥12 wk); factors not related to remission—age; age at onset; family history of depression; whether seen in primary care vs specialty clinic; number of past episodes; current alcohol or drug use; factors related to remission—participants of white ethnicity and female sex had better results (also working, wealthy, and well educated participants); clinical characteristics that predict remission—less severely depressed; shorter duration of depression; fewer psychiatric and general medical concurrent conditions, particularly anxiety; functioning well at baseline, satisfied with life, and with good quality of life; likelihood of relapse—≈40% relapse within following year; patient who goes into follow-up with full remission significantly less likely to relapse than patient who goes into follow-up with response (but not remission); important to treat to remission (better prognosis); conclusions—more than one-third of patients depressed for >2 yr; over two-thirds have concurrent general medical conditions; approximately one-third remit with SSRI; of those who remit, remission occurs after first 6 wk in almost 50% of patients; studies on remission require study periods >8 wk; of those who remit, ≈40% relapse within next year

Level 2: for participants who did not obtain optimal response, who had response but did not remit, or who had no response; randomization included switching to another treatment or augmenting with another treatment; switching strategies—patients placed on 1) another SSRI (eg, sertraline [Zoloft]), 2) another medication with different mechanism of action (eg, bupropion sustained-release [Wellbutrin SR]), 3) dual reuptake inhibitor (eg, venlafaxine extended-release [Effexor XR]), or 4) cognitive therapy; augmentation strategies—examined adding bupropion (commonly done in clinical practice), or buspirone (BuSpar; anxiolytic), or cognitive therapy; participants had choice of accepting all possibilities; only 1% allowed all 7 possibilities, indicating strong preferences of participants (mainly due to fact that psychotherapy not favored by participants, and that those participants doing well prefer not to give up citalopram but just to add another drug); augmentation not preferred by those not doing well or who had side effects; results of switching strategy—participants had 1 in 4 chance of improvement, no matter which drug switched to; pharmacologic differences between drugs did not matter; augmentation results— no statistically significant difference among choices, but somewhat better with bupropion (higher response rates, greater change in depression scores, lower end point depression score, and better tolerated); ≈40% of patients had remission with augmentation; future study to determine whether more aggressive initial strategies result in more robust remission, faster time to remission, and longer duration of remission; overall remission rate ≈50% after level 2; group who received cognitive therapy responded similarly to groups in augmentation or switching strategy (≈30% remission); conclusions—cognitive therapy not as popular as expected; switching to cognitive therapy about as effective as switching or augmenting medication; after level 2, close to 50% relapsed within following year; those who went into follow-up with full remission did better than those starting follow-up with response but not remission

Level 3: participants who did not achieve remission in level 2 switched to new antidepressant (mirtazapine [Remeron]), or tricyclic antidepressant (nortriptyline), or randomized to augmentation with lithium or thyroid hormone; disappointing results with switching strategies (only 10% achieved remission); remission rate slightly better with augmentation; only 13% had remission with lithium; thyroid hormone better tolerated and remission rate better; issues with lithium included stigma in mind of patients and fear of use on part of physicians (tendency for underdosing); study suggests thyroid hormone underutilized; follow-up showed >60% relapse in ≤1 yr, usually in ≤2 mo, especially if participants go into follow-up without full and robust remission; conclusions—thyroid augmentation better tolerated and dosed than lithium and possibly preferable; switching to nortriptyline or mirtazapine no different and had poor remission rates

Level 4: participants who failed 3 adequately delivered trials; randomized to monoamine oxidase inhibitor (MAOI; tranylcypromine) or combination of venlafaxine and mirtazapine; overall remission rates 13.8% and 15.7%; MAOI also underdosed; follow-up showed relapse rate of 70% within 1 yr (especially without full and robust remission); conclusions—tranylcypromine not aggressively dosed; combination of venlafaxine and mirtazapine had advantages over tranylcypromine (physicians and participants more comfortable with combination, better dosed, and more tolerable [fewer side effects]); overall remission rates low and relapse rates high; however, cumulative remission rate close to 70%

Overall conclusions: at least 10 to 12 wk of acute treatment necessary for remission; failure to achieve remission portends poor prognosis and dramatically increases rate of relapse over following year; when more treatment steps necessary, greater relapse rates expected; treat vigorously and as creatively as possible early on; patients with more severe depressions with more comorbidity less likely to remit and often require more treatment steps (need better strategies than currently available); more innovative treatment and research needed; currently, no algorithm for treatment

Questions and answers: whether antidepressants effective and should be used—antidepressants effective but not for all patients; studies show that, especially in adolescent depression, treatment with antidepressants may increase suicidal ideation (usually in first 2 wk and usually mild); black box warning, especially for adolescents and young adults ≥25 yr of age; however, no actual suicides seen in studies; since introduction of black box warnings, diagnosis of depression made less frequently, treatment with antidepressants less frequent, and suicide rates climbing; with black box warning, recommended that patients be seen more frequently (weekly); present use of electroconvulsive therapy (ECT)—effective mode of treatment; not used as much in California; used in patients with severe depression, often with suicidality and in depression with multiple failed vigorous attempts at treatment with antidepressants and psychotherapeutic interventions; efficacy similar to that of medications; use in long-term maintenance and episode prevention less well documented and more problematic; atypical antipsychotics in patients with no response to usual drugs—atypical antipsychotics increasingly popular; successful and failed trials with atypical antipsychotic augmentation; aripiprazole FDA-approved as augmentation agent for failed trial; can be effective; due to problems (eg, weight gain, metabolic problems, akathisia, tardive dyskinesia), not thought of by experts as first-line treatment for refractory depression; quetiapine not only possible agent in augmentation strategy, but also approval by FDA as monotherapy expected; role of social factors—genetic factors alone do not cause depression; presence of genetic polymorphism and ≥4 significant psychosocial stressors in previous year likely leads to depression; also, presence of genetic polymorphism and physical and sexual trauma in childhood extremely important in diathesis and vulnerability to depression; ongoing stress must be addressed; whether results of psychotherapy in STAR*D study related to selection bias—selection bias present because symptomatic participants entered study requesting treatment for depression and all willing to take antidepressant as first trial for 12 to 14 wk; probably selected out individuals willing to receive psychotherapy from beginning; fatigue in depression and role of methylphenidate (Ritalin)—fatigue one of common symptoms of depression; one study showed that depressed patients who have fatigue as prominent symptom possibly do well with medication that targets catecholamines (promising results with bupropion); if fatigue remains residual symptom, treatment or augmentation with bupropion or methylphenidate possibly effective; methylphenidate also used when attention deficit disorder one of comorbid conditions; evaluation of thyroid—lithium or thyroid hormone not targeted for particular subtypes in STAR*D study; thyroid hormone levels not routinely checked in study; data flimsy that euthyroid and hypothyroid individuals benefit from thyroid augmentation; thyroid augmentation possibly effective in euthyroid patients; whether history of attention-deficit/hyperactivity disorder (ADHD) factor in adult depression and stable dose of antidepressant medication—physicians most often do not obtain adequate histories of childhood ADHD in adult patients; speaker believes history of ADHD probably related to risk for adult ADHD, depression, and bipolarity; hope is that treatment in childhood prevents some long-term sequelae; adult ADHD not uncommon; medication that targets catecholamines (eg, norepinephrine, dopamine) rather than just serotonin, possibly most appropriate for adults with ADHD, especially comorbid with depression; in ≈70% of cases, dose of antidepressants that results in good response will maintain response, and therefore, not necessary to adjust dose; in minority of cases, depression recurs despite maintenance on same dose (reason unknown); one theory that long-term use of medication that boosts serotonin also has effect of decreasing dopamine (relative dopamine deficiency); strategy to increase dose or add bupropion or methylphenidate, targeting dopamine; speaker believes not due to tolerance of medication