---

Reports emerging from the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study^1-6 promise an enormous advance in the treatment of depression both in primary care and mental health settings.⁷ The results would seem to apply, with minor reservations, to older adults with depressive disorders. A brief review of the study subjects, measures, methods, and outcomes follows to describe the strengths and limitations.

Study Subjects

At the outset, the STAR*D investigators sought to provide a more generalizable sample of real-world patients seeking care and making choices at clinical centers rather than recruited through advertisement or public service announcements. The lead investigators assembled 14 regional centers representing 18 primary care and 23 mental healthcare sites, including both public and private outpatient venues. As a result, 37.9% of the sample of 2,876 people were primary care patients cared for by primary care physicians (PCPs) who would treat their depression. Participating patients ranged from 18–75 years of age, with 741 (28%) ≥51 years of age. Only 5.6% were retired. Most had comorbid general medical conditions. The most common comorbid psychiatric conditions were anxiety disorders. More than 75% met criteria for recurrent depression and approximately 18% reported a history of at least one suicide attempt. Patients with depression complicated by psychosis, dementia, or bipolar disorder were not included. Seventy-five percent of the sample were Caucasians, 17.6% were African Americans, and 13% were Hispanic. Twenty-five percent was college educated, but >33% had no health insurance.¹ In summary, the racial and ethnic character of the sample was similar to that seen among older Americans and a substantial segment of the patients received depression care from PCPs. However, patients >75 years of age were not included, and most patients were not experiencing their first episode of depression.

Measures

Depression severity measures included the 17-item version of the Hamilton Rating Scale for Depression (HAM-D₁₇) and the 16-item self-report version of Quick Inventory of Depressive Symptomatology Self Report (QIDS-SR). The HAM-D is the most widely used measure of symptom severity in antidepressant trials. The QIDS-SR was developed for ease of patient self-report with responses based more on symptom severity than frequency of occurrence. The primary outcome of interest was remission defined as a score of ≤7 on the HAM-D. The secondary outcome of interest was remission defined by a score of ≤5 on the QIDS-SR.⁸ Both outcomes represent a virtual absence of depressive symptoms and thereby minimal risk of relapse. Both measures were administered over the telephone. However, the QIDS-SR was administered every other week, with results included in a Web-based treatment monitoring system that allowed clinical research coordinators to help guide physicians in antidepressant dosing when symptoms remained elevated and side effects were negligible.⁹

Methods

STAR*D used an innovative process of allocating patients who did not experience remission with the first antidepressant trial to subsequent trials. Because each alternative choice within a new trial was considered equally desirable, subjects were sorted through a process of equipoise stratified randomization. In practice, this meant that all subjects who did not achieve a QIDS-SR–defined remission with citalopram were encouraged to accept any one of seven potential second-trial options including switching to another single drug (monotherapy), drug augmentation, and/or cognitive behavioral psychotherapy. However, as in actual practice settings, patients could opt not to accept one or more options, in which case they were randomized to the remaining “acceptability strata.” Table 1^1-6 displays the various alternatives within each of four levels of trials. For all participants, citalopram was started at 20 mg/day, raised to 40 mg/day by week 4, and to a maximum 60 mg/day by the end of week 6. The protocol recommended treatment visits at 2, 4, 6, 9, and 12 weeks. After an optimal trial based on guideline-defined dose and duration, all who did not achieve remission were encouraged to enter the subsequent randomized trial. Patients could discontinue citalopram and move to the next trial before 12 weeks if intolerable side effects required a medication change, an optimal dose increase was not possible because of side effects or participant choice, or significant symptoms (QIDS score ≥9) were present after 9 weeks or 12 weeks (QIDS >5) despite maximally tolerated doses.



Table 2 displays the pharmacologic profile of the medications employed in the trials. Regarding ease of administration, bupropion and buspirone were taken twice daily, while all others were taken once daily. Dietary precautions were necessitated for tranylcypromine. Medications chosen for first- or second-trial monotherapy included citalopram, sertraline, venlafaxine, and bupropion. This group is characterized by relatively low risk of dangerous side effects and drug interactions, and greater familiarity among PCPs. Citalopram and sertraline exhibit primarily serotonin reuptake inhibition, whereas venlafaxine shows serotonergic and noradrenergic reuptake inhibition. Bupropion is both noradrenergic and serotonergic but is neither a selective serotonin reuptake inhibitor (SSRI) or a serotonin norepinephrine reuptake inhibitor (SNRI). As a result, the choices for the first and second monotherapy trials offered different mechanisms of action with little difference in safety or ease of administration. Medications chosen for third- or fourth-trial monotherapy included nortriptyline, mirtazapine, and tranylcypromine. Nortriptyline inhibits the reuptake of norepinephrine and is an effective serotonergic and g-aminobutyric acid agonist. However, it also exhibits anticholinergic activity and is considered higher risk due to cardiovascular effects. As a monoamine oxidase inhibitor, tranylcypromine is associated with life-threatening diet and drug interactions. Mirtazapine has the disadvantage of associated weight but the advantage of sedation. 



Medications chosen for augmentation (bupropion, buspirone, mirtazapine) possessed mechanisms of action compatible with but not duplicating SSRI or SNRI properties. Similarly, the use of lithium and triiodothyronine (T₃) for augmentation reflected choices based on different theoretical mechanisms not incompatible with any monotherapy or augmentation combinations from the first or second trials. Moreover, lithium and T₃ are the most widely studied medications used to augment the effects of antidepressants.⁶ The dosing and safety monitoring for T₃ is familiar to PCPs, whereas lithium is more familiar to psychiatrists. Cognitive-behavioral therapy was an alternative in both the switch and augmentation options in the second trial, but the data have yet to be published.

Results

Table 1 provides mean dose of medication at the end of each trial as well as percentages of marked-to-moderately burdensome side effects. Also included is the percentage of patients who were medication intolerant due to either side effects or lack of benefit and discontinued the trial prior to an adequate duration. Because the studies each report exhaustive detail on side-effect frequency, intensity, and burden, intolerance and moderate-to-marked burden were chosen as representing the most informative concise measures for the reader. The remission rates in the Table are based on the QIDS-SR rather than HAM-D, in as much as the QIDS-SR is more appropriate for primary care settings.

Within the initial citalopram trial, participants who were Caucasian, female, employed, or had higher levels of education or income had higher remission rates. Participants with longer index episodes of depression, more concurrent psychiatric disorders, more general medical disorders, and lower baseline function and quality of life experienced lower remission rates. Age was not a significant predictor of remission. Of those who achieved remission, 40.3% did so only at or after 8 weeks of citalopram. Of those who were not well but could tolerate nearly 60 mg of citalopram, an additional 33% experienced remission when bupropion sustained release (SR) or buspirone were added. Bupropion SR appeared to be the better of the two choices due to tolerability. Of patients not experiencing a remission with citalopram, 25% were made well simply by switching to sertraline, bupropion SR, or venlafaxine extended release (XR). However, 56% of this group were intolerant of the side effects of citalopram. As a result, the choice of the seemingly more desirable option of augmentation was precluded. Nonetheless, intolerance of the first SSRI did not automatically mean intolerance of the second.

For patients not well after two antidepressant trials of either monotherapy or augmentation, switching to nortriptyline, mirtazapine, or tranylcypromine, or adding venlafaxine to mirtazapine made approximately one in eight well. Without placebo controls, one cannot deny the possibility that one in eight was better than the spontaneous remission rate associated with the natural history of recurrent depression. In contrast, of patients who remained symptomatic but tolerated citalopram plus bupropion SR or venlafaxine XR; or monotherapy with sertraline, bupropion SR, or venlafaxine XR; nearly 25% were made well by the addition of T₃. T₃ was significantly better tolerated than lithium.

Conclusion

The STAR*D data provide substantial guidance for the sequenced treatment of depression among older adults despite the modest number of seniors in the sample. When the first SSRI does not achieve remission, augmentation with a non-SSRI will reliably achieve remission in 33% of patients. Stopping the first medication and switching to another agent, including an SSRI, will achieve remission in 25%. Of those not well following the first and second trial of monotherapy or augmentation, subsequent augmentation with T₃ is preferable to monotherapy with nortriptyline, mirtazapine, or tranylcypromine, or augmentation with lithium or the combination of venlafaxine plus mirtazapine. Thus, drugs which heighten the risk (more serious for seniors) of cardiovascular side effects, drug interactions, and falls due to sedation do not seem worth the risk given the scant benefits. Because many in the STAR*D sample exhibited recurrent depression, the remission rates for seniors experiencing their first episode may be better than that reported in the initial citalopram trial. More rapid-dose escalation coupled with change in medication for patients experiencing little benefit by 6 weeks also might have reduced the delay in remission and the number of patients dropping out for lack of benefit. The authors of the initial citalopram study highlight the need for longer treatment duration, more vigorous medication dosing, and earlier assessment of response. Their recommendations are in concert with others who have argued that the treatment of depression in late life needs to be far more aggressive than current practice reflects.¹⁰ Missing from STAR*D are data on the choice of alternatives for patients with dementia complicated by depression not remitting with the initial SSRI trial. Nonetheless, the preferences outlined above would be equally applicable to depression in dementia based on the safety profile alone. In short, the data from STAR*D represent a major advance in the treatment of depression among adults of all ages. PP
 

References

1. Trivedi M, Rush AJ, Wisniewski SR, et al. Evaluation of outcomes with citalopram for depression using measurement-based care in STAR*D: implications for clinical practice. Am J Psychiatry. 2006;163(1):28-40.

2. Rush AJ, Trivedi MH, Wisniewski SR, et al. Bupropion-SR, sertraline, or venlafaxine-XR after failure of SSRIs for depression. N Engl J Med. 2006;354(12):1231-1242.

3. Trivedi MH, Fava M, Wisniewski SR, et al. Medication augmentation after the failure of SSRIs for depression. N Engl J Med. 2006;354(12):1243-1252.

4. Fava M, Rush AJ, Wisniewski SR, et al.  A comparison of mirtazapine and nortriptyline following two consecutive failed medication treatments for depressed outpatients: a STAR*D report. Am J Psychiatry. 2006;163(7):1161-1172.

5. McGrath PJ, Stewart JW, Fava M, et al. Tranylcypromine versus venlafaxine plus mirtazapine following three failed antidepressant medication trials for depression: a STAR*D report. Am J Psychiatry. 2006;163(9):1531-1541.

6. Nierenberg AA, Fava M, Trivedi MH, et al. A comparison of lithium and T(3) augmentation following two failed medication treatments for depression: a STAR*D report. Am J Psychiatry. 2006;163(9):1519-1530.

7. DeGruy FV 3rd. A note on the partnership between psychiatry and primary care. Am J Psychiatry. 2006;163(9):1487-1489.

8. Inventory of Depressive Symptomatology (IDS) and Quick Inventory of Depressive Symptomatology (QIDS). Available at: www.ids-qids.org. Accessed October 3, 2006.

9. STAR*D. Sequenced Treatment Alternatives to Relieve Depression. Available at: www.star-d.org.  Accessed October 3, 2006.

10. Kennedy GJ. Addressing the unmet promise of antidepressant trials in older adults. Primary Psychiatry. 2006;13(2):24-27,36.