The guideline recommends the serotonin modulator nefazodone as an effective medication for the treatment of depression. However, before initiating or continuing treatment with nefazodone, consideration should be given to the recent reports of life-threatening hepatic failure in patients treated with nefazodone (2–8). These reports have led the Food and Drug Administration (FDA) to change the drug's labeling, adding a black box warning of possible liver failure leading to death and/or a need for transplant and contraindicating the drug in patients who were withdrawn from it because of evidence of liver injury. According to the warning, the reported rate of liver failure resulting in death or transplant in the United States is about 1 case per 250,000–300,000 patient-years of nefazodone treatment. This represents a rate of about three to four times the estimated background rate of liver failure and is probably an underestimate. There are no known predictors of the development of liver toxicity and failure with nefazodone; toxicity and failure have been reported in individuals early in the course of treatment as well as in persons receiving stable dosages for many months. The FDA warns that patients with preexisting liver disease should not be treated with nefazodone and that patients should be advised to immediately report symptoms that may indicate liver dysfunction, such as jaundice, anorexia, gastrointestinal complaints, and malaise. The warning is available at http://www.fda.gov/medwatch/SAFETY/2002/serzone_deardoc.pdf.

Because patients with major depressive disorder are at greater risk of suicide, the guideline indicates that they should be assessed for suicide risk initially and over the course of treatment. The guideline also notes that "the risk of suicide in some patients recovering from major depressive disorder increases transiently as they develop the energy and capacity to act on self-destructive plans made earlier in the course of their illness" (1). The guideline describes factors associated with an increased risk of suicide in patients with major depressive disorder but notes that "the ability to predict suicide attempts and completed suicide is poor, with both many false positives (i.e., patients who appear more likely to make attempts or complete suicide but who do not) and false negatives (i.e., patients who appear less likely to make attempts or complete suicide but who do)" (1). With respect to children and adolescents, the guideline recommends caution when basing treatment decisions on adult data. All of these general principles remain true.

Recent research has raised concern about the safety of antidepressant use for children and adolescents and has led the FDA to add additional warnings, including a black box warning, to the labeling of all antidepressant medications (9). Although the APA practice guideline focused on treatment of adults, not children and adolescents, discussion of this research is appropriate in this watch. In 23 short-term (4- to 16-week) trials involving more than 4,400 patients receiving nine antidepressant drugs (selective serotonin reuptake inhibitors [SSRIs] and others), suicidal thinking or behavior was observed in 78 individuals. The average risk of suicidal thinking or behaviors was 3.8% for participants receiving a drug compared with 2.1% for those given placebo, suggesting an approximately twofold increase in risk (10). Consequently, in the new labeling, the FDA notes that pooled analyses of placebo-controlled antidepressant trials (available at http://www.fda.gov/ohrms/dockets/ac/04/briefing/2004-4065b1.htm) have shown an increased risk of suicidal thinking or behavior in children and adolescents with major depressive disorder, obsessive-compulsive disorder, and other psychiatric disorders. However, it is important to note that no suicides occurred in any of these trials. Furthermore, according to the Centers for Disease Prevention and Control, suicidal thinking and suicide attempts are common among adolescents, occurring respectively in about 17% and 8.5% of adolescents each year. As a result of such attempts, only 0.002% of girls and 0.012% of boys ultimately die (11), highlighting the fact that suicidality (i.e., suicidal thoughts and behaviors) and suicide are not equivalent.

Although the studies reviewed by the FDA showed some variations among antidepressants, it was not clear whether the increase in suicidal thinking or behavior was specific to particular antidepressants, extended to adults, or occurred with longer-term antidepressant use (i.e., beyond several months). Thus, the FDA labeling change recommends that "all pediatric patients being treated with antidepressants for any indication should be observed closely for clinical worsening, suicidality, and unusual changes in behavior, especially during the initial few months of a course of drug therapy, or at times of dose changes, either increases or decreases" (9). To facilitate this, the FDA suggests an increased frequency of face-to-face contacts (i.e., weekly for the first 4 weeks, biweekly for the next 8 weeks), with prescriptions written for "the smallest quantity of tablets consistent with good patient management, in order to reduce the risk of overdose." The FDA also recommends that patients be screened for bipolar disorder before antidepressant therapy is initiated. Medication guides that will accompany prescriptions of antidepressant medications are intended to alert families and caregivers of pediatric and adult patients to the need to monitor patients for the emergence of agitation, irritability, hostility (aggressiveness), impulsivity, anxiety, panic attacks, insomnia, akathisia, hypomania, mania, or unusual changes in behavior so that such symptoms can be reported immediately to health care professionals.

When signs or symptoms are observed that suggest an increased suicide risk, recommendations are available for assessing and managing suicidal behaviors in children and adolescents from the American Academy of Child and Adolescent Psychiatry (12) and in adults from the APA (13).

As with all decisions about treatment for individuals with psychiatric illnesses, the choice to initiate antidepressant therapy must balance the potential risks of treatment against the potential therapeutic benefit and the risks of untreated illness (14–16). The risks of untreated depression, including suicide and other suicidal behaviors, are high and have been clearly delineated in youths and in adults (1, 12, 13, 17). In adults, the evidence for antidepressant efficacy is also clear, as reviewed in the APA guideline.

In children and adolescents, the issue has been confounded by a smaller number of clinical trials as well as by negative trials. Positive evidence includes the studies of fluoxetine that led to its receiving an FDA indication for treatment of depression in children and adolescents and the recent Treatment for Adolescents With Depression Study (TADS) sponsored by the National Institute of Mental Health. This large, yearlong community effectiveness trial (18) of youths with moderate to severe major depressive disorder found that by 12 weeks of treatment, rates of response to fluoxetine alone (60.9%) were greater than those with either placebo (35%) or cognitive behavior therapy (CBT) alone (43.2%). In combination with CBT, fluoxetine was associated with an even greater response rate (71%)—twice that seen with placebo. Furthermore, clinically significant suicidal thinking, which was observed in 29% of study subjects prior to treatment, improved in all groups.

In addition to this direct evidence of antidepressant efficacy from clinical trials, the national decreases in youth suicide and overall suicide rates, during a period when the prescribing of antidepressants has increased, might provide indirect evidence for an overall preventive effect of antidepressant treatment on suicide (14, 19, 20).

Because additional information on the issue of antidepressants and suicide risk is likely to emerge, readers are urged to seek updates on the Web sites of the FDA (http://www.fda.gov), the American Academy of Child and Adolescent Psychiatry (http://www.aacap.org), and the APA (http://www.psych.org).

Two new antidepressants, escitalopram and duloxetine, have been approved by the FDA since publication of the guideline. Escitalopram, the active S-enantiomer of citalopram, is an SSRI antidepressant that has FDA approval for acute and maintenance treatment of major depressive disorder. Evidence from randomized, clinical trials suggests that escitalopram is superior to placebo in the short-term treatment of depression (21), with efficacy and tolerability comparable to that of other antidepressants, including venlafaxine (22, 23) and citalopram (21, 24).

Duloxetine, a dual serotonin-norepinephrine reuptake inhibitor, has also been approved for the treatment of depression. Randomized, clinical trials show duloxetine to be more efficacious than placebo (25–28) and comparable in efficacy to SSRI antidepressants (27, 28). The drug is generally well tolerated, with reported adverse effects of treatment (e.g., nausea, dry mouth, dizziness, somnolence, insomnia, constipation, and asthenia) varying across studies but typically being infrequent at total oral daily doses of 40–120 mg. In addition to ameliorating core symptoms of depression, duloxetine exhibits efficacy relative to placebo in treating painful physical symptoms (26, 28–31) and anxiety in depressed individuals (32). It also appears to ameliorate pain in the absence of depression (31), and it is FDA indicated for the management of pain associated with diabetic peripheral neuropathy.

Since publication of the guideline, the combination of fluoxetine and olanzapine has received FDA approval on the basis of a randomized, clinical trial supporting its use (33). Although the combination product (brand name Symbyax) is indicated for the treatment of episodes of bipolar depression, combined treatment with fluoxetine and olanzapine has also been found useful in treating episodes of major depression with psychotic features (34) and in treatment-resistant depression (35, 36).

Reboxetine, a selective norepinephrine reuptake inhibitor that was noted in the guideline to be possibly close to receiving approval, is not currently approved for use. Atomoxetine, another selective norepinephrine reuptake inhibitor, has recently become available but is approved for the treatment of attention deficit hyperactivity disorder. Although the possibility of using atomoxetine, either alone or in combination with an SSRI, for the treatment of depression has prompted interest among clinicians, the drug does not have an FDA indication for this use, and data on its use in the treatment of depression are limited (37–39). In addition, the FDA has just added a bolded warning to atomoxetine's labeling noting that the medication should be discontinued in patients who develop jaundice or laboratory evidence of liver injury (http://www.fda.gov/bbs/topics/ANSWERS/2004/ANS01335.html).

The evidence cited here on new medications is based on premarketing research. Clinicians are cautioned that real-world experiences after a drug is approved often bring unexpected findings, including less efficacy in some circumstances and unanticipated adverse events.

Since publication of the guideline, results from additional meta-analyses and well-designed trials of St. John's wort have become available. Although some additional randomized, controlled trials have shown St. John's wort to be superior to placebo (40–42) or to comparator antidepressants (43, 44) in treating mild to moderate depression, other large trials have shown it to be no better than placebo (45–47) or comparator antidepressants in efficacy (42, 46–50). Several factors continue to confound interpretation of the literature, including trial length, adequacy of comparison treatment, and the reliability, stability, and comparability of Hypericum preparations. The most recent meta-analysis noted a trend for decreasing effect sizes in trials of St. John's wort over time, suggesting that it may be less effective in treating depression than previously thought (51). Although St. John's wort is generally well tolerated in clinical trials, increasing attention has been given to its tendency to compromise the effectiveness of other medications (e.g., cyclosporine, HIV protease inhibitors, oral contraceptives, digoxin, warfarin, and theophylline) by interactions mediated through cytochrome P450 enzymes (e.g., CYP3A4, CYP2C9, CYP1A2, CYP2C19) and transport protein P-glycoproteins (52–56).

Evidence for the efficacy of electroconvulsive therapy (ECT) in treating major depressive episodes, already compelling at the time the guideline was published, has been strengthened by additional findings. Recent meta-analyses have highlighted the superior efficacy of ECT relative to sham treatment and also relative to pharmacotherapies for depression (57, 58). In depressed patients who received an acute course of ECT, data from the Consortium for Research in ECT have shown that thrice-weekly bilateral ECT is associated with rapid initial response and high rates of sustained response and remission (59). In comparison with bilateral ECT, right unilateral ECT has been associated with fewer cognitive effects (60, 61), particularly in autobiographical memory (62). However, as with bilateral ECT, cognitive effects vary with the extent to which the electrical stimulus exceeds the seizure threshold (60). In addition, the overall efficacy of unilateral ECT appears to be less than that with bilateral ECT (57). Several additional studies highlight the diminished efficacy of barely suprathreshold electrical stimulation with right unilateral electrode placement (60, 61, 63) and a corresponding need to administer right unilateral ECT at stimulus intensities that are at least six times the initial seizure threshold (64). Additional details on the clinical use of ECT in the treatment of major depression can be found in the 2001 revision of the APA's The Practice of Electroconvulsive Therapy: Recommendations for Treatment, Training, and Privileging (64).

Although other somatic treatments, including repetitive transcranial magnetic stimulation, magnetic seizure therapy, and vagal nerve stimulation, have also been studied over the past 5 years, evidence is not yet sufficient to recommend their use in routine clinical practice.