1boringoldman.com

Adolescents With Depression
tempted or completed suicides, it does lend support for theuse of fluoxetine in combination with CBT for adolescents To the Editor: The authors of the Treatment for Adoles-
with MDD and suicidal ideation. For the majority of pa- cents With Depression Study (TADS) conclude that “ . . . de- tients who have MDD without suicidal ideation, the addi- spite calls to restrict access to medications, medical man- tion of expensive and time-consuming psychotherapy does agement of MDD [major depressive disorder] with fluoxetine, not seem to be justified by this study.
including careful monitoring for adverse events, should be According to the US Census Bureau, there were 29 mil- made widely available, not discouraged.”1 We disagree with lion individuals between the ages of 12 and 18 years in 2003.2 this conclusion. Looking at the primary outcome measure Assuming a point prevalence of MDD in adolescents of 5%3 (change in the Children’s Depression Rating Scale-Revised would translate to about 1.5 million adolescents. We do not total score), fluoxetine alone resulted in minimal benefit over believe that resources can be committed for all of these ado- placebo: the placebo effects were 86% of the fluoxetine ef- lescents to receive fluoxetine in combination with CBT. For fects (change of 19.4 compared with 22.6 points). At the all adolescents with MDD to receive fluoxetine alone will same time, fluoxetine caused a significantly higher rate of require primary care physicians to be the prescribers for much harm-related adverse events, such as suicidal ideation, and of the fluoxetine. Providing CBT to all adolescents with MDD physiological effects (diarrhea, insomnia, and sedation) com- would be even more difficult, given the limited number of pared with placebo or cognitive-behavioral therapy (CBT) qualified therapists, the significant expense, and the need alone, as well as a higher rate of psychiatric adverse events for compliance with a time-consuming treatment. There- (irritability, mania, and fatigue) compared with placebo.
fore, it seems most sensible to prioritize CBT (with phar- Our own risk-benefit analysis of these results leads us to macotherapy) to adolescents who have MDD with suicidal conclude that a drug-free treatment like CBT alone, or even a psychological placebo such as exercise, should be offeredas the first-line treatment because many adolescents will ben- Arthur Rifkin, MD
efit without incurring the increased risk of psychiatric and Department of Psychiatry
Zucker-Hillside Hospital
North Shore-Long Island Jewish Health System
David Antonuccio, PhD
New Hyde Park, NY
oliver2@aol.com
William Rifkin, MD
University of Nevada School of Medicine
william.rifkin@yale.edu
Department of Medicine
David Burns, MD
Yale University School of Medicine
Stanford University School of Medicine
New Haven, Conn
Palo Alto, Calif
1. Treatment for Adolescents With Depression Study (TADS) Team. Fluoxetine,
1. Treatment for Adolescents With Depression Study (TADS) Team. Fluoxetine,
cognitive-behavioral therapy, and their combination for adolescents with depres- cognitive-behavioral therapy, and their combination for adolescents with depres- sion: Treatment for Adolescents With Depression Study (TADS) randomized con- sion: Treatment for Adolescents With Depression Study (TADS) randomized con- trolled trial. JAMA. 2004;292:807-820.
trolled trial. JAMA. 2004;292:807-820.
2. Population Division of US Census Bureau. Annual estimates of the population
by sex and five-year age groups for the United States: April 1, 2000 to July 1, 2003.
To the Editor: While the TADS trial provides valuable data
Available at: http://www.census.gov/popest/national/asrh/NC-EST2003-as.html. Accessibility verified October 26, 2004.
about a significant problem, we differ from the authors in 3. Essau CA, Petermann F. Depressive Disorders in Children and Adolescents: Epi-
their conclusion that treatment with fluoxetine should be demiology, Risk Factors, and Treatment. Northvale, NJ: Jason Aronson Inc; 1999.
made widely available, and that CBT in combination withfluoxetine should be “readily available as part of compre-hensive treatment for depressed adolescents.”1 Because 2 of GUIDELINES FOR LETTERS. Letters discussing a recent JAMA article will have
the best chance of acceptance if they are received within 4 weeks of the article’s
3 outcome measures (Clinical Global Impressions [CGI] and publication. They should not exceed 400 words of text and 5 references. Letters Reynolds Adolescent Depression Scale [RADS]) show no sig- reporting original research should not exceed 600 words and 6 references. All let-ters should include a word count. Letters must not duplicate other material pub- nificant difference between fluoxetine treatment alone and lished or submitted for publication. Letters will be published at the discretion of fluoxetine in combination with CBT, we do not see this study the editors and are subject to editing and abridgment. A signed statement for au-thorship criteria and responsibility, financial disclosure, copyright transfer, and ac- showing an advantage for the addition of CBT in all ado- knowledgment is required for publication. Letters not meeting these specifica- tions are generally not considered. Letters will not be returned unless specifically The Suicidal Ideation Questionnaire-Junior High School requested. Also see Instructions for Authors (July 7, 2004). We prefer that lettersbe submitted electronically to jama-letters@jama-archives.org. Letters may also Version showed significantly more improvement with fluox- be sent by surface mail to Letters Editor, JAMA, 515 N State St, Chicago, IL 60610, etine in combination with CBT compared with fluoxetine or by fax to (312) 464-5225 (please also send a hard copy via surface mail).
alone. Although this difference may not apply to at- Letters Section Editor: Robert M. Golub, MD, Senior Editor.
2004 American Medical Association. All rights reserved.
(Reprinted) JAMA, December 1, 2004—Vol 292, No. 21 2577
To the Editor: In interpreting the TADS results,1 the cen-
tive than treatment with fluoxetine alone and not signifi- tral issue is the benefit-to-risk ratio, which can be deter- cantly more effective than treatment with pill placebo. The mined by considering the number needed to treat (NNT), TADS Team was surprised by the 43% clinical response rate the number needed to harm (NNH), and the number needed for CBT alone, which was lower than in some other stud- to prevent.2 In this study, a categorical positive response was ies,2,3 and posited that the lower response rate may have been achieved in 71.0% of participants treated with fluoxetine in due to greater severity, chronicity, and comorbidity in the combination with CBT; in 60.6% with fluoxetine alone; in TADS trial participants compared with previous trials.
43.2% with CBT alone; and in 34.8% with placebo. Based In one of those earlier studies, we compared CBT with 2 on these outcomes, the NNT is 3.9 for fluoxetine alone com- other psychosocial treatments for mostly clinically re- pared with placebo and 3.7 for drug vs no drug. I believe ferred, depressed adolescents.2 Greater severity was associ- that these represent low NNTs (high benefit) that are clini- ated with a less robust response to CBT, and it indeed does appear that the TADS participants had more functional im- The TADS Team reported suicide-related adverse events pairment and greater chronicity of depression. However, our in 6.9% of children taking fluoxetine and in 4.0% of chil- rates of comorbidity were comparable (58.2% vs 59.5%), and dren who did not take fluoxetine; this corresponds to a NNH in fact, in our study, comorbid anxiety was a positive prog- of 34. Likewise, TADS reported suicide attempts in 6 (2.78%) nosticator of response to CBT.4 This finding was corrobo- of 216 adolescents taking fluoxetine and in 1 (0.45%) of 223 rated in another treatment study of adolescent MDD using adolescents not taking fluoxetine. The corresponding NNH CBT,5 which also found that treatment response to CBT is is 43. The NNT is far more salient than either NNH.
robust, even in the face of comorbidity with disruptive and There were no completed suicides in the TADS trial. Nev- ertheless, extrapolating from epidemiological data that in- Other explanations for the difference in response rates, be- dicate 8% of reported suicide attempts overall are lethal,3 sides greater severity and chronicity, may have to do with the the estimated NNH with an outcome of completed suicide difference in expectation, content, and format for treatment would be 535. Balancing any risk of drug-attributable sui- delivery. Because those who agreed to randomization knew cide is the prevention of disease-attributable suicide in pa- that they might get medication, it is possible that these par- tients who receive the drug. Using a conservative lifetime ticipants had different expectations about treatment than those case-fatality rate estimate of 2.2% among outpatients diag- who agreed to a study in which all of the options were psy- nosed as having MDD,4 and allocating 30% of this risk to chotherapy. There may have been differences in the method the adolescent years,5 a completed suicide rate of 0.66% of delivery and content of the CBT. With regard to format, would be expected. When the TADS-observed NNT of 3.7 in 1 of the 2 previous CBT studies,2 the delivery was less struc- is applied to these estimates, the number needed to pre- tured and in the other study3 CBT was highly structured, but vent 1 suicide is 560. Thus, there is suggestive evidence of delivered in a group format. With regard to content, both equipoise between the therapeutic outcome of preventing TADS and one of the previous studies3 focused on teaching suicide and any potential drug-related provocation of sui- multiple skills, whereas the other study focused almost mono- cide among adolescents treated for MDD with fluoxetine.
thematically on cognitive restructuring.2 Overall, these estimates of absolute risk support the con- Jeffrey A. Bridge, PhD
clusion that favorable benefit-to-risk ratios exist for treat- bridgeja@upmc.edu
ment with fluoxetine in adolescents with MDD.1 David A. Brent, MD
Western Psychiatric Institute and Clinic
Bernard J. Carroll, MB, BS, PhD
Pittsburgh, Pa
bcarroll@redshift.com
Pacific Behavioral Research Foundation
1. Treatment for Adolescents With Depression Study (TADS) Team. Fluoxetine,
Carmel, Calif
cognitive-behavioral therapy, and their combination for adolescents with depres-sion: Treatment for Adolescents With Depression Study (TADS) randomized con- 1. Treatment for Adolescents With Depression Study (TADS) Team. Fluoxetine,
trolled trial. JAMA. 2004;292:807-820.
cognitive-behavioral therapy, and their combination for adolescents with depres- 2. Brent DA, Holder D, Kolko D, et al. A clinical psychotherapy trial for adoles-
sion: Treatment for Adolescents With Depression Study (TADS) randomized con- cent depression comparing cognitive, family, and supportive therapy. Arch Gen trolled trial. JAMA. 2004;292:807-820.
2. Laupacis A, Sackett DL, Roberts RS. An assessment of clinically useful mea-
3. Clarke GN, Rohde P, Lewinsohn PM, et al. Cognitive-behavioral treatment of
sures of the consequences of treatment. N Engl J Med. 1988;318:1728-1733.
adolescent depression: efficacy of acute group treatment and booster sessions.
3. Spicer RS, Miller TR. Suicide acts in 8 states: incidence and case fatality rates
J Am Acad Child Adolesc Psychiatry. 1999;38:272-279.
by demographics and method. Am J Public Health. 2000;90:1885-1891.
4. Brent DA, Kolko DJ, Birmaher B, et al. Predictors of treatment efficacy in a clini-
4. Bostwick JM, Pankratz VS. Affective disorders and suicide risk: a reexamination.
cal trial of three psychosocial treatments for adolescent depression. J Am Acad Am J Psychiatry. 2000;157:1925-1932.
Child Adolesc Psychiatry. 1998;37:906-914.
5. Goodwin FK, Jamison KR. Manic-Depressive Illness. New York, NY: Oxford
5. Rohde P, Clarke GN, Lewinsohn PM, et al. Impact of comorbidity on a cognitive-
behavioral group treatment for adolescent depression. J Am Acad Child AdolescPsychiatry. 2001;40:795-802.
To the Editor: In their study, the TADS Team reported that
In Reply: We appreciate the opportunity to clarify the re-
the most effective treatment was fluoxetine in combina- sults of TADS. Drs Antonuccio and Burns interpret our re- tion with CBT.1 Treatment with CBT alone was less effec- sults as showing that fluoxetine offers no significant ben- 2578 JAMA, December 1, 2004—Vol 292, No. 21 (Reprinted)
2004 American Medical Association. All rights reserved.
efit over placebo, arguing (we believe incorrectly) that der of magnitude lower than those cited by Carroll. This im- exercise is a placebo treatment and is somehow preferable plies that the NNH for adolescents would be in the thou- to any of the TADS treatments. Drs Rifkin and Rifkin con- sands, not hundreds, and, hence, that the risk-to-benefit ratio clude that for the study population as a whole CBT plus fluox- is not at equipoise as Carroll suggests, but strongly favors etine offers little or no advantage over fluoxetine alone. In including medication management as one component of response, we note that TADS used the rate of improvement treatment if the aim is to prevent suicide.
and 12-week outcome on the Children’s Depression Rat- Nonetheless, given the large number of suicide attempts ing Scale, the rate of improvement and 12-week outcome in adolescents each year, any level of increased risk is rel- on the RADS, and percentage of patients much or very much evant to the public health, and we believe warrants strong improved on the CGI scale. Fluoxetine in combination with cautionary labeling language. Importantly, the TADS find- CBT was superior to placebo and to CBT alone on all 5 mea- ings suggest that combining CBT with fluoxetine may di- sures. Combined treatment was superior to fluoxetine on 2 rectly reduce suicidal ideation and, via an unknown mecha- measures. Fluoxetine alone was superior to placebo on 3 nism, also reduce the small risk for harm-related adverse measures and to CBT alone on all 5 measures. Addition- ally, fluoxetine in combination with CBT showed a large effect Finally, while we do not disagree with Rifkin and Rifkin size and fluoxetine alone showed a moderate effect size on when they suggest that CBT should be reserved for adoles- the Children’s Depression Rating Scale-Revised and on the cents showing risk for suicidality, we believe that TADS pro- CGI, whereas CBT alone was not different from placebo.
vides a potent public health argument for making CBT more Thus, we believe that the data show that fluoxetine in com- widely available for adolescents with MDD. Without a strat- bination with CBT is best, fluoxetine alone is effective but egy to speed adoption of CBT, adolescents with MDD will not not as effective as combined treatment, and that either fluox- receive the clinically meaningful increase in benefit and reduc- etine-containing treatment is superior to CBT alone.
tion in risk conferred by fluoxetine in combination with CBT.
Drs Antonuccio and Burns state that the relatively small In this context, and considering the points of Drs Bridge risk for adverse events associated with fluoxetine is unac- and Brent, it remains to be seen if the less severely ill ado- ceptable. However, on the Suicidal Ideation Question- lescent might do well with CBT alone or if CBT alone will naire, which indexes suicidal ideation rather than MDD, com- fare better at 36 weeks than it did with acute treatment.
bined treatment proved superior to fluoxetine alone, to CBT Planned analyses from the TADS group should shed fur- alone, and to placebo, but the 2 monotherapies and pla- ther light on the question of which treatment is best and cebo did not separate. Thus, for suicidal thinking, fluox- with which set of adolescent clinical characteristics.
etine in combination with CBT was the only treatment that John S. March, MD, MPH
offered benefit, whereas fluoxetine alone did not appear to for the TADS Team
induce suicidal thinking. On the other hand, the TADS re- jsmarch@acpub.duke.edu
sults are consistent with recent Food and Drug Adminis- Duke University Medical Center
tration findings in identifying an approximately 2-fold in- Durham, NC
crease in risk of self-harm behaviors associated with 1. US Food and Drug Administration. Antidepressant Use in Children, Adoles-
antidepressant medication. Fortunately, these behaviors are cents, and Adults . Available at: http://www.fda.gov/cder/drug/antidepressants/default.htm. Accessed October 22, 2004.
uncommon, occurring in the Food and Drug Administra- 2. Spicer RS, Miller TR. Suicide acts in 8 states: incidence and case fatality rates
tion analyses in approximately 4% of patients treated with by demographics and method. Am J Public Health. 2000;90:1885-1891.
3. Grunbaum JA, Kann L, Kinchen S, et al. Youth risk behavior surveillance—
an antidepressant and 2% of patients treated with placebo1; United States, 2003. MMWR Surveill Summ. 2004;53:1-96.
hence, the absolute risk increase associated with medica-tion is 2%, corresponding to an NNH of 50 patients.
Spontaneous Regression of Cancerous Tumors
Dr Carroll notes (and we presented at the Food and Drug Detected by Mammography Screening
Administration hearings on antidepressant risk for sui-cide) that the benefit (NNT) to risk (NNH) ratio strongly To the Editor: In their study, Dr Joensuu and colleagues1
favors fluoxetine in combination with CBT and, less ro- reported that cancerous tumors detected by mammogra- bustly, fluoxetine alone. We think Carroll may actually over- phy screening have a lower risk of distant recurrence and estimate the population-derived rate of suicide attempts and better survival than those detected outside of screening, in- completed suicide, and thus underestimate the degree of ben- dependent of the number of positive axillary lymph nodes, efit relative to risk. Using Spicer and Miller’s2 definition of the primary tumor size, age at cancer detection, histologi- attempters as those who make an attempt resulting in an cal grade, and other biological factors.
injury, poisoning, or overdose that had to be treated by a In several European countries with nationwide screen- physician or nurse, and data from the Youth Risk Behavior ing programs, the breast cancer rate in the invited age groups Survey,3 the attempted case rate in the adolescent popula- is about 50% higher than the background level prior to the tion is 0.6% in boys and 0.09% in girls. If this is the case, introduction of the screening program.2,3 This difference has the corresponding case-fatality rates are more than an or- so far been explained by a lead-time effect, assuming that 2004 American Medical Association. All rights reserved.
(Reprinted) JAMA, December 1, 2004—Vol 292, No. 21 2579

Source: http://1boringoldman.com/images/jama-tads.pdf