
Results
The ten studies included in this review are summarised in Table I . Eight of the ten studies included descriptions and/or tables detailing patient demographics in treatment groups versus placebo groups. Characteristics of ethnicity, sex, age and comorbid conditions were similar in the two treatment groups in the 1998 unpublished paroxetine study on the GSK Website,[12] the 2002 Emslie et al. fluoxetine study,[19] the Keller et al. paroxetine study[20] and in all four treatment groups in the study by March et al. (TADS).[18] In the first fluoxetine study by Emslie et al., placebo and treatment groups were similar in demographic and clinical features, except that subjects in the fluoxetine group had a greater lifetime incidence of comorbid anxiety disorders.[21] In the 2001 unpublished paroxetine study by Emslie at al., there was a slightly higher percentage of patients with comorbid psychiatric diagnoses in the paroxetine group versus the placebo group.[22] In Wagner et al.'s study of sertraline, treatment groups were similar with respect to race, clinical characteristics and psychosocial stressors at baseline.[23] There was a significant difference in sex: 57% of subjects treated with sertraline were female versus 45% receiving placebo. In the study by Wagner et al. of citalopram, there were no significant differences between the placebo and citalopram groups with regard to age, sex, race, weight or history of depression.[24]
Several of the studies showed significant improvement in outcome measures in SSRI-treated groups versus controls. In the earliest study by Simeon et al., subjects in the group receiving fluoxetine showed slightly greater improvement in all depressive symptoms except sleep, but the differences were not significant.[25] In the earlier study by Emslie et al., 56% of subjects receiving fluoxetine demonstrated significant improvement versus 33% receiving placebo (p = 0.02) as evidenced by the CGI-I scale at study exit and the weekly secondary measure of the CDRS-R after 5 weeks (p = 0.03).[21] In the multisite study by Emslie et al., fluoxetine-treated subjects showed significantly greater improvement in the CDRS-R (35.1 ± 13.5) than placebo-treated subjects (40.2 ± 13.5) starting at week 1 and continuing for the remainder of the study (p < 0.001).[19] Using a CGI-I rating of 1 or 2 (much improved or very much improved), 53.3% of fluoxetine-treated subjects responded versus 36.8% of placebo-treated subjects (p = 0.028). In the TADS study, paediatric subjects receiving fluoxetine with cognitive behavioural therapy (CBT) improved significantly more compared with those taking placebo (p = 0.001) or fluoxetine alone (p = 0.02), based on CDRS-R scores.[18] Fluoxetine treatment alone was also significantly superior to placebo. Response rates for the groups, using a CGI-I of 1 or 2, were as follows: fluoxetine with CBT, 71.0% (95% CI 62, 80); fluoxetine alone, 60.6% (95% CI 51, 70); CBT alone, 43.2% (95% CI 34, 52); placebo, 34.8% (95% CI 26, 44).
In the 1998 unpublished paroxetine study there were no statistically significant differences between paroxetine and placebo in either the primary efficacy variables (50% reduction in MADRS score and change in K-SADS-L depression subscale) or the secondary efficacy variables (change from baseline in MADRS total score, CGI-S score, CGI-I score, Beck Depression Inventory [BDI], and the Mood and Feelings Questionnaire [MFQ]).[12] In the 2001 unpublished paroxetine study, no statistically significant difference was demonstrated in the primary outcome measure (change in CDRS-R total score) or in the secondary outcome measure (change in CGI-S, CGI-I global assessment of functioning).[22]
When compared with imipramine and placebo, paroxetine-treated subjects in the study by Keller et al. differed from placebo on several measures.[20] The first was their primary outcome measure, a HAM-D score of <8 or a >50% reduction from baseline by the end of treatment (p = 0.02). Differences also occurred on the HAM-D depressed mood item (p = 0.001), K-SADS-L depressed mood item (p = 0.05), and CGI-I score of 1 or 2, very much or much improved, respectively (p = 0.02). In this study, imipramine treatment did not differ from placebo in any of the outcome variables.
Sertraline-treated subjects showed significantly greater improvement than those receiving placebo, as evidenced by their mean change in CDRS-R (-22.84 vs -20.19; p = 0.007), the CGI-S (p = 0.009), and CGI-I (p = 0.005).[23] Differences were seen as early as week 1 on the CGI-I. In Wagner et al.'s citalopram study, the mean CDRS-R scores decreased significantly more from baseline in the citalopram group, as early as week 1.[24] Also, by the end of the study, significantly more citalopram-treated subjects (36%) met the response criterion of a CDRS-R score of ≤28, which was decided on prospectively, versus placebo-treated subjects (24%) [p < 0.05]. While these response rates appear low, the criterion used for response in this study was full remission, and other studies have reported similar rates of remission, notably both fluoxetine studies by Emslie et al.[19,21,27] A second, unpublished, citalopram study was accessed from the UK's Medicines and Healthcare products Regulatory Agency (MHRA) website.[26] No statistically significant difference was found on the K-SADS-Present Episode total score for drug (-12.4) versus placebo (-12.7) in 233 adolescents aged 13–18 years.[26]
The SSRIs were reported as being well tolerated compared with placebo, as illustrated in Table II . Study withdrawal rates due to adverse events ranged from 0% to 9.7% and were highest for paroxetine. Some serious adverse events (SAEs) occurred in patients taking SSRIs. SAEs are defined as those requiring hospitalisation or otherwise significant psychiatric symptoms such as suicidal ideation, aggressiveness or euphoria. The SAEs for both drug and placebo are shown in Table II . In the paroxetine study, SAEs occurred in 11 of 275 subjects taking paroxetine (4%). Five subjects reported emotional lability including suicidal ideation or gestures. One subject experienced headache during discontinuation taper, two subjects experienced worsening depression, two had conduct problems or hostility, and one had an expansive mood.[20] Of the 11 total subjects with SAEs, the investigator judged only the headache to be due to paroxetine. In the placebo group, one subject had emotional lability and another had worsening depression.
In the sertraline study performed by Wagner et al., of 376 subjects, two sertraline-treated and two placebo-treated subjects attempted suicide.[23] Three additional sertraline recipients reported suicidal ideation, and one additional sertraline recipient became aggressive. One subject in the sertraline group and four in the placebo group required medical hospitalisation. In the study by Wagner et al. on citalopram,[24] of 174 total subjects, two subjects receiving citalopram discontinued the study as a result of agitation. Another two subjects receiving placebo were removed from the study because of worsening of depression.
In the TADS study, 33 (7.5%) of 439 subjects experienced harm-related adverse events, which included any type of harm to self or others.[18] There was an increased risk of harm-related adverse events in subjects receiving fluoxetine versus those not receiving it. Twenty-four of the 33 harm-related adverse events were suicide-related, comprising 5.5% of the total 439 patients. A suicide-related adverse event was defined in the study as requiring that the patient exhibit either worsening suicidal ideation or make a suicide attempt, or both. The number of actual suicide attempts (seven) was too small for comparison of statistical significance between the different treatment groups. Four of the seven who made suicide attempts were receiving fluoxetine and CBT, two had received fluoxetine only, and one had received CBT only. No completed suicides occurred in the study, but results suggest a protective effect of CBT against suicidal ideation in those taking fluoxetine. This may be an important finding, but replication is needed.
Common physical adverse effects in the studies were dry mouth, vomiting, nausea, diarrhoea, somnolence, insomnia, dizziness, tremor and agitation. Somnolence was the only statistically significant physical adverse effect versus placebo in the paroxetine study.[7] Headache was the only unsolicited adverse effect that was reported significantly more often by patients on fluoxetine versus placebo in the Emslie et al. fluoxetine study published in 2002.[19] However, there was no significant difference between treatment groups in subjects reporting headaches on the completed Side Effects Checklist. In the Wagner et al. citalopram study, seven adverse effects as listed in Table II were found to occur with a frequency of >5% and with a higher incidence than observed in the placebo group: rhinitis, nausea, abdominal pain, influenza-like symptoms, fatigue, diarrhoea and back pain.[24] In the TADS study, the most common adverse effects listed in Table II occurred in at least 2% of patients, in at least one treatment group, and with an incidence of at least twice the rate in comparison with subjects receiving placebo.[18] The symptoms are listed from most to least common.
To date, most data from unpublished studies are not available on SSRIs used in children and adolescents with depression. The data from paroxetine studies completed in the UK were not available for inclusion in this study. The UK's MHRA has issued a warning against prescribing paroxetine to depressed patients aged <18 years based on data showing higher rates of suicide-related events and suicide attempts in the paroxetine-treated group versus placebo.[23] However, no completed suicides occurred in those studies. The MHRA ordered a labeling change, contraindicating the use of paroxetine for treatment of paediatric depression.[23] In addition, the UK Department of Health went on to prohibit the prescribing of all antidepressants other than fluoxetine to patients aged <18 years. In the US, all antidepressants have a black box warning for use in children and adolescents because of the risk of suicidality.
In 2004, the FDA collated data from 24 short-term trials involving over 4400 patients using nine antidepressant drugs.[28] Children in the studies met criteria for major depressive disorder, obsessive compulsive disorder, or other psychiatric disorders. Analysis of the data demonstrated a greater risk of adverse events including suicidal ideation or behaviours during the first few months of treatment in youths taking antidepressants. Seventy-eight of the 4400 patients experienced suicidal ideation or behaviours. This constituted a risk of 4%, or double that of placebo. There were no completed suicides in any of the studies. The FDA report did not provide data on the efficacy of SSRIs, although many unpublished studies allegedly did not demonstrate drug efficacy over placebo in children and adolescents. The FDA has followed this up with a black box warning on all antidepressant labeling regarding serious risks of suicidal ideation or attempts when these medications are used in children and adolescents. Advocacy agencies, including the American Academy of Child and Adolescent Psychiatry, have expressed serious concerns that family practitioners and paediatricians practising in rural areas where child psychiatrists are not available may no longer prescribe SSRIs even to seriously depressed and suicidal children and adolescents as a result of this warning. Printer- Friendly Email This
Clin Drug Invest. 2006;26(5):247-255. ©2006 Adis Data Information BV
This is a part of article SSRIs in the Treatment of Depression in Children and Adolescents Taken from "Cheap Prozac Fluoxetine" Information Blog



