Clinical Psychopharmacology Seminar

Pediatric Psychopharmacology: Depression

Original Authors: Paul Perry, Ph.D, BCPP, Sam Kuperman, M.D.
Latest Revisers: Paul Perry, Ph.D, BCPP, Sam Kuperman, M.D., Brian C. Lund, Pharm.D.
Creation Date: 1996
Last Revised: June 2004
Peer Review Status: Internally Peer Reviewed


Pediatric Child Psychiatry Case Studies and Tests

Introduction
Pre-pubertal children and adolescents have episodes of depression that meet DSM-IV adult criteria for major depression that include symptoms such as crying, social withdrawal, hypersensitivity, and behavioral problems. The weighted mean prevalence rate for a current diagnosis of major depression (DSM-III or DSM-III-R) is estimated to be 0.3% (15/853) in prepubescents and 2.2% (112/5118) in adolescents (McCracken 1992, Roberts et al 1995). Though intuitive, data suggest that adults with a history of MDD have children that are at a 3-fold greater risk of MDD and 5-fold greater risk of alcoholism (Weissman et al 1997). Additionally, the presence or history of parental affective illness does not affect outcome in children unless the parental illness is untreated (Weintraub 1999).

The typical presentation of MDD in children and adolescents differs from that of adults. Ryan et al (1987) compared symptoms of children and adolescents who met core clinical criteria for depression. Young children had a more depressed appearance, more somatic complaints, a greater degree of psychomotor agitation, more phobias, separation anxiety, and hallucinations. Adolescents presented with more anhedonia, hypersomnia, hopelessness, weight change and drug abuse (Ryan et al 1987). The sex distribution of MDD also differs slightly between the age groups. That is, MDD occurs slightly more often in prepubertal boys and adolescent girls. Bipolar disorder is rare before puberty. Prepubertal grief reactions tend to be milder and briefer than do adolescent grief reactions. After puberty, a marked increase occurs in both attempts and completion of suicide, which are also often associated with depressive symptoms. The diagnosis is often obscured by the presence of comorbid psychiatric diagnoses. Kovacs (1996) notes that in contrast to adults, depressed children exhibit high rates of anxiety disorders, substance use disorder, and behavioral disorders. Suicide rates jump from around 2 (per 100,000) at ages 10-14 years to 10.9 at ages 15-19 (CDC 1995). From 1980 to 1992, suicide rates increased 28.3% in adolescents (15-19 years) and 120% in children (10-14 years). The largest part (81%) of the increase in the 15-19 year group was due to firearm-related suicide. Because the overall rate of attempts has not increased, it has been suggested that the increase in completed suicides is largely due to the increased lethality of attempts (CDC 1995).

Course of Illness
There is limited data available concerning the long-term outcome of childhood depression. One study performed follow-up assessment of 59 patients 1-5 years after an index episode of MDD (Emslie et al 1997a). All patients received some sort of treatment (81% drug therapy) and 98% of subjects recovered within 1 year of their initial evaluation. However, 61% had a recurrence of MDD at some time during the follow-up period. Of those subjects experiencing a recurrence, 47% had a recurrence within 1 year and 69% by 2 years. The authors also suggested that episodes of MDD might be briefer, but more frequent in children and adolescents than in adults. The risk of recurrence was 2.3 times greater if the patient was not receiving an antidepressant, i.e., fluoxetine (Emslie et al 1998).

TCAs - Prepubertal Children
There are multiple studies that have evaluated TCAs in the treatment of MDD in children and adolescents. A summary of these studies follows, with an outline presented in Table 1.

Kashani et al (1984) compared the effects of amitriptyline and placebo in depressed children diagnosed according to DSM-III criteria and the Bellevue Index of Depression (BID), in a fixed-dose double-blind crossover design study. Nine prepubertal children (8 males, 1 female, 9-12 years) were admitted to an inpatient child psychiatry service. Two child psychiatrists determined the diagnosis of major depressive disorder. Baseline evaluation occurred over 3-4 weeks and included complete physical work-up, blood and urine lab tests, EEG, and EKG. In phase 1 of the study, patients received a four-week trial of either amitriptyline or placebo. In phase 2, the patients received the alternate treatment. Differences in BID scores were evaluated after each part of the study. Amitriptyline was initiated at 1 mg/kg/day in three divided doses and increased to 1.5 mg/kg/day on the fourth day. This dose was maintained throughout the rest of the study. Daily dose ranged from 45-110 mg. Six of the nine patients showed a favorable response to amitriptyline (as evidenced by improvement in mood and an increase in level of interest), one child showed no significant response to either drug or placebo and two children responded to placebo. The BID score dropped below 20 (the cutoff) in 7 children on amitriptyline and 5 children receiving placebo (Yates correction, p = 0.62). No significant clinical side effects were observed during the study except one male who developed a hypomanic episode. There were no changes in EKG and blood pressure attributed to amitriptyline. As there was no washout period between the two phases during the study, the question of residual effect is important. An analysis of variance test was run to determine the significance of the residual effects, with a value of p=0.087, close to 0.05. The authors therefore concluded that residual effects were not significant. Mean BID scores decreased under drug and placebo, with the difference being insignificantly greater with drug treatment. If a variable dose and titration to maximal benefit approach had been used, a greater difference in efficacy between drug and placebo may have been achieved. Additionally, a relatively small dose of amitriptyline was used, which may have reduced the likelihood of response.

Preskorn et al (1987) studied the use of imipramine in the treatment of depression in prepubertal children. He enrolled 22 patients between the ages of 6 and 14 years for his study. Criteria for inclusion were hospitalization for major depressive disorder (DSM-III) for at least 30 days. The diagnosis was established both by open clinical interview and structured interview, using the Diagnostic Inventory for Children and Adolescents (DICA). All children had to score above 20 on the Children's Depressive Rating Scale-Revised (CDRS-R), indicating a moderately to severely depression. Patients were excluded if they had attention deficit disorder, organic brain disease, IQ less than 85, a medically unstable condition, and/or had psychotic symptoms. There was no difference in baseline scores for imipramine versus placebo treated patients. Following baseline assessment, patients were treated with placebo for four to seven days. They were randomly assigned to receive either placebo or 100 mg of imipramine at bedtime. During the first two weeks, plasma drug levels were drawn at days eight and twelve. If the level was outside of the 125-250 ng/ml range found to be maximally therapeutic according to previous data (Preskorn et al 1982), the laboratory adjusted the dose without the treatment team's knowledge. After appropriate adjustment the dose was maintained for the remaining four weeks. Imipramine was more effective (8/10) than placebo (2/12), with a difference being detectable within three weeks. No serious adverse effects were experienced when plasma levels of imipramine and its metabolite desipramine were in the 125-250 ng/ml range. Mean percentage change from baseline was reported at day 21 and day 42. CDI scores for imipramine were 59 and 40 respectively and for placebo were 25 and 38 (ANOVA < 0.10). The CDRS-R scores were 41 and 43 (on days 21 and 42, respectively) for imipramine and were 33 and 35 for placebo (ANOVA < 0.05). For CGI, values for imipramine were 28 and 38 and for placebo were 15 and 26 (ANOVA < 0.025). The most interesting finding of the paper involved dexamethasone suppression test (DST) results. According to the CDI, the response rate to imipramine was 69% among the DST non-suppressers and 50% among the DST suppressers. The response rate to placebo was 10% among the DST non-suppressers and 25% among the DST suppressers.

Puig-Antich et al (1987) studied imipramine in a five week, double-blind, placebo-controlled design. A two week pre-study diagnostic protocol included blind psychiatric (Kiddie-SADS-P), psychosocial, and pediatric assessments. Diagnosis was established by agreement of two psychiatrists using the Research Diagnostic Criteria for major depressive disorder. Exclusion criteria included concurrent use of medications that can produce depressive symptoms, significant medical illness, obesity or severe chronic malnutrition, seizures or neurological illness, IQ < 70, anorexia nervosa, autism, schizophrenia, Tanner III stage of genital/breast development. The study design was a two-week drug-free intensive diagnostic work-up period without placebo. Twenty percent improved of the originally recruited group improved and were dropped during this period. This was followed by a five-week treatment period. ECG and side effects were measured at baseline, day 3, day 6, day 9, and day 12. The day 1 imipramine dose was 1.5 mg/kg/day, day 3 dose was 3 mg/kg/day, day 6 dose was 4 mg/kg/day, and on day 9 was 5 mg/kg/day if tolerated. After day 9 the dosage remained constant through day 35. Mean dose was 4.35 + 0.61 mg/kg/day, with a range of 3.25 - 5.0 mg/kg/day. Thirty-eight (38) patients completed the study, 22 receiving placebo and 16 receiving imipramine. No other medications or psychotherapy were given during the study with the exception of one child with severe insomnia, who received one dose of 500mg chloral hydrate. This child was an imipramine non-responder. Side effects between the two groups were similar with a lower frequency for placebo. Flushed face on exercise was the only side effect occurring more frequently with imipramine. Nearly every child receiving imipramine showed minor ECG changes compared to baseline, which were reversible with dosage reduction. In determining clinical response, a child was classified as a responder if scores for both depressed mood and anhedonia were < 2 (K-SADS-P) or as a non-responder if one were > 3. The study showed no significant differences in clinical response between groups. Response rates were 56% (9/16) for imipramine and 68% (15/22) for placebo. A concurrent study did show, however, that the log mean maintenance plasma level of imipramine and desipramine strongly predicted clinical response. Response rates for total imipramine levels of ³ 150 ng/ml was 85% (17/20) and for < 150 ng/ml were 30% (3/10). Therefore, it seems that the imipramine dosage may have been too low. A fixed dosage limit may not be the best methodology, as some children require more than 5 mg/kg/day to achieve plasma levels in the therapeutic range. Also, the high placebo response rate made finding a statistical difference almost impossible. There was also no placebo washout period.

Geller et al (1989) investigated the effects of nortriptyline in depressed children in an eight-week randomized, double-blind, placebo-controlled study. Subjects randomized to nortriptyline were started on a dose calculated (from a single dose plasma level procedure) to place plasma levels a 60-100 ng/ml range. Children required doses between 10-140 mg. Subjects were assessed using the Children's Depression Rating Scale (CDRS) and a modified Asberg Side Effects Scale. EKG, blood pressure, and plasma nortriptyline levels were also monitored. All subjects were outpatients. Inclusion criteria were: RDC for major depressive disorder nondelusional type, DSM-III criteria for major depressive disorder nondelusional type, CDRS score > 40, 5-12 years of age, duration of illness > 2 months. Exclusion criteria were: IQ < 75, Tanner Stage >III, autism, childhood onset pervasive developmental disorder, major medical, psychiatric, or neurological illness, psychotropic drugs in the past month, substance abuse disorders, and an excessive fear of venipuncture. A rating of four or more on K-SADS-P MDD criteria items was needed. Subjects with CDRS scores of 25 or less during the first two weeks were considered placebo washout responders. Subjects were considered responders to treatment at eight weeks if the CDRS score was < 20, and item scores of one or two on MDD criteria items on K-SADS-P. Seventy-two subjects entered the study, 12 responded to placebo in the first two weeks, 50 completed the study phase, and 10 discontinued (none for adverse effects). Twenty-six subjects received active treatment and 24 subjects received placebo. No significant difference in response was observed between the drug (8 of 26, 30.8%) and placebo (4 of 24, 16.7%) groups. With such a small effect size, power analysis revealed the probability of finding a statistical difference with only 60 subjects to be 1 in 1000. No correlations were found between nortriptyline plasma levels and response, or between nortriptyline dose and response. There was no significant difference between groups in baseline or post-treatment side effect scores. The sample was largely melancholic, most had separation anxiety, and 96% had been ill for more than two years. The researchers also questioned whether primarily adrenergic agents, such as imipramine, were effective in children. Also the fixed plasma level range may have been too low for the children.

TCAs - Adolescents
Kramer et al (1981) studied the effects of amitriptyline in adolescent depression. Subjects were 13-17 year old inpatients. All were interviewed by the chief psychiatrist and rated on global impression of dysphoric symptoms. Psychological tests were administered to verify the diagnosis after the subject scored seven or above on the Psychiatric Rating Scale and global impressions indicated a pathological state of depression. Psychological tests used were the MMPI and the Depression Adjective Check List (DACL). Subjects must have been depressed for at least six months and nonfunctional in school. Twenty adolescents were randomized to placebo or amitriptyline. There was 1 male and nine females in the treatment group and six males and four females in the placebo group. The amitriptyline dose was started at 25 mg qid with increases of 25 mg until 200 mg/day was reached within 3 days. This dose was maintained for the rest of the study. Overall, the study showed no statistical difference in efficacy between groups and no difference in any single measure except the DACL. In this test, there was a statistical difference between groups, with the amitriptyline group demonstrating greater improvement (p < 0.001). Only one male was in the amitriptyline group and six males were in the placebo group. The small sample size was obviously a confounding factor in interpreting the results.

Geller et al (1990) also studied nortriptyline in depressed adolescents. Subjects were 12-17 years of age and postpubertal. All subjects met RDC and DSM-III criteria for MDD, nondelusional type. The duration of illness was greater than 2 months. Exclusion criteria were: IQ < 75, Tanner Stage < III, autism, childhood onset pervasive developmental disorder, major medical, psychiatric, or neurological illness, psychotropic drug use in the past month, substance use disorders, pregnancy, and excessive fear of venipuncture. Scales used monitor the subjects were Kiddy Global Assessment Scale (K-GAS) and CDRS. Subjects were begun on a dose calculated (from single dose plasma level determination) to place steady state nortriptyline levels in the 60-100 ng/ml range. The design was the same as the study conducted in children: 2 week single-blind placebo washout period and eight-week double-blind, placebo-controlled phase. Responders to active treatment were defined as subjects with CDRS score < 25 and score < 2 on DSM-III criteria items on K-SADS-P. Only the concentration item was allowed to be < 3. Fifty-two subjects were recruited. Seventeen were placebo washout responders, 12 were randomized to active treatment and 19 were randomized to placebo. Four subjects dropped out. Thirty-one subjects completed the study with only one subject on active medication responding. All subjects had plasma levels of nortriptyline between the range of 60-100 ng/ml and the single responder was within the range of the rest of the sample. The study showed a significant correlation between nortriptyline plasma levels and the final CDRS score. Unfortunately, the patients with higher plasma did worse. The most common side effects reported were fatigue, sleep disturbances, and headaches, all of which can also be symptoms of depression. No anticholinergic side effects were reported. No significant differences in blood pressure and EKG except an increase in heart rate in active drug were noted. Subjects all scored in the severe range for depression, were largely endogenous, melancholic, and had a high comorbidity. The results from this study were similar to Geller's (1989) previous study with prepubertal children and may have suffered from the same methodological.

Boulos et al (1991) evaluated the effectiveness of desipramine in the treatment of adolescents (15-20 yo) in depressed adolescents. All subjects met DSM-III-R criteria for MDD. Fifty-two children with minimal acceptable HAM-D scores of ³ 17 and Beck Depression Inventory (BDI) scores of ³ 16 were entered into a one-week placebo-lead-in. Co-morbid diagnoses included 47% anxiety disorders, 20% substance abuse, and 7% conduct disorder. Forty-three of the subjects were randomized to treatment with either placebo or desipramine 200 mg/d for six weeks. Response was defined as either a ³ 50% or < 50% decrease in the HAM-D score. Using an intent to treat analysis, 11 of 22 (50%) subjects randomized to desipramine responded, while 33% 7/21 of the placebo treated group responded. This was not a significant difference (c2, p<0.59). However, the response rates were quite similar to those observed for adults thereby suggesting a type II error. Unfortunately no blood levels were obtained from the patients for the data analysis. The authors posited four hypotheses to explain the lack of response of depressed adolescents to TCA. These include the following:

Kutcher et al (1994) investigated the efficacy of desipramine in a group of 15-20 year old adolescents diagnosed with MDD (DSM-III-R). Exclusion criteria included diagnoses of schizophrenia, schizoaffective disorder, bipolar affective disorder, anorexia or bulimia nervosa, drug abuse, and any organic psychiatric disorder. Inclusion criteria required a baseline HDRS of ³ 17 and a BDI of ³ 16. Eligible subjects had to complete one-week placebo lead-in and still meet entry criteria. Sixty subjects were randomized to desipramine 200 mg/d or placebo for 6 weeks. Eighteen subjects dropped out and 42 subjects completed the trial. Response was defined as a decrease in the HDRS at the end of six week of ³ 50%. There was no significant difference in response between desipramine (8/17, 48%) and placebo (9/25, 35%). There were no significant differences in the BDI change scores. The mean serum desipramine concentration was 205 ng/ml. There was no relationship between desipramine concentration and response.

TCAs - Meta-analysis
Hazell et al (1995) performed a meta-analysis on 12 randomized controlled trials that evaluated the efficacy of the TCA versus placebo in depressed subjects 6-18 years old. Their analysis concluded that TCA appear to be no more effective than placebo in the treatment of depression in children and adolescents.

We have identified multiple studies that have evaluated the effectiveness of TCAs in the treatment of childhood and adolescent depression. A summary of these studies is presented in Table 1. Only one (Preskorn et al 1987) of the eight studies found the TCA to be significantly more effective than placebo. We also performed a meta-analysis on these data using an intent to treat method. The data are summarized in Table 2. There is overall approximately a 36% response rate to TCAs in depressed children and adolescents. The placebo response rate was 31%, a difference of only 5% between the two rates. This is not a significant difference. It is important to note that all but one of the studies reviewed used standardized criteria for establishing the MDD diagnosis (e.g. DSM-III-R, DSM-III, or RDC criteria). Antidepressant trials in adult populations consistently yield greater response rates, with a similar placebo response rate. In adults, the overall response rate in inpatients and outpatients was 51% for the TCAs and 23% rate for placebo (Depression Guideline Panel 1993).

Venlafaxine
Mandoki et al (1997) treated 40 adolescents and children (8-17 years old) diagnosed with major depression (DSM-IV) with either venlafaxine plus cognitive behavioral therapy (CBT) or placebo plus CBT. The venlafaxine dose in the 8-12 year olds was 37.5 mg/d whereas in the 13-17 year olds was 75 mg/d. Based on clinician ratings (HAMD for > 12 year olds and CDRS for < 12 yo), parent ratings (Children's Behavior Checklist) and patient ratings (Children Depression Inventory) all subjects improved regardless of treatment. Since venlafaxine is classified pharmacologically as a norepinephrine-serotonin reuptake inhibitor similar to the TCA it is not surprising that the drug was ineffective. What is interesting however is the beneficial effect of CBT in these patients regardless of the presence of the drug therapy.

Lithium Augmentation of TCAs
Strober et al (1992) examined the effect of lithium augmentation of imipramine in 34 treatment refractory adolescent MDD (DSM-III or DSM-III-R) patients. Twenty-four patients were administered lithium 300 mg po tid for 3 weeks (mean = 0.89 mEq/l) while 10 patients used as controls continued their imipramine for an additional 3 weeks. Prior to the start of lithium, imipramine was administered for 6 weeks with a mean dose of 229 mg/d (mean concentration = 251 ng/ml). No significant change in the HDRS scores was observed between imipramine and the augmentation group. Clinical global improvement was observed in 10/24 (42%) of the augmentation group versus only 1/10 (10%) of the control group (p=0.11). A larger sample size of 35/group would have found this effect size significant. Although not statistically significant, it appears that lithium augmentation may be of some help in treatment refractory patients as compared to simply three additional week of antidepressant therapy.

TCA conclusions
After the review of existing data it is not possible to recommend the use of TCA in the treatment of MDD in children and adolescents. Studies with larger sample sizes are necessary. However, these studies are difficult to conduct since many centers simply do not diagnose large numbers of adolescents and children with major depression. Future studies utilizing concentration monitoring and dose adjustment may yield higher response rates with the TCA. At this time however, drug treatment for depression appears to be no more effective than placebo in children and adolescents.

Intravenous Clomipramine
Intravenous (IV) clomipramine has been shown to produce a rapid antidepressant effect in depressed adolescents (Sallee et al 1997). Sixteen patients between age 14-18 years were randomized to receive either IV saline (placebo, N=8) or IV clomipramine 200mg/d (N=8). Clomipramine administration yielded rapid reductions in HAMD scores (22.9 at baseline, 11.0 at 36 hours, 7.9 at 6 days) which became significantly different than placebo at 6 days. Using a definition of 50% reduction in HAMD, 88% of patients receiving clomipramine responded by day 6, versus only 38% of patients receiving placebo. The response rates after 14 days of treatment were 100% and 13% for clomipramine and placebo, respectively. Although not commercially available in the US, this successful therapy suggests that a TCA like clomipramine with potent serotonin agonist activity may be more likely to benefit childhood and adolescent depressions.

SSRIs
It has been hypothesized that childhood and adolescent depression are associated with CNS changes to the serotonergic system and therefore SSRI may be more appropriate drug for the treatment of this patient population. Normally, L-5-hydroxytryptophan (L-5HT) stimulation of the central serotonergic system results in an augmentation of cortisol release. Ryan et al (1992) compared the neuroendocrine response to L-5-hydroxytryptophan (L-5HT) in 37 prepubertal children who met the RDC for MDD with that in 23 normal children with no lifetime history of any psychiatric disorder and very low rates of depression in both first- and second-degree relatives. Intravenous L-5HT (0.8 mg/kg) was given over a 1-hour interval after preloading with oral carbidopa, an inhibitor of peripheral but not central L-5HT metabolism. L-5HT, a precursor of serotonin, increases serotonin turnover in the CNS when given after carbidopa. After this stimulation, cortisol secretion was compared between diagnostic groups. The depressed children secreted significantly less cortisol (effect size, 0.70). A blunted cortisol response in depressed prepubertal children exactly parallels the cortisol response observed by Meltzer et al (1990) after MK-212 administration and Weizman et al (1988) after fenfluramine administration in depressed adults. These findings are consistent with dysregulation of central serotonergic systems in childhood major depression.

SSRIs - Controlled Studies
There are five published randomized controlled trials of SSRI in the treatment of major depression in children and adolescents. There are two published randomized controlled trials of SSRIs in the treatment of major depression in children and adolescents. These trials are summarized in Table 3.

Emslie et al (1997b) conducted the first randomized controlled study that demonstrated the efficacy of a SSRI antidepressant in the treatment of major depression. Following a 3-week placebo lead-in, randomized 96 MDD (DSM-III-R) children and adolescents to an 8-week treatment trial with either fluoxetine 20 mg/d or placebo. Eighty percent of the subjects had a comorbid diagnosis of anxiety disorders, ADHD, or dysthymia. There were 14 fluoxetine dropouts and 22 placebo dropouts. According to the CGI, 56% (27/48) of the fluoxetine group and 33% (16/48) of the placebo group were rated as much or very much improved. The CDRS demonstrated a significant difference favoring fluoxetine initially at week 5. Interestingly, the children global assessment scale (KGAS) showed no difference between the treatments. This may be a function of the diagnostic comorbidity of the population. There are currently five published randomized controlled trials of SSRI in the treatment of major depression in children and adolescents. These trials are summarized in Table 3.

Jureidini et al (2004) reviewed 6 studies involving paroxetine, fluoxetine, sertraline or venlafaxine involving 477 patients who were contrasted with 464 patients treated with placebo. The authors noted that of 42 reported measures of which only 14 showed a statistical advantage for the antidepressant. Discouragingly of the patient reported or parent reported outcomes none showed a significant advantage for the antidepressant . After pooling the depression and quality of life outcome measures between the five studies the effect size between the antidepressants and placebo was a modest 0.26 (0.13-0.40, 95%CI). High rates of withdrawal occurred in all the studies, ranging from 17-32% for SSRI treated patients and from 17-46% for placebo treated patients. The authors also commented that the number of serious adverse events despite the short duration of the studies showed significant differences between the placebo treated and antidepressant group. As an example in the Keller study (2001) the 12% rate for paroxetine (11/93) vs 2% for placebo (2/87). Whittington et al (2004) performed a meta-analysis of data from randomized controlled trials that evaluated SSRI versus placebo treated patient (5-18 years old) that were published in peer-reviewed journals or were unpublished and included in a review by the UK Committee on Safety of Medicines. With respect to fluoxetine the authors concluded there was evidence of efficacy in two published studies that was confirmed in an unpublished study; no increased risk of serious adverse effects; and a favorable risk-benefit profile. With respect to paroxetine the unpublished data when added to the published data suggest the weak/equivocal clinical benefit is outweighed by the risk of its use. After combining the published and unpublished data it was concluded that there is little evidence for efficacy and serious risk for adverse events. The data from unpublished efficacy trials of citalopram and venlafaxine suggested that there is little support to demonstrate the benefit of an antidepressant effect in this patient population leading to conclusion that the risk to benefit profile was unfavorable.

SSRI conclusions
Based on published data, SSRI are the preferred antidepressant agents for the treatment of childhood/adolescent MDD. However, when unpublished data and published trials are considered in toto only fluoxetine appears to generates enough benefit to outweigh the risk of serious adverse effects. Fluoxetine is the only antidepressant that currently that can be recommended for the treatment of adolescent/childhood major depression.

MAOIs
One retrospective chart review (N=23) has found MAOI to be useful in the treatment of adolescent major depression (Ryan et al 1988). Compliance with the dietary restrictions associated with the clinical use of MAOI is always a concern, particularly in children.

Safety
In 1988, fluoxetine was introduced on to the US Market. Soon thereafter case reports of increased suicidal tendencies surfaced. A task force report published in 1993 by the American College of Neuropharmacology concluded that there was no evidence of SSRI triggering suicidal behavior. However, in 2003 the UK Department of Health Committee on the Safety of Medicines, an independent committee of scientific experts that advise the Medicines and Healthcare Products Regulatory Agency undertook a review of SSRI in the treatment of childhood/adolescent MDD (Olfson et al 2003, Whittington et al 2004). At the conclusion of their review they issued a warning against prescribing any SSRI except fluoxetine as new treatment for patients less than 18 years old for depressive illness because of increases in suicidal ideation, attempts or behavior. In 2003, ACNP repeated their review of the published literature and again concluded no increased risk of suicide was associated with the SSRI usage in the treatment population (ACNP, 2004). However, they were not privy to none unpublished studies that the UK Committee on Safety of Drugs had in their possession in their analysis of the data. The FDA (2004) followed suit in early 2004 when it issued a Public Health Advisory in which manufacturers were instructed to include a warning statement in product package insert advising clinicians to remain vigilant for worsening of patient's condition manifested by the development or worsening of suicidal ideation. Manufacturers were instructed to change labels of ten drugs to include stronger warnings and cautions regarding the potential for increased suicidal ideation in this patient population when being treated for MDD. The ten antidepressants included paroxetine, fluvoxamine, citalopram, escitalopram, fluoxetine, sertraline, venlafaxine, bupropion, nefazodone, and mirtazepine. The FDA recommended further study of this problem at this time. Thus the issue remains contentious and unresolved at this time. Two important observations need to be noted. First, there were no completed suicides in the SSRI treated patients in the published studies. Second, the suicide rates for the 15-24 year old age group in the US has decreased in the past 12 years from 13.2/100,000 in 1990 to 9.9/100,000 in 2001 (Arias et al 2003). These dates approximately coincide with the availability of SSRI antidepressants in the US.

Summary
A number of relevant clinical observations regarding the treatment of childhood and adult depressions can be distilled from the literature. Additionally, the Texas Children's Medication Algorithm Project has developed a treatment algorithm that is based on child, adolescent, and adult data. The algorithm can be viewed at the following website Hughes et al (1999): www.mhmr.state.tx.us/centraloffice/medicaldirector/mddpage.html.

Table 1. Controlled studies of antidepressants in prepubertal depressed children.

Study

Dose (mg/d)

Population

Results

Kramer et al 1981

AMT 200 mg/d x 6 weeks or placebo

No criteria dysphoria
x 6 months
15 males
5 females
13-17 yo

Moderate/Maximal improvement AMT (8/10) = placebo (6/10)

Kashani et al 1984

AMT 1.5 mg/kg/d (45-110 mg/d) & placebo x 28 days crossover

MDD (DSM-III)
8 males
1 females
9-12 yo

6/9 AMT and 4/9 PLB responders according to BID
(p=NS)
ADRs: 1 hypomanic episode

Preskorn et al 1987

IMP 100 mg/d x 2 wks, then 125-250 ng/ml x 4 wks or placebo

MDD x 30 d (DSM-III)
22 children
6-14 yo

Response; IMP (8/10) PLB (2/12) according to CGI & CDRS-R
ADRs: none significant

Puig-Antich et al 1987

IMP 3.3-5.0 mg/kg/d x 5 wks or placebo

MDD x 30 d (DSM-III)
38 children
6-14 yo

Response: IMP(9/16) PLB (15/22) according to K-SADS
Response rate was blood level related;
> or = 150 ng/ml (85%) versus < 150 ng/ml (30%)

Geller et al 1989

NTP 60-100 ng/ml, i.e., 10-140 mg/d x or placebo x 8 wks

MDD
(DSM-III)
35 males
15 females
5-12 yo

Response: NTP (8/26) PLB (4/24) according to K-SADS-P & CDRS

Geller et al 1990

NTP 60-100 ng/ml, i.e., 10-140 mg/d x or placebo x 8 wks

MDD (RDC & DSM-III) s 2 months
17 males
14 females
12-17 yo

Response: NTP (1/12) PLB (4/19) according to Kiddy-GAS & CDRS

Boulos et al 1991

DIP 200 mg/d x 6 weeks or placebo

MDD (DSM-III-R)
43 randomized 30 completers 15-20 yo

Intent to treat analysis (> or = 50% decreased HAM-D = response)
DIP 50% response vs placebo 33% response 6 DIP dropouts due to ADRs

Kutcher et al 1994

DIP 200 mg/d x 6 weeks or placebo

MDD (DSM-III-R) nondelusional 60 randomized 42 completers 18 males
42 females
6-12 yo

Response: DIP (8/17) PLB (9/25) according HDRS

HDRS: NTP = 48%; placebo 36%

Table 2. Meta-analysis data to calculate difference between TCA and placebo response rates in the treatment of childhood depressions

Study

Population

TCA

dropouts

TCA response rate

placebo response rate

Kashani 1984

children

amitriptyline

0
7/9
5/9

Puig-Antich 1987

children

imipramine

0
8/10
2/12

Preskorn 1987

children

imipramine

0*
9/16
15/22

Geller 1989#

children

nortriptyline

10
8/31
4/29

Kramer 1981

adolescents

amitriptyline

0
5/10
1/10

Geller 1990#

adolescents

nortriptyline

4
1/14
4/21

Boulous 1991

adolescents

desipramine

13
11/22
7/21

Kutcher 1994

adolescents

desipramine

18
8/30
9/30

Totalsa

48/132 (36%)
47/154 (31%)
* not noted
# dropouts noted but not noted as to which group; thus dropouts were divided between the two treatment groups
aX2 = 1.4594, p = 0.23

Table 3. Controlled studies of SSRIs in childhood depression.

Study

Dose (mg/d)

Population

Results

Simeon 1990

Fluoxetine 60 mg/d x 8 weeks or placebo

MDD (DSM-III) 16

Response on HAMD: intent-to-treat: 50% with fluoxetine. Subjects showed significant decreases in HAMD with fluoxetine, but at no point was this treatment significantly different than placebo.

Emslie 1997

Fluoxetine 20mg/d x 10 weeks or placebo

MDD (DSM-III-R)
96 randomized
70 completers
7-17 yo

Response: Intent-to-treat 56% (27/48) of the fluoxetine group and 33% (16/48) of the placebo by CGI improvement of 1 (very much improved) or 2 (very much improved)

Keller 2001

Paroxetine 20-40 mg/d, imipramine 200-300 mg/d or placebo x 8 weeks

MDD (DSM-IV)
N=275
13-18 yo

Response: HAMD (intent-to-treat) < 9, paroxetine (63%) = imipramine (50%) > placebo (46%)

Emslie 2002

Fluoxetine 10mg/d x 1 wk and 20 mg/d x 8 wk or placebo

MDD (DSM-III-R)
122 children
97 adolescents

Response

  • CDRS: fluoxetine > placebo@ wk 1-9
  • Remission rate: fluoxetine > placebo
  • Response rate: fluoxetine = placebo

Wagner 2003 (2 studies)

MDD (DSM-IV)
N=376
6-17 yo

Response

  • Sertraline = placebo
  • Sertraline = more serious ADRs 3.7% vs 3.3% and suicide attempts/ideation 2.6% vs 1.1%

REFERENCES

American College of Neuropsychopharmacology (2004). Executive Summary: Preliminary Report of the Task Force on SSRIs and Suicide Behavior in Youth,. http://www.ACNP.org/exec_summary.pdf (accessed 27 April 2004).

Arias et al (2003) National Vital Statistics Reports 52(3), DHHS Pub No 2003-1120

Boulos C, Kutcher S, Marton P, et al (1991). Response to desipramine treatment in adolescent major depression. Psychopharmacol Bull 27:59-65.

Centers for Disease Control (1995). Suicide among children, adolescents, and young adults&endash;United States, 1980-1992. MMWR Morb Mortal Wkly Rep 44:289-91.

Depression Guideline Panel (1993). Depression in Primary Care: Volume 2. Treatment of Major Depression. Clinical Practice Guideline, Number 5. Rockville, MD. U.S. Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research, AHCPR Publication No. 93-0551. April.

Duff G (2003). Selective Serotonin Reuptake Inhibitors: Use in Children and Adolescents with Major Depressive Disorder, 2003. http://medicines.mhra.gov.uk/ourwork/ pdf (accessed 27 April 2004).

Emslie GJ, Rush AF, Weinberg WA et al (1997a). Recurrence of major depressive disorder in hospitalized children and adolescents. . J Am Acad Child Adolesc Psychiatry 36:785-92.

Emslie GJ, Rush AJ, Weinberg WA, et al (1997b). A double-blind, randomized, placebo-controlled trial of fluoxetine in children and adolescents with depression. Arch Gen Psychiatry 54:1031-7.

Emslie GJ, Rush AJ, Weinberg WA, et al (1998). Fluoxetine in child and adolescent depression: Acute and maintenance treatment. Depress Anxiety 7:32-9.

Emslie GJ, Heiligenstein JH, Wagner KD, et al (2002). Fluoxetine for acute treatment of depression in children and adolescents: a placebo-controlled randomized clinical trial. J Am Acad Child Adoles Psychiatry 41:1205-15.

FDA Public Health Advisory (2004). Worsening Depression and Suicidality in Patients Being Treated with Antidepressant Medications. http://www.fda.gov/cder/drug/antidepressants/AntidepressantstPHA.htm (accessed 25 April 2004).

Geller B, Cooper TB, McCombs HG, et al (1989). Double-blind, placebo-controlled study of nortriptyline in depressed children using a "fixed plasma level" design. Psychopharmacol Bull 25,1 :101-7.

Geller B, Cooper TB, Graham DL, et al (1990). Double-blind placebo-controlled study of nortriptyline in depressed adolescents using a "fixed plasma level" design. Psychopharmacol Bull 26:85-90.

Hazell P, O'Connell D, Heathcote D, et al (1995). Efficacy of tricyclic drugs in treating child and adolescent depression: a meta-analysis. BMJ 310:897-901.

Hughes CW, Emslie GJ, Crismon ML, et al (1999) The Texas Children's Medication Algorithm Project: report of the Texas Consensus Conference Panel on Medication Treatment of Childhood Major Depressive Disorder. J Am Acad Child Adolesc Psychiatry. 38:1442-54.

Jureidini JN, Doecke CJ, Mansfield PR, et al (2004). Efficacy and safety of antidepressants for children and adolescents. BMJ 328:879-83.

Kashani JH, Shekim WO, Reid JC. (1984). Amitriptyline in children with major depressive disorder: a double-blind crossover pilot study. J Am Acad Child Psychiatry 23:348-51.

Kovacs M, Goldston, Gatsonis C (1993). Suicidal behaviors and childhood-onset depressive disorders: a longitudinal investigation. J Am Acad Child Adolesc Psychiatry 32:8-20.

Keller MB, Ryan ND, Strober M, et al (2001). Efficacy of paroxetine in the treatment of adolescent major depression: a randomized controlled trial. J Am Acad Child Adolesc Psychiatry 40:762-72

Kovacs M (1996). Presentation and course of major depressive disorder during childhood and later years of the life span. J Am Acad Child Adolesc Psychiatry 35:705-15.

Kramer AD, Feiguine RJ (1981). Clinical effects of amitriptyline in adolescent depression. J Am Acad Child Adolesc Psychiatry 20:636-44.

Kutcher S, Boulos C, Ward B, et al (1994). Response to desipramine treatment in adolescent depression: a fixed-dose. placebo-controlled trial. J Am Acad Child Psychiatry 33:686-94.

Mandoki MW, Tapla MR, Sumner GS, et al (1997). Venlafaxine in the treatment of children and adolescents with major depression. Psychopharmacol Bull 1997;33:149-54.

Meltzer HY, Schreiber M, Bastani, et al (1990). Is serotonergic function really abnormal in major depression? Presented at the XVII Congress of Collegium Internationale Neuro-Psychopharmacologicum, September 11.

Mendels J, Ramsey TA, Dyson Wl, et al (1979). Lithium as an antidepressant. Arch Gen Psychiatry 36:245-54.

McCracken JT (1992). The epidemiology of child and adolescent mood disorders. Child Adolesc Psychiatric Clin N Am 1:53-72.

McCracken Jt, Cantwell DP (1992). Management of child and adolescent mood disorders. Child Adolesc Psychiatric Clin N Am 1:229-55.

Olfson M, Shaffer D, Marcus SC, et al (2003). Relationship Between Antidepressant Medication Treatment and Suicide in Adolescents. Arch Gen Psychiatry.60;978-82.

Preskorn SH, Weller EB, Weller R (1982). Depression in children: relationship between plasma imipramine levels and response. J Clin Psychiatry 43:450-3.

Preskorn SH, Weller EB, Hughes CW, et al (1987). Depression in prepubertal children: dexamethasone nonsuppression predicts differential response to imipramine vs. placebo. Psychopharmacol Bull 23:128-33.

Puig-Antich J, Perel JM, Lupatkin W, et al (1987). Imipramine in prepubertal major depressive disorders. Arch Gen Psychiatry 44:81-9.

Rao U, Weissman M, Martin J, et al (1993). Childhood depression and risk of suicide: a preliminary report of a longitudinal study. Am Acad Child Adolesc Psychiatry 32:21-27.

Roberts RE, Lewinsohn PM, Seeley JR (1995). Symptoms of DSM-III-R major depression in adolescence: evidence from an epidemiological survey. J Am Acad Child Adolesc Psychiatry 34:1608-17.

Ryan ND, Puig-Antich J, Ambrosini P, et al (1987). The clinical picture of major depression in children and adolescents. Arch Gen Psychiatry 44:854-61.

Ryan ND, Puig-Antich J, Rabinovich H, et al (1988). MAOI in adolescent major depression unresponsive to tricyclic antidepressants. J Am Acad Child Adolesc Psychiatry 27:755-8.

Ryan ND, Birmaher B, Perel JM, et al (1992). Neuroendocrine response to L-5-hydroxytryptophan challenge in prepubertal major depression. Depressed vs normal children. Arch Gen Psychiatry 49:843-51.

Sallee FR, Vrindavanam NS, Deas-Nesmith D et al (1997). Pulse intravenous clomipramine for depressed adolescents: a double-blind controlled trial. Am J Psychiatry 154:668-73.

Simenon J, Dinicola V, Phil M, et al (1990). Adolescent depression: a placebo-controlled fluoxetine treatment study and follow-up. Prog Neuropsychopharmacol Biol Psychiatry 14:791-5.

Strober M, Freeman R, Rigali J, et al (1992). The pharmacotherapy of depressive illness in adolescence: II. Effects of lithium augmentation in nonresponders to imipramine. J Am Acad Child Adolesc Psychiatry 31:16-20.

Wagner KD, Ambrosini P, Rynn M, et al (2003) Efficacy of sertraline in the treatment of children and adlescent with major depressive disorder: two randomized controlled trials. JAMA 290-1033-41.

Weintraub P (1999). Signficance of expressed emotion in drug trial of adolescent depression. Presented at the New Clinical Drug Evaluation Unit Program (NCDEU) 39th Annual Meeting, Boca Raton, FL.

Weissman MM, Warner V, Wickramaratne P et al (1997). Offspring of depressed parents: 10 years later. Arch Gen Psychiatry 54:932-4.

Weizman A, Mark M, Gil-Ad I, et al (1988). Plasma cortisol, prolactin, growth hormone and immunoreactive beta-endorphin response to fenfluramine challenge in depressed patients. Clin Neuropharmacol 11:250-6.

Whittington CJ, Kendall T, Fonagy P (2004). Selective serotonin reuptake inhibitors in childhood depression: systematic review of published versus unpublished data. Lancet 363:1341-5.  

Title Page