Karen Dineen Wagner, M.D., Ph.D.
Steven R. Pliszka, M.D.
This chapter focuses on the use of psychopharmacology for treating psychiatric disorders in children and adolescents. It is important for clinicians to be aware of the evidence base for the use of psychotropic medications in children and adolescents. In this chapter, data from the literature, with a focus on controlled studies, are presented. On the basis of these findings, clinical recommendations regarding pharmacotherapy for childhood psychiatric disorders are offered. The appendix and tables at the end of this chapter contain specific information about dosages, monitoring, and adverse effects of psychotropic medications in children.
Prior to the initiation of psychotropic medication for children and adolescents, a comprehensive evaluation is necessary to ensure the accuracy of the diagnosis. A thorough history and careful attention to the clinical presentation are central components of the evaluation. The clinician should interview the child and parents separately so that both may have the opportunity to freely express their concerns. Extended-family members, school personnel, and school records are other potential sources of information.
Clinicians must be skilled at differential diagnosis of childhood disorders, given that there is a significant overlap of symptoms among these disorders. Knowledge of commonly occurring comorbid disorders is also necessary. Medical conditions should be considered within the differential diagnosis and adequately assessed.
Disorder-specific rating scales at baseline and during the course of treatment may be useful in assisting with the measurement of clinical outcome.
Whenever a child fails to respond to initial pharmacotherapy, several clinical issues should be addressed before initiating alternative or adjunctive medication.
The diagnosis should be reassessed. Often there is symptomatic overlap among disorders that may lead to misdiagnosis. For example, symptoms of excessive energy and distractibility are common features of both attention-deficit/hyperactivity disorder (ADHD) and bipolar disorder.
Unrecognized comorbid disorders may adversely affect treatment outcome. As an example, comorbid ADHD may lower response rates in the treatment of bipolar disorder (Pavuluri et al. 2006).
Child abuse, domestic violence, family conflict, parental psychopathology, and bullying by peers may lead to symptoms that mimic or exacerbate a preexisting psychiatric disorder.
Some children and adolescents are reluctant to take medication because of such reasons as denial of illness, perceived stigma, and side effects. For increased medication compliance, the child or adolescent, as well as the parents, must understand the youth’s disorder, course of illness, and goals of treatment. It is important for parents to participate in monitoring their child’s medication compliance.
Psychotherapy may be beneficial, either alone or in conjunction with medication. Specific psychotherapies have been found to be effective in the treatment of some childhood disorders. For example, cognitive-behavioral therapy (CBT) and interpersonal therapy have demonstrated efficacy in the treatment of adolescents with depression (Zhou et al. 2015). Similarly, CBT is commonly used for the treatment of childhood anxiety disorders (Roblek and Piacentini 2005). Behavior therapy has led to improvement in symptoms of ADHD for children (Pelham et al. 1998). Adjunctive psychoeducation with medication treatment has shown benefit in the treatment of children with bipolar disorder (Fristad et al. 2003). Social skills training can be a useful component of treatment in autism spectrum disorder (Krasny et al. 2003).
Informed consent is necessary prior to prescribing psychotropic medication to any patient, but it is particularly important in pediatric psychopharmacology because few medications have been approved by the U.S. Food and Drug Administration (FDA) for young populations. There are five recommended components of informed consent for prescribing psychotropic medications to children and adolescents (Popper 1987). The child’s parent(s) and the child or adolescent should be provided with the following information:
The purpose (benefits) of the treatment
A description of the treatment process
An explanation of the risks of the treatment, including risks that would ordinarily be described by the psychiatrist and risks that would be relevant to making the decision
A statement of alternative treatments, including nontreatment
A statement that there may be unknown risks associated with these medications (this is particularly essential for children due to the paucity of information on the potential long-term effects of psychotropic medications)
It is important for clinicians to be aware of the evidence base for medication treatment of each childhood psychiatric disorder. Clinical treatment guidelines generally rely on the strength of the available data in determining first-line agents. In most cases, clinicians should select a medication within the group of first-line agents when initiating medication treatment with a child. Additional factors that will dictate medication choice are prior medication history, medical history, side-effect profile of the drug, and adolescent and parent preferences.
The lifetime prevalence of major depression or dysthymia in adolescents is 11.7% (Merikangas et al. 2010). DSM-5 (American Psychiatric Association 2013) criteria are used to establish a diagnosis of major depression in children and adolescents. The mean length of an episode of major depression in youth ranges from 8 to 13 months, and relapse rates range from 30% to 70% (Birmaher et al. 2002). There is increasing evidence for the continuity of depression from youth into adulthood (Melvin et al. 2013).
For the acute and maintenance treatment of major depression in children and adolescents, the FDA has approved two selective serotonin reuptake inhibitor (SSRI) medications: fluoxetine for patients ages 8 years and older and escitalopram for patients ages 12 years and older. Other SSRIs that have been studied in controlled trials for children and adolescents with major depression include citalopram, sertraline, and paroxetine.
There have been three positive controlled medication trials of fluoxetine in the treatment of major depression in children and adolescents. In the first study, 96 child and adolescent outpatients (ages 8–17 years) with major depression were randomly assigned to fluoxetine 20 mg/day or placebo for an 8-week trial (Emslie et al. 1997). The fluoxetine group, with 27 youths (56%) much or very much improved, showed statistically significantly greater improvement in Clinical Global Impression (CGI) Scale scores than the placebo group, with 16 youths (33%) much or very much improved. Medication side effects leading to discontinuation in the study were manic symptoms in three patients and severe rash in one patient.
In a double-blind, placebo-controlled multicenter study of fluoxetine, 219 child and adolescent outpatients (ages 8–17 years) with major depression were randomly assigned to fluoxetine 20 mg/day or placebo for an 8-week trial (Emslie et al. 2002). Improvement in depression was statistically significantly greater for the fluoxetine group, as assessed by means of the Children’s Depression Rating Scale—Revised (CDRS-R), than for the placebo group. Fifty-two percent of patients treated with fluoxetine were rated as much or very much improved, compared with 37% of patients treated with placebo. Headache was the only side effect that was reported more frequently in the group treated with fluoxetine than in the group treated with placebo.
In a multicenter trial of 439 adolescent outpatients with a diagnosis of major depression (March et al. 2004), patients were randomly assigned to 12 weeks of fluoxetine (10–40 mg/day), fluoxetine (10–40 mg/day) with CBT, CBT alone, or placebo. Compared with placebo, the combination of fluoxetine with CBT was significantly superior, according to changes in CDRS-R scores. Combination treatment with fluoxetine and CBT was significantly superior to treatment with fluoxetine alone or CBT alone. Fluoxetine monotherapy was superior to CBT. Response rates (defined as CGI ratings of much or very much improved), were 71% for fluoxetine–CBT combination therapy, 61% for fluoxetine, 43% for CBT, and 35% for placebo.
There have been two controlled trials of escitalopram in the treatment of youth with depression, one positive and one negative. The efficacy of escitalopram in adolescents was demonstrated in a double-blind, placebo-controlled multicenter trial of 157 adolescents with major depression (Emslie et al. 2009). Patients were randomly assigned to escitalopram (dosage range=10–20 mg/day) or placebo for an 8-week trial. The group treated with escitalopram showed statistically significantly greater improvement in depression (CDRS-R scores) than the placebo group. Sixty-four percent of escitalopram-treated patients were much or very much improved, compared with 53% of placebo-treated patients. In a study that included 264 children and adolescents with major depression (Wagner et al. 2006a), there was no significant improvement on CDRS-R scores at endpoint between escitalopram and placebo. The most common adverse events with escitalopram were headache, abdominal pain, nausea, insomnia, and menstrual cramps.
There have been two controlled trials of citalopram in the treatment of depression in youth, one with positive and one with negative results. The efficacy of citalopram was demonstrated in a double-blind, placebo-controlled multicenter trial of 174 outpatient children and adolescents (ages 7–17 years) with major depression (Wagner et al. 2004b). Patients were randomly assigned to receive citalopram (dosage range=20–40 mg/day; mean daily dosage=23 mg for children, 24 mg for adolescents) or placebo for an 8-week trial. Compared with the placebo group, the citalopram group showed statistically significantly greater improvement in depression (CDRS-R scores). The most frequent adverse events were headache, nausea, rhinitis, abdominal pain, and influenza-like symptoms. A European double-blind, placebo-controlled multicenter study (von Knorring et al. 2006) of citalopram in 224 adolescents with major depression failed to show superiority of citalopram to placebo on the primary efficacy measures, the Schedule for Affective Disorders and Schizophrenia for School-Aged Children—Present Episode Version (Kiddie-SADS-P) and the Montgomery-Åsberg Depression Rating Scale (MADRS). The most commonly reported adverse events were headache, nausea, and vomiting.
The efficacy of sertraline was assessed in two identical double-blind, placebo-controlled multicenter studies of 376 outpatient children and adolescents with major depression (Wagner et al. 2003a). Patients were randomly assigned to sertraline (dosage range=50–200 mg/day; mean daily dosage=131 mg) or placebo for a 10-week trial. Compared with the placebo group, the group receiving sertraline showed a statistically significantly greater improvement in depression (CDRS-R scores). Response rates (decrease ≥40% in baseline CDRS-R scores) were 69% in the group treated with sertraline and 59% in the group treated with placebo. The most common side effects in the group treated with sertraline were headache, nausea, insomnia, upper respiratory tract infection, abdominal pain, and diarrhea.
Sertraline, CBT, and combined CBT plus sertraline were compared in the treatment of 73 adolescents with depressive disorders (Melvin et al. 2006). All treatments resulted in statistically significant improvement on all outcome measures; there were no significant advantages of combined treatment.
In the Adolescent Depression and Psychotherapy Trial (ADAPT; Goodyer et al. 2008), 208 adolescents with major depression were randomly assigned to receive an SSRI alone or an SSRI plus CBT for 12 weeks. No significant differences were found between the groups; 44% of the SSRI-alone group and 42% of the SSRI-plus-CBT group were much or very much improved on the CGI Improvement subscale (CGI-I).
There have been three double-blind, placebo-controlled trials of paroxetine for treatment of depression in children and adolescents, all of which had negative findings on the primary outcome measure. In a study of 275 adolescent outpatients (ages 12–18 years) with major depression, patients were randomly assigned to paroxetine (dosage range=20–40 mg/day; mean daily dosage=28 mg), imipramine (dosage range=200–300 mg/day; mean daily dosage=205 mg), or placebo for an 8-week trial (Keller et al. 2001). There was no statistically significant difference among the treatment groups on the primary efficacy measure of reduction in total score on the Hamilton Rating Scale for Depression (Ham-D). The most common side effects reported for paroxetine were headache, nausea, dizziness, dry mouth, and somnolence.
Two hundred six children and adolescents (ages 7–17 years) with major depression were included in an 8-week double-blind, placebo-controlled, randomized multicenter study of paroxetine treatment (Emslie et al. 2006). There was no statistically significant difference between patients treated with paroxetine and patients given placebo on change from baseline in CDRS-R total score at endpoint. Adverse events reported for paroxetine with an incidence of >5% and at least twice that of placebo were dizziness, cough, dyspepsia, and vomiting.
A 12-week international placebo-controlled multicenter trial of paroxetine in 286 adolescents with major depression failed to show superiority for paroxetine compared with placebo on change from baseline in MADRS or Schedule for Affective Disorders and Schizophrenia for School-Aged Children—Lifetime Version (Kiddie-SADS-L) total score (Berard et al. 2006).
Two double-blind, placebo-controlled multicenter studies have evaluated the efficacy of venlafaxine extended release (XR) in the treatment of major depression in 165 and 169 child and adolescent outpatients, respectively (Emslie et al. 2007). Patients were randomly assigned to venlafaxine XR (37.5–225 mg/day; flexible dosage based on body weight) for 8-week trials. Both studies were negative on the primary outcome measure of change from baseline to endpoint in the CDRS-R scores. The most common adverse events were anorexia and abdominal pain (Emslie et al. 2007).
There have been two double-blind, placebo-controlled multicenter trials of nefazodone in the treatment of major depression in youth (Rynn et al. 2002; U.S. Food and Drug Administration 2004b). These studies failed to find statistically significant improvement in baseline-to-endpoint CDRS-R scores between the nefazodone-treated group and the placebo-treated group. The most common adverse events with nefazodone were headache, abdominal pain, nausea, vomiting, somnolence, and dizziness.
There are no controlled trials of bupropion for the treatment of pediatric depression.
In an 8-week study of bupropion sustained release (dosage range=100–400 mg/day; mean daily dosage=362 mg) for treating 11 adolescents with major depression, 8 adolescents (79%) showed a 50% reduction in depression scores from baseline (Glod et al. 2000). Bupropion sustained release was assessed in an 8-week open study for the treatment of comorbid depression and ADHD in 24 adolescents (Daviss et al. 2001). Global improvement was reported in 14 subjects (58%) for both depression and ADHD and in 7 subjects (29%) for depression only. Common side effects were headache, nausea, rash, and irritability.
There have been two double-blind, placebo-controlled multicenter trials of mirtazapine in the treatment of child and adolescent outpatients with major depression. In these studies, mirtazapine was not superior to placebo on the primary efficacy measure of change from baseline to endpoint in CDRS-R scores (U.S. Food and Drug Administration 2004b).
There have been two failed 10-week trials of duloxetine in the treatment of children and adolescents with major depression (Atkinson et al. 2014; Emslie et al. 2014). Duloxetine and fluoxetine were not superior to placebo on the primary efficacy measure of change from baseline to endpoint in CDRS-R scores. The most common adverse events were nausea, vomiting, nasopharyngitis, upper abdominal pain, headache, and somnolence.
The efficacy of desvenlafaxine sustained release (SR) for the treatment of major depression in 340 children and adolescents (ages 7–17 years) was assessed in a double-blind, placebo-controlled trial (Pfizer 2015). The study had three treatment arms: desvenlafaxine SR, fluoxetine, and placebo. This was a failed study; neither desvenlafaxine SR nor fluoxetine was significantly superior to placebo.
In a placebo-controlled trial of 308 adolescents with major depression, selegiline transdermal system was not superior to placebo on the primary efficacy measure of change from baseline to endpoint in CDRS-R scores (DelBello et al. 2014). The most common adverse events for selegiline were application site reactions, headaches, and nausea.
In a combined analysis of 24 short-term placebo-controlled trials of antidepressant medications in child and adolescent major depressive disorder, obsessive-compulsive disorder (OCD), or other psychiatric disorders, the rates of suicidality (suicidal thinking and behavior) were 4% in patients given medication and 2% in patients given placebo. There were no suicides in any of the clinical trials. The FDA directed manufacturers to add a black box warning to the health professional label of antidepressant medications to describe the increased risk of suicidal thoughts and behavior in children and adolescents being treated with antidepressant medications and to emphasize the need for close monitoring of patients taking the medications (U.S. Food and Drug Administration 2004a). Parents and patients should be advised of the black box warning for antidepressant medication.
In a subsequent meta-analysis of 27 trials of antidepressants in pediatric major depression, the rates of suicidal ideation and attempts were 3% in the youth treated with antidepressants and 2% in the youth who received placebo (Bridge et al. 2007). The investigators reported that the number needed to treat was 10, whereas the number needed to harm was 112, and therefore the benefits of antidepressants outweigh the potential risk from suicidal ideation or attempt.
A number of studies, in both the United States and Europe, have failed to demonstrate an association between antidepressant use and youth suicide (Gibbons et al. 2006; Markowitz and Cuellar 2007; Simon et al. 2006; Søndergård et al. 2006). It is noteworthy that the suicide rate in youth increased following the addition of the black box warning on antidepressants (Hamilton et al. 2007). The FDA advisory has been associated with significant decreases in the rates of diagnosis and treatment of pediatric depression (Libby et al. 2007). The risk of deliberate self-harm was found to be higher in youth whose treatment was initiated with high-dosage SSRIs than in those whose treatment began with modal-dosage SSRIs (Miller et al. 2014).
In the National Institute of Mental Health (NIMH)–funded multisite Treatment of SSRI-Resistant Depression in Adolescents (TORDIA) trial (Brent et al. 2008), 334 adolescents with SSRI-resistant depression were randomly assigned to one of four treatments for 12 weeks: 1) switch to an alternate SSRI (paroxetine, citalopram, or fluoxetine), 2) switch to an alternate SSRI plus CBT, 3) switch to venlafaxine, or 4) switch to venlafaxine plus CBT. CBT plus a medication switch (to venlafaxine or to an alternate SSRI) produced the highest rate of response (54.8%). Response rates (CGI-I score ≤2 and CDRS-R ≥50% reduction) were similar for switching to an alternate SSRI and switching to venlafaxine (47% and 48.2%, respectively). Skin problems and increases in diastolic blood pressure and pulse were more frequently experienced during venlafaxine treatment than during SSRI treatment.
An evidence-based consensus medication algorithm for the treatment of childhood major depression is available (Texas Children’s Medication Algorithm Project [TMAP]; Hughes et al. 2007). Based on research evidence and panel discussion, four stages of medication treatment are identified:
Stage 1: SSRI
Stage 2: Alternate SSRI
Stage 2A (if partial response to SSRI): SSRI+lithium, bupropion, or mirtazapine
Stage 3: Different class of antidepressant medication (venlafaxine, bupropion, mirtazapine, duloxetine)
Stage 4: Reassessment, treatment guidance
An additional recommendation is that antidepressants should be continued for 6–12 months after symptom remission. At the time of discontinuation of an antidepressant, the dosage should be tapered slowly (i.e., no more than 25% per week). The typical tapering and discontinuation period is 2–3 months.
The prevalence of bipolar disorder in a community sample of adolescents was found to be 1% (Merikangas et al. 2010). Although DSM-5 criteria are used to diagnose bipolar disorder in youth, the clinical features in children may differ from those in adolescents and adults. Children with bipolar disorder frequently exhibit mixed mania and rapid cycling (B. Geller et al. 2000). One-year recovery rates of 87% and relapse rates of 64% have been reported in children with bipolar disorder (B. Geller et al. 2004).
Six medications have FDA approval for the acute treatment of bipolar I disorder, mixed or manic episode, in youth: lithium (≥12 years old), aripiprazole (≥10 years old), asenapine (≥10 years old), olanzapine (≥13 years old), risperidone (≥10 years old), and quetiapine (≥10 years old). Ziprasidone has been studied for the treatment of bipolar disorder in youth, but it does not have FDA approval for that indication.
The efficacy of lithium for the treatment of bipolar I disorder, manic or mixed episode, was assessed in an 8-week double-blind, placebo-controlled trial involving 81 youths (ages 7–17 years) (Findling et al. 2015b). The mean lithium serum level at endpoint was 0.98 mEq/L. Lithium was superior to placebo on the primary outcome measure of change in Young Mania Rating Scale (YMRS) score from baseline to endpoint. On CGI-I scores, 47% of lithium-treated and 21% of placebo-treated patients were much or very much improved. The most common adverse events with lithium were vomiting, nausea, and headache. There was a statistically significant increase in thyrotropin concentration with lithium.
In an NIMH-funded study, 153 children and adolescents with bipolar I disorder, mixed or manic episode, were randomly assigned to treatment with lithium, divalproex, or placebo in an 8-week trial (Kowatch et al. 2007). Target lithium serum levels were 0.8–1.2 mEq/L. Lithium was not significantly superior to placebo.
There is one small double-blind, placebo-controlled study of lithium treatment for adolescent bipolar disorder and DSM-III-R (American Psychiatric Association 1984)–defined substance dependence (B. Geller et al. 1998). Twenty-five adolescent outpatients were randomly assigned to either lithium (mean serum level=0.97 mEq/L) or placebo for a 6-week trial. There was significantly greater improvement in global functioning with lithium than with placebo. Side effects in the group treated with lithium were polyuria, thirst, nausea, vomiting, and dizziness.
The efficacy of aripiprazole was assessed in a 4-week double-blind, placebo-controlled trial that included 296 youths with bipolar I disorder, mixed or manic episode (Findling et al. 2009). Both low-dosage aripiprazole (10 mg/day) and high-dosage aripiprazole (30 mg/day) were significantly superior to placebo in reduction of YMRS scores. The response rate (≥50% reduction in YMRS score) was 44.8% for low-dosage aripiprazole, 63.6% for high-dosage aripiprazole, and 26.1% for placebo. The most common adverse events with aripiprazole were somnolence, extrapyramidal side effects (EPS), and tremor, which were more frequent in the high-dosage aripiprazole group.
The efficacy of asenapine was evaluated in a 3-week, double-blind, placebo-controlled trial of 403 youths (ages 10–17 years) with bipolar I disorder, mixed or manic episode (Actavis 2017). Dosages ranged from 2.5 to 10 mg twice daily. Asenapine was significantly superior to placebo on outcome measures of change in YMRS and CGI Bipolar (CGI-BP) Severity of Illness scores from baseline to endpoint. The most comment adverse events were somnolence, dizziness, oral hypoesthesia, headache, fatigue, and increased appetite.
There is one reported double-blind, placebo-controlled multicenter study of olanzapine (2.5–20 mg/day) in the treatment of adolescent outpatients with bipolar I disorder, mixed or manic episode (Tohen et al. 2007). Adolescents were randomly assigned to receive olanzapine (n=107) or placebo (n=54) for 3 weeks. Response rates (defined as ≥50% decrease in YMRS score) were significantly greater for the olanzapine group (44.8%) than for the placebo group (18.5%). Remission rates (defined as YMRS score <12) were significantly greater for the olanzapine group (35.2%) than for the placebo group (11.1%). Adverse effects in the olanzapine group were hyperprolactinemia, weight gain (mean=3.7 kg), somnolence, and sedation.
In a 3-week double-blind, placebo-controlled trial, the efficacy of risperidone was assessed in 169 children and adolescents with bipolar I disorder, manic or mixed episode (Haas et al. 2009). Both low-dosage risperidone (0.5–2.5 mg/day) and high-dosage risperidone (3–6 mg/day) were significantly superior to placebo in reduction of YMRS scores. The response rate (≥50% reduction in baseline YMRS score) was 59% for low-dosage risperidone and 63% for high-dosage risperidone, compared with a placebo response rate of 26%. The most common adverse events with risperidone were somnolence, headache, and fatigue. EPS were more frequent in the high-dosage risperidone group than in the low-dosage group.
In a 3-week double-blind, placebo-controlled trial, the efficacy of quetiapine was assessed in 277 youths with bipolar I disorder, manic or mixed episode (DelBello et al. 2007). Both low-dosage quetiapine (400 mg/day) and high-dosage quetiapine (600 mg/day) were significantly superior to placebo in reduction of YMRS scores. The response rate (≥50% reduction in baseline YMRS score) was 64% for low-dosage quetiapine, 58% for high-dosage quetiapine, and 37% for placebo. The most common adverse events with quetiapine were somnolence, sedation, dizziness, and headache.
The efficacy of ziprasidone was assessed in a 4-week double-blind, placebo-controlled trial that included 150 youths with bipolar I disorder, manic or mixed episode (DelBello et al. 2008). Ziprasidone dosages ranged from 80 to 160 mg/day. Ziprasidone was significantly superior to placebo in reduction of YMRS scores. The response rate (≥50% reduction in YMRS score) was 62% for ziprasidone and 35% for placebo. The most common side effects with ziprasidone were sedation, somnolence, nausea, fatigue, and dizziness.
A 4-week double-blind, placebo-controlled multicenter trial of 150 youths (ages 10–17 years) with bipolar I disorder, mixed or manic episode, did not show a significant difference in YMRS scores from baseline to endpoint for those patients given divalproex extended release (ER) and those given placebo (Wagner et al. 2009). The mean modal dose of divalproex ER was 1,286 mg. There were no statistically significant differences in adverse-event incidents between the divalproex ER and placebo groups. Gastrointestinal symptoms were more commonly reported in divalproex ER than in placebo groups.
In an open-label study of carbamazepine ER for 137 youths (ages 10–17 years) with bipolar I disorder, mixed or manic episode, there was a statistically significant reduction in scores on the YMRS from baseline to endpoint (Findling and Ginsberg 2014). At endpoint, the most prevalent dosage of carbamazepine ER was 1,200 mg/day. The most common adverse events were rash, decreased white blood cell count, nausea, and vomiting.
The only double-blind, placebo-controlled multicenter trial of oxcarbazepine for the treatment of children and adolescents with bipolar I disorder, current episode mixed or manic, failed to show superiority of oxcarbazepine over placebo. The researchers randomly assigned 116 youths (ages 7–18 years) to receive oxcarbazepine (mean dosage=1,515 mg/day) or placebo for a 7-week trial (Wagner et al. 2006b). There was no significant difference in YMRS scores at endpoint between the oxcarbazepine and placebo groups. The most common side effects in the oxcarbazepine-treated patients were dizziness, nausea, somnolence, diplopia, fatigue, and rash.
A double-blind, randomized, placebo-controlled multicenter study assessing the efficacy of topiramate treatment in children and adolescents with acute mania was designed as a 200-patient study but was terminated after randomization of 56 patients (ages 6–17 years) when adult mania trials failed to show efficacy (DelBello et al. 2005). Dosages were titrated to 400 mg/day (mean dosage=278 mg/day). Over a 4-week period, no significant difference was found between the topiramate and placebo groups. The most common adverse events in the topiramate group included decreased appetite, nausea, diarrhea, paresthesias, and somnolence.
In a 12-week open-label trial, the efficacy of lamotrigine was assessed in 30 children and adolescents with bipolar spectrum disorders (Biederman et al. 2010). The mean lamotrigine dosage at endpoint was 160.7 mg/day. Significant improvement in mean YMRS scores was reported; however, only about half the participants completed the trial, and seven participants discontinued lamotrigine because of rash.
The Treatment of Early Age Mania (TEAM) study compared the effectiveness of risperidone, lithium, and divalproex in the treatment of 279 youths (ages 6–15 years) with bipolar I disorder, manic or mixed episode (B. Geller et al. 2012). Response was defined as a CGI-BP-Improvement Mania score of 2 or lower. The mean lithium level was 1.09 mEq/L, the mean divalproex level was 113.6 μg/mL, and the mean risperidone dosage was 2.57 mg/day. The risperidone response rate (68.5%) was significantly higher than response rates for lithium (35.6%) and divalproex (24%). There was no significant difference between lithium and divalproex response rates. Increased weight gain, body mass index (BMI), and prolactin levels occurred significantly more frequently with risperidone than with lithium or divalproex.
Similarly, a comparator analysis of the efficacy of antipsychotics and mood stabilizers for the treatment of pediatric bipolar disorder showed significantly greater improvement in YMRS scores for patients given antipsychotics than for those given mood stabilizers (Correll et al. 2010). Effect sizes were 0.65 for antipsychotics and 0.24 for mood stabilizers.
The comparative efficacy of lithium, divalproex, and carbamazepine was assessed in a 6-week randomized open-label trial involving 42 children and adolescents with bipolar disorder (Kowatch et al. 2000b). There were no significant differences in response rates (≥50% reduction in YMRS score) among the groups given lithium (38%), divalproex (53%), and carbamazepine (38%).
The comparative efficacy of risperidone and divalproex was assessed in an 8-week double-blind, randomized trial in 66 children and adolescents with bipolar disorder (Pavuluri et al. 2010). Significantly higher response rates (≥50% reduction in YMRS) were found for risperidone (78.1%) than for divalproex (45.5%), and improvement was more rapid in risperidone-treated patients than in divalproex-treated patients.
The comparative efficacy of quetiapine and divalproex was assessed in a 4-week double-blind, placebo-controlled trial involving 50 adolescents with bipolar I disorder, manic or mixed episode (DelBello et al. 2006). No significant group differences were found in YMRS scores during the trial.
In an 8-week open-label trial, the efficacies of olanzapine and risperidone were compared in preschool-age children with bipolar disorder (Biederman et al. 2005). There were no significant differences in response rates between risperidone (69%) and olanzapine (53%).
In a 6-week placebo-controlled trial of valproic acid versus risperidone for children (ages 3–7 years) with bipolar disorder, manic, mixed or hypomanic episode, risperidone was superior to placebo on reduction in YMRS scores from baseline to endpoint. There was no significant difference found between valproic acid and placebo (Kowatch et al. 2015).
Some children with bipolar disorder may not respond to initial monotherapy treatment or may need combination treatment over the course of the illness. For example, in a study by Kowatch et al. (2000a), following acute 6-week treatment with one mood stabilizer, 20 of 35 youths (58%) required additional psychotropic medication over the next 16 weeks. The response rate to combination treatment with two mood stabilizers was high (80%) for those youths who did not respond to monotherapy.
The FDA has approved the use of quetiapine or aripiprazole as an adjunct to lithium or valproate treatment in children ages 10 years and older with bipolar I disorder, mixed or manic episode.
The effectiveness of combination treatment with lithium and divalproex was assessed in an open trial (Findling et al. 2003). Ninety youths (ages 5–17 years) with bipolar I or II disorder were treated for up to 20 weeks with divalproex (mean blood level=79.8 μg/mL) and lithium (mean blood level=0.9 mmol/L). The clinical remission rate (defined as contiguous weekly ratings of YMRS scores ≤12.5, CDRS-R scores ≤40, Children’s Global Assessment Scale [CGAS] scores ≥51, clinical stability, and no mood cycling) was 42%.
The efficacy of risperidone in combination with lithium or divalproex was assessed in a 6-month open-label trial (Pavuluri et al. 2004). Thirty-seven youths (ages 5–18 years) with bipolar I disorder, manic or mixed episode, received either risperidone (mean dosage=0.75 mg/day) plus divalproex (mean serum level=106 μg/mL) or risperidone (mean dosage=0.70 mg/day) plus lithium (mean serum level=0.9 mEq/L). Response rates (≥50% reduction in baseline YMRS scores) were similar for both combinations: 80% for divalproex plus risperidone, and 82.4% for lithium plus risperidone. There were no significant differences between the groups in safety and tolerability.
Risperidone augmentation for lithium nonresponders was assessed in a 1-year open-label study (Pavuluri et al. 2006). Twenty-one of 38 youths (ages 4–17 years) who failed to respond to lithium monotherapy or relapsed after initial response were given risperidone (mean dosage=0.99 mg/day) for 11 months. Response rates in the lithium plus risperidone group were 85.7%.
In a double-blind, placebo-controlled study of quetiapine, 30 adolescents with bipolar disorder received divalproex (20 mg/kg) and were randomly assigned to adjunctive quetiapine (mean daily dosage=432 mg) or placebo for 6 weeks (DelBello et al. 2002). Response rates (≥50% reduction in baseline YMRS score) were significantly higher in the group receiving divalproex and quetiapine (87%) than in the group receiving divalproex and placebo (53%).
Sixty youths who had responded to a combination of lithium and divalproex in a 20-week trial were randomly assigned in a double-blind trial to either lithium or divalproex for 18 months (Findling et al. 2005). There was no significant difference in the time to relapse between the groups (median days: divalproex 112, lithium 114).
Treatment guidelines were developed by expert consensus and review of the available treatment literature for children and adolescents (ages 6–17 years) with bipolar I disorder, manic or mixed episode (Kowatch et al. 2005). Six stages were identified:
Stage 1: Monotherapy with mood stabilizer or atypical antipsychotic
Stage 2: Switch monotherapy agent (drug class not tried in stage 1)
Stage 3: Switch monotherapy agent (drug class not tried in stage 1 or 2) OR combination treatment (2 agents)
Stage 4: Combination treatment (2 agents) OR combination treatment (3 agents)
Stage 5: Alternative monotherapy (drugs not tried in stages 1, 2, 3)
Stage 6: Electroconvulsive therapy (adolescents) or clozapine
If a child fails to respond to treatment in one stage, the clinician should move to the next stage of treatment. For treatment of bipolar I disorder, manic or mixed episode with psychosis, the recommendation for initial treatment is a mood stabilizer plus an atypical antipsychotic. A minimum of 4–6 weeks at therapeutic blood levels and/or adequate dosages for each medication is recommended. Following sustained remission of at least 12–24 months, medication taper should be considered.
DSM-5 anxiety disorders include generalized anxiety disorder, social anxiety disorder, panic disorder, selective mutism, agoraphobia, specific phobia, and separation anxiety disorder. Obsessive-compulsive disorder and posttraumatic stress disorder are now in separate DSM-5 categories (Obsessive-Compulsive and Related Disorders and Trauma- and Stressor-Related Disorders, respectively.
The prevalence of generalized anxiety disorder (GAD) in children and adolescents is estimated to range from 2.9% to 7.3% (Anderson et al. 1987; Kashani and Orvaschel 1988; Merikangas et al. 2010). Children with GAD have excessive anxiety and worry about several events or activities (e.g., school performance), have difficulty controlling the worry, and have at least one associated symptom, such as restlessness, fatigue, concentration difficulties, irritability, muscle tension, and sleep disturbance (American Psychiatric Association 2013). The course of GAD in youth tends to be chronic (Keller et al. 1992).
Duloxetine has been approved by the FDA for the treatment of GAD in youths ages 7–17 years.
The efficacy of duloxetine was assessed in a 10-week double-blind, placebo-controlled trial for 272 youths (ages 7–17 years) with GAD (Strawn et al. 2015). Duloxetine was flexibly dosed from 30 to 120 mg/day. The primary efficacy measure was the Pediatric Anxiety Rating Scale (PARS). Duloxetine was significantly superior to placebo on improvement in PARS scores from baseline to endpoint. Response (defined as 50% improvement on PARS severity for GAD) and remission (defined as a PARS severity for GAD ≤8) were significantly greater for the duloxetine group (59% and 50%, respectively) than the placebo group (42% and 34%, respectively). Adverse events reported with significantly greater frequency for the duloxetine group than for the placebo group were nausea, vomiting, decreased appetite, oropharyngeal pain, dizziness, cough, and palpitations.
The efficacy of venlafaxine XR in children and adolescents with GAD (N=320) was evaluated in two double-blind, placebo-controlled trials (Rynn et al. 2007). Venlafaxine was given at dosages up to 225 mg/day. In one study, venlafaxine XR was superior to placebo on primary and secondary measures; however, in the other study, the results were negative.
Twenty-two children and adolescents (ages 5–17 years) with GAD were randomly assigned to sertraline or placebo in a 9-week double-blind trial (Rynn et al. 2001). The maximum dosage of sertraline was 50 mg/day. Significant differences in favor of sertraline over placebo were observed on the Hamilton Anxiety Scale (Ham-A) scores and on CGI Severity of Illness (CGI-S) and CGI-I ratings. Side effects found to be more common (but not statistically significantly so) with sertraline than with placebo were dry mouth, drowsiness, leg spasm, and restlessness.
The efficacy of buspirone was evaluated for 559 youths (ages 6–17 years) with GAD who participated in a 6-week randomized placebo-controlled trial (Bristol-Myers Squibb 2010). Buspirone dosages ranged from 15 to 60 mg/day. There was no statistically significant difference in outcome between buspirone and placebo.
The prevalence of social anxiety disorder (social phobia) in adolescents is reported to be 9.1% (Merikangas et al. 2010). The DSM-5 diagnostic criteria for social anxiety disorder are the same for children and adolescents as for adults. Social anxiety disorder in youth is a chronic condition, and it increases the risk of depression (Stein et al. 2001).
Paroxetine. The efficacy and safety of paroxetine were evaluated in a 16-week double-blind, placebo-controlled multicenter trial in 322 outpatient children and adolescents (ages 8–17 years) with social anxiety disorder (Wagner et al. 2004a). Paroxetine was significantly superior to placebo, with rates of response (defined as CGI-I score=1 or 2) of 77.6% and 38.3%, respectively. Side effects more common with paroxetine than with placebo were insomnia, decreased appetite, and vomiting.
Fluoxetine. The efficacy of fluoxetine was evaluated in a 12-week randomized trial of fluoxetine (up to 40 mg/day), Social Effectiveness Therapy for Children (SET-C), and placebo for 80 youths (ages 7–17 years) with social anxiety disorder (Beidel et al. 2007). Both fluoxetine and SET-C were superior to placebo in reducing social stress and behavioral avoidance and increasing general functioning.
Sertraline. Fourteen outpatient children and adolescents (ages 10–17 years) with a diagnosis of social anxiety disorder received sertraline (dosage range=100–200 mg/day; mean daily dosage=123 mg) in an 8-week open trial (Compton et al. 2001). Five of the patients (36%) were much or very much improved, and four of the patients (29%) had a partial response by the end of the 8-week trial. Sertraline was well tolerated, and no patient developed significant behavioral disinhibition or mania (Compton et al. 2001).
Escitalopram. Twenty children with social anxiety disorder participated in a 12-week open-label study of escitalopram (Isolan et al. 2007). Sixty-five percent of participants were much or very much improved.
Citalopram. Chavira and Stein (2002) investigated the effectiveness of a combined psychoeducational and pharmacological treatment program for youth with social anxiety disorder. Twelve children and adolescents (ages 8–17 years) with social anxiety disorder received citalopram (mean daily dosage=35 mg) and eight 15-minute counseling sessions over a 12-week period. On the basis of clinical global ratings of change, five of the patients (41.7%) were very much improved, and five of the patients (41.7%) were much improved.
In a 16-week double-blind, placebo-controlled trial, 293 children and adolescents with social anxiety disorder were randomly assigned to venlafaxine XR (dosage range=37.5–225 mg) or placebo (March et al. 2007). Venlafaxine XR was significantly superior to placebo in reducing ratings of social anxiety. Response rates (CGI-I score ≤2) were 56% for the venlafaxine XR group and 37% for the placebo group.
The prevalence of panic disorder in children and adolescents ranges from 0.6% to 5.0% in the community and from 0.2% to 9.6% in clinical settings (Masi et al. 2001). The DSM-5 diagnostic criteria for panic disorder in children and adolescents are the same as those for adults. Panic disorder in youth is a chronic condition, and there is continuity between pediatric and adult panic disorder (Biederman et al. 1997).
In an open-label trial, 12 children and adolescents (ages 7–17 years) with panic disorder were treated with an SSRI for 6–8 weeks (Renaud et al. 1999). Mean daily dosages of SSRIs were fluoxetine 34 mg, paroxetine 20 mg, and sertraline 125 mg. Adjunctive benzodiazepines were used for 8 patients. Seventy-five percent of patients showed much to very much clinical improvement while receiving treatment with SSRIs. At the end of the trial, 8 patients (67%) no longer fulfilled panic disorder criteria.
Paroxetine. A chart review was conducted of 18 child and adolescent outpatients (ages 7–16 years) with a diagnosis of panic disorder who received monotherapy with paroxetine (dosage range=10–40 mg/day; mean daily dosage=23 mg) (Masi et al. 2001). The mean paroxetine treatment duration was 11.7 months. Fifteen patients (83%) had a CGI score of much or very much improved. The most common side effects were nausea, tension-agitation, sedation, insomnia, palpitations, and headache.
Citalopram. Three youths (ages 9, 13, and 16 years) with panic disorder and school phobia were treated with citalopram (up to 20 mg/day) over an 8- to 15-month period. All patients experienced resolution of panic attacks during the course of citalopram treatment (Lepola et al. 1996).
In a 2-week open trial, four adolescents with panic disorder were treated with clonazepam (0.5 mg twice daily). A significant reduction in panic attacks (from 3 attacks per week to 0.25 per week) was reported (Kutcher and MacKenzie 1988).
Fluvoxamine. One hundred twenty-eight outpatient children and adolescents (ages 6–17 years) with GAD, social anxiety disorder, or separation anxiety disorder (who had received 3 weeks of open treatment with supportive psychoeducational therapy without improvement) were randomly assigned to fluvoxamine (up to 300 mg) or placebo for an 8-week trial (Research Unit on Pediatric Psychopharmacology Anxiety Study Group 2001). The group treated with fluvoxamine had a significantly greater reduction in scores on the PARS than did the group treated with placebo. The response rate (CGI-I score ≤3) was 76% in the group being treated with fluvoxamine and 29% in the group receiving placebo. After completion of the 8-week placebo-controlled study, the 128 patients entered a 6-month open-label treatment phase (Walkup et al. 2002). Anxiety symptoms remained low in 33 of 35 of the subjects (94%) who initially responded to fluvoxamine. Of 14 fluvoxamine nonresponders switched to fluoxetine, anxiety symptoms significantly improved in 10 patients (71%). Among 48 placebo nonresponders, 27 (56%) showed significant improvement in anxiety on fluvoxamine.
Fluoxetine. Seventy-four youths (ages 7–17 years) with GAD, separation anxiety disorder, and/or social phobia were randomly assigned to fluoxetine (20 mg/day) or to placebo for 12 weeks (Birmaher et al. 2003). Sixty-one percent of fluoxetine-treated patients and 35% of placebo-treated patients were much or very much improved.
Fluoxetine’s efficacy in long-term treatment of children with GAD, separation anxiety disorder, and/or social phobia was assessed in a 1-year open treatment (Clark et al. 2005) following the acute-phase study (Birmaher et al. 2003). Compared with youth taking no medication, those taking fluoxetine (n=42) showed significantly superior outcome in anxiety measures.
The comparative efficacy of fluoxetine and clomipramine was evaluated in a 12-week double-blind, randomized, placebo-controlled trial for 36 youths (ages 7–17 years) with GAD, separation anxiety, and/or social anxiety disorder (da Costa et al. 2013). Although all groups showed significant improvement, no significant difference was found between fluoxetine and placebo or between clomipramine and placebo.
Sertraline. The comparative efficacy of sertraline, CBT, sertraline plus CBT, and placebo was evaluated in a 12-week randomized controlled trial (the Child Anxiety Multimodal Study [CAMS]) involving 488 children and adolescents with a diagnosis of GAD, separation anxiety disorder, or social phobia (Walkup et al. 2008). Combination treatment was significantly superior to sertraline alone or CBT alone. Rates of response (CGI-I score ≤2) were 80.7% for combined treatment, 59.7% for CBT, and 54.9% for sertraline. All treatments were significantly superior to placebo (23.7%).
Long-term outcomes from the CAMS study have been reported (Piacentini et al. 2014). At 24 and 36 weeks, most (≥80%) of the acute responders maintained a positive response. Combination treatment maintained an advantage over CBT and sertraline alone.
In regard to childhood anxiety disorders, SSRIs are the first-line treatment (Reinblatt and Walkup 2005). Duloxetine and venlafaxine also have demonstrated efficacy for the treatment of childhood GAD. Other treatment options include buspirone, tricyclic antidepressants, and benzodiazepines (Bernstein et al. 1996). However, benzodiazepines should be used only on a short-term basis (i.e., weeks) because of the potential for abuse and dependence in youth (Riddle et al. 1999).
OCD has a prevalence rate of 2%–4% in youth (Douglass et al. 1995; Zohar 1999). The DSM-5 criteria for OCD are the same in children and adults. The course of OCD in youth is chronic.
Four medications have received FDA approval for the treatment of OCD in children and adolescents: sertraline (≥6 years old), fluoxetine (≥7 years old), fluvoxamine (≥7 years old), and clomipramine (≥10 years old). Citalopram and paroxetine have been studied for treatment of OCD in youth but do not have FDA approval.
In a double-blind, placebo-controlled multicenter study, 187 children and adolescents (ages 6–17 years) with OCD were randomly assigned to sertraline or placebo (March et al. 1998). Sertraline dosages were titrated to a maximum of 200 mg/day during the first 4 weeks of the trial, and these dosages were maintained for an additional 8 weeks. The mean dosage of sertraline was 167 mg/day at endpoint. Compared with patients receiving placebo, patients receiving sertraline showed significantly greater improvement on the CY-BOCS, the NIMH Global Obsessive Compulsive Rating Scale (NIMH GOCS), and the CGI-S and CGI-I subscales. Forty-two percent of patients in the sertraline group and 26% of patients in the placebo group were rated as very much or much improved. Side effects of insomnia, nausea, agitation, and tremor occurred significantly more often in the group receiving sertraline than in the group receiving placebo.
In an assessment of the long-term safety and effectiveness of sertraline for pediatric OCD, 137 patients who completed the 12-week double-blind, placebo-controlled sertraline study (March et al. 1998) were given open-label sertraline (mean dosage=120 mg/day) in a 52-week extension study. Significant improvement was found on CY-BOCS, NIMH GOCS, and CGI scores. Rates of response (defined as >25% decrease in CY-BOCS and a CGI-I score of 1 or 2) were 72% for children and 61% for adolescents (Cook et al. 2001). Full remission (defined as a CY-BOCS score <8) was achieved in 47% of patients, and an additional 25% achieved partial remission (CY-BOCS score >8 but <15) (Wagner et al. 2003b). The most common side effects were headache, nausea, diarrhea, somnolence, abdominal pain, hyperkinesias, nervousness, dyspepsia, and vomiting.
The relative and combined efficacy of sertraline and CBT was assessed in a 12-week trial for 112 children and adolescents (ages 7–17 years) with OCD (Pediatric OCD Treatment Study [POTS] Team 2004). Patients were randomly assigned to sertraline, CBT, combined sertraline and CBT, or placebo. Combined treatment was significantly superior to CBT alone and sertraline alone, which did not differ from each other.
The efficacy of sequential sertraline and CBT compared with CBT and placebo was assessed in an 18-week trial for 47 youths (ages 7–17 years) with OCD (Storch et al. 2013). No significant difference was found between the groups on improvement in OCD symptoms.
The safety and efficacy of fluoxetine were assessed in a 13-week double-blind, placebo-controlled multicenter trial (D.A. Geller et al. 2001). One hundred three children and adolescents (ages 7–17 years) with OCD were randomly assigned in a 2:1 ratio to either fluoxetine (dosage range=10–60 mg/day; mean daily dosage=24.6 mg) or placebo. The group treated with fluoxetine showed a statistically significant reduction in OCD severity compared with the group treated with placebo, as determined by changes in CY-BOCS scores. Rates of response (defined as >40% reduction in CY-BOCS score) were 49% in the fluoxetine group and 25% in the placebo group. There were no significant differences in treatment-emergent adverse events between the fluoxetine and placebo groups.
Fluoxetine was compared with placebo in a controlled trial in 43 children and adolescents with OCD (Liebowitz et al. 2002). After 16 (but not 8) weeks of treatment, the fluoxetine group had significantly lower CY-BOCS scores than the placebo group.
The safety and efficacy of fluvoxamine were evaluated in a double-blind, placebo-controlled multicenter study (Riddle et al. 2001). One hundred twenty outpatient children and adolescents (ages 8–17 years) with OCD were randomly assigned to receive fluvoxamine (dosage range=50–200 mg/day; mean daily dosage=165 mg) or placebo for a 10-week trial. Patients who did not respond after 6 weeks could discontinue the double-blind phase and enter an open-label trial of fluvoxamine. Mean scores on the Children’s Yale-Brown Obsessive Compulsive Scale (CY-BOCS) were significantly different for the fluvoxamine and the placebo groups at weeks 1, 2, 3, 4, 6, and 10. Response rates (>25% reduction in CY-BOCS scores) were 42% in the group being treated with placebo. Adverse events occurring at a placebo-adjusted frequency of greater than 10% were insomnia and asthenia.
In an assessment of the safety and effectiveness of fluvoxamine in the long-term treatment of pediatric OCD, 99 patients who completed the acute double-blind, placebo-controlled fluvoxamine study by Riddle et al. (2001) participated in a 1-year open-label extension study (Walkup et al. 1998). Fluvoxamine dosages were titrated to 200 mg/day over the first 4 weeks. Patients experienced a 42% reduction in CY-BOCS scores by the end of long-term treatment. Clinical improvement plateaued at about 6 months of treatment. The most common side effects of fluvoxamine were insomnia, asthenia, nausea, hyperkinesias, and nervousness.
Clomipramine has been shown to be efficacious in the treatment of pediatric OCD in two double-blind, placebo-controlled trials. In the first study (Flament et al. 1985), 19 children (ages 10–18 years) with OCD were randomly assigned to clomipramine (dosage range=100–200 mg/day; mean daily dosage=141 mg) or placebo for 5 weeks. Significant improvement in observed and self-reported obsessions and compulsions was found for patients who received clomipramine. The most common side effects with clomipramine were tremor, dry mouth, dizziness, and constipation. One patient had a grand mal seizure.
In an 8-week double-blind, placebo-controlled multicenter study of 60 children and adolescents (ages 10–17 years) with OCD, it was found that patients who received clomipramine (up to 200 mg/day) had significantly greater reductions in scores on the CY-BOCS than the placebo group (37% and 8%, respectively). Forty-seven patients continued in a 1-year open-label extension trial, and effectiveness was maintained with long-term treatment. The most frequent side effects with clomipramine were dry mouth, somnolence, dizziness, fatigue, tremor, headache, constipation, and anorexia (DeVeaugh-Geiss et al. 1992).
Twenty-three child and adolescent outpatients (ages 9–18 years) with OCD were administered open-label citalopram (dosage range=10–40 mg/day; mean daily dosage=37 mg) in a 10-week trial (Thomsen 1997). There was a statistically significant improvement in CY-BOCS scores from baseline to endpoint. Adverse effects were minimal and transient.
In an 8-week open-label citalopram study of 15 youths (ages 6–17 years) with OCD, 14 patients showed significant improvement in CY-BOCS scores from baseline to endpoint (Mukaddes and Abali 2003).
In a long-term open study of 30 adolescents with OCD, citalopram (dosage range=20–70 mg/day; mean daily dosage=46.5 mg) was administered for 1–2 years (Thomsen et al. 2001). There was a significant reduction in CY-BOCS scores from baseline to assessment at 2 years. No serious adverse events were reported, and the most common side effects were sedation, sexual dysfunction, and weight gain.
The efficacy and safety of paroxetine were assessed in a double-blind, placebo-controlled multicenter study of 203 outpatient children and adolescents (ages 7–17 years) with OCD (D.A. Geller et al. 2004). Patients were randomly assigned to paroxetine (dosage range=10–50 mg/day; mean daily dosage=23 mg) or placebo for a 10-week trial. There was a statistically significant greater reduction in CY-BOCS scores from baseline to endpoint in patients treated with paroxetine than in patients treated with placebo. Response rates (>25% reduction in CY-BOCS scores) were 64.9% in the paroxetine-treated patients and 41.2% in the placebo-treated patients. The most common adverse effects in the paroxetine group were headache, abdominal pain, nausea, respiratory disorder, somnolence, hyperkinesias, and trauma.
The efficacy of paroxetine in 335 outpatients (ages 7–17 years) with OCD was assessed in a 16-week open-label multicenter study of paroxetine (10–60 mg/day), followed by double-blind randomization of responders to paroxetine or placebo for an additional 16 weeks (Emslie et al. 2000). No significant differences in response rates were found between the group receiving paroxetine and the group receiving placebo in the randomization phase.
Adjunctive risperidone (≤2 mg daily) was investigated in an open trial for 17 adolescents with OCD who failed to respond to two serotonin reuptake inhibitor monotherapy trials. A significant reduction in CY-BOCS scores was reported (Thomsen 2004).
Aripiprazole augmentation of CBT was found to be effective in the case of an adolescent who had a partial response to combined CBT and sertraline (Storch et al. 2008).
In a naturalistic sample of 220 children and adolescents with OCD, 43 children were treated with an atypical antipsychotic as an augmenting agent (Masi et al. 2009). Twenty-five of these youths (58.1%) responded to treatment.
The efficacy of aripiprazole augmentation of an SSRI was assessed in 39 youths with OCD who had not responded to two trials with SSRI monotherapy (Masi et al. 2010). Fifty-nine percent of the youths responded to treatment.
Augmentation of CBT with D-cycloserine was evaluated in a double-blind, placebo-controlled trial for 30 youths (ages 8–17 years) with OCD. No significant differences were found between treatment groups (Storch et al. 2010).
D-Cycloserine-augmented CBT was assessed in a double-blind, placebo-controlled pilot trial for 17 youths with difficult-to-treat OCD (Farrell et al. 2013). The augmented CBT resulted in greater improvement in OCD symptoms than CBT plus placebo.
Sixty children and adolescents with treatment-resistant OCD participated in a 12-week double-blind, placebo-controlled trial of add-on riluzole or placebo with current treatment (Grant et al. 2014). The riluzole final dosage was 100 mg/day. All participants significantly improved on CY-BOCS ratings, and there was no significant difference between riluzole and placebo on any outcome measures. Adverse events reported for patients given riluzole included one case of pancreatitis and five instances of slight increase with transaminases.
The authors of a case report (Yazici and Percinel 2014) and a case series (Yazici and Percinel 2015) describe the use of N-acetylcysteine (NAC) augmentation for treatment-resistant OCD in children and adolescents. NAC dosages were initiated at 600 mg/day and increased up to 2,400–3,000 mg/day. Improvement in OCD symptoms was found in some cases.
SSRIs are the medication treatment of choice for OCD in children and adolescents (American Academy of Child and Adolescent Psychiatry 2012). Clomipramine is also effective in the treatment of this disorder; however, it is not a first-line treatment because of its adverse-event profile. A 12-week trial at an adequate dosage is indicated to determine whether a child with OCD will respond to an SSRI. If a child fails to respond to one SSRI, switching to another SSRI is a reasonable strategy. Clomipramine, either as monotherapy or as augmentation of an SSRI, may be a third treatment option. Other possible SSRI augmentation strategies are clonazepam and atypical antipsychotics; however, these agents have not received systematic study. Some children may require long-term medication maintenance; however, it is reasonable to attempt medication discontinuation 1 year after symptom stabilization. Medication should be tapered gradually to assess for relapse. Two or three relapses should lead to long-term treatment.
The prevalence of posttraumatic stress disorder (PTSD) in adolescents is reported to be 5% (Merikangas et al. 2010). The criteria for diagnosing PTSD in children older than age 6 years are the same as those used for adults (American Psychiatric Association 2013). PTSD symptoms in children tend to vary over time, and although the disorder is chronic, the course is prolonged with greater severity of the stressor (Clarke et al. 1993).
The efficacy of sertraline was evaluated in a 10-week double-blind, placebo-controlled trial in 131 children and adolescents with PTSD (Robb et al. 2010). Sertraline dosages ranged from 50 to 200 mg/day. There was no difference between sertraline-treated patients and placebo-treated patients in scores on the primary outcome measure, the University of California, Los Angeles Post-Traumatic Stress Disorder Index for DSM-IV (UCLA PTSD-I).
The benefit of adding sertraline versus placebo to trauma-focused CBT was assessed in a controlled 12-week trial (Cohen et al. 2007). Although both groups showed significant improvement in symptoms of PTSD, there was no significant advantage from adding sertraline rather than placebo to CBT.
Eight adolescents with PTSD received citalopram in a fixed daily dosage of 20 mg in a 12-week open-label study (Seedat et al. 2001). Core PTSD symptoms of reexperiencing, avoidance, and hyperarousal showed statistically significant improvement at week 12, with a 38% reduction in total score on the Clinician-Administered PTSD Scale—Child and Adolescent Version (CAPS-CA). Citalopram was well tolerated, and the most common side effects were increased sweating, nausea, headache, and tiredness.
In a larger 8-week open trial, Seedat et al. (2002) treated 24 children and adolescents with citalopram (dosage range=20–40 mg/day; mean daily dosage=20 mg). Both the children and the adolescents had a significant reduction in CAPS-CA scores at endpoint. Common side effects of citalopram were drowsiness, headache, nausea, and increased sweating.
The effectiveness of guanfacine extended release (GXR) was assessed in an 8-week open-label trial for 19 youths (ages 6–18 years) with current traumatic stress symptoms (Connor et al. 2013). GXR dosages ranged from 1 to 4 mg/day. On parent report, symptom clusters of reexperiencing, avoidance, and overarousal significantly improved.
Seven preschool children (ages 3–6 years) with a diagnosis of PTSD received open treatment with clonidine at a dosage range of 0.05–0.15 mg/day (Harmon and Riggs 1996). To decrease sedation, oral clonidine was subsequently converted to a clonidine patch. The majority of children showed at least moderate improvements in hyperarousal, hypervigilance, insomnia, nightmares, and mood lability.
Twenty-eight children and adolescents (ages 8–17 years) with a diagnosis of PTSD received carbamazepine (dosage range=300–1,200 mg/day) for an average of 35 days. Twenty-two patients (78%) became asymptomatic, and the remaining six patients were significantly improved during the course of treatment (Looff et al. 1995).
The efficacy of propranolol initiated within 12 hours after emergency department admission in preventing PTSD in 29 injured youths was assessed in a 10-day double-blind, placebo-controlled trial (Nugent et al. 2010). There was no significant difference between the propranolol and placebo groups on PTSD symptoms at 6 weeks.
Eleven children (ages 6–12 years) with a diagnosis of PTSD participated in an off-on-off medication design of 4 weeks of propranolol treatment (Famularo et al. 1988). Propranolol was initiated at 0.8 mg/kg/day and titrated to a maximum of 2.5 mg/kg/day. A significant improvement in PTSD symptoms was found during the treatment period. Side effects included sedation and mildly lowered blood pressure and pulse.
In case studies (Oluwabusi et al. 2012; Strawn and Keeshin 2011; Strawn et al. 2009), prazosin 1–3 mg/day was reported to decrease symptoms of PTSD in children and adolescents.
In a controlled trial, 57 youths (ages 7–18 years) with PTSD were randomly assigned to receive D-cycloserine and CBT or placebo and CBT (Scheeringa and Weems 2014). Although both groups had significant reductions in PTSD symptoms, there was no significant difference between the groups.
In case reports and a small open-label trial, risperidone has shown some benefit in reducing symptoms of PTSD in children and adolescents (Horrigan and Barnhill 1997; Keeshin and Strawn 2009; Meighen et al. 2007).
In a case series, six adolescents with PTSD reported improvement in symptoms after 6 weeks of low-dosage (50–200 mg/day) quetiapine (Stathis et al. 2005).
For childhood trauma- and stressor-related disorders, SSRIs are the first-line treatment (Reinblatt and Walkup 2005). Other possible medications for PTSD are α- and β-adrenergic blocking agents, non-SSRI antidepressants, and mood-stabilizing agents (Cohen et al. 2010).
The prevalence of ADHD in children and adolescents is estimated to range from 5% to 12%, whereas about 4% of adults in the general population meet criteria for ADHD (Kessler et al. 2006). In addition to demonstrating the core behavioral features of inattention, hyperactivity, and impulsivity, children with ADHD often have significant impairment in social and academic functioning (Barkley 2005). Of all of the childhood psychiatric disorders, ADHD has the greatest number of pharmacological treatment studies.
The classes of psychostimulants include methylphenidate, dexmethylphenidate, dextroamphetamine, mixed amphetamine salts, and L-lysine-D-amphetamine (lisdexamfetamine). By the 1980s, there were already hundreds of randomized controlled trials showing the efficacy of stimulants in the treatment of ADHD in school-age children (Greenhill et al. 1999). Arnold (2000) reviewed studies in which subjects underwent a trial of both amphetamine and methylphenidate. This review suggested that approximately 41% of subjects with ADHD responded equally to methylphenidate and amphetamine, whereas 44% responded preferentially to one of the classes of stimulants. This finding suggests that the initial response rate to stimulants may be as high as 85% if both stimulants are tried (in contrast to the finding of 65%–75% response when only one stimulant is tried). In contrast, placebo response rates in stimulant trials are rarely above 20%–30%, making the effect size of the stimulants (0.8–1.0) one of the largest among all the psychotropics. At present, however, no method is available to predict which stimulant will produce the best response in a given patient. The past decade has seen the emergence of the long-acting stimulants; studies of these agents have been the focus of major reviews (Biederman and Spencer 2008; Pliszka and AACAP Work Group on Quality Issues 2007).
Initial research with stimulants was carried out in school-age children, but more recent controlled trials of stimulants have focused on adolescents (Spencer et al. 2006b; Wilens et al. 2006a) and adults (Biederman et al. 2006a; Weisler et al. 2006). These studies show that older individuals’ rates of response to stimulants are similar to those of children, with adequate response for most subjects being obtained with 70–100 mg/day of methylphenidate or 40–60 mg/day of amphetamine.
Preschoolers with ADHD have also been the focus of investigation. In the NIMH Preschool ADHD Treatment Study (PATS), 183 children (ages 3–5 years) underwent an open-label trial of methylphenidate, with 165 of these subjects subsequently randomly assigned to a double-blind, placebo-controlled crossover trial of methylphenidate lasting 6 weeks (Greenhill et al. 2006; Wigal et al. 2006). The mean optimal dosage of methylphenidate was found to be 0.7 ±0.4 mg/kg/day, which is lower than the mean of 1.0 mg/kg/day found to be optimal in school-age children. Eleven percent of subjects discontinued methylphenidate because of adverse events. Compared with school-age children, the preschool group showed a higher rate of emotional adverse events, including crabbiness, irritability, and proneness to crying. The conclusion was that the dosage of methylphenidate (or any stimulant) should be titrated more conservatively in preschoolers than in school-age patients, and lower mean dosages may be effective.
The appendix at the end of this chapter shows the recommended dosage ranges for these agents. The appendix also discusses adverse events, particularly growth suppression, that occur with psychostimulants.
Atomoxetine is a noradrenergic reuptake inhibitor that has indirect effects on dopamine reuptake in the cortex but not in the striatum (Bymaster et al. 2002). Numerous double-blind, placebo-controlled trials have demonstrated the medication’s efficacy in the treatment of ADHD in children, adolescents, and adults (Michelson et al. 2001, 2002, 2003). Given its pharmacokinetic half-life of 5 hours, it is generally dosed twice a day. Although open trials comparing methylphenidate with atomoxetine showed the two agents to have similar efficacy (Kratochvil et al. 2002), double-blind, placebo-controlled trials comparing atomoxetine with amphetamine (Biederman et al. 2006b; Wigal et al. 2005) and methylphenidate (Newcorn et al. 2008) have shown the stimulants to be more efficacious.
Atomoxetine is effective in treating ADHD in patients with comorbid tics and may also reduce tics (Allen et al. 2005). It is also useful in children with ADHD who have comorbid anxiety, showing effectiveness in treating anxiety and inattention (Sumner et al. 2005). Atomoxetine is well tolerated in long-term use. In a global multicenter study, 416 children and adolescents who responded to an initial 12-week open-label period of treatment with atomoxetine were randomly assigned to continued atomoxetine treatment or placebo for 9 months under double-blind conditions. Relapse (defined as a return to 90% of baseline symptom severity) occurred significantly less often with atomoxetine (22.3%) than with placebo (37.9%) (Michelson et al. 2001). Data from 13 (6 double-blind, 7 open-label) atomoxetine studies were pooled for youths (ages 12–18 years) with ADHD (Wilens et al. 2006b). Of the 601 atomoxetine-treated subjects in this meta-analysis, 537 (89.4%) completed 3 months of acute treatment. At the time of the article’s publication, 259 subjects (48.4%) were continuing atomoxetine treatment; 219 of these subjects had completed at least 2 years of treatment. Symptoms remained improved for up to 24 months without dosage escalation. During the 2-year treatment period, 99 subjects (16.5%) discontinued treatment due to lack of effectiveness, and 31 subjects (5.2%) discontinued treatment due to adverse events. No clinically significant abnormalities in height, weight, blood pressure, pulse, mean laboratory values, or electrocardiography parameters were found.
A review of the literature from 1980 to 1999 found 39 studies regarding the use of clonidine for symptoms of childhood ADHD, and 11 of these studies had sufficient data to be included in a meta-analysis (Connor et al. 1999). Of the 150 subjects in these studies, 42 received clonidine for ADHD, and the others received clonidine for ADHD comorbid with tic disorders (n=67), developmental disorders (n=15), or conduct disorders (n=26). The mean daily dosage of clonidine was 0.18 mg, and the average length of treatment was 10.9 weeks. Clonidine showed a moderate effect size of 0.58 on symptoms of ADHD, which is smaller than the effect size (0.82) reported for stimulant treatment of ADHD (Swanson et al. 1995b).
An extended-release (ER) form of clonidine was more recently developed, and this formulation was evaluated in an 8-week double-blind, placebo-controlled trial (Jain et al. 2011) in which patients (N=236) were randomly assigned to receive clonidine ER 0.2 mg/day, clonidine ER 0.4 mg/day, or placebo. Improvement in ADHD symptoms was significantly greater in the clonidine groups relative to the placebo group, with this difference apparent at week 2 and greatest at week 5. Somnolence was the most common side effect, and the rate of withdrawal due to adverse events was higher in the clonidine ER 0.4-mg/day group than in the other groups. There were no serious adverse events, and bradycardia was the most common cardiovascular side effect.
Clonidine ER has been added to stimulant medication in an effort to improve the initial stimulant response in the treatment of ADHD (Kollins et al. 2011). Children and adolescents with ADHD who had an inadequate response to their initial stimulant regimen (n=198) were randomized to receive placebo or clonidine extended release added to their stable stimulant dosage for 8 weeks. Clonidine ER was flexibly dosed. At weeks 4 and 5, of the patients within the group receiving clonidine ER plus stimulant, 3%, 15%, 68%, and 14% received clonidine ER 0.1 mg/day, 0.2 mg/day, 0.3 mg/day, and 0.4 mg/day, respectively. Reduction in ADHD Rating Scale IV (ADHD-RS-IV) scores was greater for the group receiving clonidine ER than for the group receiving placebo in weeks 2–7 but not at week 8. Oddly, although addition of clonidine ER to amphetamine led to significantly greater improvement, addition to methylphenidate did not. No serious adverse events were reported.
In a study by Hunt et al. (1995), 13 children and adolescents with ADHD received guanfacine (mean daily dosage=3.2 mg) for 1 month. Significant improvements in hyperactivity and inattention were found. In an 8-week double-blind, placebo-controlled trial, 34 children and adolescents (ages 7–14 years) with ADHD and tic disorder were randomly assigned to receive guanfacine (dosage range=1.5–3.0 mg/day) or placebo (Scahill et al. 2001). A 37% improvement in ADHD symptoms was reported for children treated with guanfacine, compared with an 8% improvement for children who received placebo. The most common side effects of guanfacine were sedation and dry mouth. There were no significant changes in pulse or blood pressure with guanfacine treatment.
An extended-release formulation of guanfacine (GXR) is also used for the treatment of ADHD (Biederman et al. 2008; Sallee et al. 2009). In a double-blind, placebo-controlled Phase III multicenter trial (Melmed et al. 2006), children and adolescents ages 6–17 years were randomly assigned to placebo or 2 mg, 3 mg, or 4 mg/day of GXR. All three dosages of GXR were superior to placebo in reducing symptoms of ADHD. The most commonly reported side effects of GXR were headache, somnolence, and fatigue. No serious adverse events were reported. In healthy young adults (ages 19–24 years), abrupt discontinuation of 4 mg/day of GXR did not lead to increases in blood pressure or electrocardiogram (ECG) abnormalities (Kisicki et al. 2006).
GXR has also been assessed as add-on therapy when children with ADHD have only a partial response to stimulants (Wilens et al. 2012). In a 9-week double-blind, placebo-controlled dosage optimization study, patients (N=461) continued their stable dosage of psychostimulant given in the morning and were randomized to receive GXR in the morning (GXR AM), GXR in the evening (GXR PM), or placebo. At endpoint, compared with the group receiving placebo plus psychostimulant, each guanfacine treatment group showed significantly greater improvement from baseline on ADHD-RS-IV total scores. Results did not differ between the GXR AM and GXR PM groups in either efficacy or adverse events. There were no serious or unexpected adverse events.
Based on the strength of clinical trial data, stimulants should be regarded as the first line for treatment of ADHD, and generally a different stimulant class should be used if the first stimulant prescribed has failed. Either atomoxetine or guanfacine is an appropriate second-line agent. Stimulants have been combined with atomoxetine in an open trial, with results suggesting that the two agents are superior to atomoxetine alone (Wilens et al. 2009). Placebo-controlled studies have shown that the addition of either guanfacine (Wilens et al. 2012) or clonidine (Kollins et al. 2011) can further reduce symptoms of ADHD after a partial response to psychostimulants. Thus, α2-adrenergic receptor agonists are increasingly used in this adjunctive role.
The stages of pharmacological treatment of ADHD can be summarized as follows:
Stage 1: Psychostimulant
Stage 2: Alternative psychostimulant
Stage 3: Atomoxetine or α2 agonist
Stage 4: Combination of α2 agonist or atomoxetine with stimulant
Oppositional defiant disorder (ODD) and conduct disorder (CD) are highly comorbid with ADHD, particularly in younger children (Pliszka et al. 1999). ODD and CD are syndromes, whereas aggression is a symptom. Although many children and adolescents with CD are aggressive, a child can meet criteria for CD without being aggressive. Also, aggression may present as a problematic symptom in children with depression, psychosis, or bipolar disorder without the child meeting criteria for CD. Thus, the clinician must be clear whether the target of treatment is the syndrome of ODD or CD or the symptom of aggression, because studies have addressed the problems separately. Treatments for ADHD have been used to target ODD and CD, whereas mood stabilizers and antipsychotics have been used in patients with severe aggressive outbursts, regardless of diagnosis (Pappadopulos et al. 2006).
In a 5-week double-blind, placebo-controlled trial of methylphenidate in 84 youths (ages 6–15 years) with CD (with and without ADHD), ratings of antisocial behaviors specific to CD were significantly reduced by methylphenidate treatment (up to 60 mg/day) (Klein et al. 1997). The severity of the ADHD did not affect the response of CD symptoms in the stimulant study. Since that study, multiple double-blind, placebo-controlled trials have shown that ODD responds to stimulant medication, yielding an effect size similar to that for the ADHD symptoms (Spencer et al. 2006a).
Children and adolescents (ages 8–18 years) with ADHD were treated for approximately 8 weeks with placebo or atomoxetine under randomized, double-blind conditions. Of the 293 subjects, 39% were diagnosed with comorbid ODD and 61% were not (Newcorn et al. 2005). Treatment group differences and differences between patients with and without comorbid ODD were examined post hoc for changes on numerous clinical measures. Treatment response was similar in youth with and without ODD, although the comorbid group may require higher dosages to achieve response than those with ADHD alone.
In general, a child with ODD or CD should be treated with a stimulant or atomoxetine before proceeding to the use of other psychotropic agents. The use of more potent agents (mood stabilizers, antipsychotics) is generally reserved for those with severe aggression, and then only after a behavioral treatment has failed.
In a meta-analysis of the literature from 1970 to 2001 that examined 28 studies to determine the effect size for stimulants on overt and covert aggression-related behaviors in children with ADHD, it was found that the mean effect size for aggressive behaviors was similar to that for core behaviors of ADHD (Connor et al. 2002).
A significant body of research has accumulated showing the effectiveness of risperidone in the treatment of aggression, although most of these studies involve patients with subaverage intelligence (Snyder et al. 2002); 80% of subjects had comorbid ADHD. Risperidone dosages ranged from 0.02 to 0.06 mg/kg/day. The risperidone-treated subjects showed a significant (P<0.001) reduction (47.3%) in conduct problems compared with placebo-treated subjects (20.9%). The effect of risperidone was unaffected by diagnosis, presence versus absence of ADHD, psychostimulant use, and IQ status. Risperidone produced no changes in the cognitive variables, and the most common side effects were somnolence, headache, appetite increase, and dyspepsia. Somnolence did not predict response of aggressive symptoms. Side effects related to EPS were reported in 7 (13.2%) and 3 (5.3%) of the subjects in the risperidone and placebo groups, respectively (P=0.245).
Other double-blind, placebo-controlled trials of risperidone in children and adolescents with disruptive behavior disorders (and subaverage IQ) have yielded similar results, with no negative trials reported (Aman et al. 2002; Buitelaar et al. 2001; LeBlanc et al. 2005). Weight gain was a significant side effect in these studies, but there has not been evidence of adverse neuropsychological effects (Günther et al. 2006). Addition of risperidone to a stimulant does not appear to increase rates of side effects and enhances treatment of hyperactivity (Aman et al. 2004). Indeed, adding risperidone to a stimulant to control aggression has become a common practice, although a controlled study showed that aggression was equally reduced when either placebo or risperidone was added to psychostimulant medication (Armenteros et al. 2007). The sample in this study was small (N=25), but the study should caution clinicians that aggression can respond to psychosocial events, such as the expectations of a study.
Pandina et al. (2006) wrote a full review of all studies of risperidone in the treatment of childhood aggression. This review pooled adverse-event data from these studies, showing the most common side effects of risperidone to be somnolence (33%), weight gain (20%), hyperprolactinemia (10.2%), and fatigue (10%). In the pooled studies, there was an excess mean weight gain (over normal growth) of 6.0±7 kg after 35–43 weeks of treatment. Of the 688 patients, 651 were free of dyskinetic movements at baseline, and only 1 patient developed new dyskinetic movements during the follow-up period (these symptoms resolved even though risperidone was continued). There was no worsening of dyskinetic movements in those patients with such preexisting symptoms. Rates of EPS were low throughout the long-term follow-up period. It should be noted that the dosages of risperidone used in these studies were quite low (1–2 mg/day); thus, these results may not apply to dosages in the 6-mg/day range.
Aman et al. (2014) tested the efficacy of adding risperidone to concurrent psychostimulant treatment and parent training (PT) in behavior management in children with ADHD and severe comorbid aggression. Children ages 6–12 years (mean age 8.89 years) with severe physical aggression were randomly assigned to a 9-week trial of PT, stimulant, and placebo (basic treatment; n=84) or of PT, stimulant, and risperidone (augmented treatment; n=84). Children received a psychostimulant for 3 weeks, titrated for optimal effect, while parents received PT. If there was room for improvement at the end of week 3, placebo or risperidone was added for an additional 3 weeks. Both groups showed substantial reductions in aggressive behavior during the first 3 weeks of stimulant plus PT treatment. In the second phase, the risperidone group showed significantly greater reduction in rating scale measures of aggressive behavior, but CGI-I scores did not discriminate the groups. Adverse events were as expected, with the risperidone group experiencing greater weight gain and increases in serum lipid and prolactin levels than the placebo group. The study clearly showed the need to adequately treat ADHD in the aggressive child before prescribing second-generation antipsychotics.
There have been only small open trials and case reports of the use of quetiapine, aripiprazole, olanzapine, and ziprasidone in the treatment of aggression (Findling et al. 2006; Hazaray et al. 2004; Khan and Mican 2006; Rugino and Janvier 2005; Staller 2004; Stephens et al. 2004; Valicenti-McDermott and Demb 2006). In most of these studies, children had primary psychiatric diagnoses other than ODD or CD, such as mood disorders or developmental disorders.
The efficacy of lithium in the treatment of CD in youth has been demonstrated in three double-blind, placebo-controlled studies. Haloperidol, lithium, and placebo were compared in a double-blind, randomized trial in 61 hospitalized children (ages 5–12 years) with aggression and CD (Campbell et al. 1984). The optimal dosages of haloperidol ranged from 1 to 6 mg/day; the optimal dosages of lithium ranged from 500 to 2,000 mg/day. Both haloperidol and lithium were found to be significantly superior to placebo in reducing aggression. However, there were more adverse effects associated with haloperidol than with lithium, including excessive sedation, acute dystonic reaction, and drooling. Stomachache, headache, and tremor were more common with lithium than with haloperidol.
In a subsequent study, Campbell et al. (1995) conducted a 6-week double-blind, placebo-controlled trial of lithium treatment for 50 hospitalized children (ages 5–12 years) with aggression and CD. The mean optimal daily dosage of lithium was 1,248 mg, and the mean serum level was 1.12 mEq/L. Lithium was significantly superior to placebo in reducing aggression. The most common lithium side effects were stomachache, nausea, vomiting, headache, tremor, and urinary frequency.
Eighty-six inpatients (ages 10–17 years) with CD were randomly assigned to treatment with lithium (mean daily dosage=1,425 mg; mean serum level=1.07 mmol/L) or placebo in a 4-week double-blind trial. Aggression ratings decreased significantly for the group treated with lithium, compared with the group treated with placebo. More than 50% of patients in the lithium group experienced nausea, vomiting, and urinary frequency (Malone et al. 2000).
In contrast, Rifkin et al. (1997) found no significant differences between lithium and placebo in aggression ratings in a 2-week double-blind study of 33 inpatients with CD. The short duration of treatment may have accounted for the lack of efficacy, suggesting that a 4- to 6-week trial is necessary to show response.
Severe mood dysregulation, referred to in DSM-5 as disruptive mood dysregulation disorder, is often characterized by severe aggressive outbursts. Dickstein et al. (2009) studied the effect of lithium in youths (ages 7–17 years) with severe mood dysregulation who were tapered off their medications. Those who continued to have significant mood dysregulation after a 2-week single-blind placebo run-in were randomized to a 6-week double-blind trial of either lithium (n=14) or placebo (n=11). Interestingly, 20 of 45 youths (45%) with severe mood dysregulation were not randomized due to significant clinical improvement during the placebo run-in. Among randomized patients, there were no significant between-group differences in any clinical measures. The placebo group showed no improvement at all (a true negative trial), with no evidence that an increase in sample size would have changed the results.
Twenty outpatient children and adolescents (ages 10–18 years) with CD or ODD were randomly assigned to divalproex (dosage range=750–1,500 mg/day; mean blood level=82 μg/mL) or placebo in a 6-week double-blind, placebo-controlled crossover study. Of the 15 patients who completed both phases, 12 (80%) had a statistically significant superior response to divalproex. Increased appetite was the only significant side effect (Donovan et al. 2000). Steiner et al. (2003) randomly assigned 71 adolescents with CD in a residential facility for juvenile offenders to two groups, which received either a therapeutic dosage or a low dosage of divalproex for 7 weeks; both subjects and outcome raters were blind to treatment status. Reduction in aggression severity (P=0.02), improvement in impulse control (P<0.05), and global improvement (P=0.0008) were greater in the group with therapeutic divalproex levels than in the low-dosage condition. As reported in another article from the same study, serum level and “immature defenses” (as assessed by the Weinberger Adjustment Inventory) predicted response to divalproex, but psychiatric comorbidity did not (Saxena et al. 2005).
Blader et al. (2009) treated 74 children with ADHD and aggression with open stimulant treatment during a lead-in phase that averaged 5 weeks. Children whose aggressive behavior persisted at the conclusion of the lead-in phase (n=30) were randomly assigned to receive double-blind, flexibly dosed divalproex or a placebo adjunctive to stimulant for 8 weeks. Families received weekly behavioral therapy throughout the trial. A significantly higher proportion of the children randomly assigned to divalproex met remission criteria (8 of 14; 57%) than of those randomly assigned to placebo (2 of 13; 15%).
Clonidine has often been combined with stimulants to treat comorbid aggression in children with ADHD. In a 2-month randomized comparison of clonidine, methylphenidate, and clonidine combined with methylphenidate in the treatment of 24 children and adolescents (ages 6–16 years) with ADHD and CD or ODD, it was found that all three treatment groups showed significant improvement in oppositional and CD symptoms (Connor et al. 2000). No significant ECG changes were noted.
Children (ages 6–14 years) with ADHD currently taking methylphenidate were randomly assigned to receive clonidine syrup 0.10–0.20 mg/day (n=37) or placebo (n=29) for 6 weeks (Hazell and Stuart 2003). Analysis showed that significantly more clonidine-treated children than control subjects were responders on the Conduct subscale (21 of 37 vs. 6 of 29; P<0.01) of the parent-report Conners Behavior Checklist, but not on the Hyperactive Index subscale (13 of 37 vs. 5 of 29). Compared with placebo, clonidine was associated with a greater reduction in systolic blood pressure measured standing and with transient sedation and dizziness. Clonidine-treated individuals had a greater reduction in a number of unwanted effects associated with psychostimulant treatment compared with placebo. The findings supported the use of clonidine in combination with psychostimulant medication to reduce conduct symptoms associated with ADHD.
The Center for the Advancement of Children’s Mental Health at Columbia University joined with the New York State Office of Mental Health to develop guidelines for treatment of aggression, which led to the Treatment Recommendations for the Use of Antipsychotics for Aggressive Youth (TRAAY; Pappadopulos et al. 2003; Schur et al. 2003). These recommendations call first for a thorough psychiatric evaluation of the child with severe aggression. Next, a psychosocial intervention should be used first when the aggression is the primary problem, such as in ODD, CD, or intermittent explosive disorder. The clinician should then treat any primary condition, such as ADHD, psychosis, or mood disorder that may be causing or contributing to the aggression. If the aggression does not respond to these steps, then an atypical antipsychotic should be used. Different atypical antipsychotics should be tried as monotherapy before moving to polypharmacy (e.g., adding a classic mood stabilizer such as lithium or divalproex to the antipsychotic). Monitoring of weight and laboratory measures of glucose, cholesterol, and triglycerides is mandatory (Correll and Carlson 2006).
α2-Adrenergic receptor agonists may be used in children with milder aggression or temper outbursts because the effect size of these medications on aggression is more modest (Hazell and Stuart 2003).
The prevalence of Tourette syndrome is estimated to be 0.7% in children (Comings et al. 1990). Tourette syndrome is characterized by multiple motor tics and by one or more vocal tics that occur frequently for longer than 1 year. More commonly children suffer from chronic motor or vocal tics, but treatment is the same as for Tourette syndrome. A meta-analysis has been performed examining the effects of α2-adrenergic receptor agonists and antipsychotics in the pharmacological treatment of tics (Weisman et al. 2013).
Weisman et al. (2013) identified six placebo-controlled studies examining the effect of clonidine (Du et al. 2008; Leckman et al. 1991; Singer et al. 1995; Tourette’s Syndrome Study Group 2002) or guanfacine (Cummings et al. 2002; Scahill et al. 2001) in patients with tics. A meta-analysis demonstrated a significant effect of an α2 agonist relative to placebo, with a standardized mean difference of 0.31 (95% confidence interval [CI], 0.15–0.48). This is a fairly modest effect. Of note, the analysis showed that clonidine had a significantly larger effect against tics when patients had comorbid ADHD than when patients did not have ADHD.
Weisman et al. (2013) reviewed five placebo-controlled studies of haloperidol and pimozide (Sallee et al. 1997; Shapiro et al. 1989), ziprasidone (Sallee et al. 2000), and risperidone (Dion et al. 2002; Scahill et al. 2003). Two studies directly compared haloperidol and risperidone without a placebo control (Bruggeman et al. 2001; Gilbert et al. 2004), one study compared risperidone and clonidine (Gaffney et al. 2002), and one study compared haloperidol and the clonidine patch (Kang et al. 2009). However, the Kang et al. (2009) study, unlike the others, was not blinded. In the meta-analysis, all antipsychotics were superior to placebo, with a standardized mean difference of 0.61 (95% CI, 0.36–0.86). There was no difference between the various antipsychotic agents with regard to their efficacy for reducing tics. Gaffney et al. (2002) found clonidine and risperidone to be equivalent in efficacy. Studies of aripiprazole were not included in the Weisman et al. (2013) meta-analysis.
Murphy et al. (2005) reported six cases of children and adolescents (age range= 8–19 years; mean age=12.1 years) who had comorbid tic disorder and OCD and were treated with aripiprazole (mean dosage=11.7 mg/day; range=5–20 mg/day) for 12 weeks. The subjects experienced a mean reduction of 56% in the severity of their tics as assessed by the Yale Global Tic Severity Scale (YGTSS). Similarly, Yoo et al. (2006) treated 15 children and adolescents who had tic disorder with aripiprazole (mean dosage=10.89 mg/day; range=12.5–15 mg/day) and reported a mean reduction of 40% in YGTSS scores; side effects were minimal. Two subjects experienced nausea, one experienced weight gain, and one experienced sedation. The sedation responded to dosage reduction.
In a case series, 11 consecutive patients with Tourette syndrome (mean age=7 years; age range=7–50 years) were treated with aripiprazole; the symptoms of the majority of these patients had been refractory to previous treatments with other antipsychotics (Davies et al. 2006). Ten of the 11 patients who were treated with aripiprazole improved, although to variable degrees. In the majority of patients, response was sustained with aripiprazole dosages ranging from 10 to 20 mg/day. Side effects were mild and transient.
Zheng et al. (2016) reviewed six studies of the treatment of tics with aripiprazole. This study included two randomized controlled trials, but in these studies aripiprazole was compared only with tiapride, an antipsychotic not available in the United States. The other four were open trials. The review found aripiprazole to be well tolerated and effective in reducing tics relative to baseline levels.
If a tic is not severe or socially impairing, observation may be in order because tics typically wax and wane; in general, tics improve over time (Leckman 2002). Often, psychosocial treatment such as habit reversal training is highly effective at reducing tics (Piacentini and Chang 2006). If conservative treatment fails, use of an α2 agonist would be desirable, due to the risk of weight gain and dyslipidemia with atypical antipsychotics. The atypical antipsychotics are preferred to typical antipsychotics because the latter have lower efficacy and a higher risk of tardive dyskinesia. Haloperidol and pimozide should be used only as a last resort when several atypical agents have failed.
In children with comorbid ADHD, a stimulant can be used, but a nonstimulant is indicated if the stimulant exacerbates tics (Pliszka et al. 2006a). Stimulants often must be combined with α2 agonists or antipsychotics to control both the ADHD and the tics (Pliszka et al. 2006a; Tourette’s Syndrome Study Group 2002).
Cases of schizophrenia in children younger than age 13 years are very rare; however, the prevalence rises in adolescence, with peak onset between ages 15 and 30 years (McClellan and Werry 2001). The clinical features of the disorder are similar in youth and adults, and the same DSM-5 criteria are used to establish a diagnosis. The outcome of childhood-onset schizophrenia is reported to be poor (Eggers and Bunk 1997).
Five atypical antipsychotics have FDA approval for the treatment of schizophrenia in adolescents: olanzapine (for ages ≥13 years), risperidone (≥13 years), aripiprazole (≥13 years), quetiapine (≥13 years), and paliperidone ER (≥12 years). Ziprasidone, asenapine, and clozapine have been studied for treatment of schizophrenia in youth but do not have FDA approval.
In a double-blind, placebo-controlled multicenter study of olanzapine (mean dosage=11.1 mg/day) for the treatment of adolescents with schizophrenia (Kryzhanovskaya et al. 2009), 107 adolescents were randomly assigned to receive olanzapine (n=72) or placebo (n=35) for a 6-week trial. Olanzapine-treated adolescents had significant improvements on the Brief Psychiatric Rating Scale for Children (BPRS-C) and on CGI-S scores compared with the placebo group. There was no significant difference in response rate (defined as a ≥30% decrease in BPRS-C and a CGI-S score ≤3) between the olanzapine (37.5%) and placebo (25.7%) groups. Significantly more olanzapine-treated adolescents had high aspartate aminotransferase/serum glutamic oxaloacetic transaminase (AST/SGOT), alanine aminotransferase/serum glutamic pyruvic transaminase (ALT/SGPT), and prolactin levels, as well as low bilirubin levels and hematocrit values, during treatment. There was a significant increase in fasting triglycerides at endpoint in the olanzapine-treated adolescents. Early response at 2 and 3 weeks has been shown to predict ultimate response and remission at week 6 (Stentebjerg-Olesen et al. 2015).
Positive findings were reported in a double-blind, placebo-controlled multicenter trial of risperidone in the treatment of adolescents with schizophrenia (Haas et al. 2007). One hundred sixty adolescents were randomly assigned to risperidone 1–3 mg/day (n=55), risperidone 4–6 mg/day (n=51), or placebo (n=54) for a 6-week trial. Both dosage ranges of risperidone were significantly superior to placebo on the primary efficacy measure, the Positive and Negative Syndrome Scale (PANSS) total change score at endpoint. The most common adverse events in the risperidone 1- to 3-mg/day group were somnolence (24%), agitation (15%), and headache (13%). EPS (16%), dizziness (14%), and hypertonia (14%) were the most common adverse events in the risperidone 4- to 6-mg/day group. The investigators concluded that the overall risk–benefit ratio favored the lower dosage range of risperidone.
In a large double-blind, placebo-controlled multicenter trial of aripiprazole for the treatment of schizophrenia in adolescents (Findling et al. 2008), 302 adolescents were randomly assigned to receive aripiprazole 10 mg/day, aripiprazole 30 mg/day (after 5- or 11-day titration), or placebo over a 6-week period. Both dosages of aripiprazole showed statistically significant differences from placebo on the PANSS total score at week 6. The most common adverse events associated with aripiprazole were EPS, somnolence, and tremor. Early response to aripiprazole at weeks 2 and 3 has been shown to predict later response and remission (Correll et al. 2013).
The efficacy of quetiapine in the treatment of schizophrenia in adolescents was evaluated in a 6-week double-blind, placebo-controlled trial in which subjects (N=222) were randomly assigned to quetiapine 400 mg/day, quetiapine 800 mg/day, or placebo (Findling et al. 2012). Both quetiapine dosages were significantly superior to placebo in reduction of total PANSS scores. The most common adverse events with quetiapine were somnolence, headache, and dizziness. Mean changes in body weight, total cholesterol, and total triglycerides were greater in the quetiapine group than the placebo group.
Paliperidone ER was evaluated in a 6-week double-blind, placebo-controlled trial in 201 adolescents (ages 12–17 years) with schizophrenia (Singh et al. 2011). Paliperidone ER fixed dosing was weight based; youth weighing 29 kg to less than 51 kg received 1.5 mg, 3 mg, or 6 mg/day, whereas patients weighing at least 51 kg received 1.5 mg, 6 mg, or 12 mg/day. With weight-based treatment, only the medium dosage range (3–6 mg/day) showed significant improvement compared with placebo. The fixed dosages of 3 mg/day, 6 mg/day, and 12 mg/day were also superior to placebo. Somnolence, akathisia, insomnia, headache, and tremor were the most common adverse events in the paliperidone ER group.
The long-term safety of paliperidone ER was evaluated in a 2-year open-label study for 400 adolescents with schizophrenia (Savitz et al. 2015). The most frequently reported treatment-emergent adverse events were increased weight, headache, insomnia, nasopharyngitis, akathisia, schizophrenia exacerbation, and tremor. There were no clinically significant changes in weight or BMI. Hyperprolactinemia was found in 56% of patients, and 9.3% had prolactin-related adverse events. The most common EPS–related adverse events were parkinsonism (15.5%) and hyperkinesia (13.8%).
The efficacy of ziprasidone was evaluated in a 6-week double-blind, placebo-controlled trial for 283 adolescents ages 13–17 years with schizophrenia (Findling et al. 2013). Ziprasidone was flexibly dosed (40–160 mg/day). There was no significant difference between ziprasidone and placebo on change from baseline to endpoint on the BPRS-Anchored. The most common adverse events in the ziprasidone group were somnolence and EPS. During the 26-week open-label follow-up, there were no clinically significant changes in metabolic indices and laboratory measures.
The efficacy of asenapine was assessed in an 8-week double-blind, placebo-controlled trial for adolescents ages 12–17 years with schizophrenia (Findling et al. 2015a). Asenapine dosing was 2.5 mg bid or 5 mg bid. Asenapine was not significantly superior to placebo on the primary efficacy measure of change from baseline to endpoint on the PANSS total score. Weight gain ≥7%, somnolence, sedation, and hypersomnia were more common in the asenapine group than the placebo group. Adverse events of akathisia, fasting glucose elevation, and EPS were more common in the asenapine 5 mg bid group than the placebo group. Those youth who showed improvement during the acute phase maintained improvement in the 26-week open-label extension.
A 6-week double-blind, placebo-controlled comparison of clozapine and haloperidol was conducted in 21 children and adolescents (ages 6–18 years) with schizophrenia (Kumra et al. 1996). Clozapine (mean dosage=176 mg/day; range=25–525 mg/day) was significantly superior to haloperidol (mean dosage=16 mg/day; range=7–27 mg/day) in reducing positive and negative symptoms of schizophrenia. Clozapine improved interpersonal functioning and enabled patients to live in a less restrictive setting. Side effects, however, were significant with clozapine. One patient had a seizure, and three patients were given anticonvulsants after they became more irritable and aggressive and experienced epileptiform changes on electroencephalogram (EEG). Mild to moderate neutropenia, weight gain, and sinus tachycardia were the other major side effects.
In a 12-week open-label trial, risperidone (mean dosage=1.6 mg/day) was compared with olanzapine (mean dosage=8.2 mg/day) in the treatment of 25 children with schizophrenia (Mozes et al. 2006). Both treatment groups showed similar significant improvement as measured by PANSS total and subscale scores.
In an 8-week double-blind study, 50 children and adolescents (ages 8–19 years) with psychotic disorders were randomly assigned to receive risperidone (mean dosage=4 mg/day), olanzapine (mean dosage=12.3 mg/day), or haloperidol (mean dosage=5 mg/day) (Sikich et al. 2004). Eighty-eight percent of patients treated with olanzapine, 74% treated with risperidone, and 53% treated with haloperidol met response criteria (CGI-I scores of much or very much improved and at least a 20% reduction in BPRS-C total score).
Clozapine was compared with olanzapine in an 8-week double-blind randomized trial (Shaw et al. 2006). Twenty-five youths (ages 7–16 years) with schizophrenia that was resistant to treatment with at least two antipsychotics participated in the trial. Clozapine (mean dosage=327 mg/day) showed significant improvement on all outcome measures, whereas olanzapine (mean dosage=19.1 mg/day) showed improvement on some outcome measures. Improvement in negative symptoms was significantly greater for the clozapine group.
In an 8-week double-blind trial comparing olanzapine and risperidone against the typical antipsychotic molindone in 116 children and adolescents with early-onset schizophrenia or schizoaffective disorder (Sikich et al. 2008), no significant differences were observed in rates of response (CGI-I score ≤2, and ≥20% reduction in PANSS total score) among the treatment groups (risperidone 46%, olanzapine 34%, and molindone 50%). Significantly greater weight gain occurred with olanzapine and risperidone than with molindone.
A double-blind maintenance study followed this acute controlled trial in children and adolescents with early-onset schizophrenia or schizoaffective disorder (Findling et al. 2010). Of the 54 youths eligible for the maintenance phase, 14 (26%) completed 44 weeks of treatment. No significant differences were found among olanzapine, risperidone, and molindone in reduction of symptoms or time to discontinuation.
In an 8-week double-blind trial comparing paliperidone ER and aripiprazole in 228 adolescents with schizophrenia (Savitz et al. 2015), there was no significant difference between the paliperidone ER group and the aripiprazole group on change in PANSS scores from baseline to endpoint. Both medication groups showed improvement in symptoms and functional outcomes.
Haloperidol has been compared with placebo and other typical antipsychotics in controlled trials in youth. In a 10-week double-blind, placebo-controlled crossover study, the safety and efficacy of haloperidol were assessed in 12 hospitalized children (ages 5–12 years) with schizophrenia. Haloperidol (optimal dosage range=0.5–3.5 mg/day) was significantly superior to placebo in improving overall clinical functioning and reducing ideas of reference, delusions, and hallucinations. Common side effects were acute dystonic reaction, drowsiness, and dizziness (Spencer et al. 1992).
Haloperidol and loxapine were compared in a 4-week double-blind, placebo-controlled study of 75 adolescent inpatients with schizophrenia. Both haloperidol and loxapine were significantly superior to placebo, and there was no significant difference in efficacy between the two medications. Response rates (based on CGI-I scores) were 87.5% for loxapine, 70% for haloperidol, and 36.4% for placebo. Common side effects were sedation, EPS, and somnolence (Pool et al. 1976).
Thiothixene was compared with thioridazine in a 6-week single-blind study in 21 adolescent inpatients with schizophrenia. Thiothixene (optimal mean dosage=16.2 mg/day) and thioridazine (optimal mean dosage=178 mg/day) were equally effective in controlling symptoms, although most of the adolescents continued to be quite impaired. Thiothixene was less sedating than thioridazine (Realmuto et al. 1984). Thiothixene was also compared with trifluoperazine in an 8-week double-blind study of 16 children (ages 8–15 years) with schizophrenia (Wolpert et al. 1967). The effects of both medications were similar in terms of decreasing avoidance behavior, reducing stereotypic behavior, and increasing peer socialization.
Both typical and atypical antipsychotics have demonstrated effectiveness in the treatment of schizophrenia in youth, although the sample sizes have been small in the trials of typical antipsychotics. Because fewer EPS and instances of tardive dyskinesia have been reported with atypical antipsychotics, it would be reasonable to initiate treatment with an atypical antipsychotic for a child with schizophrenia. It is important to monitor weight and metabolic parameters for children who receive atypical antipsychotics. Clozapine should be considered for treatment-resistant schizophrenia (McClellan et al. 2013).
Antipsychotics should be administered at adequate dosages for a period of 6 weeks to assess efficacy. If there is no response or if intolerable side effects occur, a trial of a different antipsychotic should be initiated (American Academy of Child and Adolescent Psychiatry 2012).
No data are available to guide maintenance treatment. Because the majority of youth will have a second psychotic episode within 5–7 years of stabilization, there is a significant risk of relapse with medication withdrawal (Kumra 2000). Maintenance treatment should be provided to most youth with schizophrenia to improve functioning and prevent relapse (American Academy of Child and Adolescent Psychiatry 2012). The aim is to use the lowest effective dosage of medication to reduce the risk of adverse events. For those youth with prolonged remission, it may be possible to discontinue medication, but long-term monitoring will be necessary because of possible reemergence of psychotic symptoms.
Autism spectrum disorder is a new DSM-5 disorder that encompasses the previous DSM-IV (American Psychiatric Association 1994) categories of autistic disorder (autism), Asperger’s disorder, and pervasive developmental disorder not otherwise specified. Symptoms of autism spectrum disorder represent a continuum of mild to severe impairments in two core domains: 1) deficits in social communication and social interaction and 2) restricted repetitive patterns of behavior, interests, and activities (American Psychiatric Association 2013). Associated behavioral features include hyperactivity, stereotypies, attentional problems, self-injurious behavior, aggression, mood lability, anxiety, obsessions, and compulsions. Autism spectrum disorder is estimated to have a prevalence of up to 18.7 per 10,000 population (Howlin 2000). The majority of children with the disorder will continue to have significant social and communication impairments throughout adulthood (Buitelaar and Willemsen-Swinkels 2000).
For the treatment of irritability associated with autism spectrum disorder, the FDA has approved risperidone (for patients ≥5 years old) and aripiprazole (≥6 years old). Olanzapine, quetiapine, and ziprasidone have been studied for autism spectrum disorders but do not have FDA approval for use in youth.
One hundred one children (ages 5–17 years) with DSM-IV–defined autistic disorder participated in an 8-week double-blind, placebo-controlled trial of risperidone (dosage range=0.5–3.5 mg/day; mean dosage=1.8 mg/day) (McCracken et al. 2002). A significantly greater positive response (defined as a 25% decrease on the irritability subscale of the Aberrant Behavior Checklist [ABC] and a rating of much or very much improved on the CGI-I) was found for the risperidone group (69%) than the placebo group (12%). Adverse events of increased appetite, fatigue, drowsiness, dizziness, and drooling were more common in the risperidone group than in the placebo group. Mean weight gain was 2.7 kg in the risperidone group and 0.8 kg in the placebo group. An 18-month follow-up showed that the majority of subjects who responded to risperidone during intermediate-length treatment continued to show improvement (McDougle et al. 2005).
In a 24-month follow-up of 84 of these youths, there was continued improvement in maladaptive behavior from baseline (Aman et al. 2015). Social skills improved and irritability decreased for those youths who were taking risperidone at follow-up. Risperidone treatment was associated with enuresis, excessive appetite, and weight gain.
In an 8-week double-blind, placebo-controlled trial, 79 children (ages 5–12 years) with a DSM-IV diagnosis of autistic or other pervasive developmental disorder were randomly assigned to receive either placebo or risperidone (mean dosage=1.5 mg/day). Risperidone-treated patients exhibited a 64% improvement over baseline irritability, compared with a 30.7% improvement in subjects receiving placebo (Shea et al. 2004).
In a 6-month placebo-controlled study of 40 children (ages 2–9 years) with autism, risperidone (1 mg/day) decreased aggressiveness, hyperactivity, and irritability and improved social responsiveness and nonverbal communication. Appetite increase, weight gain, sedation, and transient dyskinesias were reported in the risperidone-treated children (Nagaraj et al. 2006).
The long-term effects of risperidone were assessed in youths (ages 5–17 years) with autism spectrum disorder (Troost et al. 2005). Twenty-four youths received risperidone for 6 months, followed by a double-blind discontinuation to placebo or continued risperidone. Risperidone was superior to placebo in preventing relapse, with relapse rates of 25% and 75%, respectively.
In a double-blind, placebo-controlled risperidone dosing study for 96 youths (ages 5–17 years), high-dosage risperidone (1.25 mg/day for youths weighing 20 to <45 kg; 1.75 mg/day for youths weighing ≥45 kg) was significantly superior to placebo in baseline-to-endpoint change in score on the ABC Irritability subscale (Kent et al. 2013b). There was no significant difference, however, between scores of patients given low-dosage risperidone (0.125 mg/day for youths weighing 20 to <45 kg; 0.175 mg/day for youths weighing ≥45 kg) and scores of patients given placebo. In an open-label extension study with flexibly dosed risperidone, all groups showed additional improvement in efficacy scores (Kent et al. 2013a).
The comparative efficacy of risperidone and haloperidol was assessed in an 8-week double-blind trial that included 30 children and adolescents with DSM-IV autistic disorder (Miral et al. 2008). Risperidone produced significantly greater reductions in scores on the ABC as well as on other scales used to assess symptoms of autism. An open-label continuation study (Gencer et al. 2008) of this controlled trial showed that risperidone-treated patients had greater improvement than haloperidol-treated patients on CGI ratings and on the ABC.
In a 24-week randomized trial, combined treatment of risperidone and PT was compared with risperidone alone for the treatment of severe behavioral problems in 124 children with autism spectrum disorder (Scahill et al. 2012). Combined treatment was superior to medication alone in reduction of noncompliant behavior as assessed by the Home Situation Questionnaire (HSQ). In a 1-year follow-up, there was no significant difference between treatment groups on noncompliant behavior (Arnold et al. 2012).
The efficacy of aripiprazole in the treatment of irritability was evaluated in 218 children and adolescents with DSM-IV autistic disorder in an 8-week double-blind, placebo-controlled trial (Marcus et al. 2009). Aripiprazole dosages were 5, 10, or 15 mg/day. Compared with placebo, all aripiprazole dosages produced significantly greater improvement on mean scores on the ABC Irritability subscale. The most common adverse events were sedation, tremor, and somnolence. The efficacy of aripiprazole was assessed in another 8-week double-blind, placebo-controlled trial that included 98 children and adolescents with DSM-IV autistic disorder (Owen et al. 2009). Aripiprazole dosages were 5, 10, or 15 mg/day. Significantly greater improvements in mean scores on the ABC Irritability subscale were seen with aripiprazole versus placebo. Rates of EPS-related adverse events were 14.9% with aripiprazole and 8% with placebo.
Long-term maintenance treatment with aripiprazole was examined in a double-blind, placebo-controlled relapse-prevention trial for 86 youths with autism spectrum disorder (Findling et al. 2014). There was no statistically significant difference in time to relapse between aripiprazole and placebo. Relapse rates at week 16 were 35% for the aripiprazole group and 52% for the placebo group.
The efficacy of aripiprazole and risperidone were compared in a 2-month randomized, double-blind trial that included 59 children and adolescents with autism spectrum disorder (Ghanizadeh et al. 2014). Both aripiprazole and risperidone groups had lower scores on the ABC; there was no statistically significant difference between these medications.
The efficacy of olanzapine was evaluated in an 8-week double-blind, placebo-controlled trial that included 11 children and adolescents with pervasive developmental disorders (Hollander et al. 2006). Rates of response (CGI-I score ≤2) were 50% for olanzapine-treated patients and 20% for placebo-treated patients. Olanzapine-treated patients experienced significantly greater weight gain (mean 7.5 lbs for olanzapine vs. 1.5 lbs for placebo-treated patients).
In a 12-week open-label study of olanzapine (mean dosage=7.8 mg/day) in eight patients (ages 5–42 years) with DSM-IV autistic disorder or pervasive developmental disorder not otherwise specified, the six patients who completed the trial showed much or very much global improvement (Potenza et al. 1999). Significant improvements were found in hyperactivity, social relatedness, affectual responses, sensory responses, language usage, self-injurious behavior, aggression, irritability, anxiety, and depression. The most significant adverse effects were increased appetite and weight gain in six patients and sedation in three patients.
In a 3-month open study of olanzapine (dosage range=1.25–20 mg/day) in 25 subjects (ages 6–16 years) with pervasive developmental disorder, significant global improvement was reported. The most common side effect was weight gain (mean=4.8 kg) (Kemner et al. 2002).
Olanzapine was compared with haloperidol in a 6-week open trial in 12 children (ages 4–11 years) with DSM-IV autistic disorder (Malone et al. 2001). Both the olanzapine treatment (mean dosage=7.9 mg/day) and the haloperidol treatment (mean dosage=1.4 mg/day) reduced symptoms of social withdrawal and stereotypies and improved speech and object relations.
The effectiveness of quetiapine (dosage range=100–350 mg/day) was assessed in a 16-week open-label trial in six children with DSM-IV autistic disorder (Martin et al. 1999). No significant behavioral improvements were found from baseline to endpoint.
In a 12-week open-label study of quetiapine in nine adolescents (mean age=14.6 years) with DSM-IV autistic disorder, only two patients were much or very much improved at study endpoint (Findling et al. 2004).
The efficacy of ziprasidone was evaluated in a 6-week open-label study in 12 adolescents with autism (Malone et al. 2007). Ziprasidone dosages ranged from 20 mg/day to 160 mg/day (mean dosage=98.3 mg/day). Of the 12 patients, 9 (75%) were considered treatment responders (based on CGI-I scores ≤2). The mean change in QTc from baseline to endpoint was 14.7, which was a statistically significant increase. There was no significant weight change.
The efficacy and safety of ziprasidone in children, adolescents, and young adults with autism were evaluated in an open-label study in which 12 patients (ages 8–20 years) were treated with ziprasidone (mean daily dosage=59.23 mg) for at least 6 weeks (McDougle et al. 2002). Fifty percent of patients were responders based on a CGI rating of much improved or very much improved. Transient sedation was the most common side effect.
Haloperidol has been the most widely studied typical antipsychotic for the treatment of autism in children and adolescents. In double-blind, placebo-controlled studies, haloperidol has been shown to be significantly superior to placebo in reducing maladaptive behaviors and facilitating learning on discrimination tasks (Campbell et al. 1982); in increasing retention of discrimination learning and decreasing maladaptive behaviors in the classroom (Anderson et al. 1984); in decreasing occurrence of stereotypies and increasing orienting reactions of children (Cohen et al. 1980); and in decreasing hyperactivity, temper tantrums, withdrawal, and stereotypies and increasing relatedness (Anderson et al. 1989). Optimal dosages of haloperidol in these studies ranged from 0.25 to 4 mg/day. The most common side effects were sedation, increased irritability, and acute dystonic reactions. Weight gain was modest (0.2 kg) in autistic children who received haloperidol 0.25–3.5 mg/day for a 6-month period (Silva et al. 1993).
The long-term efficacy of haloperidol was assessed in 48 children (ages 2–8 years) with autism who received haloperidol for 6 months (Perry et al. 1989). Haloperidol remained effective throughout the 6-month treatment period, and it was equally effective whether it was given continuously or on a discontinuous schedule consisting of 5 days on haloperidol and 2 days on placebo. Children who had symptoms of irritability, angry and labile affect, and uncooperativeness were the best responders to haloperidol.
The efficacy of liquid fluoxetine in treating repetitive behaviors in autism was assessed in 45 children and adolescents with autism spectrum disorder. Subjects were randomly assigned to two 8-week acute phases in a double-blind crossover study (Hollander et al. 2005). Low-dosage liquid fluoxetine (mean dosage=9.9 mg/day) was superior to placebo in reducing repetitive behaviors.
In an open study of fluoxetine (dosage range=20 mg every other day to 80 mg/day) in 23 patients (ages 7–28 years) with autistic disorder, 15 patients (65%) experienced significant clinical global improvement (Cook et al. 1992). The most common side effects were restlessness, hyperactivity, agitation, decreased appetite, and insomnia.
A double-blind, placebo-controlled study of fluvoxamine treatment (mean dosage=106.9 mg/day) in 34 children and adolescents with DSM-IV autistic disorder did not find significant clinical improvement with fluvoxamine (McDougle et al. 2000).
Open-label sertraline (dosage range=25–50 mg/day) was administered to nine children with DSM-IV autistic disorder (Steingard et al. 1997). Eight of the nine patients showed clinically significant improvement in ability to tolerate changes in their routine or environment without displaying symptoms of anxiety, irritability, or agitation.
The efficacy of citalopram was evaluated for treatment of repetitive behavior in children with autism spectrum disorder (King et al. 2009). One hundred forty-nine children and adolescents with autism spectrum disorder were randomly assigned to receive citalopram or placebo in a 12-week controlled trial. The mean dosage of citalopram at endpoint was 16.5 mg/day. There was no significant difference in rates of response (CGI-I score ≤2) between citalopram-treated patients (39.2%) and placebo-treated patients (34.2%). No difference was found in reduction in CY-BOCS scores between the citalopram group and the placebo group. Adverse events that were significantly more common in the citalopram group were increased energy level, impulsiveness, hyperactivity, stereotypy, decreased concentration, diarrhea, insomnia, and dry skin or pruritus.
In a 10-week open-label study, 28 children and adolescents (ages 6–17 years) with pervasive developmental disorder received escitalopram. There was significant improvement in irritability and clinical global functioning. Twenty-five percent of youths responded at escitalopram daily dosages less than 10 mg, and 36% of youths responded at dosages greater than or equal to 10 mg (Owley et al. 2005).
Controlled trials with clomipramine in the treatment of autism spectrum disorder have yielded mixed results. Clomipramine and haloperidol were compared in a placebo-controlled crossover study for 7 weeks with active treatment (Remington et al. 2001). Thirty-six patients (ages 10–36 years) with DSM-IV autistic disorder were randomly assigned to clomipramine (mean dosage=128.4 mg/day; range=100–150 mg/day), haloperidol (mean dosage=1.3 mg/day; range=1–1.5 mg/day), or placebo. A significant advantage for haloperidol was found on global measures of autism symptom severity and on specific measures of irritability and hyperactivity. Clomipramine was comparable to haloperidol only among patients who were able to complete a full therapeutic trial. However, significantly fewer patients receiving clomipramine versus haloperidol were able to complete the trial (37.5% vs. 69.7%, respectively) for reasons related to inefficacy, side effects, or behavioral problems.
In an open-label study of mirtazapine (dosage range=7.5–45 mg/day; mean=30.3 mg/day) in 26 patients (ages 3–23 years) with pervasive developmental disorders, 9 patients (34.6%) were judged much or very much improved in symptoms of aggression, self-injury, irritability, hyperactivity, anxiety, depression, and insomnia (Posey et al. 2001). Mirtazapine did not improve symptoms of social or communication impairment. Common side effects included increased appetite, irritability, and transient sedation.
The effectiveness of venlafaxine was assessed in an open retrospective study of 10 patients (ages 3–21 years) with pervasive developmental disorders (Hollander et al. 2000). Six of 10 patients who received venlafaxine (mean dosage=24.4 mg/day; range=6.25–50 mg/day) over an average of 5 months were much or very much improved. Improvements were observed in repetitive behaviors, restricted interests, social deficits, communication and language function, inattention, and hyperactivity. Side effects of venlafaxine included behavioral activation, nausea, inattention, and polyuria.
Eleven adolescents with autism spectrum disorder with depressive and ADHD symptoms were treated with reboxetine (maximum dosage 4 mg/day) in a 12-week open-label trial (Golubchik et al. 2013). Significant, but modest, decreases in depressive and ADHD symptoms were found with reboxetine treatment. Irritability and insomnia were adverse events.
Twenty-eight children (ages 3–11 years) with DSM-IV autistic disorder participated in a double-blind, placebo-controlled study of lamotrigine (mean main-tenance dosage=5 mg/kg/day) for a 12-week study period (Belsito et al. 2001). There were no significant differences between the lamotrigine and placebo groups on severity of behavioral symptoms. Insomnia and hyperactivity were the most frequently reported side effects. No children in the study were withdrawn because of rash.
Three double-blind, placebo-controlled trials have been conducted to assess the efficacy of valproate in the treatment of autism spectrum disorder. In one 8-week trial with 13 children, divalproex was superior to placebo in improvement in repetitive behaviors (Hollander et al. 2006). In another 8-week trial that included 27 youths with autism spectrum disorder, divalproex significantly reduced symptoms of irritability compared with placebo (Hollander et al. 2010). However, valproate was not superior to placebo in reduction of aggression and irritability for 30 youths with pervasive developmental disorders who participated in an 8-week trial (Hellings et al. 2005).
In a retrospective case series of 30 youths (ages 5–21 years) with autism spectrum disorder, 14 patients (47%) who were treated with oxcarbazepine (mean final dosage=1,360 mg/day) had CGI-I scores of 2 or lower on ratings of irritability/agitation symptoms (Douglas et al. 2013).
A 10-week double-blind, placebo-controlled trial was conducted to assess the efficacy of levetiracetam in the treatment of 20 children with autism (Wasserman et al. 2006). The mean maximum dosage of levetiracetam was 862.50 mg/day. There were no significant differences between levetiracetam and placebo on measures of global improvement of autism, aggression and affective instability, and impulsivity and hyperactivity.
In a retrospective chart review of 30 children with autism spectrum disorder who were treated with lithium (mean blood level 0.70 mEq/L), 13 youths (43%) were rated as improved on the CGI-I (Siegel et al. 2014). Vomiting, tremor, fatigue, irritability, and enuresis were the most common adverse effects.
A double-blind, placebo-controlled crossover study of transdermal clonidine (0.005 mg/kg/day or placebo by a weekly transdermal patch) in nine patients (ages 5–33 years) with autistic disorder was conducted for a total active period of 8 weeks (Fankhauser et al. 1992). Significant improvement with clonidine, compared with placebo, was found on measures of social relationship, affectual responses, and sensory responses. In a double-blind, placebo-controlled crossover trial of clonidine in eight children with autistic disorder, clonidine was found to be modestly effective in reducing irritability and hyperactivity (Jaselskis et al. 1992).
The efficacy of GXR was assessed in an 8-week randomized, placebo-controlled trial that included 62 children with autism spectrum disorder (Scahill et al. 2015). The GXR modal dose at week 8 was 3 mg/day. The GXR group showed significantly greater decline (43.6%) in scores on the ABC Hyperactivity subscale compared with the placebo group (13.2%). The most common adverse events were drowsiness, fatigue, and decreased appetite.
A meta-analysis of four methylphenidate trials for treatment of ADHD symptoms in children with pervasive developmental disorders showed an effect size of 0.67 (Reichow et al. 2013). The most likely adverse events were decreased appetite, insomnia, depressive symptoms, irritability, and social withdrawal. Us of an extended-release methylphenidate formulation has also been shown to reduce hyperactive and impulsive behavior in children with autism spectrum disorder and ADHD symptoms (Pearson et al. 2013).
Atomoxetine as a treatment for symptoms of ADHD in children and adolescents with autism spectrum disorder was examined in a double-blind, placebo-controlled 8-week trial that included 97 youths (Harfterkamp et al. 2012). The atomoxetine dosage was 1.2 mg/kg/day. There was a statistically significant difference between atomoxetine and placebo on change from baseline to endpoint ADHD-RS-IV scores. Adverse effects of nausea, decrease in appetite, fatigue, and early morning awakening were more common in the atomoxetine group than in the placebo group. In a subsequent analysis, atomoxetine did not improve social functioning, but there was some improvement on stereotyped behaviors, inappropriate speech, and fear of change (Harfterkamp et al. 2014).
In a placebo-controlled crossover trial that included 16 youths with autism spectrum disorder and ADHD symptoms, atomoxetine was superior to placebo in reduction of hyperactivity and impulsivity symptoms as measured on the ABC (Arnold et al. 2006).
Long-term efficacy and tolerability of atomoxetine were assessed in a 20-week follow-up of an 8-week controlled trial for 88 youths with autism spectrum disorder and autism (Harfterkamp et al. 2013). Continued treatment with atomoxetine resulted in further improvement of ADHD symptoms. Adverse events, particularly nausea and fatigue, diminished over time with continued treatment.
The efficacy of NAC in the treatment of behavioral disturbance was assessed in 33 children with autism in a 12-week double-blind, placebo-controlled trial (Hardan et al. 2012). The NAC dosage was titrated up to 900 mg three times daily. Compared with placebo, NAC resulted in significant improvement on scores on the ABC Irritability subscale. The most common adverse effects were constipation, nausea and vomiting, and diarrhea. One participant treated with NAC had worsening agitation and irritability and was removed from the study.
The efficacy of bumetanide for the treatment of 60 children with autism spectrum disorder was evaluated in a 3-month double-blind, placebo-controlled trial (Lemonnier et al. 2012). The bumetanide dosage was 1 mg/day. Bumetanide was significantly superior to placebo in reduction of symptoms on the primary outcome measure, the Childhood Autism Rating Scale (CARS). Occasional mild hypokalemia was an adverse event.
Intranasal oxytocin has been examined as a treatment for autism spectrum disorder in controlled trials and a long-term open-label study. Oxytocin nasal spray (18 or 24 international units [IU]) or placebo was administered to 16 youths were autism spectrum disorder in a double-blind, placebo-controlled crossover design (Guastella et al. 2010). Compared with placebo, oxytocin significantly improved emotion recognition.
In a double-blind, placebo-controlled trial, 36 male youths with autism spectrum disorder received 24 or 12 IU of oxytocin or placebo over a 4-day period (Dadds et al. 2014). There were no significant differences between intranasal oxytocin and placebo on measures of social interaction skills, repetitive behaviors, and emotion recognition.
In a 7-month open-label study, intranasal oxytocin was administered (8- to 24-IU dose every 2 months) to eight male youths with autism spectrum disorder (Tachibana et al. 2013). Six of eight youths showed improvement in communication and social interaction on the Autism Diagnostic Observation Schedule—Generic (ADOS-G).
The efficacy of arbaclofen was assessed in a 12-week double-blind, placebo-controlled trial that included 150 children, adolescents, and young adults (Delahunty et al. 2013). Arbaclofen was titrated to a maximum of 10 mg three times daily for youths ages 5–11 years and 15 mg three times daily for youths ages 12–21 years. There was no significant difference between arbaclofen and placebo in improving lethargy or social withdrawal.
Memantine as adjunctive treatment to risperidone for autism spectrum disorder was assessed in a 10-week double-blind, placebo-controlled trial for 40 children (Ghaleiha et al. 2013a). The dosage of memantine was titrated to 20 mg/day, and the risperidone dosage was titrated to 3 mg/day. Compared with the placebo group, the group that received adjunctive memantine had a significantly greater reduction in the Irritability score on the ABC—Community version. There was no significant difference in adverse effects between the groups.
Amantadine as adjunctive treatment to risperidone for treatment of 40 children with autism spectrum disorder was evaluated in a 10-week double-blind, placebo-controlled trial (Mohammadi et al. 2013). Amantadine was titrated to 100–150 mg/day, and risperidone was titrated to 1–2 mg/day. Compared with the placebo group, the adjunctive amantadine group had a significantly greater reduction in Hyperactivity and Irritability scores on the ABC—Community version. There were no significant adverse effects among the groups.
The efficacy of riluzole as adjunctive treatment to risperidone was assessed in a 10-week double-blind, placebo-controlled trial for 40 children with autism spectrum disorder (Ghaleiha et al. 2013b). Riluzole was titrated to 50–100 mg/day, and risperidone was titrated to 2–3 mg/day. Significantly greater improvement in irritability, as assessed by the ABC—Community version, was found for the adjunctive riluzole group than the placebo group. Increased appetite and weight gain were more common in the riluzole group than the placebo group.
Buspirone as adjunctive treatment to risperidone was evaluated in an 8-week double-blind, placebo-controlled trial for 40 youths with autism (Ghanizadeh and Ayoobzadehshirazi 2015). The mean dosage of buspirone was 6.7 mg/day. Compared with the placebo group, the adjunctive buspirone group showed a significantly greater reduction in the Irritability subscale score of the ABC—Community version. The most common adverse events in the buspirone group were increased appetite, drowsiness, and fatigue.
Adjunctive treatment with celecoxib to risperidone was evaluated in a 10-week double-blind, placebo-controlled trial that included 40 children with autism (Asadabadi et al. 2013). Celecoxib was titrated to 300 mg/day and risperidone to 3 mg/day. Adjunctive celecoxib was superior to placebo in reducing irritability, social withdrawal, and stereotypy as measured on the ABC—Community version.
The efficacy of adjunctive pentoxifylline to risperidone was assessed in a 10-week double-blind, placebo-controlled trial that included 40 children with a DSM-IV-TR (American Psychiatric Association 2000) diagnosis of autistic disorder (Akhondzadeh et al. 2010). Pentoxifylline was titrated to 600 mg/day, and risperidone was titrated to 3 mg/day. Adjunctive pentoxifylline was superior to placebo in reducing scores on the ABC—Community version subscales for irritability, lethargy/social withdrawal, stereotypic behavior, hyperactivity/noncompliance, and inappropriate speech. There was no significant difference between the groups in adverse events.
NAC as adjunctive treatment to risperidone for treatment of irritability in autism spectrum disorder has been evaluated in two double-blind, placebo-controlled studies. In a 10-week study that included 40 children with autism spectrum disorder, the NAC dosage was 600–900 mg/day (Nikoo et al. 2015). In an 8-week study that included 40 children with autism spectrum disorder, the NAC dosage was 1,200 mg/day (Ghanizadeh and Moghimi-Sarani 2013). Adjunctive NAC was superior to placebo in the reduction of irritability as assessed by the ABC in both studies.
Although there is no evidence that pharmacotherapy is effective in treating the core social and communication deficits in autism spectrum disorder, medications have been shown to be useful in treating associated symptoms, such as hyperactivity, inattention, stereotypies, self-injurious behavior, tantrums, aggression, mood lability, and anxiety. Antipsychotics may decrease withdrawal, stereotypies, and aggression and may facilitate learning. To date, most available data support the use of risperidone or aripiprazole for treating irritability, aggression, self-injurious behavior, temper tantrums, and mood lability associated with autism spectrum disorder in children and adolescents. Serotonin reuptake inhibitors and other antidepressants have been shown to reduce compulsions, anxiety, and depression in children with autism spectrum disorder. Other agents that may be beneficial to treat associated symptoms are α2 agonists, stimulants, mood stabilizers, and norepinephrine reuptake inhibitors (Volkmar et al. 2014).
Limited data are available on the long-term use of pharmacotherapy in children with autism spectrum disorder. After receiving an intermediate-length (4- to 6-month) course of treatment with risperidone, children withdrawn from the medication through placebo substitution had high relapse rates (Research Units on Pediatric Psychopharmacology Autism Network 2005; Troost et al. 2005). Therefore, clinicians must weigh the risk–benefit ratio of maintenance medication treatment in this population and carefully monitor children for side effects.