The stimulants, especially the short- and long-acting forms of methylphenidate and amphetamine, are first-line treatments for attention deficit hyperactivity disorder (ADHD). In its classic form, ADHD is characterized by inattention, impulsiveness, and hyperactivity, though current convention includes primarily inattentive and impulsive/hyperactive subtypes (26). Epidemiological studies indicate that ADHD is relatively common in childhood, affecting 2% to 10% of school-aged children (27,28). Boys
are affected more often than girls. It persists into adolescence in a majority of cases (29), and into adulthood in as many as 30% to 40% of cases (30). Studies of clinical populations indicate that the symptoms of ADHD are among the most common reasons for referral of children to mental health agencies (3). Nonetheless, there also is a substantial number of affected children who are not receiving treatment (31). Given the high prevalence of ADHD in school-age children, the potential for long-term functional disability and the high health-related costs associated with the disorder (32), ADHD is a major public health concern.

The most commonly used stimulants for the treatment of ADHD include methylphenidate, dextroamphetamine, and the mixed preparation of D,L-amphetamine (33). Immediate-release methylphenidate has been studied more carefully than the other stimulants and remains the most commonly used agent in clinical practice— though the longer acting preparations are growing in market share. Although less well studied, the amphetamine products and extended-release formulations of methylphenidate have short-term efficacy and safety profiles that are comparable to methylphenidate (34,35) (Table 6.1.2.1).

The empirical basis for the use of stimulants in children with ADHD rests on findings from hundreds of short-term, randomized, placebo-controlled studies conducted over the past 30 to 40 years (33). Results from controlled studies over the last decade provide additional information about dose response, similarities and differences in response across stimulant preparations, and the importance of regular clinical monitoring to achieve optimal response. Rapport and colleagues (36) conducted a dose-response study in 76 subjects (66 boys and 10 girls) between the ages of 6 and 11 years. The 5-week study used 4 dose levels of methylphenidate (5 mg, 10 mg, 15 mg, and 20 mg given bid) and placebo given in random order in a crossover design. All dose levels of active medication were superior to placebo. In addition, there was a clear linear trend, such that classroom behavior and the number of completed assignments improved with each increase in dose. However, the incremental improvement was smaller as the dose moved above the 10 mg dose level.

In order to compare methylphenidate and d-amphetamine, a placebo-controlled crossover study was conducted in 48 boys between 6 and 12 years of age (37). Subjects were treated with methylphenidate, d-amphetamine, or placebo in random order for 3 weeks in each treatment condition. During each condition the dose was increased on a weekly basis (e.g., methylphenidate doses for children under 30 kg of body weight were 12.5 mg, 20 mg, and 35 mg at breakfast and lunchtime for the respective weeks). Based on a global rating of response, 79% of the subjects showed a positive response to methylphenidate and 88% showed a positive response to d-amphetamine. Only 2 of the 48 subjects failed to respond to one or the other stimulant.

Pelham and colleagues (35) compared the extended methylphenidate release product, Concerta, to immediate release methylphenidate and placebo in a crossover trial. Sixty-eight children between 6 and 12 years old were assigned to receive placebo, immediate release methylphenidate, or the extended release methylphenidate in random order. All children were on methylphenidate prior to enrollment. Thus, each child was assigned to a dosage level that was similar to the pretrial dose. The three dose levels for immediate release methylphenidate were 5 mg three times per day, 10 mg three times per day or 15 mg three times per day. Each dose level of immediate release methylphenidate was matched to a similar, though slightly higher, dose of extended release. Specifically, the dose levels of extended release were: 18 mg, 36 mg and 54 mg given as a single morning dose. Each treatment (immediate release, matching dose of extended release, or placebo) was given in random order for 1 week. Both active treatments were superior to placebo, achieving showed approximately 50% improvement on both parent and teacher measures of ADHD symptoms. There were no detectable differences between immediate release and extended release methylphenidate preparations with respect to efficacy or adverse effects.

To date, only methylphenidate has been evaluated in long-term studies (3). With its sample size of 576 children, the Multimodal Treatment Study of children with attention-deficit/hyperactivity disorder (MTA) provides convincing evidence for the long-term benefits of methylphenidate. In the MTA study, children between 7 and 10 years of age were randomly assigned to one of 4 treatment groups: medication management (N = 144; primarily methylphenidate administered in a systematic fashion with close monitoring); an intensive behavioral treatment program (N = 144); combined medication management and the same behavioral treatment program (N = 145); or community care, which served as the control group (N = 146). The MTA research sites provided treatment to three of these groups, including the medication management group, the behavioral therapy only group, and the combined medication plus behavioral therapy group. The community care group received treatment from self-selected practitioners. In most cases (84 of 146, or 58%), community care consisted of methylphenidate given on a twice-daily schedule. After 14 months of treatment, all four groups showed improvement compared to baseline. Comparisons across the four groups showed that the combined treatment group and the medication management group did significantly better than the community care group and the behavioral treatment only group across a range of outcomes.

Several potentially important differences emerged when community care was compared to medication management provided by the MTA sites. First, community practitioners most often administered methylphenidate on a twice-daily schedule, compared to three times per day in the MTA-treated groups. Among the children randomly assigned to community care, 33% (N = 48) stopped taking medication during the study period. Not surprisingly, this group showed the least improvement on the primary outcome measures. By contrast, only 3% (N = 18) of the research medication management groups (both medication only and medication plus behavioral treatment) discontinued medication during the study. Finally, in the MTA medication management groups, follow-up visits were more frequent and parent and teacher ratings were used in a systematic way to inform clinical decision-making.

Taken together, the results of these studies suggest that stimulants are effective for short- and long-term treatment of children with ADHD. When considering group effects, stimulants appear to be equally beneficial, but individual patients may respond better to one preparation over another. The modest effectiveness of stimulants observed in the community care group of the MTA suggests that close clinical monitoring with dose adjustments based on systematic assessment of therapeutic and side effects contributes to compliance and optimal results.

Although stimulants have become the standard treatment for ADHD, their mechanism of action is not clearly understood. In addition, the mechanism of action for amphetamines and methylphenidate may be slightly different (38). Methylphenidate promotes release of stored dopamine and blocks the return of dopamine at presynaptic dopamine transporter sites. Amphetamines also block dopamine reuptake at the transporter, but appear to promote the release of newly synthesized dopamine more selectively. These combined effects enhance dopamine function in striatum and, at least indirectly, in the prefrontal cortex. It is also clear that both methylphenidate and the amphetamines affect the norepinephrine system (39). For example, both compounds decrease the firing rate in the locus coeruleus (LC),

though amphetamine appears to be more potent in this action. Whether the effect on the norepinephrine system is facilitory or inhibitory is not clear at present. Nonetheless, these combined effects appear to be essential to the clinical effects of the stimulants, as drugs with more selective action (guanfacine or desipramine) tend to have smaller clinical effects.

Immediate release formulations of methylphenidate, dextroamphetamine, and the combined levo- and dextroamphetamine preparation are readily absorbed and show behavioral effects 30 to 60 minutes after ingestion. The peak level of immediate release methylphenidate occurs approximately 90 to 150 minutes after ingestion and the clinical effects last 3 to 5 hours. The immediate release amphetamine products achieve peak levels between 1 and 3 hours, with duration of action of 5 to 7 hours. Based on blinded studies in a research classroom setting, the D,L-amphetamine preparation appears to have a slightly longer duration of action than standard D-amphetamine (40). Methylphenidate and amphetamine are metabolized in the liver, but by quite different pathways (see previous chapter for more details). For the immediate release formulations of these stimulants, the parent compound and metabolites are excreted in the urine within 24 hours. Several newly developed sustained release products have been introduced, offering a range of options for clinical management of children and adolescents with ADHD. Individual manufacturer’s materials should be reviewed before prescribing these agents and some trial and error may be needed in some cases.

There has been considerable debate over whether stimulant dose should be weight based or a fixed dose (33,41). At least in part, this controversy can be traced to early research suggesting that lower doses, such as 0.3 mg/kg/dose, were optimal for enhancing cognitive performance, whereas higher doses (0.6 mg/kg/dose or higher) were more effective for behavioral control (42). Subsequent studies have not supported this view. For example, a convincing linear dose-response was demonstrated across a range of outcomes in a placebo-controlled, crossover study involving 76 children and four dose levels of methylphenidate (36). Nonetheless, individual children may indeed show variability in response across a range of dose levels. Other children may show an orderly dose-response up to a threshold, above which there is little additive benefit. Thus, the mg/kg calculation can be used as a crude guide to calculate the starting dose of 0.3 mg/kg/dose and to a usual ceiling dose (e.g., 0.8 mg/kg/dose).Thereafter, the dose can be increased to establish an optimal response. For example, school-age children can be started on 5 mg tid (just before breakfast, just before lunch and 3 to 4 PM). The dosage may be increased to 10 mg bid (morning and noon) and 5 mg after school after 5 to 7 days. Subsequent increases are based on clinical response and emergence of adverse effects. The third dose typically remains half (or even less) of the first and second doses so as to minimize rebound effects and possible interference with sleep (41).

To determine the optimal daily dose of methylphenidate, it is essential to get feedback from both parents and teachers. In the MTA study, children were seen weekly when starting the medication to monitor progress and side effects. The study used daily ratings to assist with the assessment of response. Although this is not always feasible in clinical practice, clinicians may elect to pace dose increases with the collection of parent and teacher ratings. For example, in the first month of treatment, clinicians may increase the dose on a weekly basis. Collection of parent and teacher ratings prior to each increase would allow comparisons across dose levels. This information could be integrated with side effect data in order to select the optimal dose.

A similar approach can be used with D-amphetamine and D,L-amphetamine. The dosing of these two drugs is similar and both have approximately two-fold greater effect compared to methylphenidate. Thus the amphetamines are administered at half the methylphenidate dose. Moreover, due to their slightly longer duration of action, the amphetamines are typically given twice a day— morning and noon. The initial dosage may be a single 2.5-mg dose in younger children or a 5-mg dose in older children. After 5 to 7 days, the medication may be raised 5 mg bid in younger children and 10 mg bid in older children. Thereafter the dosage may be raised every 5 to 7 days to a total of 15 to 20 mg per day in younger children, and 40 mg per day in older children.

Until recently, methylphenidate sustained release, and D-amphetamine spansules (as well as pemoline, which has been taken off the U.S. market) were the only available long-acting formulations. Several new extended release products— formulated with methylphenidate and one with D,L-amphetamine— have entered the marketplace (Concerta, Focalin XR, Meladate CR, Ritalin RA). These longer acting preparations are as effective as the immediate release compounds and show similar side effect profiles (34,35,43). The advantage of the long-acting preparations is that children do not need to take a dose in school, which may enhance compliance. The longer duration formulations may minimize the behavioral rebound often seen with immediate release formulations. Originally developed for childern who could not swallow pills, Daytrana (methylphenidate transdermal patch) has been tested clinically on youth from 7 to 16 years with the patch adhered to intact skin only on their hips.

Growth retardation, presumed to be secondary to stimulant-induced appetite suppression, has been a common concern among clinicians and families alike. Based on data from a large cohort of clinic cases treated with stimulants, Spencer and colleagues contend that slowed growth may be temporary, and that children with ADHD may be shorter than their age mates before puberty, but “catch up” in adolescence (44). Followup data from the MTA study show that children who remained on medication from the end of the 14-month study to the 24-month followup did not gain as much in height or weight when compared to the children who were never started stimulant medication. The group who did not receive medication (N = 106) was about 1 cm taller and 1 kg heavier at the 24-month followup than the group (N = 222) that was treated with stimulant medication for the entire 2-year period (4). Appetite suppression can often be managed by giving stimulant medications with food, or immediately after meals. Height and weight should be monitored regularly in children treated with stimulants, and tracked during long-term maintenance on population-normed weight- and height-for-age charts.

Other common side effects include sleep disturbance, depressed mood, stomachaches, headaches, overfocusing on details, tics and mannerisms, and picking at skin. Insomnia can be difficult to sort out, as many children with ADHD have sleep difficulties prior to receiving stimulant medications. Rebound effects associated with stimulant withdrawal may compound preexisting sleep problems. Thus, the child’s sleep history should be documented prior to treatment and monitored throughout. As noted, it is common practice for the third dose of methylphenidate to be lower than the first two in order to minimize a possible rebound effect. The use of clonidine as an aid for sleep has been proposed, yet remains a controversial practice (45). A multisite trial conducted by the Tourette Syndrome Study Group (46) showed that complaints of insomnia were indeed lower in subjects randomly assigned to combined treatment with methylphenidate and clonidine compared to methylphenidate also.

Results from case reports and controlled studies suggest that exposure to stimulants can be associated with the emergence of tics (47) or the worsening of preexisting tics (48). However, several studies have also shown that tics do not invariably worsen when children with ADHD and comorbid tic disorders are treated with stimulants (and if fact about one third get better) (46,49,50). Nonetheless, children with tic disorders should be monitored carefully when treated with stimulants. Dose reduction may be sufficient, but discontinuation may be warranted in some cases (48).

All of the SSRIs in current use are approved for use in the treatment of adults with OCD. With the exception of fluvoxamine, the SSRIs are also approved for use in adults with major depression. More recently, paroxetine and sertraline have been approved for adults with anxiety disorders, and sertraline for PTSD. In pediatric populations, fluvoxamine and sertraline have been approved by the FDA for the treatment of OCD, and fluoxetine for the treatment of depression.

The introduction of the SSRIs, starting with fluoxetine in the late 1980s, has had a dramatic impact on the practice of pediatric psychopharmacology. Compared to the TCAs, monotherapy with the SSRIs is relatively simple. As a group, these medications are generally well tolerated, can typically be given once a day, and do not require blood level monitoring or ECGs. Following the early clinical trials with clomipramine and fluoxetine in children and adolescents (51,52,53), several large placebo-controlled clinical trials with sertraline (54) and fluvoxamine (55) in OCD; with fluoxetine (56) and paroxetine (57) in depression; and with fluvoxamine in non-OCD anxiety disorders (10) have been conducted in pediatric populations. In each of these studies, the SSRI was superior to placebo in the primary outcome measure of interest. More recently, trials comparing the relative efficacy of SSRIs, cognitive behavioral therapy (CBT), and their combination, have been completed with fluoxetine for depression (12), and with sertraline for OCD (24).

Sertraline and fluvoxamine have been evaluated in randomized, multisite, placebo-controlled trials of parallel groups. Using the Children’s Yale–Brown Obsessive-Compulsive Scales (CYBOCS) as the primary outcome measure, both drugs were superior to placebo in improving obsessive-compulsive symptoms. Sertraline was evaluated in 187 subjects ranging from 6 to12 years of age. In that study, sertraline, at an average daily dose of 167 mg, was associated with at least a 25% improvement in the CYBOCS score in 53% of subjects, compared to 37% for placebo (p = 0.03) (54). In 120 children between the ages of 8 and 17 years, fluvoxamine at a mean daily dose of 165 mg was effective in 42% of children, compared to 26% among those treated with placebo (p = 0.06) (55).

An open-label study of paroxetine in pediatric OCD revealed not only promising results (58), but also interestingly, reductions in thalamic volume (59) and caudate glutamate levels (60), which paralleled clinical response. Two small placebo-controlled studies provide additional support for the efficacy of fluoxetine in OCD (15,61). Finally, one open-label study with citalopram has been done in children with OCD (62). In that study, 23 subjects were treated with 10 to 40 mg per day of citalopram for 10 weeks in an open-label fashion. Eleven of 23 subjects showed a clinically meaningful positive response (30% improvement or more). Five patients showed little or no response, and the remaining seven showed a partial response.

Taken together, these data suggest that the SSRIs are effective for the treatment of OCD in children and adolescents. However, the magnitude of response may not be large, as has been shown in a metaanalysis of all studies published through 2002 (63). In addition, some children with OCD may show only a partial response to an adequate trial of an SSRI. For example, approximately 40% of the subjects in the multisite sertraline study showed less than a 25% improvement in obsessive-compulsive symptoms (54). This observation indicates that clinicians should remind parents and patients not to have unreasonably high expectations for SSRI treatment. The problem of partial response raises questions about whether to switch to another SSRI or clomipramine, or to embark on one of several augmentation medication strategies. Although not well studied in children (64), two studies have shown that the addition of low-dose haloperidol (65) or risperidone (66) to an SSRI can be effective in adults with refractory OCD. A recent metaanalysis provides support for augmentation of SSRIs with haloperidol and risperidone in the treatment of adults with OCD (67). In view of the inadequate support for combined pharmacotherapy in children with refractory OCD, other interventions should be considered, particularly CBT.

Indeed, the Pediatric Obsessive Compulsive Treatment Study (POTS) (23) showed superiority of CBT over SSRI treatment alone. In that study, 112 children (mean age about 11.5 years, 56 boys and 56 girls) with OCD enrolled across three treatment sites were randomly assigned to sertraline, CBT, their combination, or pill placebo. Statistical analyses on the Children’s Yale–Brown Obsessive-Compulsive Disorder Scales (CYBOCS) indicated a significant advantage for CBT alone (P = .003), sertraline alone (P = .007), and combined treatment (P = .001) compared with placebo. Combined treatment showed a 53% improvement on the CYBOCS

compared to 46% for CBT alone, 30% for sertraline alone and 15% for placebo. These results suggest that CBT, if readily available, may be the preferred first treatment for OCD, either alone or in combination with an SSRI. The two groups who were treated with active medication each started on a dose of 25 mg per day with gradual increases to a maximum of 200 mg per day. The combined treatment group received an average of 133 mg per day compared to 170 mg per day for the sertraline only group.

Based on the demonstrated efficacy and safety of the SSRIs in children and adolescents with OCD, these medications are commonly used to reduce repetitive behavior in pervasive developmental disorders (PDDs) (68). Despite their common use in clinical practice, however, the SSRIs have not been well studied in children with PDD. Moreover, the best available evidence indicates that the SSRIs may only be moderately effective. Hollander and colleagues (69) compared fluoxetine to placebo in a crossover trial of 39 subjects age 5 to 17. At an average dose of approximately 10 mg per day, the drug was well tolerated with a low frequency of activation. The low frequency of activation appears to be due the low starting dose and gradual increase.

Although the report notes that the active drug was superior to placebo, fluoxetine showed an average 10% improvement over baseline on a clinician measure of repetitive behavior. Clearly, more study is needed. To address questions of efficacy and safety of the SSRIs for repetitive behavior in PDD, an NIH-sponsored, multisite study by the STAART Group is comparing citalopram to placebo in a 12-week parallel group trial. When completed, the study will enroll 144 subjects, which may provide guidance for clinicians concerning which children with PDD are appropriate for SSRI treatment.

Fluoxetine, sertraline, paroxetine, and citalopram have each been studied for the treatment of depression in children and adolescents. The landmark fluoxetine study by Emslie and colleagues (56) was the first to show superiority of an antidepressant over placebo for the treatment of depression in children and adolescents and was subsequently followed by a replication study (70). A placebo-controlled, multicenter trial comparing paroxetine, imipramine, and placebo (71) indicated that paroxetine was superior to placebo, achieving a 63% response rate, compared to 46% in the placebo group. By contrast, the response rate in the imipramine group was 50%, which was not statistically different from placebo. Imipramine was associated with common TCA side effects and with a high rate of premature discontinuations. Two identical, industry-sponsored trials of sertraline showed superiority over placebo when combined into a single report (72), but failed to differentiate from placebo when analyzed separately. A randomized clinical trial of citalopram has also shown superiority over placebo (73).

The largest and most important study to date in this area is the Treatment for Adolescents with Depression Study (TADS) (12). In it, 439 adolescents 12–17 of age, were randomly assigned to fluoxetine alone, CBT alone, their combination, or pill placebo. They were followed for 12 weeks in the acute phase and for a 6-month extension. Results of the acute phase (12-week trial) showed that combined treatment had the highest rate of positive response (71% for the combined treatment, compared to 61% for medication alone, 43% for CBT alone, and 35% for placebo). The combined treatment group also had a slightly lower rate of suicidal ideation (5.6%) compared to the fluoxetine-only group (8.3%). Fluoxetine alone and fluoxetine with CBT were both superior to placebo. However, combined treatment with fluoxetine and CBT was not significantly better than medication only and CBT alone was not superior to placebo.

Concerns over safety of the use of the SSRIs in children and adolescents have become paramount, initially garnering substantial attention in the media and lay press, and following extensive review by British and American regulatory agencies, eventually led to the removal (in the U.K.) and the introduction of an FDA-mandated black-box warning (in the U.S.). The history and full implications of this series of concerns is explored in detail in a recent review (74) and continues to be a source of controversy, shifting policy, and clinical recommendations. In the context of this general overview of pediatric psychopharmacology, the following are important high points of the discussion to date: 1) A review of all clinical trials (both published and unpublished) using SSRIs in the treatment of children and adolescents with depression and other indications (N>4,400 subjects, across 26 controlled trials, 16 of them for depression) was commissioned by the FDA. It revealed an increase risk in new onset suicidal ideation between SSRI- and placebo-treated individuals (occurring at respective rates of 4% and 2%, for a risk ratio of 1.95 (95% confidence interval, 1.28–2.98) (75); 2) all reported events referred to suicidal ideation, rather than suicidal acts or completed suicides; 3) there is compelling pharmacopidemiological data to suggest that paralleling the widespread use of SSRIs in the US, the suicide rate among those ages 15 to 19 fell from about 11 per 100,000 in 1990 to 7.3 per 100,000 in 2003 and synchronous with the FDA Black Box warning on antidepressants and the likely reduction in antidepressant usage in 2004, the suicide rate climbed 18 percent for those younger than 20, from 1,737 deaths to 1,985 (76); and 4) based on these data, it seems most parsimonious to recommend the judicious use of antidepressants in children and adolescents, particularly if other interventions have failed or are not available. When treatment with an SSRI is opted for, fluoxetine may be generally recommended as the first-line agent because of its FDA indication for depression and a lower reported rate of incident suicidal ideation. Guidelines from the FDA and the American Academy of Child and Adolescent Psychiatry (AACAP) call for intensive monitoring during the early phases of treatment: as often as weekly for the first 4 weeks, every other week for the next month, and monthly thereafter. While such recommendations are clinically sensible, such guidelines may potentially lead to the perverse outcome of increased suicide rates in that practitioners may hesitate resorting to these agents if unable to provide monitoring as intense as is being called for. This may become especially problematic in underserved areas, where nonspecialists may be reluctant to prescribe antidepressants.

To date, only fluvoxamine has been evaluated in the treatment of children and adolescents with non-OCD anxiety disorders. In a multisite study sponsored by the National Institute of Mental Health (10), 128 subjects between the ages of 6 and 17 years were randomly assigned to placebo or fluvoxamine after a 3-week psychoeducational intervention. The primary outcome measure was the Pediatric Anxiety Rating Scale (PARS), a new scale developed specifically for the trial. After 8 weeks of treatment, children in the fluvoxamine group showed a 52% improvement (mean decrease in PARS from 18.7 to 9.0) compared to 16% improvement (mean decrease from 19.0 to 15.9) for the placebo group (p < 0.001). The dose began at 25 mg per day, with a planned increase to 25 mg twice a day after 4 days. The dose schedule continued upward in 25 mg increments every 4 to 5 days as tolerated. The findings from this study provide support for the efficacy and large effect size of fluvoxamine in the treatment of generalized anxiety disorder, social phobia, and separation anxiety. This study also paves the way for the study of the other SSRIs and combination treatments (medication and psychotherapy) in non-OCD anxiety disorders. Indeed, the Children and Adolescents Anxiety Multimodal Treatment study (CAMS) is currently underway. With a projected sample size of 320 children and adolescents (ages 7–17), it will compare the efficacy of sertraline, CBT, or their combination in
the treatment of generalized or separation anxiety or social phobia (77).

The SSRIs interfere with the return of serotonin into the presynaptic neuron by blocking the serotonin transporter located on presynaptic nerve terminals. Over time, this blockade leads to a desensitization of the serotonin autoreceptors, which typically exert an inhibitory influence on serotonin release. With continued blockade of the transporter, the desensitized autoreceptors do not exert their usual inhibitory influence and serotonergic function is enhanced. Based on a series of animal studies, Blier and colleagues suggest that the main location of the enhanced serotonergic function appears to be the hippocampus in depression, and the orbital frontal cortex in OCD (78).

All SSRIs have relatively long half-lives, permitting single daily dosing. Fluvoxamine, which has the shortest half-life, is sometimes given on a twice-daily schedule. A recent pharmacokinetic evaluation of paroxetine showed that children metabolize the medication faster than adults (79). Despite the shorter half-life in the pediatric population, these investigators still recommend once-daily dosing for paroxetine. At low doses of sertraline, 50 mg per day or less, children may require bid dosing (80). The pharmacokinetic profiles of the other SSRIs in pediatric populations have not been documented. In adults, fluoxetine and citalopram have the longest half-lives of currently available SSRIs, with estimates of 48 to 72 and of 33 hours, respectively. In addition, fluoxetine has an active metabolite (norfluoxetine) with an elimination half-life of 7 to 14 days. Both fluoxetine (primarily norfluoxetine) and paroxetine are potent inhibitors of CYP 2D6. Because both paroxetine and fluoxetine are 2D6 substrates, they inhibit their own metabolism, resulting in nonlinear kinetics at higher doses. Fluvoxamine also has nonlinear pharmacokinetics, which may be related to its inhibiting its own metabolism.

Fluoxetine (Prozac) is available in a 10-mg scored tablet, a 20-mg capsule and in a liquid preparation (20 mg per 5 ml). A typical starting dose for school-age children is 5 to 10 mg per day; smaller children may start at 2.5 mg per day. Given its long half-life, fluoxetine should be increased slowly (weekly or even at 2-week intervals) to avoid “overshooting” the optimal dose. The usual dose range for children and adolescents is 5 to 40 mg per day, though some children and adolescents may require higher doses (81).

Sertraline (Zoloft) is available in 25-, 50-, and 100-mg tablets that can be easily broken in half; it is also available as an oral suspension in a 20 mg/ml strength. Treatment might start with a 12.5 to 25-mg dose, with similar weekly increments to a range of 50 to 150 mg in children. Higher dosages may be required in older adolescents. Clinicians should review therapeutic response during the dose adjustment phase to determine whether additional increases are needed, rather than using an automatic dose schedule.

Fluvoxamine (marketed under the generic name) is available in 25-, 50-, and 100-mg scored tablets. Treatment usually begins at 12.5 to 25 mg per day and is increased by 25 mg on a weekly basis. The typical dose range is 50 to 200 mg per day. Although the double-blind trial in children and adolescents with OCD used a rapid dose escalation with increases every 3 days (55), the more recent RUPP anxiety study used a slower upward adjustment (10). This study started with 25 per day and increased to 25 mg twice a day within the first week. Thereafter, the dose was increased in 25 mg steps each week as tolerated.

Paroxetine (Paxil) is available in 10-, 20-, and 30-mg tablets that can be broken in half, as well as in an oral suspension (10 mg per 5 ml). The typical starting dose is 5 to 10 mg per day, with weekly increases to a total daily dose of 10 to 40 mg. As noted before, paroxetine is no longer recommended for the routine use of children and adolescents. Exposure to paroxetine during the first trimester of pregnancy may increase the risk for congenital malformations, especially cardiac ones.

Citalopram (Celexa) is available in 10-, 20-, and 40-mg scored tablets, as well as in a liquid preparation. Based on experience with the other SSRIs, a reasonable starting dose would be 5 mg per day, with increases on weekly or 2-week intervals, to a maximum of 40 mg per day.

Escitalopram (Lexapro) is available in 10- and 20-mg scored tablets.

As a group, the SSRIs are generally well tolerated, and potentially serious side effects such as alterations in cardiac conduction times or seizures have not been reported in the usual dose range. In addition to their propensity for cytochrome P450-based drug interactions (as reviewed in the preceding chapter), common side effects of the SSRIs in children and adolescents appear to be behavioral activation and GI complaints such as nausea or diarrhea. Signs of behavioral activation include motor restlessness, insomnia, impulsiveness, disinhibited behavior, and garrulousness. It may occur early in treatment, with dose increases (82), or following the addition of drugs that inhibit the metabolism of the SSRIs (e.g., cimetidine). The potential for behavioral activation early in treatment underscores the importance of starting at low doses and moving upward slowly. As with other antidepressants, hypomania and mania have also been reported, and peripubertal children may be at especially heightened risk (7). Other adverse effects include diarrhea, nausea, heartburn, decreased appetite, and fatigue. Sexual side effects, such as erectile dysfunction, delayed ejaculation, or anorgasmia, all of which are relatively common in adults, should also be considered in sexually active adolescents.

The controversy over suicidal ideation is briefly presented above, and in greater detail in a recent review (74). As with all antidepressants, particularly when treating depression, clinicians should monitor suicidal thought and self-injurious potential in any child or adolescent treated with an SSRI.

A flu-like syndrome characterized by dizziness, moodiness, nausea, vomiting, myalgia, and fatigue occurring in association with the withdrawal or acute discontinuation of shorter acting SSRIs such as paroxetine, fluvoxamine, and sertraline has been described (83). Recently, a controlled discontinuation study in 220 adults compared the withdrawal effects of fluoxetine, paroxetine, and sertraline: Paroxetine and sertraline were associated with irritability, agitation, fatigue, insomnia, confusion, dizziness and nervousness upon abrupt withdrawal, but fluoxetine was not (84). The long half-life of norfluoxetine presumably results in a gradual “auto-taper,” even when the oral dose is stopped abruptly. Based on these results, a slow withdrawal of the shorter acting SSRIs is warranted. Citalopram has a 33-hour half-life, but no known active metabolites. In the absence of data on adverse effects following abrupt withdrawal, sudden discontinuation of citalopram should also be avoided.

Another clinical issue that often arises in the course of treating children and adolescents with an SSRI concerns the duration of treatment. Studies of adults with depression suggest that an episode of depression typically lasts 9 months to a year. Based on this evidence, the duration of treatment for depression can be set at a 1-year minimum. For OCD and anxiety disorders, however, there are no data upon which to base duration of treatment. A review on OCD suggests discontinuation after a relatively symptom-free period of 8 to 12 months (85). A long-term followup study of 54 children and adolescents with OCD found that 70% (N = 39) remained on medication for more than 2 years (86), and persistence of OCD through late adolescence and adulthood is estimated to occur in about 40% of subjects (87). Given the potential for chronicity in
OCD, children and parents should be informed that symptoms may return following a planned SSRI discontinuation.

Due to their multiple clinical applications, ease of use and perceived safety, the SSRIs are increasingly common in clinical practice. In addition, the use of combined psychotropic medications seems to be on the rise. These trends underscore the importance of monitoring drug–drug interactions in clinical practice. SSRIs vary in their potential for such interactions at particular P450 cytochromes. As illustrated previously, inhibition of the P450 cytochrome responsible for metabolizing an additive drug raises its serum level, thereby enhancing its beneficial or deleterious effects. For example, oculogyric crises and other dystonic reactions have been reported in youngsters when an antipsychotic was added to ongoing treatment with paroxetine, probably the result of the latter’s inhibition of CYP2D6, which is a major pathway of risperidone metabolism (88). (For a more detailed discussion of drug interactions, interested readers are referred to the previous chapter).

Tricyclic antidepressants (TCAs) have been used to treat several psychiatric disorders of childhood over the past three decades, including depression, ADHD, OCD, separation anxiety disorder, and enuresis. Although TCAs have been used frequently in clinical practice, evidence for their efficacy in treating children with these psychiatric disorders is inconsistent. For example, a series of carefully conducted controlled trials have consistently failed to show the superiority of any TCA over placebo in the treatment of child- and adolescent-onset depression (89). This poor track record stands in marked contrast to the more compelling results in adult depression (90). The use of TCAs in non-OCD anxiety is equivocal. One study found imipramine superior to placebo in the treatment of separation anxiety (91), but another failed to replicate an earlier report of efficacy (92). In contrast to these respectively disappointing or inconclusive findings in depression or separation anxiety, double-blind, placebo-controlled studies have demonstrated the efficacy of desipramine in children with ADHD (93,94,95), and of clomipramine for the treatment of OCD in children and adolescents (51). Tricyclic agents continue to have a limited but important role in the treatment of enuresis and of treatment-refractory ADHD and OCD (Table 6.1.2.2b).

To varying degrees, all TCAs inhibit the reuptake of norepinephrine by presynaptic neurons. Over time, this pharmacologic effect is presumed to enhance noradrenergic neurotransmission. Among the TCAs, desipramine is the most selective in its capacity to block norepinephrine reuptake. This highly selective property of desipramine plays a role in its efficacy in ADHD. Based on the favorable effects of desipramine in ADHD, interest developed in other compounds, such as atomoxetine, with highly selective norepinephrine reuptake inhibiting properties (96). Unlike desipramine, however, atomoxetine does not appear to prolong cardiac conduction times. Clomipramine is unique among the TCAs in that it is a potent inhibitor of serotonin reuptake. This property explains its superiority over desipramine for the treatment of OCD.

Tertiary amine TCAs such as imipramine (or amitryptiline, rarely used in psychiatry, but useful in medicine for the treatment of neuropathic pain) have highly anticholinergic profiles. Because of this, they are less often used as a first-line intervention, with the exception of enuresis, for which imipramine continues to be used in clinical practice. By contrast, the secondary amines desipramine and nortriptyline (derived from their patent compounds imipramine and amitryptiline, respectively) are less likely to cause orthostatic hypotension, constipation, or urinary retention, and are thus generally preferred in clinical practice.

Due to genetic differences in P450 cytochrome enzyme activity, serum levels of TCAs can show wide variation across individuals taking the same oral dose. Thus, therapeutic levels for the TCAs are not well established in pediatric populations. Serum levels may be useful, however, to identify children with low or ultrarapid metabolic activity, to rule out toxicity, and to assess compliance. As a general rule of thumb, blood levels of nortriptyline are close in absolute value to daily oral dosage among normal metabolizers (thus, 75 mg/day would be expected to yield a steady-state trough level of ∼ 75 ng/ml). When major discrepancies are seen to this pattern, the clinician can anticipate noncompliance or ultrafast metabolism of CYP2D6 (in the case of low levels), or CYP2D6 inhibition from a medication or another chemical compound or slow metabolism (in the case of high levels).

An ECG, pulse, and blood pressure should be obtained prior to starting any of the TCAs. A medical and family history that focuses on syncope in the child, as well as in episodes of syncope or sudden death in close relatives may be informative. Evidence of a normal physical examination within the past year should also be documented. The typical dose range for TCAs in children is up to 5 mg/kg/day for imipramine, and 2.5 mg/kg/day for nortriptyline and perhaps somewhat higher for clomipramine (3 mg/kg/day). Imipramine may be started at a dosage of 25 mg and increased every 4 or 6 days in similar increments to 100 to 150 mg per day. In younger children, nortriptyline is typically introduced with a 10-mg dose, with increases every 4 to 6 days to a range of 50 to 75 mg per day in divided doses. Clomipramine is usually started at a dose of 25 mg, with gradual increases every 4 or 6 days to a maximum of 100 mg per day in younger children and 150 mg in older children. For all of the TCAs, repeat vital signs and ECGs should be obtained during the dose adjustment phase and when the maintenance dose has been achieved. As part of the informed consent process, potential cardiac effects and the reason for repeat ECG monitoring should be discussed with the family and with the child in a developmentally appropriate manner. A corrected QT interval (QTc) above 450 ms, a QRS complex longer than 120 ms, or a PR interval greater than 200 ms (97) warrant dose reduction followed by a repeat ECG. Exceeding these parameters should prompt treatment reevaluation and perhaps discontinuation. For cases showing clinical benefit and persistent ECG abnormalities, consultation with a pediatric cardiologist is in order.

The TCAs are associated with a range of adverse effects, including sedation, dizziness, dry mouth, excessive sweating, weight gain, urinary retention, tremor, and agitation. In addition to these largely anticholinergic-based side effects, TCAs can have dose-dependent adverse effects on cardiac conduction (which can be tracked with an expectable dose-dependent prolongation of the QTc) as well as on the seizure threshold. With regard to seizures, clomipramine may have the highest vulnerability to lower the seizure threshold, so that its dose and potential drug interactions need to be monitored closely.

For most adverse effects, lowering the dose, changing dose schedules, or switching from a tertiary to a secondary amine can often help manage symptoms. For example, to deal with sedation, the medication could be given twice a day, with the higher dose in the evening. Switching between TCAs can be helpful at times: For example, imipramine can be changed to nortriptyline in an effort to minimize sedation or constipation. Despite the evidence showing the efficacy of clomipramine for

OCD and desipramine for ADHD, the TCAs appear to be declining in use. This trend is largely due to the side effect profile and the potential for serious adverse effects, with the rare possibilities of sudden death related to tachyarrhythmias such as torsade de pointes being the most ominous. Originally reported in cases treated with desipramine, the series of case reports that accrued over ensuing years has prompted many experts to recommend avoiding TCAs in the pediatric population (98). The combination of: 1) stepwise dosing within clear weight-adjusted margins; 2) careful ECG monitoring; 3) full disclosure of the risk-to-benefit ratio in the treatment planning process; and 4) their selective use in nonresponders to first-line agents provides a rational basis for keeping these compounds as potential treatment options.

As with most other psychotropic drugs, the TCAs are metabolized by hepatic enzymes in the cytochrome P450 system (CYP 450). Several psychotropic (fluoxetine, fluvoxamine, paroxetine) and nonpsychotropic drugs (ketoconazole, cimetidine, clarithromycin) inhibit the action of one or more of these hepatic enzymes. Inhibition of the enzyme specific for metabolizing the TCA can result in toxicity. (For a more detailed discussion of drug–drug interactions, see previous chapter).

Bupropion is unrelated to all other available antidepressants. Although its mechanism of action is unclear, it appears to have both dopaminergic and noradrenergic effects. It is approved for the treatment of depression and smoking cessation in adults. Although bupropion has not been studied for depression in children or adolescents, it has been evaluated in controlled studies for the treatment of ADHD. A placebo-controlled trial of bupropion in 72 children with ADHD showed its superiority over placebo, although the treatment effect was smaller than that usually seen with stimulants (99). The dose of bupropion (3–6 mg/kg/day) ranged from 50 to 200 mg per day in divided doses. The findings of this study are consistent with previous placebo-controlled studies, and a direct comparison with methylphenidate (100). In an open-label study conducted in 24 adolescents (ages 11 to 16) with ADHD and depression, sustained release bupropion was associated with improvements in both conditions in 58% (N = 14), in depression only in 29% (N = 7), and in ADHD alone in 4% (N = 1), suggesting that further studies of this monotherapy appear warranted for these commonly co-occurring conditions (Table 6.1.2.2c)(101).

Side effects of bupropion include agitation, insomnia, skin rashes, nausea, vomiting, constipation, and tremor. Bupropion may also reduce the seizure threshold in a dose-dependent fashion. The seizure liability of bupropion was first described in a group of female patients with bulimia nervosa. Because of this, regardless of diagnosis, it is recommended that daily doses not exceed 300 mg in children, and no single dose be higher than 150 mg. Bupropion (Wellbutrin) is available in 75- and 100-mg tablets, in 100- and 150-mg sustained release (SR) tablets, and 150 mg and 300 mg Bupropion XL tablets. Treatment is usually on a tid basis for the immediate release formulation given the agent’s short half-life; bid dosing is possible with the SR preparations and once daily is recommended with XL preparations.

Venlafaxine is an agent that selectively inhibits serotonin reuptake at lower doses (<150 mg/d) and acts on both norepinephrine and serotonin reuptake at the higher dose range. To date, there is one placebo-controlled trial in children with depression. The study, which included 32 youngsters, found that the drug was no better than placebo in relieving depression (102). The lack of a significant results were replicated in a larger scale, as yet unpublished, industry-sponsored study. These negative results, coupled with the fact that venlafaxine had the single highest association with incident suicidal ideation (75), advise against the routine use of this agent in pediatric psychopharmacology.

Venlafaxine is available in 25-, 37.5-, 50-, 75-, and 100-mg tablets, and in 37.5-, 75-, and 150-mg extended release (XR) capsules. Dosing is started with the smallest dose given at bedtime, and with attention to early sedation and dizziness, before moving to a twice-daily regimen. At higher doses (>150 mg/d), venlafaxine can be associated with diastolic hypertension, an effect that is clearly dose-dependent in nature.

These agents are potent 5HT_2a postsynaptic antagonists and moderate serotonin and norepinephrine reuptake inhibitors. This novel mechanism of action initially raised great interest in these compounds, but lackluster efficacy results among adults, coupled with rare but serious adverse effects (priapism for trazodone, hepatotoxicity for nefazodone) has led to the selective use of trazodone as an adjunct for insomnia in females only. Trazodone is available in 50-,100-, 150-, and 300-mg pills, and is prescribed in HS dosing for insomnia. Due to concern about hepatic toxicity, serzone but not nefazodone has been taken off the market.

Several other new antidepressants have entered into the marketplace, including mirtazapine, a combined norepinephrine pre- and serotonin post-synaptic antagonist with a characteristic side effect profile of drowsiness, increased appetite and weight gain, though uncommon in the lower dosage range. Few data on the use of mirtazapine in the pediatric population are available, but an unpublished industry-sponsored randomized clinical trial did not show efficacy in depression in youth (103).

Atomoxetine was originally developed for the treatment of depression, but it has never been marketed as an antidepressant. As noted previously, atomoxetine is a selective norepinephrine reuptake inhibitor— a property that it shares in common with desipramine. Unlike desipramine, however, atomoxetine does not appear to prolong cardiac conduction times. Given the efficacy of desipramine for the treatment of ADHD, atomoxetine was evaluated as a treatment for ADHD. Following the completion of several placebo-controlled trials (104,105), atomoxetine was approved by FDA as a safe and effective treatment of children and adolescents with ADHD. In fact, it is the only approved, nonstimulant medication for the treatment of ADHD. It may also be useful in the treatment of comorbid oppositional defiant disorder (106,107).

The serum half-life of atomoxetine is approximately 4 hours, which implies that the drug would be given at least twice a day. Indeed, initial studies used a twice daily regimen, but subsequent trials evaluated the efficacy of once a day dosing (104,105). Although both dosing strategies have been shown to be superior to placebo, twice a day dosing showed a similar magnitude of improvement than once daily dosing (30% improvement on a rating of ADHD symptoms scored by a clinician following a semistructured interview) (104,105). Improvement in ADHD symptoms are not as robust as stimulants (108). The total daily dose is likely to fall between 0.8 and 1.2 mg/kg. Doses greater than 1.2 mg per kg per day are unlikely to produce greater benefit (107). Based on these data, clinicians might consider starting with once a day dosing

with gradually increasing doses. If the child encounters adverse effects or the benefits are inadequate, a twice a day schedule could be considered.