Clinical Applications

Stimulants have been in clinical use since 1937, when it was observed that a group of children in residential treatment showed marked improvement in their behavior with benzedrine (d– and l-amphetamine) [Bradley, 1937]. Since then, the criteria for ADHD have been refined, and other stimulant medications have been evaluated and consistently found efficacious in numerous placebo-controlled studies [Jensen et al., 2001]. Approximately 70–80 percent of school-age patients with ADHD have a positive response to stimulant medication. Although behavioral treatment may be considered, in head-to-head studies medication alone is more beneficial than behavioral therapy alone in children 5 years of age and older. The combination of medication and behavioral therapy is specifically helpful for oppositional behavior and children with ADHD and anxiety [Jensen et al., 2001].

Based on clear evidence of efficacy, the stimulant medications have been used to address attention, hyperactivity, and impulsivity in those with pervasive developmental disorders and in younger children. For many years, stimulant medications were viewed as poorly effective or contraindicated for hyperactivity in children with autism and pervasive developmental disorders. However, in the first large, randomized, controlled trial in this population, methylphenidate demonstrated efficacy. The response rate was less robust than in typically developing children, and rates of adverse events were higher, with nearly 1 in 5 youngsters having to stop treatment [RUPP, 2005]. A single test dose of stimulant may be useful to identify children for whom a longer stimulant trial is contraindicated [Di Martino et al., 2004]. Children who experience a significant worsening of ADHD (e.g., hyperactivity and irritability) or other symptoms (i.e., tics or stereotypies) with a single dose may be excluded from further stimulant treatment [Di Martino et al., 2004].

Preschoolers with ADHD represent another specialized population for whom the use of stimulants may be considered. The Preschoolers with ADHD Treatment Study (PATS), a large randomized, placebo-controlled trial of methylphenidate in children of ages 3–5 years, demonstrated that methylphenidate at doses greater than 2.5 mg three times a day were effective in reducing ADHD in youngsters unresponsive to psychosocial intervention. Again, response rates were less robust than that observed in school-age children, and adverse effects were more common [Greenhill et al., 2006]. Furthermore, methylphenidate may reduce growth rates, even in short-term treatment, and requires careful monitoring of weight and height [Swanson et al., 2006].

Despite substantial data supporting the efficacy of stimulants, their use in children remains controversial. The lay public and media have expressed concerns about the potential for overdiagnosis of ADHD and overuse of stimulant medications in children. However, despite the dramatic increases in stimulant use in the past 10 years [Zito et al., 2003], there is little evidence to support this claim. According to the U.S. Surgeon General, most children with psychiatric disorders are not assessed or treated [Satcher, 2001], and the ratio of stimulant use to prevalence of ADHD is less than 1:1, suggesting that undertreatment is a more prevalent and important issue [Jensen et al., 1999]. Stimulant medications have also come under public scrutiny because of concerns about safety. Concerns about worsening tics, growth suppression, enhancing risk for addiction and, more recently, sudden death, although not substantiated by the evidence, remain important issues for families considering stimulant medications in their child. Some studies actually suggest that children with ADHD who are treated with stimulant early in life have a lower risk of substance abuse than children with ADHD who are not treated [Wilens et al., 2003], but such evidence may not be as reassuring to parents as one would hope. Prudent practice therefore necessitates an evaluation that leads to confidence in the diagnosis, fully informed consent, and use of appropriate doses with close monitoring, combined with effective and available psychosocial treatments. Detailed discussion, in the assessment and treatment phases, with the family and teachers, backed up by full documentation in the medical record, is essential. During the dose adjustment phase, children often are monitored for side effects, including blood pressure, pulse, height, and weight. After a maintenance dose is achieved, visit intervals can vary from 1–4 months.

Pharmacology

Stimulants are sympathomimetic drugs that directly stimulate α- and β-adrenergic receptors. They also stimulate the release of dopamine from presynaptic nerve terminals and inhibit dopamine reuptake. The exact mechanism for efficacy on attention and hyperactivity in ADHD is unknown. Stimulant medications are available in immediate-release and sustained-release preparations (see Table 49-2).

Clinical Management

Adverse Effects

Although most children tolerate stimulant medications well, some do not. The most common side effects of stimulants are insomnia and nervousness [Ritalin, 2004]. Some children with ADHD experience sleep disturbance before exposure to stimulants. The child’s pre-existing sleep history should be obtained, including other factors that may contribute to sleep disturbance (e.g., poor sleep hygiene, caffeine use, oppositional behavior, separation anxiety). Ironically, some children experience difficulty falling asleep related to hyperactivity, and they may actually benefit from a small dose of stimulant to help them stay in bed and fall asleep. In group comparisons, dosing three times daily with short-acting stimulants versus dosing two times daily was not found to increase sleep disturbance [Kent et al., 1995; Stein et al., 1996]; however, new-onset sleep delay after starting stimulants may be associated with a stimulant dose that is too high overall or with dosing stimulants too late in the day. Difficulty with falling asleep because of stimulants can be addressed with a combination of dose adjustment, improved sleep hygiene (i.e., straightforward ritual for falling asleep, and removal of distractions, such as televisions, video games, homework, and night lights, from the room), cooler room temperatures, room-darkening shades, avoidance of late-afternoon or evening doses of stimulants, switching to a different stimulant, or addition of a sedating medication with ADHD treatment effects, such as clonidine [Brown and Gammon, 1992; Prince et al., 1996] or guanfacine.

Short-term reductions in expected weight gain and height have been reported in many studies [MTA, 2004; Swanson et al., 2006], although this has not been universally reported [Spencer et al., 2006; Zeiner, 1995]. Overall, the risk of growth suppression appears mild in school-age children but is moderate in preschool-age children. The Preschool ADHD Treatment Study (PATS) reported an annual reduction in growth rate of 20 percent for height and 50 percent for weight in preschoolers treated with methylphenidate compared to those that were not on stimulants [Swanson et al., 2006]. Because the impact on weight is likely mediated by stimulant-induced appetite suppression, administering stimulant doses immediately after meals [Swanson et al., 1983], allowing the child to eat later in the evening after the appetite suppressant effect of the stimulant has abated, or allowing drug holidays on weekends or evenings may mitigate these effects.

The possibility of growth suppression with long-term use of stimulants remains a controversial topic. Longitudinal studies of children with ADHD treated with stimulants followed into adulthood have not revealed any reduction in height or weight compared to adults who were never exposed to stimulants [Hechtman et al., 1984; Klein and Mannuzza, 1988]. The apparent lack of impact on adult height was thought to result from the discontinuation of stimulants in adolescence. An alternative theory proposes that this pattern of growth is inherent in children with ADHD and not related to stimulant treatment. The dysregulation of several neurotransmitter systems associated with ADHD may alter neuroendocrine function, including those involving growth [Spencer et al., 1998]. It is not clear whether growth suppression identified in the NIMH Multimodal Treatment Study of ADHD (MTA) or PATS studies will continue and have a long-term impact, resolve with stimulant discontinuation, or resolve on its own [Spencer et al., 1998].

Historically, stimulant treatment has been associated with the emergence of tics [Borcherding et al., 1990] or the exacerbation of pre-existing tics [Law and Schachar, 1999]. Children at risk appeared to be those with a personal history or family history of tic disorders. However, the effect on tics is not absolute. In placebo-controlled trials, low to moderate doses of methylphenidate improve attention and behavior in children with chronic tic disorders without significantly worsening tics [Castellanos et al., 1997; Gadow et al., 1995, 1999]. In the Treatment of ADHD in Children with Tics (TACT) study, tic increases categorized as an adverse event occurred in nearly equal numbers of subjects on placebo (22 percent), clonidine (26 percent), and methylphenidate (20 percent). These data suggest that 20–26 percent of youngsters with tics will experience tic worsening shortly after initiation of any medication treatment, including placebo, lending support to the hypothesis that tic increases observed after starting stimulants reflect the natural waxing and waning of tic severity.

Although the product information for the stimulants warns of possible decreases in seizure threshold, no increase in seizure frequency was observed in children treated with stimulants who had co-occurring ADHD and seizure disorder [Feldman et al., 1989; Wroblewski et al., 1992]. There is no increased risk of addiction from appropriate treatment with stimulants for ADHD. Adolescents with ADHD treated with stimulants were less likely to abuse stimulants [Faraone and Wilens, 2003; Hechtman, 1985].

Concerns that stimulants may increase the risk of sudden unexplained death in children followed the publication of case reports and small case series in the early 1990s [Nissen, 2006]. A recently published case-control study comparing 564 cases of sudden death in children of 7–19 years with matched controls who died as passengers in motor vehicle traffic accidents over an 11-year period in the U.S. revealed an association between unexplained sudden death and use of stimulants [Gould et al., 2009]. This finding has enhanced concern that sudden death may be a rare side effect of stimulant use in children. However, the study design does not permit the establishment of causality. The possibility that having ADHD itself confers additional risks for sudden death must be considered, as should the co-occurring increased incidence of risk-taking behavior and substance abuse [Vitiello and Towbin, 2009]. Strategies for managing the cardiac risk of stimulants have been published [Vetter et al., 2008].

Despite the impact on a broad range of ADHD symptoms, including irritability, stimulant medications can be associated with undesirable changes in mood and behavior [Gadow et al., 1992]. Some of these changes occur when the stimulant is reaching peak serum concentrations, or they may occur when serum concentrations are waning – so-called rebound effects. Effects associated with peak concentrations include dysphoria, anxiety, agitation, or the “zombie” effect (i.e., over-focused and passive). Rebound effects include overactivity, talkativeness, excitability, irritability, and insomnia [Rapoport et al., 1978; Zahn et al., 1980]. It is often difficult to determine whether such rebound symptoms are the re-emergence of ADHD symptoms once the patient is off stimulants or true withdrawal or rebound symptoms. These adverse behavioral effects can be managed in a variety of ways, including dose reduction or medication discontinuation, and by improving late-afternoon coverage by adding a low dose of immediate-release stimulant or by switching to a long-acting stimulant [Pliszka, 2007]. Youngsters who experience these side effects on one of the stimulants do not necessarily have the same behavioral effects on another stimulant, and switching to another stimulant or alternative medication can be useful. Clinicians should be comfortable using all stimulant preparations to optimize treatment effects and minimize side effects.

Pemoline (Cylert©), a long acting stimulant with a novel structures has been associated with rare cases of acute hepatic failure [Shevell, 1997]. In 2005, the FDA withdrew labeling of pemoline for use in the treatment of ADHD, and subsequently the pharmaceutical companies voluntarily agreed to stop sales and manufacture of pemoline [Hogan, 2000].

Drug Interactions

If stimulant medications are administered concomitantly or within 2 weeks of receiving monoamine oxidase inhibitors, there is a risk of precipitating a hypertensive crisis. Methylphenidate may inhibit the metabolism of coumarin anticoagulants, some anticonvulsants (e.g., phenobarbital, phenytoin, carbamazepine, primidone), and tricyclic antidepressants (e.g., imipramine, desipramine, clomipramine). Foods and beverages with caffeine may increase both efficacy and adverse effects. Amphetamines inhibit adrenergic blockers. Gastrointestinal acidifying agents, including citrus juices, lower absorption of amphetamines. Urinary acidifying agents increase urinary excretion of methylphenidate. Both types of acidifying agents can lower blood levels and efficacy of amphetamines.

The safety of combining stimulants, especially methylphenidate, with clonidine has been controversial. Clonidine has been helpful in managing sleep disturbance related to ADHD or stimulant treatment [Wilens et al., 1994]. The combination appears to help children with ADHD and co-occurring oppositional and conduct disorder [Hunt et al., 1990]. The combination also appears useful in the treatment of ADHD in children with tic disorders [TSSG, 2002; Wilens et al., 2003]. Concerns about the interaction developed after the report of four deaths of children who apparently received clonidine and methylphenidate, as well as other medications in some cases [Fenichel, 1995]. Although the cases did not have a clear pattern of causality [Sallee et al., 2000a; Wilens et al., 1999], clinicians are more cautious about prescribing this combination. When combining clonidine and stimulants, careful screening is recommended for a personal and family history of cardiac abnormalities, arrhythmias, and nonvasovagal syncope, and a baseline and follow-up EKG may be warranted.

Clinical Applications

Atomoxetine is a nonstimulant medication that is approved by the FDA for the treatment of ADHD in adults and in children 6 years old and up. It is a selective norepinephrine reuptake inhibitor that is structurally more similar to fluoxetine than to stimulant medications.

Atomoxetine potentially offers several advantages over stimulant medications: longer duration of action, lower misuse or abuse potential, lower risk of rebound effects, lower risk of precipitating tics or psychosis, and ease of prescribing because duplicate or paper prescriptions are not required and multiple refills are possible. Effectiveness of atomoxetine in children has been supported by placebo-controlled trials [Kelsey et al., 2004; Spencer et al., 2002]. Atomoxetine doses of 1.2 and 1.8 mg/kg/day administered in divided doses were superior to placebo and atomoxetine in a dose of 0.5 mg/kg/day; there was no clear superiority of 1.8 versus 1.2 mg/kg/day. Early clinical trials used twice-daily dosing, but once-daily dosing also appears effective [Michelson et al., 2004]. Longer-term treatment (9 months) with atomoxetine appears to be safe and well tolerated [Michelson et al., 2004]. Children maintained therapeutic effect and demonstrated superior psychosocial functioning compared to children who received placebo [Kratochvil et al., 2002]. When compared with methylphenidate treatment, atomoxetine was associated with comparable therapeutic effects [Michelson et al., 2003]. Efficacy of atomoxetine in adults with ADHD was demonstrated at doses of 60–120 mg/day [Strattera, 2010].

Pharmacology

Atomoxetine is well absorbed through the gastrointestinal tract and is metabolized by the liver, predominantly by cytochrome P-450 2D6 (CYP2D6). A small percentage of the population (approximately 7 percent of whites and 2 percent of African Americans) has reduced activity of the cytochrome P-450 isoenzyme system. These individuals are considered to be “poor metabolizers” of medications whose predominant method of metabolism is this isoenzyme. Poor metabolizers can be expected to achieve higher than expected plasma concentrations on a given dose and a prolonged half-life compared with those with normal CYP2D6 activity. The half-life of atomoxetine in poor metabolizers is 21.6 hours, compared with 5.2 hours in normal metabolizers. Atomoxetine is 98 percent bound to plasma proteins, primarily serum albumin [Strattera, 2010].

Clinical Management

Children should undergo standard psychiatric and medical assessment for ADHD (see “Clinical Management of Stimulant Medication”). No laboratory screening is required, although obtaining baseline and follow-up liver function studies should be considered in view of reports of severe acute liver dysfunction [Lim et al., 2006]. Baseline values for weight, heart rate, and blood pressure should be obtained. Concomitant use of medications with cardiovascular effects and medications that inhibit cytochrome P-450 may necessitate dosage adjustment. Although the determination of cytochrome P-450 metabolizer status is available, it appears to be unnecessary because regular dosing parameters provide the opportunity for assessment of outcome and adverse events that may be experienced by poor metabolizers.

In children, the usual starting dose of atomoxetine is 0.5 mg/kg/day, given in the morning or divided into morning and late-afternoon doses. Starting with a split dosing regimen may decrease gastrointestinal side effects and irritability. Decreasing side effects on initiation will ultimately lead to better compliance. The dose can be consolidated after a target dose is achieved. The dose may be increased every 3 days to a target dose of 1.2 mg/kg/day. For adults or for children and adolescents weighing more than 70 kg, atomoxetine may be initiated at 40 mg/day. The dose may be increased gradually every 3 days to a maximum dose of 80 mg. If there is an inadequate response after a 2- to 4-week trial, the dose may be increased to 100 mg/day. Patients taking atomoxetine concomitantly with potent inhibitors of CYP2D6 should be maintained at the initial dose for at least 4 weeks before a dosage increase is considered. Dosage reduction is required in patients with hepatic impairment [Strattera, 2010].

In clinical practice, combination therapy may be used during initiation and titration of atomoxetine, especially in children with severe ADHD symptoms. The effectiveness and tolerability of combining atomoxetine with methylphenidate in children who have not responded to monotherapy have been reported in a few patients [Brown, 2004]. Discontinuation of atomoxetine does not require dose tapering. Abrupt discontinuation of atomoxetine has not been associated with an acute discontinuation syndrome [Wernicke et al., 2003].

Adverse Effects

The most common side effects of atomoxetine in children and adolescents are upset stomach, decreased appetite, nausea, vomiting, dizziness, tiredness, and mood swings. Atomoxetine was associated with increases in heart rate of 6 beats per minute, and increases in systolic and diastolic blood pressure of 1.5 mmHg compared with placebo. In an analysis of short-term and long-term treatment with atomoxetine, increased systolic blood pressure was observed in adults, and increased diastolic blood pressure occurred in children and adolescents. Heart rate increased in both groups, and no prolongation of the QTc interval was observed [Strattera, 2010]. Seizures and prolonged QTc intervals were reported after an overdose with atomoxetine [Sawant and Daviss, 2004]. During post-marketing experience, several cases of significant liver injury in children treated with atomoxetine have been reported [Lim et al., 2006; Stojanovski et al., 2007]. Lilly Research Laboratories subsequently reviewed all hepatobiliary events during clinical trials and through post-marketing voluntary adverse events reporting, and identified three cases of probable severe, reversible liver injury related to atomoxetine use [Bangs et al., 2008]. Strattera labeling now warns that severe liver injury is possible and states that atomoxetine should be discontinued and not restarted when laboratory tests show liver injury or when there is clinical evidence of jaundice. Use in patients with liver disease probably should be avoided.

Drug Interactions

Inhibitors of CYP2D6 may increase serum concentrations of atomoxetine. Atomoxetine does not have significant effects on the cytochrome P-450 system. Combination with monoamine oxidase inhibitors may precipitate a hypertensive crisis. Atomoxetine has potential interactions with cardiovascular agents and adrenergic agonists. Albuterol’s tendency to increase heart rate and blood pressure may be potentiated by atomoxetine. No increase in the cardiovascular effects of methylphenidate was observed when atomoxetine was added, and no interactions between atomoxetine and other protein-bound medications have been observed [Strattera, 2010].

Clinical Applications

Two α2 agonists, clonidine (Catapres®, Kapvay®) and guanfacine (Tenex®, Intuniv®) are prescribed in the treatment of ADHD, Tourette’s syndrome, aggressive or self-inujurious behavior, and the physiological symptoms of anxiety. In 2009, extended release guanfacine (Intuniv®) was the first α2-agonist to be approved by the FDA for treatment of ADHD in children and adolescents. More recently extended release clonidine (Kapvay®) has also been found effective and given FDA approval for treatment of ADHD in children and adolescents. In clinical practice, clonidine and guanfacine are second-line treatments for ADHD.

Clonidine is available in immediate-release and extended-release (ER) formulations. Immediate-release clonidine has been shown to improve behavior of children with ADHD; however, the degree of its effect in a meta-analysis was less than that of stimulants, and it was associated with many side effects [Connor et al., 1999]. The largest randomized controlled study to compare clonidine, methylphenidate, and the combination directly confirmed this earlier analysis [Palumbo et al., 2008]. Greatest improvement is seen with hyperactivity and impulsivity [Hunt, 1987; Hunt et al., 1982] and with frustration tolerance [Hunt, 1987] than with distractibility [Palumbo et al., 2008]. Children have been maintained on immediate-release clonidine for up to 5 years with continued benefit [Hunt et al., 1990, 1991]. Recently, clonidine ER was found effective in 2 randomized controlled trials (RCTs) in patient age 6-17 years (Kapvay, 2010, Kollins et al, 2011). One RCT used clonidine ER monotherapy and one RCT used clonidine ER as adjunctive therapy to a stimulant. To date, only the RCT with adjunctive therapy has been published in a peer reviewed journal (Kollins et al, 2011).

The combination of clonidine and methylphenidate has been helpful in adolescents with co-occurring ADHD and oppositional or conduct disorder [Hunt et al., 1990]. This combination allowed the dose of methylphenidate to be reduced by 40 percent [Hunt et al., 1990]. In children with co-occurring ADHD and tic disorders, those treated with clonidine experienced improvements in both conditions [Steingard et al., 1993]. A randomized controlled trial of clonidine, alone and in combination with methylphenidate, in children with ADHD and tics (TACT Trial) showed that clonidine was equally effective as methylphenidate but that the combination was most effective for treating ADHD symptoms [TSSG, 2002]. Clonidine has been used to treat sleep disturbances in children with ADHD [Wilens et al., 1994], and it appears to be helpful in managing severe aggression [Kemph et al., 1993]. Studies of clonidine in youngsters with Tourette’s syndrome show mixed results [Leckman et al., 1982], but significant improvements in tics and behavior have been reported [Cohen et al., 1980; Comings et al., 1990; Leckman et al., 1991].

Guanfacine is available in immediate-release and extended-release (ER) formulations through different manufacturing companies. Limited data exist on immediate-release guanfacine in children. One open trial demonstrated its effectiveness in ADHD [Hunt et al., 1995]. Findings for patients with ADHD and Tourette’s syndrome have been mixed [Chappell et al., 1995; Horrigan and Barnhill, 1995]. Results of more recent studies have been more promising [Scahill et al., 2001]. More data are available on the extended-release formulation (Intuniv®). Randomized, double-blind, placebo-controlled studies have demonstrated efficacy of guanfacine ER for treatment of symptoms of ADHD in short-term treatment [Biederman et al., 2008; Sallee et al., 2009b]. Continued efficacy is seen in longer-term treatment of up to 24 months [Sallee et al., 2009a]. Review of data suggests that guanfacine ER may be most effective in younger children and children with ADHD, combined type [Biederman et al., 2008].

Pharmacology

Clonidine and guanfacine immediate release are well absorbed from the gastrointestinal tract. Clonidine reaches peak plasma levels 3–5 hours after oral administration and has a half-life of 12–16 hours. Clonidine ER reaches peak plasma levels about 5 hours later than clonidine immediate release and has a similar half-life (Kapvay, 2010) Guanfacine reaches peak plasma levels between 1 and 4 hours and has a half-life of about 16 hours. Guanfacine ER reaches peak plasma levels at about 5 hours and has a slightly longer half-life of 18 hours [Intuniv, 2011]. Clonidine is also available in a transdermal patch.

Clonidine and guanfacine exert agonist effects on presynaptic α₂-adrenergic receptors in the sympathetic nuclei of the brain, resulting in decreased release of norepinephrine from presynaptic nerve terminals. In the central nervous system (CNS), α₂-adrenergic agonists are thought to regulate noradrenergic activity in the locus ceruleus [Arnsten et al., 1988].

Clinical Management

Before the initiation of clonidine or guanfacine, baseline EKG, heart rate, and blood pressure should be obtained. A history of syncope or cardiovascular disease is a relative contraindication. Heart rate and blood pressure should be monitored regularly during therapy for hypotension and bradycardia.

Clonidine is initiated at a 0.05-mg dose taken at bedtime, gradually titrating the dose over several weeks to 0.15–0.3 mg/day in divided doses. Clonidine is also available in a transdermal patch, which may have the advantage of increased medication compliance and more stable blood levels. After a therapeutic dose of oral clonidine is achieved, the equivalent patch can be substituted [Catapres, 2010]. Clonidine ER is initiated at 0.1 mg at bedtime and increased weekly by 0.1 mg to a maximum dose of 0.4 mg divided twice daily (Kapvay, 2010).

Guanfacine immediate release offers the advantage over clonidine of a longer half-life and therefore less frequent dosing, and it may cause less sedation than clonidine [Hunt et al., 1995]. It is initiated at 0.5 mg at bedtime and is titrated up to a maximum of 3 mg/day. Guanfacine ER has the greatest ease of administration and allows once-daily dosing. It is initiated at 1 mg daily and increased weekly to a maximum daily dose of 4 mg/day [Intuniv, 2011].

Adverse Effects

The most common side effects of clonidine and guanfacine include sedation, dizziness, fatigue, dry mouth and eyes, nausea, hypotension, and constipation [Catapres, 2010; Iintuniv, 2011]. Syncope has been reported in 1 percent of children on guanfacine ER [Intuniv, 2011]. Clonidine in patch form may be associated with contact dermatitis, which may be reduced by changing its location on the body. Because of the risk of rebound hypertension with abrupt discontinuation, these medications should be prescribed only to patients with reliable medication compliance. Likewise, the dose should be tapered gradually when discontinuing the medication [Catapres, 2010; Intuniv, 2011].

Case reports of sudden death in children treated with clonidine has raised concerns about the cardiovascular safety of clonidine by itself or in combination with methylphenidate [Fenichel, 1995]. The most commonly reported cardiovascular side effect is bradycardia in patients taking clonidine either alone or in combination with methylphenidate [Chandran, 1994; Connor et al., 2000; Daviss et al., 2008], although this has not been universally reported [Kofoed et al., 1999; Leckman et al., 1991; Wilens et al., 2003]. One study reported prolongation of the PR interval that was not clinically significant [Connor et al., 2000], and one study reported bradycardia accompanied by a variety of EKG changes, which was confounded by concomitant treatment with other medications with potential for cardiovascular side effects [Chandran, 1994]. Some studies have suggested that the risk for bradycardia is higher when clonidine is given concomitantly with methylphenidate [Connor et al., 2000], but others have shown no additional risk [Daviss et al., 2008; Leckman et al., 1991]. To date, studies of guanfacine ER have not reported any EKG abnormalities considered to be serious adverse events, although a few subjects have been noted with clinically asymptomatic bradycardia [Biederman et al., 2008; Sallee et al., 2009a]. Practitioners should monitor bradycardia in patients taking clonidine.

Drug Interactions

Clonidine may potentiate the CNS-depressant effects of alcohol and sedating medications. The manufacturer advises caution in combining clonidine with medications that affect cardiac conduction and sinus node function because of potential additive effects, such as bradycardia and atrioventricular block [Catapres, 2010].

Clinical Applications

Tricyclic antidepressants are approved for the treatment of depression in adults and adolescents (Table 49-3). Placebo-controlled studies have demonstrated the efficacy of tricyclic antidepressants in the treatment of depression in adults, but studies enrolling children and adolescents have yielded inconsistent results. Some open studies show no significant response to tricyclic antidepressants [Kashani et al., 1984; Kramer and Feiguine, 1981]. Others suggest efficacy of desipramine, imipramine, and nortriptyline [Boulos et al., 1991; Geller et al., 1986; Ryan et al., 1986]. Placebo-controlled studies, however, do not support efficacy [Geller et al., 1989, 1990; Puig-Antich et al., 1987]. The lack of efficacy in placebo-controlled trials may be related to methodological problems, such as a small number of subjects, types of subjects enrolled (with depression that is either too mild or too severe), and the high placebo response rates in many studies. The use of tricyclic antidepressants for the treatment of depression in children and adolescents has declined because of their equivocal efficacy and safety in this population and the availability of antidepressants with a more favorable side-effect profile (i.e., selective serotonin reuptake inhibitors). Clomipramine has been approved for use in children 10 years of age or older for OCD, following publication of several randomized, controlled trials [DeVeaugh-Geiss et al., 1992; Leonard et al., 1989]. However, clomipramine is used less commonly for OCD than selective serotonin reuptake inhibitors and is considered second-line treatment in clinical practice, in part because of a less favorable side-effect profile.

Although amitriptyline is still commonly used in neurologic settings for adults, its side-effect profile in doses effective for the treatment of anxiety and depression is often prohibitive – particularly sedation and weight gain. Nortriptyline, the primary metabolite of amitriptyline, is often better tolerated at treatment doses and has the benefit of an established target blood level for antidepressant efficacy that is unique among the tricyclic antidepressants. Because clinical trials do not support the efficacy of any tricyclic antidepressant for anxiety or depression in children or adolescents, they should not be considered first-line treatments. However, tricyclic antidepressants may be considered an alternative for children with anxiety and depression who cannot tolerate the selective serotonin reuptake inhibitors.

Tricyclic antidepressants have been prescribed as third-line treatments for patients with ADHD who have failed adequate trials of stimulants, who could not tolerate the adverse effects of stimulants, and who had co-occurring conditions such as depression, anxiety, tic disorders [Wilens et al., 1993], or enuresis. Studies have demonstrated significant improvement in ADHD symptoms with imipramine, nortriptyline [Saul, 1985; Wilens et al., 1993], and desipramine [Biederman et al., 1986, 1989; Gastfriend et al., 1984]. However, tricyclic antidepressants appear to be less effective than methylphenidate [Rapoport et al., 1974], and are associated with a large dropout rate over time due to loss of efficacy or side effects [Quinn and Rapoport, 1975].

Studies of imipramine in the treatment of separation anxiety disorder and school refusal have yielded inconsistent findings [Bernstein et al., 2000; Gittelman-Klein and Klein, 1971; Klein et al., 1992]. The largest double-blind, placebo-controlled study of adolescents with school refusal and comorbid depression and anxiety found greatest improvement in school attendance in youngsters who received a combination of imipramine and cognitive behavioral therapy when compared to youngsters receiving cognitive behavioral therapy alone [Bernstein et al., 2000]. Case reports describe the usefulness of tricyclic antidepressants in the treatment of panic disorder [Ballenger et al., 1989; Black and Robbins, 1990], but no placebo-controlled trials have been carried out.

Imipramine appears to be effective in treating enuresis as an initial agent.[Fritz et al., 2004], and in children who have failed nonpharmacological interventions and desmopressin [Gepertz and Neveus, 2004]. Efficacy appears to be correlated with serum concentration [Fritz, 1994].

Clomipramine is approved for the treatment of OCD in adults and children older than 10 years. In short-term clinical trials, clomipramine has demonstrated superiority over placebo [DeVeaugh-Geiss et al., 1992; Leonard et al., 1989] in reducing symptoms of OCD and preventing relapse [Leonard et al., 1991]. Improvement in clomipramine-treated subjects was sustained during 1-year open treatment [DeVeaugh-Geiss et al., 1992].

Pharmacology

Tricyclic antidepressants are divided into two groups, based on the number of methyl groups bonded to the nitrogen atom of the side chain. Tertiary amines (i.e., imipramine, amitriptyline, clomipramine, trimipramine, and doxepin), which have two methyl groups, are metabolized into secondary amines. Several secondary amines, which have one methyl group, are marketed as independent entities, including desipramine (metabolite of imipramine), nortriptyline (metabolite of amitriptyline), and protriptyline (metabolite of trimipramine).

Tricyclic antidepressants differ from each other in their ability to block the reuptake of norepinephrine and serotonin (5HT), and act as competitive antagonists at muscarinic, histaminic (H₁), and adrenergic (α₁, α₂) receptors. Their therapeutic mechanism of action is thought to result from their effect on norepinephrine and serotonin systems in the brain. Clomipramine distinguishes itself from other tricyclic antidepressants in its relatively potent inhibition of serotonin reuptake in addition to its noradrenergic activity. Tricyclic antidepressant activity at central and peripheral muscarinic, histaminic, and adrenergic receptors is not thought to be involved in the clinical response, but rather is associated with side effects.

Tricyclic antidepressants are well absorbed in the gastrointestinal tract. They are metabolized in the liver by the cytochrome P-450 system. Although multiple isoenzymes are involved in their metabolism (i.e., CYP1A2, CYP2D6, CYP3A4, and CYP2C), the CYP2D6 isoenzyme is of particular importance. About 7 percent of the population are poor metabolizers, with reduced activity of CYP2D6. This results in increased plasma concentrations of tricyclic antidepressants and other medications metabolized by CYP2D6 compared with most of the population.

Co-administration of the tricyclic antidepressants with medications that inhibit cytochrome P-450 (e.g., some selective serotonin reuptake inhibitors) can result in clinically significant interactions. These drug interactions may result in discernible changes in efficacy or exacerbation of common tricyclic antidepressant side effects. Specifically, such interactions may result in elevated blood levels of tricyclic antidepressants, leading to a silent but potentially significant impact on cardiac conduction (i.e., prolonged QTc interval). EKG monitoring is required for patients taking such drug combinations.

Clinical Management

Adverse Effects

Side effects of tricyclic antidepressants include sedation, weight gain, orthostatic hypotension, and anticholinergic side effects (i.e., dry mouth, constipation, blurred vision, and urinary retention). Serious anticholinergic effects may occur, including paralytic ileus, exacerbation of narrow-angle glaucoma, and the anticholinergic syndrome (“mad as a hatter, dry as a bone, blind as a bat, hotter than a hare”).

Eight sudden deaths have been reported in children and adolescents taking tricyclic antidepressant medications [Popper and Zimnitzky, 1995; Riddle et al., 1993; Varley and McClellan, 1997]. Six were taking desipramine and two were taking imipramine. Although the role of the tricyclic antidepressants in these tragic deaths is unclear, it has been speculated that the cause of death was cardiac in nature, most likely due to malignant arrhythmias. After the reports, child psychiatrists have become more cautious about prescribing tricyclics in general, carefully screening patients at risk for adverse effects and monitoring cardiac status with serial EKGs [Elliot and Popper, 1990/1991]. It is not clear whether the vulnerability to such medication effects can be determined from routine monitoring of cardiac function. Those who experience increased QTc above that which is considered safe while on tricyclic antidepressants should be monitored closely, and the dose should be adjusted downward until resolution of QTc prolongation.

Behavioral side effects include early and acute increases in anxiety or depression. Tricyclic antidepressants may also induce manic episodes or rapid cycling in patients with a history of bipolar disorder. A family history of bipolar disorder may be a risk factor for this effect, but such a history is not a contraindication for the use of antidepressants. A discontinuation syndrome (i.e. nausea, headache, and malaise), hypomania, or mania may result from abrupt cessation of antidepressants [Haddad, 2001] and may be associated with manic reactions [Narayan and Haddad, 2010]. The FDA mandated a black box warning regarding increases in suicidal ideation for all antidepressants (see “Selective Serotonin Reuptake Inhibitors”).

Drug Interactions

Tricyclic antidepressants have multiple drug interactions. The product information for each of the tricyclic antidepressants provides details of specific drug interactions. A combination of tricyclic antidepressants and monoamine oxidase inhibitors can precipitate a hypertensive crisis requiring immediate medical treatment. Early signs and symptoms include headache, stiff neck, palpitations, sweating, nausea, and vomiting.

Medications that inhibit CYP2D6 activity, such as fluoxetine, paroxetine, and haloperidol (refer to prescribing information for other inhibitors), result in increased serum concentrations of tricyclic antidepressants and increased risk for tricyclic antidepressant side effects. Decreased serum concentrations may be caused by cigarette smoking, lithium, ascorbic acid, ammonium chloride, barbiturates, and primidone. Oral contraceptives may induce hepatic enzymes and decrease tricyclic antidepressant serum concentrations. Tricyclic antidepressants may block the activity of antihypertensives (e.g., guanethidine, propranolol, clonidine). Tricyclic antidepressants may increase the sedative effects of other drugs, such as over-the-counter cold medications, alcohol, opioids, anxiolytics, and hypnotics.

Clinical Applications

The selective serotonin reuptake inhibitors, including citalopram, escitalopram, fluvoxamine, fluoxetine, paroxetine, and sertraline, have FDA approval for a number of mood, anxiety, and eating disorders in adults (Table 49-4). In children and adolescents, fluoxetine, fluvoxamine, and sertraline have FDA approval for OCD; fluoxetine also has FDA approval for major depression. Escitalopram has FDA approval for major depression in adolescents (12 years and older). These indications are based on large-scale, randomized, controlled trials for OCD [Geller et al., 2001; March et al., 1998; Riddle et al., 2001] and major depression [Emslie et al., 1997, 2002, 2009; TADS, 2004]. Other large, randomized, controlled trials of the agents have been completed, but not all have been published. Published trials demonstrating efficacy include citalopram for depression [Wagner et al., 2004b]; fluoxetine [Birmaher et al., 2003], fluvoxamine [RUPP, 2001], and sertraline [Walkup et al., 2008] for separation, social, and generalized anxiety; sertraline for depression [Wagner et al., 2003] and generalized anxiety [Rynn et al., 2001]; paroxetine [Wagner et al., 2004a] and venlafaxine [March et al., 2007] for social phobia, and fluoxetine for stereotypies in children with autistic spectrum disorder [Hollander et al., 2001].

The use of the selective serotonin reuptake inhibitors for depression and anxiety disorder increased significantly in the 1990s, with upwards of 1–2 percent of teens being prescribed these medications for any indication [Olfson et al., 2002a]. Despite the marked increase in use, the prevalence of disorders in children and adolescents that are potentially responsive to the selective serotonin reuptake inhibitors (e.g., depression and anxiety disorders) is much greater (minimum prevalence of 3–5 percent) than the number of children who are taking these medications (1–2 percent), suggesting overall under-utilization of these evidenced-based treatments.

In 2004, the FDA issued a black box warning for increased risk of suicidal ideation and behavior in children and adolescents using all antidepressants. This warning was based on the analysis of data from 24 short-term, placebo-controlled trials of nine antidepressant drugs, involving a total of over 4400 children and adolescents being treated for major depressive disorder, OCD, and other psychiatric disorders. Results of this analysis showed an increased risk of suicidal thoughts and behavior in the treatment groups compared to placebo (approximately 4 percent versus 2 percent). Since this time, prescriptions of antidepressants in pediatric populations have fallen off and recent reports suggest that rates of youth suicide increased after many years of decline [Gibbons et al., 2007]. There is no evidence that the prevalence of mood disorders in children has diminished over this same time period, or that patients who are identified as being depressed are being offered alternative treatments to medication. This is especially ironic, as the current evidence base suggests the limited value of nonpharmacological approaches to the treatment of depression [TADS, 2004].

Pharmacology

Clinical Management

Adverse Effects

Drug Interactions

The product information for each of the selective serotonin reuptake inhibitors provides detailed information regarding specific drug interactions. Because the number of individual interactions is too high to describe here, general principles are discussed. Adverse drug–drug interactions occur most commonly with highly protein-bound medications and with medications that significantly inhibit or induce cytochrome P-450 isoenzymes. Drug–drug interactions can result in loss (e.g., anticonvulsant combinations) or increase in benefit (e.g., cyclosporine and ketoconazole), increase in noticeable side effects (e.g., sedation), and an increase in side effects that can only be identified with careful monitoring (e.g., QTc prolongation). As a class, the selective serotonin reuptake inhibitors have more and more frequent drug–drug interactions because most of the selective serotonin reuptake inhibitors inhibit cytochrome P-450 isoenzymes. Most result in increased benefit of the co-administered agent or increase in noticeable side effects. The caution with using the selective serotonin reuptake inhibitors lies when they are combined with medications that have clinical effects, such as QTc prolongation, that require active monitoring to identify. Ironically, medications used in neuropsychiatry, such as the antipsychotics (e.g., pimozide) and tricyclic antidepressants (e.g., desipramine), have a potential for producing cardiac conduction effects. Careful monitoring is indicated when using selective serotonin reuptake inhibitors in combination with medication that may affect the QTc interval.

Clinical Applications

The serotonin-norepinephrine reuptake inhibitors, which include duloxetine, venlafaxine, and desvenlafaxine, have FDA approval for treatment of depression and anxiety disorders in adults. Duloxetine also has FDA approval for treatment of fibromyalgia and diabetic neuropathic pain in adults. Venlafaxine has also shown promise in the treatment of ADHD in adults [Findling et al., 1996], and in co-occurring major depressive disorder and ADHD in adults [Hornig-Rohan and Amsterdam, 2002]. The use of serotonin-norepinephrine reuptake inhibitors in children and adolescents is off-label. At present, the only published randomized controlled studies of serotonin-norepinephrine reuptake inhibitors in pediatric populations include a positive trial of venlafaxine ER for social phobia [March et al., 2007], and equivocal trials for generalized anxiety disorder [Rynn et al., 2007] and depression [Emslie et al., 2007]. Two small open-label studies of venlafaxine in children and adolescents with ADHD suggest potential benefit for impulsivity and hyperactivity [Findling et al., 2007; Olvera et al., 1996] but were complicated by a high dropout rate due to adverse effects, including behavioral activation [Olvera et al., 1996]. A few case reports have been published on the successful use of duloxetine in youngsters with chronic pain and comorbid major depressive disorder [Meighen, 2007]; essentially, no data are available on use of desvenlafaxine. This review will limit its discussion to venlafaxine use in children and adolescents, given the limited data for duloxetine and desvenlafaxine.

Pharmacology

Adverse Effects

In adults, common adverse events associated with venlafaxine include asthenia, sweating, nausea, constipation, anorexia, vomiting, somnolence, dry mouth, dizziness, nervousness, anxiety, tremor, blurred vision, and sexual dysfunction. In addition, venlafaxine has been associated with increases in diastolic blood pressure that appear dose-dependent, especially with doses above 300 mg daily [Effexor, 2009]. In children and adolescents, treatment-emergent side effects include asthenia, anorexia, nausea, weight loss, abdominal pain, dizziness, somnolence, and nervousness [Emslie et al., 2007; March et al., 2007; Rynn et al., 2007]. Small increases in diastolic blood pressure have been reported in randomized clinical trials but were not considered clinically significant. Venlafaxine appears to have one of the highest rates of treatment-emergent suicide-related behaviors (relative risk of 8.84 percent) compared to other antidepressant medication in children and adolescents [Hammad et al, 2006], and has been associated with other treatment-emergent behavioral abnormalities, including hostility, overactivity, and agitation.

Drug Interactions

Venlafaxine or venlafaxine ER is contraindicated to be used in conjunction with a monoamine oxidase inhibitor and requires a 14-day washout period. Additionally, venlafaxine should only be used cautiously in conjunction with triptans due to the increased risk of serotonin syndrome. Although venlafaxine is metabolized by the CYP2D6 isoenzyme, the combined serum concentration of venlafaxine and its active metabolite does not appear to be altered by poor metabolizers [Effexor, 2009].

Bupropion

Mirtazapine

Clinical Applications

Benzodiazepines are classified as sedative-hypnotic medications. They act as sedatives or anxiolytics at low doses and sleep-inducing hypnotics at high doses. Although rapidly effective in relieving symptoms, they carry a risk of tolerance, dependence, and withdrawal with long-term use. These effects may be avoided by using only moderate doses for short-term therapy (1–2 weeks) [Sadock et al., 2007]. Benzodiazepines differ from each other with regard to relative potency and half-life. In contrast to long-half-life benzodiazepines, short-half-life benzodiazepines are associated with less daytime sedation, more frequent dosing, earlier and more severe withdrawal syndromes, rebound insomnia, and anterograde amnesia. Long-half-life benzodiazepines require less frequent dosing and have a less severe withdrawal syndrome, but high plasma concentrations of the drug may accumulate and cause signs of toxicity over time [Sadock et al., 2007].

Case reports and small open trials indicate that clonazepam and alprazolam may be helpful in treating anxiety disorders in children and adolescents [Biederman, 1987; Kutcher and MacKenzie, 1988; Simeon and Ferguson, 1987]. However, placebo-controlled studies have not supported their use [Graae et al., 1994; Simeon et al., 1992].

With the availability of more effective anxiolytics (e.g., selective serotonin reuptake inhibitors, tricyclic antidepressants), benzodiazepines are not considered first-line treatment for anxiety disorders in children and adolescents. However, benzodiazepines can be useful for the short-term management of specific anxiety-provoking situations, such as medical procedures, or as short-term adjunctive treatment with selective serotonin reuptake inhibitors in the initial treatment of severe anxiety disorders. Chronic administration of benzodiazepines for anxiety should be undertaken only after patients have failed more effective treatments for anxiety, including other medications and behaviorally oriented psychotherapy.

Pharmacology

Benzodiazepines are well absorbed from the gastrointestinal tract. They are extensively metabolized in the liver and excreted in urine. Benzodiazepine plasma concentrations are sensitive to medications that inhibit or induce cytochrome P-450 isoenzymes (see “Drug Interactions”). Benzodiazepines activate gamma-aminobutyric acid (GABA) benzodiazepine binding sites, which open chloride channels, resulting in a reduced rate of neuronal firing and muscle activity (Table 49-6) [Sadock, 2007].

Clinical Management

In considering benzodiazepine therapy for a patient, the clinician should inquire about any possible history of substance abuse, as combined use of benzodiazepines and other CNS depressants can be lethal [Moody, 2004], and benzodiazepines can also be abused or diverted [Longo and Johnson, 2000]. Family history of substance abuse may identify patients who are at risk for drug abuse or dependence, and household members with substance use histories may indicate a risk for drug diversion. Concomitant use of medications and co-occurring medical conditions that may affect the metabolism of benzodiazepines should be identified. The patient and family should be informed about the risks of tolerance, dependence, and withdrawal effects, as well as possible effects on cognitive and motor functioning. During treatment, the clinician must ensure that medications are not overused, abused, or diverted, and should monitor for possible adverse cognitive or behavioral side effects. After the treatment trial is complete, the medication should be tapered slowly because of the potential for withdrawal and rebound anxiety [Coffey, 1993; Kutcher et al., 1992].

Adverse Effects

The most frequent adverse effects in children and adolescents are sedation, drowsiness, and decreased mental acuity [Biederman, 1991]. Benzodiazepines may cause paradoxical behavioral disinhibition, which manifests as increased agitation, aggression, and hyperactivity. This has been described with clonazepam in children [Graae et al., 1994] and adolescents [Reiter and Kutcher, 1991]. Anterograde amnesia also has been associated with benzodiazepines, particularly high-potency benzodiazepines [Uzun et al., 2010].

Benzodiazepine therapy carries a risk of tolerance, dependence, and withdrawal effects. Abrupt discontinuation or rapid dose reduction has been associated with seizures, delirium, and withdrawal symptoms [Uzun et al., 2010].

Drug Interactions

Additive CNS depression may occur when benzodiazepines are administered concomitantly with other CNS depressants, such as alcohol. Disulfiram may decrease the clearance of diazepam and chlordiazepoxide. Because benzodiazepines are metabolized by cytochrome P-450 isoenzymes, inhibitors and inducers of these isoenzymes can significantly alter the plasma concentration of benzodiazepines. Clonazepam and phenytoin decrease each other’s plasma concentrations. Valproate may increase plasma concentrations of lorazepam [Depakote, 2009].

Clinical Applications

Buspirone (BuSpar®) is approved for use in the treatment of anxiety disorders and for short-term relief of symptoms of anxiety [Buspar, 2007]. It is typically used in the treatment of generalized anxiety disorder. Buspirone does not appear to be effective as monotherapy in other anxiety disorders, such as post-traumatic stress disorder, panic disorder [Sheehan et al., 1988], or social phobia [van Vliet et al., 1997]. Although randomized, controlled trials do not support the use of buspirone to augment selective serotonin reuptake inhibitors in partial responders, some patients with depression [Dimitriou and Dimitriou, 1998] and OCD [Jenike et al., 1991] may benefit.

Open trials of buspirone in adolescents with generalized anxiety disorder [Kutcher et al., 1992] and children and adolescents with varied anxiety disorders [Simeon, 1991] demonstrate benefit. Additionally, buspirone appeared to improve anxiety and irritability in children with pervasive developmental disorders [Buitelaar et al., 1998] and may improve anxiety and aggression symptoms in some hospitalized children [Pfeffer et al., 1997]. However, two large unpublished, industry-sponsored, randomized, controlled trials in children and adolescents with generalized anxiety disorder did not demonstrate benefit [Buspar, 2007]. Buspirone offers several potential benefits over benzodiazepines, including lack of euphoric effects, performance impairment, abuse potential, and withdrawal symptoms with discontinuation. Although buspirone appears to be as effective as benzodiazepines in decreasing psychic symptoms of anxiety, it appears to be less effective for somatic symptoms of anxiety, and the onset of therapeutic effect may be delayed 2–4 weeks [Kutcher et al., 1992].

Pharmacology

Buspirone differs from benzodiazepines in that it has no activity at the GABA-benzodiazepine receptor complex. Buspirone acts as a partial agonist or mixed agonist/antagonist at 5HT_1A receptors, as an agonist at postsynaptic dopamine receptors, and as an antagonist at presynaptic autoreceptor and postsynaptic sites in the CNS. Buspirone is rapidly absorbed from the gastrointestinal tract. Only 4 percent of the dose reaches systemic circulation because of extensive first-pass metabolism. Buspirone is metabolized in the liver, primarily by the CYP3A4 isoenzyme. Its primary metabolite, 1-pyrimidinylpiperazine (1-PP), has clinical activity. The elimination half-life is about 2–4 hours in all age groups and is prolonged by renal or hepatic impairment. Unlike adolescents or adults, prepubertal children show the highest concentration of the active metabolite, 1-PP, after oral administration of buspirone at doses of 15–60 mg/day, which may account for higher incidence of side effects in this group [Salazar et al., 2001]. The onset of clinical response may appear within the first 2 weeks of therapy, but optimal therapeutic effect is usually delayed [Kutcher et al., 1992].

Clinical Management

Laboratory screening is not required before prescribing buspirone. The typical starting dose is 5 mg taken 2–3 times daily, with gradual increases to 15 mg taken three times daily. Younger children may be started at 2.5–5 mg/day and gradually increased to a total dose of 20–30 mg/day, administered in divided doses 2–3 times daily [Scahill and Martin, 2002].

Adverse Effects

The most common adverse effects of buspirone are dizziness, headache, drowsiness, and lightheadedness [Buspar, 2007]. Side effects in adults appear to be dose-related and to decrease over time [Feighner and Cohn, 1989]. CNS depression is less common than in benzodiazepines. Behavioral activation, including agitation, aggression and euphoric mania, has been described in prepubertal children [Pfeffer et al., 1997].

Drug Interactions

Combination of buspirone and monoamine oxidase inhibitors may result in a hypertensive crisis, requiring emergency medical attention. Because of reports of liver transaminase elevations with the concomitant use of buspirone and trazodone, the manufacturer recommends monitoring liver function when prescribing this combination. Because buspirone is highly protein-bound, it may displace other medications that are protein-bound and increase their free plasma level. Inducers of CYP3A4 may decrease plasma levels of buspirone. Alternatively, inhibitors of CYP3A4 (e.g., fluvoxamine, erythromycin, grapefruit juice) may increase the serum concentration of buspirone. Increased serum concentrations of haloperidol have been reported when combined with buspirone [Kudo and Ishizaki, 1999].

Clinical Applications

Lithium is FDA-approved for use in the treatment of bipolar affective disorder. It is effective in managing acute manic and depressive episodes, preventing or diminishing the intensity of subsequent episodes in maintenance treatment, and decreasing mood instability between episodes [Goodwin and Jamison, 1990]. Data in children and adolescents with bipolar disorder are limited and offer mixed results. Several earlier studies showed good response rates [DeLong and Aldershof, 1987; Strober et al., 1990; Varanka et al., 1988], although this was not universally demonstrated [Carlson et al., 1992]. These studies were largely uncontrolled and naturalistic, varied in diagnostic criteria, and included mixed diagnostic groups (e.g., bipolar I, bipolar II, bipolar not otherwise specified (NOS)), making interpretation of results difficult. A more recent large open-label study has provided additional support for the use of lithium in adolescents with bipolar disorder [Kafantaris et al., 2003], although the need for larger randomized, controlled studies of children and adolescents in both the acute and maintenance phases is recognized and protocols are currently in development [Findling et al., 2008]. Lithium also has been used with positive results in the treatment of severe aggression in adults [Sheard, 1975] and children [Campbell et al., 1984; DeLong and Aldershof, 1987; Malone et al., 2000].

Pharmacology

Lithium is an alkali metal, similar to sodium and potassium, and it is available in salt form as lithium carbonate or lithium citrate. Lithium has effects at the cellular level (i.e., cell membrane ion channels and cyclic adenosine monophosphate second-messenger cellular processes) and in neurotransmitter systems. However, the specific mechanism of action is unclear [Lenox and Wang, 2003].

Lithium is available in immediate-release and sustained-release preparations. Ninety percent of immediate-release lithium is absorbed by the gastrointestinal tract; 60–90 percent of extended-release lithium is absorbed. Immediate-release lithium reaches peak serum levels in 1–1.5 hours (4–12 hours for the extended-release form), and it is almost completely excreted in urine, with the remainder excreted in the feces and breast milk. Lithium has a half-life of about 20 hours and reaches steady-state levels after 1 week of administration [FDA, 2010].

Clinical Management

Before starting lithium therapy, the patient should be screened for conditions that may contraindicate or complicate lithium therapy, such as pregnancy, renal disease, or diabetes mellitus. A medical evaluation should be completed, including an EKG, laboratory screen for electrolytes, kidney and thyroid function tests, CBC, urinalysis, and urine pregnancy screen.

Lithium is started at dosages ranging from 300 to 600 mg/day. Formulas [Weller et al., 1986] and nomograms [Alessi et al., 1999; Geller and Fetner, 1989] have been developed to guide lithium dosing in pediatric populations. Trough lithium levels (10–12 hours after last dose) are drawn when steady state is achieved, at least 5 days after initiating lithium or changing the dose. The recommended therapeutic serum levels range from 0.6 to 1.2 mEq/L in adults, although therapeutic level has not been firmly established in children. Lithium levels between 0.8 and 1.0 mEq/L may offer the best prophylaxis against mania, whereas lower levels may be adequate for prophylaxis against depression [Severus et al., 2005, 2009]. Lithium levels above the therapeutic range increase the risk of toxicity. Early therapeutic effects may be detected at 10–14 days after achieving a therapeutic level; however, there is often a significant lag of 4–6 weeks [Kowatch et al., 2000]. Between 4 and 6 weeks of lithium at therapeutic serum concentrations constitutes an adequate trial. Because several days are required to achieve therapeutic plasma concentrations, antipsychotic medications may be necessary to stabilize severe mania acutely, especially if psychotic symptoms are present.

After mood symptoms are stabilized, lithium levels should be routinely checked every 3–6 months. Other indications to obtain a lithium level include symptoms of toxicity, worsening side effects, and breakthrough mood symptoms. Urinalysis, renal evaluation [Jefferson, 2010], and thyroid function tests should be repeated every 3–6 months to monitor for the emergence of subclinical renal and thyroid dysfunction (Table 49-7).

Stabilization of manic symptoms may be prolonged up to several years [Biederman et al., 1998]. Lithium should be continued for at least 18 months after manic symptoms have been stabilized [APA, 2002] because discontinuation of lithium therapy, especially abrupt discontinuation, is associated with increased relapse rates [Strober et al., 1990].

Adverse Effects

Side effects of lithium are polydipsia, polyuria or enuresis, gastric distress (i.e., nausea, vomiting, and diarrhea), weight gain, tremor, fatigue, leukocytosis, acne, and mild cognitive impairment [Alessi et al., 1999; Silva et al., 1992]. Serious renal side effects are rare and include nephrotic syndrome, features of distal renal tubular acidosis, nonspecific interstitial fibrosis, and renal failure [Alessi et al., 1999; Gelenberg and Schoonover, 1991]. Lithium therapy can induce alterations in thyroid function tests and produce hypothyroidism [Alessi et al., 1999]. Lithium can have rare cardiac side effects, including first-degree atrioventricular block, irregular sinus rhythm, and increased premature ventricular contractions [Gelenberg and Schoonover, 1991]. Reversible conduction abnormalities in children have been demonstrated on their EKGs [Campbell et al., 1972]. Studies examining the long-term effects of lithium treatment of children are lacking. Intrauterine exposure to lithium during the first trimester has been associated with teratogenic effects, most commonly Ebstein’s anomaly of the tricuspid valve [Giles and Bannigan, 2006]. Lithium has been associated with pseudotumor cerebri in children [Kelly et al., 2009].

Drug Interactions

Lithium levels are affected by medications or conditions that can alter the body’s fluid and electrolyte status. Dehydration and extremely low salt intake can result in elevated lithium levels. Likewise, an extremely high salt intake can result in lowered lithium levels. Many medications, such as diuretics, nonsteroidal anti-inflammatory drugs, and tetracycline, can cause an increase in lithium levels by influencing renal clearance [Finley et al., 1995].

Lithium is commonly used in combination with other psychotropic medications in the treatment of bipolar disorder. However, a number of case reports have described neurotoxicity occurring rarely when combined with other mood-stabilizing medications, including antipsychotic medications [Finley et al., 1995].

Clinical Applications

Valproic acid is approved for the treatment of seizures, migraine prophylaxis, and the manic phase of bipolar affective disorder in adults [Depakote, 2009]. The initial support for use of valproic acid in child psychiatric practice came from the adult literature. The evidence for its use in the treatment of juvenile bipolar disorder, however, is limited. Although some support for the use of valproic acid was derived from earlier case series and open trials [Henry et al., 2003; Kowatch et al., 2000; Wagner et al., 2002], a large, double-blind, randomized, placebo-controlled trial did not demonstrate superiority of divalproex extended release over placebo in the acute phase management of bipolar disorder [Wagner et al., 2009]. Divalproex sodium has also been shown in small studies to benefit children and adolescents with explosive temper and mood lability [Donovan et al., 2000] and for treatment of core symptoms of autism and associated symptoms of affective instability, aggression, and impulsivity [Hollander et al., 2001].

Pharmacology

The mechanism of action of valproic acid is unidentified, although it is known to increase brain concentrations of the inhibitory neurotransmitter GABA. Valproic acid is available in several preparations: valproate (Depakene® capsules); valproate syrup (Depakene® syrup); divalproex (Depakote®); divalproex capsules with coated particles (Depakote Sprinkle®); divalproex sodium, delayed-release tablets (Depakote®); and divalproex sodium in extended-release tablets (Depakote ER®).

Valproate rapidly converts to valproic acid in the stomach. Divalproex dissociates into valproic acid in the gastrointestinal tract. Valproic acid is rapidly and almost entirely absorbed from the gastrointestinal tract. Peak plasma concentrations are achieved 1–4 hours, 3–5 hours, and 7–14 hours after oral administration of valproate, divalproex, and extended-release tablets, respectively. Food intake can retard the time to peak plasma concentrations, although this has little clinical consequence once steady-state concentrations have been achieved. Valproic acid binds plasma proteins; therefore, the presence of other protein-binding drugs can increase the unbound or free fraction of valproic acid. Conversely, the addition of valproate with another protein-bound drug can increase the unbound or free fraction of the other drug. Renal and hepatic insufficiency can reduce protein binding, which can increase unbound valproate. Additionally, plasma protein binding of valproic acid is dependent on drug concentration. Concentrations of free valproic acid increase at serum levels above 100 mg/mL. Peak and trough concentrations with delayed-release tablets and capsules containing coated particles may differ from the equivalent dose of valproic acid. Valproic acid is metabolized in the liver and excreted predominantly in the urine [Malone et al., 2000; Depakote, 2009]. No therapeutic serum concentrations for the treatment of behavioral or psychiatric disorders have been determined.

Clinical Management

In addition to the routine psychiatric or neurologic history, patients should be asked about risk factors affecting treatment with valproate, including pregnancy, disorders of metabolism, hepatic or renal disease, concomitant medications, and coagulation disorders. Laboratory screening should include liver function tests and a CBC, and a pregnancy test in females [Danielyan and Kowatch, 2005]. Parents should be educated about potential signs of hepatic or pancreatic toxicity and polycystic ovary disease.

Recommended initial dosing of valproic acid is 10–15 mg/kg/day, given in divided doses two or three times per day. Dosage may be increased by 5–10 mg/kg/day at weekly intervals [Malone et al., 2000; Depakote, 2009]. The manufacturer’s maximum recommended dosage is 60 mg/kg/day. However, starting at lower doses and using slower upward adjustments of dose may be required for tolerability. Although therapeutic levels for treating mania have not been established, achieving a plasma concentration of 85–110 mg/L [Kowatch et al., 2000] is needed to consider a trial adequate for determining outcome in an individual patient. Divalproex extended release allows for once-daily dosing.

In adult psychiatry, an oral loading dose of valproate is sometimes used to achieve a therapeutic plasma concentration and to gain rapid control over mania. The tolerability of an oral loading dose of divalproex sodium was examined in children and adolescents [Good et al., 2001]. A loading dose of 15 mg/kg/day resulted in therapeutic plasma levels by day 5 of treatment and it was well tolerated. Dosage adjustment was required for overweight children, in order to avoid supratherapeutic drug levels [Good et al., 2001]. Total valproic acid levels should be checked when steady state is reached, after any dosage change, after conversion to a different preparation of valproate, for signs of toxicity, or for a change in clinical status.

Maintenance involves monitoring of the therapeutic response and side effects and requires laboratory screening. Liver enzymes, CBC, and serum valproic acid levels should be periodically monitored, although the exact frequency of testing has not been determined. Initially, monitoring may be done every 3–6 months for the first year and semiannually thereafter (see Table 49-7).

Adverse Effects

Adverse effects include sedation, nausea, vomiting, gastrointestinal upset, increased appetite, weight gain, and hair loss. Rare side effects include thrombocytopenia, platelet dysfunction, prolonged bleeding time, elevation in liver transaminase levels, and lactate dehydrogenase, and can occur without clinical symptoms. Children younger than 2 years appear to have an increased risk of hepatotoxicity, and extreme caution should be used in this population. This heightened risk may result from an undiagnosed metabolic or mitochondrial disorder. In particular, children with urea cycle defects, such as ornithine transcarbamylase deficiency, should not receive this medication. Carnitine supplementation should be considered when prescribing this medication, as this may facilitate the processing of valproate in the mitochondrial β-oxidation pathway. Valproate has also been associated with rare cases of hyperammonemic encephalopathy [Barrueto and Hack, 2001; Elgudin et al., 2003], which occurs more commonly with polypharmacy and with associated liver dysfunction [Verrotti et al., 1999]. One case reported in the child psychiatry literature manifested as increased aggression and confusion, an EEG with symmetric 5- to 6-Hz waves, elevated serum ammonia levels, and normal levels of serum liver transaminases [Yehya et al., 2004]; the situation resolved with valproic acid discontinuation. Elevated ammonia levels have also been observed in the absence of encephalopathy, but the clinical relevance of such laboratory abnormalities is unclear. Other reports suggest that oral carnitine can be useful in reducing ammonia levels [Bohan et al., 2001]. Rare cases of pancreatitis have been reported [Malone et al., 2000; Depakote, 2009]. Cases of polycystic ovary disease have been reported in some women treated with valproate for seizure disorder [Isojarvi et al., 1993]. It is unclear whether this is related to the medication or is an endocrine abnormality related to the epilepsy [Davis et al., 2000]. Intrauterine exposure to valproic acid during the first trimester is associated with neural tube defects [Malone et al., 2000; Depakote, 2009].

Drug Interactions

As it is metabolized by the cytochrome-P450 system, valproic acid has many interactions with medications that are also metabolized by this system. Medications that may increase valproic acid concentrations include erythromycin, selective serotonin reuptake inhibitors, guanfacine, and aspirin. Medications that may decrease concentrations of valproate include carbamazepine. Additionally, valproic acid may result in increases in concentration of other anticonvulsants (phenytoin, phenobarbital, primidone, carbamazepine, and lamotrigine), tricyclic antidepressants, diazepam, and warfarin. Great caution should be utilized when valproic acid is used in combination with lamotrigine due to enhanced risk of Stevens–Johnson syndrome. When valproic acid is combined with antipsychotics, CNS depressant effects and extrapyramidal side effects may increase. Valproic acid may increase lithium-associated tremor [Danielyan and Kowatch, 2005].

Clinical Applications

Limited data are available on the efficacy of carbamazepine for the treatment of bipolar disorder children and adolescents. Justification for its use in clinical practice has rested on the adult literature, which is difficult to interpret. Carbamazepine was superior to placebo and comparable to lithium [Lerer et al., 1987; Small et al., 1991] in improving manic symptoms. However, it was not as effective as valproate, and patients on carbamazepine required adjunctive medication to achieve stabilization [Vasudev et al., 2000]. Similarly, in children, carbamazepine was as effective as lithium but less so than valproic acid [Kowatch et al., 2000]. Carbamazepine may be helpful in the depressive phase of bipolar disorder [Dilsaver et al., 1996].

Through the induction of hepatic enzymes carbamazepine can have multiple drug interactions that complicate combined pharmacotherapy in the management of bipolar disorder. Clinicians and patients may prefer carbamazepine to the other mood stabilizers because of its reduced tendency to cause weight gain.

Pharmacology

Carbamazepine is structurally related to tricyclic antidepressants. It reduces activity at sodium and calcium channels, resulting in reduced synaptic transmission. It is available in immediate-release forms (e.g., Tegretol® tablets, chewable tablets, suspension) and extended-release capsules (i.e., Tegretol XR®). It is slowly absorbed from the gastrointestinal tract. The plasma half-life gradually decreases over 3–5 weeks from 25–65 hours to approximately 12–17 hours due to autoinduction. Steady-state concentrations are achieved in 2–4 days. Between 75 and 90 percent of the drug is bound to plasma proteins. It is metabolized in the liver and excreted in the urine [Sachs et al., 1994; Tallian et al., 1996].

Clinical Management

Patients should be evaluated for medical conditions that may contraindicate use of carbamazepine, including liver disease, hematologic disorders, immune disorders, cardiac disease, and pregnancy. Laboratory screening should include a urine pregnancy screen, liver function tests, and a CBC. Because laboratory screening can fail to predict impending hepatic and hematologic complications, parents should be aware of signs and symptoms indicating serious adverse effects, such as rash, symptoms of aplastic anemia or agranulocytosis, and hepatitis or cholestasis. Sexually active patients should be informed of its teratogenic effects and the possible decrease in effectiveness of oral contraceptives [Sachs et al., 1994; Tallian et al., 1996].

Dosage is initiated at 100–300 mg/day, increasing as tolerated while monitoring blood levels. Therapeutic blood levels in the treatment of mania have not been established. Recommended levels are the same as those for seizure disorder, 4–12 mg/mL [Rosenberg et al., 1994]. Blood levels should be checked 5 days after initiating or changing doses, and again in 4 weeks because of an anticipated reduction in plasma levels due to autoinduction. Although it is recommended that liver function tests, bilirubin levels, and CBCs be routinely checked every 3 months, the utility of such testing in predicting serious hepatic or hematologic complications is questionable (see Table 49-7). Hyponatremia resulting from carbamazepine use is rare in children, and if it is clinically suspected, sodium levels should be obtained.

Adverse Effects

The most common adverse effects are nausea, vomiting, sedation, and ataxia [Sachs et al., 1994; Tallian et al., 1996]. Carbamazepine can cause benign elevations in liver transaminases; hepatitis, indicated by elevated liver transaminase levels greater than three times normal; and cholestasis, indicated by elevations in bilirubin and alkaline phosphatase. Carbamazepine therapy has been associated with leukopenia in 1–2 percent of patients, and agranulocytosis or aplastic anemia in 1 of 250,000 patients. Although blood dyscrasias are not always predicted by leukopenia, CBCs are recommended at 3- to 6-month intervals. Other adverse effects observed with carbamazepine include cognitive or behavioral disturbance [Carpenter and Vining, 1993], CNS toxicity, altered cardiac conduction, and life-threatening dermatologic conditions (i.e., exfoliative dermatitis, toxic epidermal necrolysis, and Stevens–Johnson syndrome). Intrauterine exposure has been associated with neural tube defects and other fetal malformations [Rosa, 1991]. Case reports of manic reactions to carbamazepine may be related to its antidepressant structure [Myers and Carrera, 1989].

Drug Interactions

Carbamazepine is an inducer of CYP3A4, leading to decreased plasma concentrations of medications metabolized by this isoenzyme. Several medications that are commonly combined with carbamazepine in psychiatric practice may be affected, including anticonvulsants (e.g., lamotrigine, clonazepam, valproic acid), antidepressants (e.g., bupropion, clomipramine, desipramine, imipramine, nefazodone), and antipsychotics (e.g., clozapine, fluphenazine, haloperidol, olanzapine, ziprasidone). Carbamazepine levels can be increased by fluoxetine, lamotrigine, valproate, cimetidine, and erythromycin. Coadministration of carbamazepine with nefazodone results in inadequate levels of nefazodone and its active metabolite, and is therefore contraindicated. Combination with lithium or antipsychotics may increase the risk of adverse neurologic effects (e.g., drowsiness, dizziness, ataxia) [Sachs et al., 1994; Tallian et al., 1996].

Lamotrigine

Oxcarbazepine

Gabapentin

Topiramate