In this study, adverse events

included drowsiness, irritability, and enuresis, as well as diarrhea, constipation, and

social withdrawal; additional adverse events from an open-label study of guanfacine

in children with PDD include sleep disturbance (insomnia or midsleep awakening).

49

Doses in both studies ranged from 1 to 3 mg/day.

Overall, a small collection of literature suggests that clonidine and guanfacine may

be beneficial for the treatment of ADHD symptoms in children with ASD/ID, with

similar adverse events as reported in neurotypical children, though more evidence is

needed to draw any more extensive conclusions about the use of these medications in

the ID/ASD population.

Atomoxetine

Atomoxetine is a medication for ADHD that acts by inhibiting the reuptake of

norepinephrine, therefore increasing the amount of norepinephrine in the synaptic

cleft. Dosing is generally once daily, with adjustments after a minimum of 3 days to a

target daily dose. Clinical benefit has been noted within the first 1 to 2 weeks.

50

Reichow et al.’s

40 meta-analysis found two RCTs comparing atomoxetine to

placebo in children with PDD, though only the larger of the two (n = 97)

51 showed

statistically significant benefits for the treatment of global ADHD symptoms and

hyperactivity (effect sizes calculated by Reichow et al.

40 of γ = 0.83 and γ 0.80,

respectively). In this study by Harfterkamp et al.,

51 changes in the subscales of

inattention and oppositional behavior did not reach statistical significance for

children on atomoxetine compared to placebo.

Children with PDD in the Harfterkamp et al.

51 study demonstrated increased rates

of nausea, decreased appetite, and early morning awakening compared to study

participants taking placebo. Mean dosage in the two studies was 1.2 mg/kg/day

51 and

44.2 mg/day.

52 This dosing is perhaps only slightly lower than other large studies of

atomoxetine in children with ADHD without developmental disorders (mean final

atomoxetine dose was 1.45 mg/kg/day, 53.0 mg/day

53

), though documentation from

the drug manufacturer’s studies, of presumably mostly neurotypical children with

ADHD, indicates no additional benefits for doses above 1.2 mg/kg/day.

50

Similar to the α2 agonists, it seems that atomoxetine shows benefits in the literature

for addressing ADHD symptoms in children with PDD, particularly for hyperactive

symptoms, with similar dosing and adverse events to children with ADHD without

developmental disorders.

Other Medications

Tricyclic antidepressants (TCA), which primarily act as serotonin–norepinephrine

reuptake inhibitors (SNRIs), had been a part of older treatment algorithms for the

treatment of ADHD in typically developing children

54 and, as a result, have also been

studied in children with ASD. In Gordon et al.’s crossover RCT of clomipramine and

desipramine versus placebo

55

(5 weeks each intervention), both drugs were

significantly superior to placebo for reducing the hyperactivity domain of the

Children’s Psychiatric Rating Scale (CPRS) Autism Relevant Subscale, though no

different from each other. TCAs are known for adverse effects, primarily

antimuscarinic in nature, which includes dry mouth, blurred vision, decreased

gastrointestinal motility, and urinary retention. TCAs can also cause irregular heart

rhythms, tachycardia and hypotension, and should be used in caution in patients with

preexisting cardiac conditions and can be fatal with just small excesses in dose.

TCAs are also highly metabolized by the cytochrome P450 hepatic enzymes,

including 3A4 and 2D6. Therefore, inhibitors of these enzymes can lead to increased

concentrations and higher risk of side effects including cardiac abnormalities.

Although antipsychotic medications are not typically part of the treatment

algorithm for ADHD in children with typical development,

54,56 studies of risperidone

and aripiprazole for the treatment of children with ASD and irritability/aggression

symptoms did show statistically significant improvement in hyperactivity subscales

(Risperidone

32–34

; Aripiprazole

57,58

). As a result, atypical antipsychotics are part of a

treatment algorithm developed by the Autism Speaks Autism Treatment Network

Psychopharmacology Committee Medication Choice Subcommittee for symptoms of

ADHD in individuals with ASD, if they have not shown adequate improvement from

stimulants, atomoxetine, and α2 agonists.

45

Other medications, which have not yet been included in standard practice, though

have shown some promising evidence for improvement in hyperactivity symptoms in

RCTs of children with ASD, include omega-3 fatty acids,

59

tianeptine,

60 and

adjunctive pentoxifylline

61 and topiramate,

62 when each added to risperidone.

CASE 88-1, QUESTION 2: L.B., now 5 years of age, has been receiving ABA services for nearly 2 years

and has been progressing well. At home though, L.B. has become increasingly disobedient and hyperactive. In

school, his teachers note that he is inattentive, wandering around the classroom and staring out the windows. In

addition, during story time and craft time, he is disruptive and hyperactive, running around the room. In

response, L.B.’s pediatrician facilitates an increase in the number of hours of in-home ABA services, whose

providers also work with L.B.’s parents’ teaching behavioral management techniques. The pediatrician also

recommends certain classroom modifications to help with attention. This, however, does not reduce the

outbursts or inattentive and hyperactive behavior, both at home and school. The pediatrician believes that a trial

of pharmacotherapy may be helpful. The pediatrician consults you, the clinical pharmacist, to discuss an

appropriate treatment plan.

What pharmacotherapy and formulation would be most appropriate for L.B.’s target symptoms?

Clinical trials have shown that stimulants can be a very effective treatment option

for children with attention-deficit/hyperactivity disorder (ADHD). The American

Academy of Child and Adolescent Psychiatry (AACAP) and the American Academy

of Pediatrics (AAP) recommend the use of methylphenidate, over that of

amphetamines or nonstimulant medication, in preschool children where behavioral

modification therapy has not been adequate. The metabolism of methylphenidate in

preschool children is slower than that of older children, and therefore the starting

dose is lower, and the optimal dose is likely lower than in older children.

Based on this information, you recommend starting L.B. on methylphenidate IR

solution 1.25 mg twice daily, with an increase to 2.5 mg twice daily on day 3, with

subsequent dose increases every 3 days to reduce target symptoms, up to a maximum

of 7.5 mg 3 times daily. You recommend that the parents keep a journal of the

responses and possible experienced adverse effects during this period with a report

back to the pediatrician at the end of the week.

Irritability/Aggression

Irritability/aggression is the pharmacologic target behavior with the most robust RCT

evidence in children with ASD. Most of this evidence comes from four large RCTs

of either risperidone or aripiprazole.

32,33,57,58

In this context, risperidone and

aripiprazole are the only two medications with FDA approval for the

p. 1856

p. 1857

use in individuals with developmental disabilities. Specifically, each has an FDAapproved indication for the treatment of autistic disorder-associated irritability in

children and adolescents (risperidone—age 5–17 years; aripiprazole—age 6–17

years).

Risperidone

Both McCracken et al.’s

32 and Shea et al.’s

33 studies of risperidone were 8 weeks in

duration, in a population of children and adolescents with either solely autistic

disorder,

32 or mostly autistic disorder

33

(70% of study population had autistic

disorder; rest were diagnosed with either PDD NOS, Asperger disorder, or

childhood disintegrative disorder), and with mostly comorbid ID (64% of study

population

33

; 81% of study population

32

). Both showed risperidone to have

statistically significant improvements, compared to placebo, in the Irritability

subscale of the Aberrant Behavior Checklist (ABC-I) (ABC-I is a 15-item subscale

that includes items such as “injures self,” “physical violence to self,” “aggressive to

other children and adults,” “irritable,” “temper outbursts,” “depressed mood,”

“mood changes,” and “yells” or “screams” inappropriately; individual items are

rated on a scale ranging from 0—not at all a problem, to 3—severe),

58 as well as the

ABC subscales for hyperactivity and stereotypy. The subscales for social withdrawal

and inappropriate speech only showed statistically significant improvements for

risperidone compared to placebo in Shea et al.’s study.

33 The definition of

“response” was different in each study. In McCracken et al.,

32 positive response was

defined as at least a 25% decrease in the ABC-I score and a rating of much improved

or very much improved on the Clinical Global Impression rating scale (CGI),

improvement subscale (CGI-I). With this definition, 69% were responders in the

risperidone group and 12% in the placebo group. In Shea et al.,

33 a responder was

defined as having a 50% or greater decrease from baseline in at least two of the five

ABC subscales with none of the other subscales presenting a 10% or larger increase.

With this definition, 69% of subjects in the risperidone group were responders,

compared to 40% in the placebo group.

In McCracken et al.,

32 study participants in the risperidone group, compared to

placebo, experienced significantly higher rates of increased appetite, fatigue,

drowsiness, dizziness, and drooling, with trends of increased rates of tremor,

tachycardia, and constipation (P = 0.06). In a 16-week open-label extension of

McCracken et al.’s

32 study, the most common adverse events were excessive

appetite, enuresis, tired during day, dry mouth, excess saliva, rhinitis, coughing, and

anxiety.

63 Risperidone-treated participants in Shea et al.’s study

33 experienced

significantly greater increases in weight, pulse rate, and systolic blood pressure and

had a significantly higher rate of reported somnolence (73% risperidone vs. 8%

placebo). Extrapyramidal symptoms (EPS) were endorsed by 28% of the

risperidone-treated participants in Shea et al.,

33 compared to 13% in the placebo

group. Mean weight gain in both 8-week studies for those taking risperidone was 2.7

kg. Mean final daily risperidone dose was 1.8

32 and 1.48 mg,

33 which is in contrast to

the mean final daily risperidone dose of 2.8 mg in a large RCT of the antipsychotic

treatment of early-onset schizophrenia and schizoaffective disorder in children and

adolescents (ID was an exclusion criteria).

64

Although prolactin was not mentioned in either 8-week study of risperidone, a

long-term follow-up of the McCracken et al.

32 study confirmed that risperidone

treatment was associated with a 2- to 4-fold increase in serum prolactin levels.

65

McDougle et al.’s

66 study of risperidone in adults with ASD showed similarly

beneficial findings for the reduction in aggressive behavior, compared to placebo,

and common adverse events also included sedation and weight gain. The mean final

daily dose of risperidone was 2.9 mg, which is in contrast to the mean daily dose of

risperidone of 3.9 mg in a large RCT of adults with schizophrenia.

67

Aripiprazole

Both Marcus et al.’s

58 and Owen et al.’s

57 studies of aripiprazole versus placebo in

children and adolescents with autistic disorder were 8 weeks in duration. Prevalence

of comorbid ID was not mentioned in either study. Both studies showed statistically

significant improvement in children treated with aripiprazole, compared to placebo,

for ABC-I, CGI-I, as well as ABC subscales of hyperactivity and stereotypy.

Although Owen et al.’s

57

flexible dosing schedule showed statistically significant

improvements in the ABC subscale for inappropriate speech with aripiprazole versus

placebo, in Marcus et al.’s

58

fixed-dose study, only the 15 mg/day dose of

aripiprazole showed statistically significant improvement in inappropriate speech.

Neither study showed statistically significant improvement for aripiprazole, versus

placebo for the ABC lethargy/social withdrawal subscale. Both studies defined

response as at least a 25% reduction from baseline to endpoint in the ABC-I subscale

and a CGI-I score of 1 (very much improved) or 2 (much improved) at endpoint.

With this definition, response rate was 52.2% for aripiprazole in Owen et al.’s

study

57

(14.3% for placebo) and although response rates for each dose of

aripiprazole (5 mg—55.8%*, 10 mg—49.2%, 15 mg—52.8%) were higher than

placebo (34.7%) in Marcus et al.’s study,

58 only the response rate at 5 mg/day

separated from placebo with statistical significance.

In Marcus et al.’s study,

58

the three most common adverse events leading to

discontinuation of aripiprazole were sedation, drooling, and tremor, none of which

were reported in the placebo group. The most common adverse events reported in the

aripiprazole group of Owen et al.’s study

57 were fatigue, somnolence, vomiting,

increased appetite, and sedation. In Owen et al.,

57 14.9% of aripiprazole-treated

subjects (vs. 8% for placebo) endorsed EPS symptoms, with the most common being

tremor (8.5%). In Marcus et al.,

58 EPS symptoms were reported at roughly twice the

rate in all aripiprazole doses (5 mg/day—23.1%, 10 mg/day—22.0%, and 15 mg/day

—22.2%) compared to placebo (11.8%), the most common of which were tremor

and extrapyramidal disorder. Aripiprazole was associated with statistically

significant gains in weight compared to placebo after 8 weeks, with the mean weight

gain on aripiprazole in Owen et al.

57 being 2.0 kg, and for Marcus et al.,

58 1.3 kg (5

mg/day), 1.3 kg (10 mg/day), and 1.5 kg (15 mg/day). Interestingly, both studies

showed aripiprazole to have statistically significant reductions in prolactin levels

compared to placebo. Mean final daily dosing for aripiprazole in Owen et al.

57 was

8.9 mg/day. In a 52-week open-label follow-up study of participants from both Owen

et al.

57 and Marcus et al.,

58 as well as newly enrolled subjects, mean final daily

dosing of aripiprazole was 10.6 mg/day.

68

In a 2-month RCT of aripiprazole versus

risperidone in children with ASD, the mean final daily dose of aripiprazole was 5.5

mg.

69 Manufacturer recommended daily dosing of aripiprazole in pediatric patients

with schizophrenia or bipolar disorder is 10 mg/day.

70

Overall, the literature shows that risperidone and aripiprazole, in doses likely

lower than that used in children with schizophrenia or bipolar disorder, can be used

effectively to reduce symptoms of irritability and aggressive behavior in children and

adolescents with ASD, though treatment should be monitored closely for relatively

common adverse events of sedation, weight gain, and EPS. Of note, although

risperidone has been shown to consistently increase prolactin levels, aripiprazole

was shown to consistently lower prolactin levels.

Although risperidone and aripiprazole are the two antipsychotic medications with

FDA approval for the treatment of irritability in children with ASD, other typical and

atypical antipsychotics have

p. 1857

p. 1858

also demonstrated benefits for treating symptoms of irritability in children with

ASD. Haloperidol, dosed in the range of 0.5 to 4.0 mg/day, has shown benefits in

four RCTs and two long-term follow-up studies, though with significant rates of

dyskinesia (34%).

71 One RCT supports the benefits of pimozide, one RCT supports

the use of olanzapine, and either open-label studies or case reports exist to support

the benefits of olanzapine, clozapine, quetiapine, ziprasidone, and paliperidone.

72 A

head-to-head comparison RCT of risperidone and haloperidol in children with

autistic disorder showed risperidone to be significantly more effective than

haloperidol at reducing aberrant behavior and symptoms of PDD.

73 An RCT though

comparing risperidone, haloperidol, and placebo for the treatment of aggression in

adults with ID showed all three treatment arms to decrease aggression, though no

significant difference between any of the three treatments.

74

Other Medications

VALPROATE

Two RCTs of similar size (n = 30

75 and n = 27

36

) looked at valproate versus placebo

in children and adolescents with ASD and irritability/aggression with somewhat

conflicting results. Hellings et al.

75

found no significant difference in improvement of

irritability or aggression (including ABC-I and CGI-I) between valproate and

placebo after 8 weeks, whereas Hollander et al.,

36 after 12 weeks, showed valproate

to be significantly superior to placebo at improving CGI-I and the rate of

improvement of ABC-I. While in Hollander et al.,

36 no significant differences existed

in adverse events between valproate and placebo, in Hellings et al.,

75

there was a

significantly increased rate of endorsed appetite increase in the valproate group, and

two individuals in the valproate group developed elevated ammonia levels, one with

clinically associated symptoms. Both studies had target blood valproate levels (at

least 50 mcg/mL,

36 70–100 mcg/mL

75

).

Additional medications with either limited RCT data, open-label studies, or case

reports to support its use for irritability in individuals with ASD include buspirone,

clomipramine, clonidine, levetiracetam, memantine, mirtazapine, pioglitazone,

topiramate (when added to risperidone), riluzole, sertraline, and trazodone,

72 as well

as fluvoxamine

76 and lithium.

77

In Reichow et al. recent meta-analysis of medications

for the treatment of ADHD symptoms in children with PDD, methylphenidate had

moderate, though not statistically significant, benefits in treating irritability.

40

CASE 88-2

QUESTION 1: C.Y., a 9-year-old girl with ASD has been well maintained on risperidone for the last 3 years.

At the most recent checkup, the prescriber notes that she has gained ∼25 lbs over the last 2 years (note that

average weight gain for a 9-year-old child is approximately 5 lbs/year). The majority of weight gained seems to

be in the abdominal area. Blood tests note an elevated hemoglobin A1C of 8.4%. The provider discusses diet

and exercise, but C.Y.’s mother notes no change over the last few years. She believes her weight gain started

with the addition of risperidone, which is now at 4 mg/day.

The prescriber asks you for a recommendation based on your knowledge of the metabolic profiles of atypical

antipsychotics. You make the following recommendation.

Although risperidone and aripiprazole are not necessarily interchangeable, both

are indicated for the treatment of irritability in children with ASD, and aripiprazole

may convey less of a risk for weight gain than risperidone. This potential benefit may

outweigh the potential risk of aripiprazole being less clinically efficacious than

risperidone in some patients. In this context, you may recommend for the prescriber

to decrease risperidone by 1 mg every other week, while starting aripiprazole 2

mg/day for 7 days followed by an increase to 5 mg daily. Additional 5 mg dose

increases can be made weekly up to a maximum of 15 mg/day. C.Y. should be

monitored for tolerability during both the cross-taper/titration period and beyond.

REPETITIVE BEHAVIORS

Numerous different types of medications have been studied for the treatment of

repetitive behaviors in individuals with ASD. Selective serotonin reuptake inhibitors

(SSRIs) and TCAs have been studied likely due to perceived similarities between

the repetitive behaviors and compulsions exhibited in children with obsessive–

compulsive disorder, as well as the clinical observation that repetitive behaviors

may increase in children with PDD due to underlying anxiety. However, the literature

is mixed to support the use of SSRIs or TCAs for this indication.

Selective Serotonin Reuptake Inhibitors

Two RCTs looking at SSRIs for children with ASD and repetitive behavior

demonstrate conflicting results. Hollander et al.’s

78 crossover trial (n = 39, mean age,

8.2 years) showed that low-dose liquid fluoxetine was superior to placebo for

treating repetitive behaviors, as measured by the Children’s Yale-Brown Obsessive

Compulsive Scale (CY-BOCS) compulsion subscale, with a medium to large effect

size (0.76), though did not separate from placebo for improvements in other aspects

of global autism severity. In King et al.’s

79

large study of citalopram (n = 73) versus

placebo (n = 76) to target repetitive behaviors in children with ASD, there were no

statistically significant differences between citalopram and placebo in improvements

in overall symptomatology (CGI-I) or repetitive behaviors on the modified CYBOCS-PDD rating scale. An RCT of fluvoxamine versus placebo in adults with ASD

(n = 15 each group), however, did show fluvoxamine to be significantly superior to

placebo for improving repetitive thoughts and behavior.

76

[See below in

anxiety/depression for more details on SSRI prescribing in children with ASD].

TCAs

Clomipramine was studied in Gordon et al.’s

55 RCT, in which it was shown to be

significantly superior to both placebo and desipramine for the treatment of repetitive

behavior in children with ASD as shown by the Modified CPRS OCD subscale, the

modified National Institute of Mental Health (NIMH) OCD Scale, and the modified

NIMH Global OCD and Anxiety scales. The mean final week daily dose of

clomipramine was 152 mg; the mean final week clomipramine blood level was 235

ng/mL and desmethyl clomipramine level was 422 ng/mL. Overall adverse events

were relatively minor, though in those taking clomipramine, one person had a seizure,

and two had cardiac adverse events, one with a prolonged corrected QT interval

(0.45 seconds) and the other with severe tachycardia (160–170 beats/minute), both of

which resolved with reduction in dose. In Remington et al.’s

80 crossover RCT (7

weeks each intervention) comparing clomipramine to haloperidol in children with

autistic disorder, the most striking finding was that only 37.5% of participants taking

clomipramine finished the study (compared to 69.7% haloperidol and 65.6% for

placebo), with adverse events contributing to discontinuation including behavioral

problems, fatigue or lethargy, tremors, tachycardia, insomnia, diaphoresis, nausea or

vomiting, and decreased appetite. The mean daily dose of clomipramine was 128.4

mg, range 100 to 150 mg.

p. 1858

p. 1859

Overall, for the reduction of repetitive behaviors, SSRIs, specifically fluoxetine,

and TCAs, specifically clomipramine, may be helpful in children with ASD, though

adverse events may be particularly limiting for clomipramine. For adults with ASD,

SSRIs, specifically fluvoxamine, may be helpful in reducing repetitive behaviors.

Antipsychotics

Antipsychotics have shown the most robust evidence in reducing repetitive behaviors

in children with ASD. Nearly all of the data to support antipsychotics for this

purpose come from the same studies that targeted irritability. Both aripiprazole

57,58

and risperidone

32,33 were superior to placebo for reducing the score on the ABC

Stereotypy subscale. In McDougle et al.’s

81 secondary analysis of the 8-week RUPP

Risperidone trial,

32