Aripiprazole in Children With Autism: A Pilot Study

I. SPECIFIC AIM

To obtain pilot data regarding the safety, dosing, and efficacy of open-labeled aripiprazole
in children, aged 12 to 18 years, diagnosed with autism. A larger, more controlled trial will
be planned for this population if the drug is safe and tends to be efficacious.

II. BACKGROUND AND SIGNIFICANCE

Autism is a serious disorder that first presents in childhood, usually before three years of
age (American Psychiatric Association, 1994). Though the incidence of autism is low
(approximately 3-6 cases per 10,000 live births), the consequences of the disorder are
severe. As the term "pervasive" implies, the symptoms affect most areas of functioning in the
developing child, and the disabilities associated with the disorder tend to persist
throughout life. Symptoms of autism include abnormalities in social relatedness including
extreme withdrawal; communication including no or little speech; and a sameness or
repetitiveness in behaviors and interests. Children diagnosed with autism can have severe
problems including hyperactivity, lability of mood, irritability, withdrawal, and
stereotypical behavior. Because of these severe problems, treatment of symptomatology often
includes pharmacotherapy.

The most studied drug for treating children and adolescents with autism is the typical
neuroleptic, haloperidol (Anderson et al, 1989; Anderson et al, 1984; Campbell et al, 1978;
and Cohen et al, 1980). The main public health concern regarding the use of haloperidol is
the risk of inducing dyskinesias (Campbell et al, 1988, 1997), though the dyskinesias have
been reversible.

Evidence also suggests that atypical neuroleptics, such as olanzapine (Horrigan et al, 1997b;
Malone et al, 2001; Potenza et al, 1999) and risperidone (Demb, 1996; Findling et al, 1997;
Fisman and Steele, 1996; Hardan et al, 1996; Horrigan et al, 1997a; Malone et al, 1999b;
McDougle et al, 1997; Nicholson et al, 1998; Perry et al, 1997) are effective. But the weight
gain associated with these agents can be a considerable health risk (Pi-Sunyer, 1993;
Sjostrom, 1992). For example, olanzapine use has reportedly been associated with increased
blood glucose (Fertig et al, 1998; Gatta et al, 1999; Ober et al, 1999; Wirshing et al, 1998;
Yazici et al, 1998). Ziprasidone is associated with increases in QTc, including in children
(Malone, unpublished data).

Aripiprazole, a recently released atypical agent is likely to be as effective as other
neuroleptics in autism, but may have the considerable health advantage of not causing weight
gain. However, because there is a lack of safety and efficacy data with aripiprazole in
children with autism, an open pilot study of aripiprazole in this population is indicated.

III. DESIGN AND METHODS

A. Setting: The setting for this study will be the Specialty Clinic for Pervasive
Developmental Disorders, part of the Child and Adolescent Psychiatry Outpatient Clinic of
Drexel University College of Medicine located at Friends Hospital.

B. Subjects: Subjects will be 15 children and adolescents who meet the DSM-IV Criteria for
Pervasive Developmental Disorder.

C. Design: This is a 6-week pilot study employing open treatment with aripiprazole. Subjects
will be rated at baseline and thereafter assessed weekly.

D. Medication: Subjects will be treated openly with aripiprazole. Should untoward effects
occur, the dosage of the medication will be reduced or the medication will be discontinued,
as clinically appropriate. Subjects will not receive other concomitant psychotropic
medication during the study. Medication will be taken concomitant with food.

Aripiprazole Dosing Strategy: Aripiprazole tablets (10, 15, 20, and 30 mg) will be employed
for the study. The dosage range for aripiprazole will be 5 mg/day to 30 mg/day. Every effort
will be made to reach and maintain a therapeutic dosage by week four of the treatment phase.

1. For subjects who weigh less than 25 kg, starting dosage will be 5 mg/every other day.
After 3 days, the dosage can be increased to 5 mg. By week 2, subjects can be increased
up to 20 mg/day. Thereafter, dosage increases can be made in up to 10 mg increments
weekly.

2. For subjects who weigh greater than 25 kg, the starting dosage will be 5 mg/day. After
three days, the dosage can be increased to 10 mg/day. By week two, subjects can be
increased to 20 mg/day. Thereafter, dosage increases can be made in up to 10 mg
increments weekly as needed.

Note: Should 5 mg tablets not be available, the dosing schedule will employ ½ of a 10 mg
tablet and dosage increases will be made in 10 mg increments. We are starting at a 5 mg
dosage to avert excessive sedation at onset of drug administration.

F. Measures

1. Primary efficacy measure: The Clinical Global Impressions (CGI; Psychopharmacology
Bulletin, 1985). The efficacy of treatments can be judged on their ability to enhance
global functioning, an assessment that is particularly relative in a pilot study. The
CGI consists of three global scales measuring severity of illness, global improvement,
and drug effect, and has been useful in measuring drug treatment effect in this
population (see Campbell and Palij, 1985). The PI and another trained rater will
complete this scale. Although only members of the research team generally complete this
scale, it will additionally be completed by the subject's parent/caretaker. Having it
completed by the parent/caretaker will take only minutes and may result in useful data.
This measure will be completed weekly beginning at baseline.

2. Secondary efficacy measure: Children's Psychiatric Rating Scale (Psychopharmacology
Bulletin, 1985). The CPRS was developed by the Psychopharmacology Branch of the NIMH to
rate childhood psychopathology. Each of the items is rated from "1" (not present) to "7"
(extremely severe). The first 28 items of this scale require no verbal response on the
part of the subject, making them appropriate for rating children and adolescents with
Pervasive Developmental Disorder. Of these 28 items, 14 are particularly relevant in
autism (see Campbell and Palij, 1985): they assess the symptoms for which drug treatment
is indicated including hyperactivity, aggression, self-abusive behavior, temper
tantrums, lability of mood, irritability, social withdrawal, and stereotypies. A
composite sum of these 14 items, the CPRS-14, will be constructed and employed as a
secondary outcome measure, a procedure used by a number of investigators (Anderson et
al, 1989; Campbell et al, 1986, 1989, 1993; Findling et al, 1997; Sanchez et al, 1996).
We analyzed our data using the CPRS-14; the ICC was 0.8978. In addition, 4 factors
derived from these 14 items will be examined. They include: autism,
anger/uncooperativeness, hyperactivity, and speech deviance (Overall and Campbell,
1988). This measure will be completed at baseline and at the end of treatment.

3. Safety measures: (1) A physical exam will be completed at baseline. (2) Height, weight,
blood pressure, and pulse will be obtained and recorded at baseline and at each visit
during the study. (3) The following laboratory measures will be obtained at baseline and
repeated at the end of the treatment phase: complete blood count with differential,
liver functions, and electrocardiogram. Serum prolactin will be obtained at baseline and
at the end of the treatment phase. Any other clinically appropriate tests and
evaluations will also be completed whenever needed.

Untoward effects will be measured and recorded at each visit employing the following
measures: (1) Dosage Record and Treatment Emergent Symptom Scale (DOTES); (2) Treatment
Emergent Symptoms Scale (TESS); (3) Abnormal Involuntary Movement Scale (AIMS) (all from
Psychopharmacology Bulletin, 1985); and (4) the Neurologic Rating Scale (Simpson and Angus,
1970). The DOTES and TESS measure a wide range of possible untoward effects. The AIMS
measures dyskinesias. The Neurologic Rating Scale measures other forms of extrapyramidal
effects that can occur with neuroleptics such as dystonias, parkinsonian effects, and
akathisia.

G. Procedures: All patients appropriate to the study and their parent/caretakers will be
approached and informed consent and assent (in subjects under 14 years) will be obtained.
Subjects meeting the Inclusion Criteria, but not the Exclusion Criteria, will enter the
baseline period of the study.

Baseline (week 0): Subjects will be rated at baseline employing the CGI and the CPRS.

End of Treatment (week 6): At the end of the treatment period, subjects will again be rated
employing the CGI and the CPRS (selected items).

In addition, each subject will be rated on the CGI at each visit so that data from the last
visit is available should the subject terminate the study prematurely. If it is known that a
subject will terminate the study at a visit before week 6, the subject will be rated with the
CGI and the CPRS at that visit.

All safety measures will be completed at each visit. Laboratory measures and EKG will be
obtained in the morning.

H. Analysis: This is a pilot study whose purpose is to get initial safety and efficacy data
with aripiprazole in children with autism. We will perform an ANOVA, repeated measures, for
CGI severity scores, the CPRS-14 and the CPRS factors. Similar analyses will be performed for
safety measures including for weight, prolactin and other laboratory measures, and EKG
indices.

Inclusion Criteria:

- Diagnosis of autism (DSM-IV) made by a boarded child and adolescent psychiatrist.

- Males and females.

- Aged 12 to 18 years.

- Clinical judgment that medication treatment for autism is indicated.

Exclusion Criteria:

- Major medical problems including cardiac, liver, endocrine, or renal diseases.

- Uncontrolled seizures.

- Baseline QTC greater than 425 msec.

- Concomitant treatment with psychotropic medication.

- History of prior exposure to aripiprazole.