Pharmacotherapeutics of Tourette Syndrome and Stereotypies in Autism Thilinie Rajapakse, MD and Tamara Pringsheim, MD, MSc

Tourette syndrome (TS) and stereotypy in autism spectrum disorders (ASDs) are 2 common movement disorders in childhood. The objective of this review was to summarize random- ized controlled trials published over the past 5 years as an update of the current pharma- cotherapeutic options for the treatment of tics, TS, and motor stereotypies in children with ASD. We searched MEDLINE (2005-May 2010) for randomized controlled trials of medi- cations used for the treatment of these disorders. For the treatment of tics in TS, 2 trials suggest that levetiracetam is not effective, whereas 1 trial found that topiramate was effective. Single clinical trials of metoclopramide, atomoxetine, and ondansetron were of limited quality, preventing conclusions to be made regarding the usefulness of these treatments for tic disorders. For the treatment of stereotypy in children with ASD, risperi- done has been shown in both a Cochrane review in 2006 and 2 subsequent randomized control trials to be effective. The addition of pentoxifylline to risperidone may have added benefit. Haloperidol did not improve stereotypy and was poorly tolerated. There is good evidence that aripiprazole is effective in the treatment of sterotypies in children with ASD. A large randomized trial of citalopram did not show any improvement in stereotypy. Single trials of levetiracetam, guanfacine, and atomoxetine suggest they are not useful in the reduction of stereotypy in children with ASD. Semin Pediatr Neurol 17:254-260 © 2010 Published by Elsevier Inc.

ourette syndrome (TS) is a childhood-onset movement many times a day (usually in bouts) nearly every day or Tdisorder defined by the presence of multiple motor and intermittently over more than 1 year, during which time phonic tics. TS is frequently associated with other neuropsy- there must not have been a tic-free period of more than 3 chiatric symptoms, including obsessive-compulsive behav- consecutive months; (3) the age at onset is younger than 18 iors, inattention, and impulsivity. Tics vary across a wide years; and (4) the disturbance is not caused by the direct spectrum of intensity from mild to severe in form. physiological effects of a substance or a general medical con- The impact of tics and related psychiatric comorbidity on dition. children and their family members can be substantial.1 Sig- In 2006, the Tourette Syndrome Association convened an nificant effects can be felt in the realms of education, occu- expert panel to summarize current pharmaceutical options pation, and peer relationships. As a result, the clinical man- and provide recommendations for the management of tics agement of TS is often multifaceted, requiring attention not and TS. Haloperidol, pimozide, and risperidone were the only to severity of the tics but also to associated features and treatments with the highest level of evidence (receiving a overall functioning. grade A), with several randomized controlled trials (RCTs) The diagnostic and statistical manual IV-text revision showing efficacy for treating tics.3 First-generation antipsy- (DSM-IV-TR) cites the diagnostic criteria for TS (307.23) as chotics have been associated with adverse effects, including 2 the following: (1) both multiple motor and one or more Parkinsonism, dystonia, dyskinesia, akathisia, and tardive vocal tics must be present at some time during the illness dyskinesia, fueling the search for other medications with although not necessarily concurrently; (2) the tics occur which to treat tics.4 Several nonantipsychotic medications were also recommended for the treatment of tics although From the Department of Clinical Neurosciences and Pediatrics, University of most have not been as extensively or as rigorously studied. Calgary, Calgary, Canada. Clonidine, guanfacine, pergolide, and botulinum toxin all Address reprint requests to Tamara Pringsheim, MD, MSc, Department of received a grade B designation.3 Clinical Neurosciences and Pediatrics, University of Calgary, Alberta Children’s Hospital, C4-431, 2888 Shaganappi Trail NW, Calgary, Al- Motor stereotypies are another common pediatric move- berta AB T3B 6A8, Canada. E-mail: [email protected] ment disorder seen both in normal children and those with

254 1071-9091/10/$-see front matter © 2010 Published by Elsevier Inc. doi:10.1016/j.spen.2010.10.008 Tourette syndrome 255 autism spectrum disorders (ASD). Motor stereotypies are re- Table 1 Criteria for Assigning Levels of Quality to Random- petitive and rhythmic movements that have a predictable ized Trials* pattern and location, seem purposeful but serve no obvious Good Studies are graded “good” only if all of the function, tend to be prolonged, and can be suppressed. Com- following are met: mon examples include movements such as arm flapping, Comparable groups assembled hand waving, head nodding, and body rocking. Follow-up at least 80% (80% contribute at least The underlying pathophysiology of stereotypies is un- one postbaseline efficacy measurement) known, and hypotheses range from psychological to neuro- Interventions are clearly stated biological mechanisms. A similar stereotyped behavior All important outcomes are considered Measurement instruments acceptable and applied known as punding can be seen in patients with Parkinson equally disease receiving excessive dopaminergic replacement ther- Outcome assessment is blinded apy. In punding, it has been suggested that combined activa- Appropriate attention to confounders in analysis tion of both dopamine and N-methyl-D-aspartic acid recep- Intention to treat is used tors is required.5 Antipsychotic medications have been Concealment: adequate measures to conceal efficacious in reducing the frequency of stereotypy in autistic allocation to study groups from those responsible children, suggesting similar mechanisms. for assessing patients for entry in the trial The purpose of this review was to summarize RCTs pub- Fair Studies are graded fair if any of the following lished over the past 5 years as an update of the current phar- problems occur, without the fatal flaws listed in macotherapeutic options for the treatment of tics, TS, and the “poor” category: motor stereotypies in children with ASD. We chose to focus Generally comparable groups or some minor problems with follow-up on tic disorders and motor stereotypies because these are the Some but not all important outcomes are most common movement disorders encountered in clinical considered practice in child neurology. Some but not all important confounders are accounted for Method of randomization not stated in methods Search Methods and Criteria Poor Studies are graded poor if any of the following fatal We searched MEDLINE (2005-May 2010) for clinical trials of flaws exist: medications used for the treatment of tics, TS, and motor Groups assembled are not comparable stereotypy in children with ASD. Randomized, double-blind, Unreliable or invalid measurements are used or are not applied equally controlled trials of any pharmacologic treatment for these Lack of blinding to outcome assessment disorders were included. Only trials that included children Key confounders are not addressed were examined. Trials for symptoms of autism that did not Intention-to-treat analysis is lacking assess the effect on motor stereotypies were excluded. Both Inadequate power of study parallel group and crossover study designs were included. *Based on the current methods of the US Preventive Services Task Force.6 Data Collection and Analysis Articles meeting the search criteria were discussed between motor and vocal tics. The YGTSS also has an impairment the 2 authors before dismissal or consideration for this review score of 50 points. The impairment and TTS are added to- based on fulfillment of inclusion criteria. Data were ab- gether to give a total possible score of 100. The Clinical stracted independently by 1 author onto standardized forms. Global Impression Scales (CGI) for severity (CGI-S) and im- Levels of quality were assigned to RCTs using criteria (Table provement (CGI-I) are used to assess both the severity of 1) based on guidelines from the US Preventive Services Task illness and clinical improvement by comparing the condi- Force.6 tions of the person standardized against other people with the same diagnosis or in the same individuals before and after Results therapy. Levetiracetam 2005 to 2010 Smith-Hicks et al9 studied the effectiveness of levetiracetam Pharmaceutical Treatment of Tics for the treatment of moderate to severe tics. They designed a Six RCTs for the treatment of tics were published during the fair-quality double-blind, randomized, placebo-controlled, search period of this review. One trial each evaluated leveti- crossover trial involving 22 patients. Subjects were random- racetam, clonidine versus levetiracetam, metoclopramide, ized to 28 days of levetiracetam or placebo given twice daily atomoxetine, topiramate, and ondansetron. The primary out- and crossed over to the alternate treatment after a tapering comes used in most trials were tic severity as measured using and washout period of 15 days. Standardized outcome scales the Yale Global Tic Severity Scale (YGTSS)7 or the Clinical used included the YGTSS and CGI-I. Global Impression Scales.8 The YGTSS has a total tic score Twenty of 22 subjects completed the study at an average (TTS) of 50 possible points, which is divided equally into daily dose of levetiracetam of 1563 mg/d. None of the pri- 256 T. Rajapakse and T. Pringsheim mary outcomes showed a significant treatment effect. No were rated every 2 weeks along with adverse events. Assess- evidence of a carryover effect was observed. Side effects re- ment for comorbid neuropsychiatric disease was made at ported with levetiracetam included irritability, hyperkine- baseline and every 2 weeks via the Yale Brown Obsessive sias, insomnia, tiredness, sadness, verbal aggression, anxiety, Compulsive Scale and the Conners Parent Rating Scale-Re- reduced school participation, and headache. Placebo com- vised. Twenty-four of 27 subjects completed the study. One plaints included headache, irritability, aggression, low frus- subject on metoclopramide withdrew because of increased tration tolerance, insomnia, tiredness, sadness, worry, hy- crying, and 2 subjects on placebo withdrew because of in- perkinesias, anxiety, and dry mouth. creased tic severity. This study suggests that levetiracetam is not more benefi- After 8 weeks of treatment, subjects receiving metoclopra- cial than placebo in suppressing tics. Limitations of the study mide showed a 39% reduction in their TTS on the YGTSS, include a short treatment protocol of 4 weeks that may have whereas placebo subjects only experienced a 13% reduction missed some responders as well as a very brief washout pe- (P ϭ .001). There was no significant difference on the Yale riod between treatment arms. Brown Obsessive Compulsive Scale or Conners Parent Rating Hedderick et al10 compared the efficacy of clonidine and Scale-Revised. levetiracetam in the treatment of tics. Twelve subjects with The most frequently reported adverse events in the meto- moderate to severe tics were enrolled in a 15-week random- clopramide group were increased appetite and sedation. No ized, double-blind, flexible-dose crossover protocol. After a subjects showed any evidence of extrapyramidal symptoms baseline visit, there was a 1-week placebo “run-in” treatment or dyskinesias. There were no significant group differences in period to exclude high placebo responders. Levetiracetam any cardiac conduction parameters or liver function tests. dosing was increased by 5 to 10 mg/kg/d weekly to a maxi- One subject taking metoclopramide had a 30-fold increase in mum of 50 mg/kg/d (or 2500 mg/d) as needed for tic sup- prolactin that resolved with medication discontinuation at pression. A 2-week washout period was implemented be- the end of the trial. tween each of the 6-week treatment phases. Ten subjects Overall, this trial was of poor quality. The process of ran- completed the study. domization is not stated in the methods, and an intention-to- The baseline to posttreatment change in TTS of the YGTSS treat analysis was not conducted. A key confounder between did not improve significantly in the levetiracetam group ver- the groups was the number of patients who had previously sus the clonidine group (22.7 vs 23.6, P ϭ .655). There was received medical therapy for their tics. It is uncertain whether no significant difference in the TTS related to drug order or the 2-week washout period was adequate and if there were carryover. Secondary outcomes measuring neuropsychiatric any lingering effects of prestudy medications that could com- comorbidities at baseline and posttreatment showed no im- promise study results. As with many other studies, small provement with levetiracetam. Side effects most commonly sample sizes and brief duration were also limiting factors. reported with levetiracetam were irritability and anxiety. Initial clonidine dosing was 0.05 mg twice daily and in- Atomoxetine creased as needed for tic suppression by 0.05 to 0.1 mg/wk to Spencer et al12 examined the changes in severity of tics and a maximum dose of 0.4 mg/d. With clonidine, the mean TTS attention-deficit hyperactivity disorder (ADHD) during ato- improved 3.4 points from baseline to posttreatment (P ϭ moxetine treatment in children with ADHD and TS.12 This .013) with an effect size of 0.57. The secondary outcomes article provided a post hoc subgroup analysis of data from an (mean YGTSS score and CGI scale) did not show significant 18-week parallel group, placebo-controlled, double-blind improvement with clonidine. The most commonly reported study. Subjects met DSM-IV criteria for ADHD and concur- side effect for clonidine was sedation. rent TS. Patients were randomly assigned to double-blind This crossover study was of poor quality because an inten- treatment with either placebo or atomoxetime (0.5-1.5 mg/ tion-to-treat analysis was lacking. Nevertheless, the study kg/d) for 18 weeks. Subjects were evaluated using the YGTSS, results suggest that treatment with clonidine but not leveti- CGI, and Tic Symptom Self-Report scales. A total of 148 were racetam resulted in a small reduction in TTS, with an effect assigned to treatment, and 145 provided data for the primary size of 0.57. As with any crossover study, it remains unclear efficacy analysis. The baseline characteristics between treat- whether the lasting effects of previously dosed medications, if ment groups were not significantly different. any, may have long-term consequences that can affect study On the primary measure of efficacy (ie, the YGTSS total results. In summary, 2 RCTs of fair and poor quality do not score), the atomoxetine treatment group showed a signifi- support the use of levetiracetam for the treatment of tics. cantly greater decrease from baseline in tic severity relative to placebo (P ϭ .027), and this was associated with a small to Metoclopramide moderate effect size (ES ϭ 0.40). This difference was reached Nicolson et al11 studied metoclopramide for the treatment of in the atomoxetine group by week 1 and continued to be- children with tic disorders. Twenty-seven patients were en- come highly significant (P Ͻ .001) at the endpoint of the rolled in an 8-week, double-blind, randomized, placebo- study. Scores on the Tic System Self-Report showed improve- controlled trial of metoclopramide after a 2-week washout ment in tic severity in both groups with atomoxetine showing period of previous medications. Metoclopramide was started a larger but not significant (P ϭ .134) improvement relative at 5 mg daily and titrated every 3 days to a maximum dose of to placebo despite a small-to-moderate effect size (ES ϭ 40 mg daily as required (mean dose, 32.9 Ϯ 5.1 mg). Tics 0.37). On the CGI scale, improvement in the atomoxetine Tourette syndrome 257 group at endpoint was significant compared with placebo Further evidence in the form of RCTs is required to increase and was associated with a moderate to large effect size (ES ϭ support for the efficacy of topiramate for the treatment of tics. 0.63). Adverse events reported in the atomoxetine group in- Ondansetron cluded headache (21.3%), vomiting (16.4%), decreased ap- Toren et al15 evaluated the efficacy of ondansetron in allevi- petite, and nausea (18%). Treatment group differences were ating tics in a placebo-controlled trial of 30 subjects (mean significant in the decreased appetite and nausea categories. age, 22 years; 15 subjects younger than 18 years). In this poor The parent study was initially designed as a noninferiority quality trial, patients who had previously failed treatment analysis with the hypothesis that atomoxetine causes no with haloperidol were randomized to ondansetron (up to 24 greater worsening of tic symptoms than placebo. The most mg/d) or placebo for a period of 3 weeks. At baseline, patients noteworthy limitation in this study is the post hoc subgroup randomized to ondansetron had significantly lower tic sever- analysis. There is a potential for type I errors (false-positives) ity scores than those randomized to placebo as measured because of multiple subgroup analyses, and the authors do 13 using the YGTSS and the Tourette Syndrome Global Scale. not discuss how this issue was addressed. In addition, Analysis controlling for baseline differences in severity found power calculations for the study were made based on the a significant improvement in tics in the ondansetron-treated overall subject sample and not for the subsample analyzed. subjects on the Tourette Syndrome Global Scale but not the Finally, bias may have been introduced with the measure- YGTSS. Overall, because of the poor quality of the study, it is ment of ADHD symptom improvement. A clear improve- difficult to draw any conclusions about the efficacy of ondan- ment in ADHD symptoms would likely have been apprecia- setron as a treatment for tics. ble to parents and investigators and may have led to potential unblinding of the randomization. Overall, however, this is a 2005 to 2010 Pharmaceutical randomized control trial of fair quality. Treatment of Stereotypy in Autism Topiramate There were 10 trials and 1 systematic review published dur- Jankovic et al14 investigated the effects of topiramate on tics. ing the search period of this review that assessed motor ste- They designed a multicenter, randomized, double-blind, pla- reotypies in children with autism. These publications in- cebo-controlled parallel group study. Twenty-nine subjects cluded evaluations of risperidone, aripiprazole, haloperidol, with TS ranging from ages 7 to 65 years (mean age, 16.5 pentoxifylline, omega-3 fatty acids, citalopram, atomoxetine, years) were enrolled. Topiramate was increased from a min- levetiracetam, and guanfacine. imum of 50 mg/d over a 6-week period up to 200 mg/d as Rating Scales needed followed by a 4-week maintenance phase, with a Various rating scales were used in the 13 studies to describe subsequent 12-day tapering phase. Efficacy analyses were the severity of motor stereotypies. The following scales were based on an intention-to-treat analysis that included all sub- used: jects who had taken at least 1 dose of study medication and had at least 1 postbaseline efficacy evaluation. The last obser- 1. The Nisonger-Child Behaviour Rating Form consists of vation carried forward method was used to analyze data in 60 items subdivided among 6 scales: conduct problem, the case of those subjects with early withdrawal. Twenty out insecure/anxious, hyperactive, self injury/stereotypic, of the 29 patients completed the double-blind phase of the self isolated/itualistic, and overly sensitive. study. The primary outcome was the TTS, which improved 2. The Ritvo-Freeman Real Life Rating Scale (RF-RLRS) is by 14.29 points from baseline to visit 5 (day 70) with topi- an in vivo observational measure for a variety of symp- ramate (mean dose, 118 mg) compared with a 5-point toms of autism. It is comprised of 5 subscales. For the change in the placebo group (P ϭ .0259). purposes of this review on stereotypy, we report statis- There were no differences in adverse events observed be- tically significant changes in “subscale I—sensory mo- tween the 2 treatment groups. Most commonly reported ad- tor (rocking, flapping).” verse effects included headache, abdominal pain and diar- 3. Aberrant behavior checklist (ABC) consists of 58 items subdivided among 5 scales: irritability, lethargy and rhea, and drowsiness. One child in the topiramate group social withdrawal, stereotypic behavior, hyperactivity, reported cognitive slowing, and one had a kidney stone. and inappropriate speech. There was also a mean weight loss of 2.1 kg from baseline to 4. Chouinard Extrapyramidal Symptoms Rating Scale endpoint in the topiramate group. (ESRS) asks parents about children’s dystonia, parkin- Overall, this was a fair quality trial with promising data on sonism, and dyskinesia. the use of topiramate in the treatment of TS. This may be an 5. Repetitive behavior scale-revised (RBS-R) is a parent- especially useful treatment option in children in whom rated scale comprising 43 items across 6 subscales as- weight gain may pose a significant health problem. Limita- sessing repetitive behaviors. tions include the inclusion of adults in this study, which challenges the applicability to children alone. Some of the Risperidone commonly observed side-effects of topiramate such as som- A systematic review of risperidone for ASD was published in nolence, cognitive slowing, and weight loss may not have the Cochrane Database of Systematic Reviews in 2006.16 been evident because of the short trial duration of 70 days. Three RCTs were included in this review. The ABC subscale 258 T. Rajapakse and T. Pringsheim on stereotypic behavior was examined in 2 of the studies, also had a more tolerable side effect profile and notably fewer including 178 subjects. Meta-analysis of these studies found reports of extrapyramidal symptoms. that subjects treated with risperidone had lower stereotypy Pentoxifylline subscale scores than those treated with placebo, with a mean Pentoxifylline is a methylated xanthine derivative that in- difference of Ϫ1.71 points (95% confidence interval, Ϫ2.97 Ϫ creases red blood cell deformability, reduces blood viscosity, to 0.45). and decreases the potential for platelet aggregation and 17 Pandina et al performed a subgroup analysis of 55 chil- thrombus formation. This medication is indicated for the dren enrolled in an 8-week, randomized, double-blind, pla- treatment of intermittent claudication on the basis of chronic cebo-controlled trial of risperidone for pervasive develop- occlusive arterial disease of the limbs. Akhondzadeh et al19 mental disorders (PDD). Subjects were randomized to investigated the effects of the immune-modifying agent pen- ϭ ϭ risperidone (n 27) or placebo (n 28). The ABC subscale toxifylline as an adjunctive treatment with risperidone in the for stereotypic behavior and the Nisonger-Child Behaviour management of autism in children. This study was a 10- Rating Form Self-Injury/Stereotypic Subscale showed no sig- week, parallel group, randomized, placebo-controlled trial. ϭ ϭ nificant change from baseline to endpoint (P .053 and P Forty patients with the DSM-IV-TR diagnosis of autistic dis- .183) although the ABC subscale result very closely ap- order were randomized to pentoxifylline ϩ risperidone or proached statistical significance. Adverse effects were re- placebo ϩ risperidone, and all patients completed the trial. ported in 100% of the risperidone group and 71% of the The dose of risperidone was increased up to 2 mg/d for chil- placebo group. The most common adverse effect in the ris- dren weighing 10 to 40 kg and 3 mg/d for children weighing peridone group was somnolence. Although 2 patients on above 40 kg. The dose of pentoxifylline was titrated up to 400 risperidone gained weight, this was not deemed statistically mg/d for children between 10 and 40 kg and up to 600 mg for significant. The mean ESRS scores with respect to extrapyra- children weighing over 40 kg. midal side effects were low throughout the trial. This trial was The data showed a significant treatment difference be- of fair quality and that data suggested that risperidone may be tween pentoxifylline ϩ risperidone versus placebo ϩ risperi- effective in reducing stereotypy in children with autistic dis- done in the ABC stereotypic behavior subscale (P Յ .0001) order; the result was very close (P ϭ .053) but not statistically by the effect of groups-by-time interaction. Commonly re- significant. Also of concern was the high percentage of can- ported adverse events in the pentoxifylline group included didates with reports of somnolence. Limitations of this study drowsiness, increased appetite, and weight gain, but these include the small sample size and brief study duration. Also, results were not significantly different from the placebo data were analyzed as a subgroup analysis of a larger study group. This trial was of fair quality. The results of this study initially aimed at testing the hypothesis that risperidone suggest that the combination of atypical antipsychotic med- would significantly reduce disruptive behavior in children ications, such as risperidone along with pentoxifylline, may with PDD. have synergistic effects in the treatment of stereotypy in chil- Miral et al18 conducted a study on the safety, efficacy, and dren with autism. Limitations of this study include the small tolerability of risperidone compared with haloperidol in the sample size, short duration of the study, and the lack of treatment of autistic disorder. The 12-week trial enrolled 30 immune response measurements before and after the inter- subjects. Twenty-eight of the 30 subjects completed the vention. In addition, the evaluations of the therapeutic effects study. The mean daily dosages of both haloperidol and ris- of pentoxifylline without an additional neuroleptic drug were peridone were 2.6 mg. Results indicated that the risperidone not permitted by the study country’s national ethics commit- group, but not the haloperidol group, had a statistically sig- tee. Further RCTs with pentoxifylline in children with autism nificant change from baseline to endpoint in the sensory mo- need to be performed, especially to study the safety profiles of tor subscale for stereotypy of the RF-RLRS. There was no this medication from an immune system perspective. significant change from baseline to endpoint in the ESRS Aripiprazole scores in the risperidone group; however, the haloperidol There are 2 RCTs of aripiprazole in children with autistic group showed a significant increase on this measure, indicat- disorder of similar design.20,21 The Owen trial20 included 98 ing an increased incidence of extrapyramidal symptom com- children and was rated “good” in quality, whereas the Marcus plaints in this group. There were no significant weight gain trial21 included 218 children and was rated “fair.” Both trials differences between the 2 groups at baseline or the 12-week compared aripiprazole with placebo for a period of 8 weeks endpoint of the study. Both groups had significant increases for the treatment of irritability as measured using the ABC. in weight and prolactin at endpoint compared with baseline Dosages of aripiprazole ranged from 2 to 15 mg/d. The effect measurements. However, there was a larger increase in pro- on aripiprazole on stereotypies was included as a secondary lactin in the risperidone group compared with the haloperi- outcome. Meta-analysis of the results of the 2 studies for the dol group (P ϭ .0128). ABC Stereotypic Behavior Subscale revealed a mean differ- Overall, the Miral et al study18 is a randomized, double- ence of Ϫ2.66 (95% confidence interval, Ϫ3.55 to Ϫ1.77, blind, placebo-controlled trial of fair quality. This study P Ͻ .00001) favoring aripiprazole over placebo. Side effects found that risperidone was superior to haloperidol on the that occurred more commonly in aripiprazole-treated pa- RF-RLRS sensory motor subscale in significantly reducing tients included weight gain, sedation, drooling, and tremor. stereotypic behaviors in the population studied. Risperidone The results of these 2 studies support the use of aripiprazole Tourette syndrome 259 in decreasing the severity of stereotypies in children with Table 2 Summary of Clinical Randomized Controlled Studies autism. Treatment of tics and TS Omega-3 Fatty Acids Levetiracetam is not effective in the treatment of tics in TS.9,10 Amminger et al22 conducted a 6-week pilot trial to look at the Topiramate was effective in the treatment of tics14* effects of 1.5 g/d of omega-3 fatty acids (0.84 g/d of eicosa- Single clinical trials of metoclopramide,11 atomoxetine,12 pentaenoic acid, 0.7 g per d of docosahexaenoic acid supple- and ondansetron15 were of limited quality† mentation) in 13 children with autistic disorder. Results of Treatment of stereotypy in children with ASD the ABC subscales showed no significant difference between Risperidone is effective in the treatment of stereotypy in treatment groups in the stereotypic behavior subscale of the children with ASD.16-18 The addition of pentoxifylline to ABC. This trial was of fair quality. Limitations include the risperidone may have added benefit.19 small number of subjects, the short intervention period, lack Haloperidol, studied in comparison to risperidone, did of laboratory testing, and side effect surveillance. not improve stereotypy in children with autism and was poorly tolerated.18 Citalopram Aripiprazole is effective in the treatment of sterotypies in King et al23 examined the efficacy and safety of citalopram for children with ASD.20,21 repetitive behavior in children with autistic disorder. This Citalopram in children with autism did not show any study was a randomized, double-blind, placebo-controlled improvement in stereotypy.23 study involving 149 subjects (citalopram, n ϭ 73; placebo, Levetiracetam,25 guanfacine,26 and atomoxetine24 each n ϭ 76) with autistic disorder. Most subjects completed the have 1 RCT, suggesting they are not useful in the 12-week trial, and results were analyzed on an intention-to- reduction of stereotypy in children with autism spectrum disorder. treat principle. All participants began with 2.5 mg/d, and the maximum dosage was 20 mg/d. Results showed no signifi- *The use of topiramate for the treatment of tics may be of special cant difference in stereotypic behavior between groups in the consideration for patients in whom weight gain associated with ϭ the use of other medications may pose a significant health risk. RBS-R stereotyped behavior subscale (P .75) or the ABC- †Firm conclusions could not be made regarding the usefulness of Community Version stereotypy score (P ϭ .57). The most these treatments for tic disorders. commonly reported side effects in the citalopram group included increased energy level, impulsiveness, decreased concentration, hyperactivity, stereotypy, insomnia, and diarrhea. The King trial is of good quality. Its results, however, pleted the trial. There was no significant difference in the do not support the use of citalopram for the treatment of repet- ABC stereotypic behavior subscale for the levetiracetam itive behavior in children and adolescents with ASD. In ad- group as reported by parents (P ϭ .967) or teachers (P ϭ dition, the significant side effects of impulsivity and hyper- .688). Adverse events most commonly reported included ag- activity with decreased attention and concentration suggest gression, agitation, hyperactivity, and impulsivity in the le- that this medication should only be tried with caution in vetiracetam group. This study was of fair quality, and the data children with comorbid ADHD. suggest that levetiracetam is not effective in treating stereo- Atomoxetine typy in children with autism. A notable limitation of the Arnold et al24 explored the efficacy and safety of atomoxetine study is its small sample size. for ADHD symptoms in a crossover study of 16 children with ASD. Each medication was trialed over 6 weeks with a Guanfacine 1-week washout period between treatments. The mean ato- Handen et al26 conducted a crossover trial of guanfacine in 11 moxetine dose was 44.2 mg/d. Study data revealed no signif- children with autism and/or intellectual disabilities. The icant change in the atomoxetine group on the ABC stereo- 6-week trial involved a maximum dose of 3 mg/d of guan- ϭ typic behavior subscale (P .08) or the RBS-R stereotypy facine titrated over a 19-day period. After remaining on the ϭ subscale (P .11). The most commonly reported adverse highest dose for 8 days, subjects underwent a 6-day washout effects in the atomoxetine group were upset stomach, nausea/ phase followed by an 8-day placebo phase. The alternate vomiting, tiredness/fatigue, and racing heart. This trial was of possible drug order began with the placebo phase followed fair quality. A limitation of this study includes the use of a by the guanfacine of similar durations. The study indicated crossover design with a short washout period and a medica- no significant change within the groups with respect to the tion with known considerable pharmacodynamic properties. stereotypic behavior subscale of the parent ABC (P ϭ .484) or Other drawbacks include the small sample size and allow- the teacher ABC (P ϭ .348). The most commonly reported ance of concomitant psychotropic medication. side effects in the guanfacine group included drowsiness and Levetiracetam lethargy. This was a trial of fair quality. Its results suggest that Wasserman et al25 conducted a study to determine the safety guanfacine is not useful in the treatment of stereotypy in and efficacy of levetiracetam in the treatment of 20 children children with autism and/or intellectual disabilities. Limita- with autism in a 10-week trial. Levetiracetam was started at tions of the study include the small study size, short duration 125 mg/d and titrated to 20 to 30 mg/kg/d based on treat- of treatment, and the diversity of the patient population par- ment response and tolerability. Seventeen patients com- ticipating in the trial. 260 T. Rajapakse and T. Pringsheim

Conclusions children with comorbid Tourette syndrome. J Atten Disord 11:470-481, 2008 The last five years have produced a promising amount of 13. Wang R, Lagakos SW, Ware JH, et al: Statistics in medicine—Reporting literature regarding the pharmacotherapy of tics, Tourette of subgroup analyses in clinical trials. N Engl J Med 357:2189-2194, 2007 syndrome and stereotypy in children with ASD (summarized 14. Jankovic J, Jimenez-Shahed J, Brown LW: A randomised, double-blind, in Table 2). While some medications hold promise in man- placebo-controlled study of topiramate in the treatment of Tourette aging these conditions, further trials confirming their results syndrome. J Neurol Neurosurg Psychiatry 81:70-73, 2010 would provide greater confidence in their effect estimates. In 15. Toren P, Weizman A, Ratner S, et al: Ondansetron treatment in light of recent data confirming the risk of metabolic syn- Tourette’s disorder: A 3 week, randomized, double-blind, placebo con- drome in children on atypical antipsychotic medications,27 it trolled study. J Clin Psychiatry 66:499-503, 2005 16. Jesner OS, Aref-Adib M, Coren E: Risperidone for autism spectrum becomes essential that our efforts to treat these movement disorder. Cochrane Database Syst Rev 1:CD005040, 2007 disorders in children do not result in secondary harm. 17. Pandina GJ, Bossie CA, Youssef E, et al: Risperidone improves behav- ioral symptoms in children with autism in a randomized, double-blind, References placebo-controlled trial. J Autism Dev Disord 37:367-373, 2007 18. Miral S, Gencer O, Baykara B, et al: Risperidone versus haloperidol in 1. Pringsheim T, Lang A, Kurlan R, et al: Understanding disability in children and adolescents with AD: A randomized, controlled, double- Tourette syndrome. Dev Med Child Neurol 51:468-472, 2009 2. American Psychiatic Association, Diagnostic and Statistical Manual of blind trial. Eur Child Adolesc Psychiatry 17:1-8, 2008 Mental Disorders, 4th edition, Text Revision. Washington, DC, Amer- 19. Akhondzadeh S, Fallah J, Mohammadi M, et al: Double-blind placebo- ican Psychiatric Association, 2000 controlled trial of pentoxifylline added to risperidone: Effects on aber- 3. Scahill L, Erenberg G, Berlin CM, et al: Contemporary assessment of rant behavior in children with autism. Prog Neuropsychopharmacol pharmacotherapy of Tourette syndrome. Neurorx 2:192-206, 2006 Biol Psychiatry 34:32-36, 2010 4. Sethi K, Morgan J: Drug-induced movement disorders, in Jankovic J, 20. Marcus RN, Owen R, Kame L, et al: A placebo-controlled, fixed-dose Tolosa E (eds): Parkinson’s Disease and Movement Disorders. Philadel- study of aripiprazole in children and adolescents with irritability asso- phia: Lippincott Williams and Wilkins, 2007, pp 394-408 ciated with autistic disorder. J Am Acad Child Adolesc Psychiatry 48: 5. Wolf DS, Singer HS: Pediatric movement disorders: An update. Curr 1110-1119, 2009 Opin Neurol 21:491-496, 2008 21. Owen R, Sikich L, Marcus RN, et al: Aripiprazole in the treatment of 6. Harris RP, Helfand M, Woolf SH, et al: Current methods of the U.S. irritability in children and adolescents with autistic disorder. Pediatrics Preventive Services Task Force a review of the process. Am J Prev Med 124:1533-1540, 2009 20:21-35, 2001 22. Amminger GP, Berger GE, Schäfer MR, et al: Omega-3 fatty acids sup- 7. Leckman JF, Riddle MA, Hardin MT, et al: Global Tic Severity Scale: plementation in children with autism: A double-blind randomized, Initial testing of a clinician-rated scale of tic severity. J Am Acad Child placebo-controlled pilot study. Biol Psychiatry 61:551-553, 2007 Adolesc Psychiatry 28:566-573, 1989 23. King BH, Hollander E, Sikich L, et al: Lack of efficacy of citalopram in 8. Guy W: Clinical global impressions, in ECDEU Assessment Manual for children with autism spectrum disorders and high levels of repetitive Psychopharmacology, Revised. Rockville, MD, Dept of Health, Education, behavior: Citalopram ineffective in children with autism. Arch Gen and Welfare, National Institute of Mental Health, 1976, pp 217-222 Psychiatry 66:583-590, 2009 9. Smith-Hicks CL, Bridges DD, Paynter NP, et al: A double blind ran- 24. Arnold LE, Aman MG, Cook AM, et al: Atomoxetine for hyperactivity in domized placebo control trial of levetiracetam in Tourette syndrome. autism spectrum disorders: Placebo-controlled crossover pilot trial. Mov Disord 22:1764-1767, 2007 J Am Acad Child Adolesc Psychiatry 45:1196-1205, 2006 10. Hedderick EF, Morris CM, Singer HS: Double-blind, crossover study of 25. Wasserman S, Iyengar R, Chaplin WF, et al: Levetiracetam versus pla- clonidine and levetiracetam in Tourette syndrome. Pediatr Neurol 40: cebo in childhood and adolescent autism: a double-blind placebo- 420-425, 2009 controlled study. Int Clin Psychopharmacol 21:363-367, 2006 11. Nicolson R, Craven-Thuss B, Smith J, et al: A randomized, double- 26. Handen BL, Sahl R, Hardan AY: Guanfacine in children with autism blind, placebo-controlled trial of metoclopramide for the treatment of and/or intellectual disabilities. J Dev Behav Pediatr 29:303-308, 2008 Tourettes disorder. J Am Acad Child Adolesc Psychiatry 44:640-646, 27. Correll CU, Manu P, Olshanskiy V, et al: Cardiometabolic risk of sec- 2005 ond-generation antipsychotic medications during first-time use in chil- 12. Spencer TJ, Sallee FR, Gilbert DL: Atmoxetine treatment of ADHD in dren and adolescents. JAMA 28:765-773, 2009