The multimodal treatment plan should address (1) establishing realistic goals for educational efforts, (2) identifying behavioral target symptoms for intervention, (3) prioritizing target symptoms and comorbid conditions for intervention, (4) using specific methods of outcome monitoring of functional domains (behavioral, adaptive skills, academic skills, social interaction skills, communication skills), and (5) monitoring for efficacy and potential adverse effects of medication (if used). The National Institutes of Health (NIH) suggests that evidence-based treatment strategies include both the use of educational/behavioral therapies and medications.55 An effective, well-designed, multimodal treatment plan that is consistently executed has the potential to positively shape the autistic individual's interaction with the environment and improve the quality of life of the patient and his or her family.
After a thorough diagnostic evaluation, treatment planning for the individual with autism is critical to assure consistency and efficacy of interventions. With the often severe nature of the behavioral and adaptive problems, it is not surprising that many potential treatment modalities have been proposed for persons with autism. Although there have been considerable advances in early intervention strategies, research regarding the comparative efficacy of the various interventions (education, psychosocial, psychopharmacology) is lacking. The two treatment approaches for autism with the most significant body of evidence-based support and clinical consensus are behavioral/psychoeducational therapies and psychoactive medication intervention.56 All stakeholders (family, educators, and clinical professionals) should be involved in the treatment planning process. Treatment decisions should be evidence based and focused on the specific identified needs of the individual. Overreliance on patient verbal reporting for assessment of psychopathology should be avoided because of the potential for communication deficits. A multifaceted approach to information gathering should include direct observation, as well as interviews with parents/family, caregivers, teachers, and review of the medical record, including any behavioral rating scale information.2
Available evidence suggests that appropriately designed, consistently implemented educational services positively impact the acquisition of social, communicative, self-care, and cognitive skills, each of which facilitates the person's long-term success. Services, such as occupational therapy, physical therapy, and speech pathology, are often required as integral aspects of an overall educational plan. Because of the pervasive need for sameness in routine, ongoing and consistent year-round educational programming is more effective than intermittent, episodic interventions. Training of teachers and parents in appropriate behavioral management techniques, as well as collaboration and consistency among these groups in management approaches is essential to program success. Effective language and communication training can lead to generalized improvements in social skills, repetitive behaviors, and thus positively impact other nonspecific maladaptive behavioral problems such as noncompliance, self-injury, and aggression.56
Behavior modification techniques based upon principles of applied behavior analysis have demonstrated decreased inappropriate behaviors and fostered social interaction and language skills acquisition. This includes structuring the environment, providing consistent responses to behaviors, rewarding positive behaviors, not rewarding negative behaviors (negative reinforcement), application of an adverse stimulus to deter an unwanted response (punishment), and reinforcing closer and closer approximations to the desired behavior (shaping).52,57,58
In addition to the core symptoms of autism, many persons with this disorder exhibit other significant maladaptive behaviors, such as self-injury, aggression toward others, and severe tantrums in response to routine demands. These behavioral issues can interfere with day-to-day activities and are challenging for families, caregivers, and educators to effectively manage.59 Current research on the neurobiologic basis of autism is centered on the serotonergic, peptidergic, dopaminergic, and noradrenergic systems. Despite limited evidence-based support, psychoactive medications have been widely used to minimize the frequency and intensity of these behaviors. It is important that clinicians identify and carefully monitor specific behavioral target symptom response to avoid the practice of overprescribing psychoactive medications. Risperidone and aripiprazole are currently Food and Drug Administration (FDA) approved to treat the behavioral symptoms associated with autism.60 However, off-label use of FDA-approved medications (i.e., use of an approved drug for an unapproved use) is an acceptable clinical practice when there is evidence-based support for the use of the medication and informed consent is obtained.
Most research on pharmacotherapy in autism has been directed at management of behavioral symptoms. Many studies of psychopharmacologic interventions in persons with autism have been limited by problems in experimental design, sample size, loose or poorly defined diagnostic criteria, and many clinical outcomes have been limited in duration or of dubious significance.61 Psychopharmacologic medications can be helpful, as part of a multifaceted treatment program, in minimizing behavioral and/or psychiatric problems that interfere with educational interventions or in ameliorating symptoms that limit social interaction.52,62 Dopamine-blocking agents and SSRIs have clinically significant effects on some aspects of the core symptoms of autism when examined in randomized controlled trials (RCT).56
Prevalence studies of psychoactive medication use in this population have consistently reported higher rates of medication use in persons living in more restricted supportive living arrangements (institutions or sheltered living environments), older age, more severe autism, and more severe mental retardation.62,63 A survey of psychoactive medication use in this population in 1993 by Aman and colleagues found that 34% of the sample was prescribed some psychoactive agent, and antipsychotic agents were the most prevalent. Other medications cited in this seminal analysis were psychostimulants, sedative/hypnotics, and antidepressants, in descending frequency of use.64 In a follow-up study of 3,228 families involved in the Autism Society of North Carolina in 2001, the largest study of psychoactive medication use in this population to date, changes in the patterns of psychoactive medication use were reported. The rate of antidepressant use (21.4%) reflected a 250% increase compared to the 1993 evaluation, reflecting a shift in practice for intervention with antidepressants to treat depressive symptoms as well as perseverative, stereotypic, and aggressive behaviors.62 The use of atypical antipsychotics in this study (16.8% of patients) comprised 85% of all antipsychotic agents used, in part because of the presumed lower risk of extrapyramidal symptoms and tardivedyskinesia with the atypical agents.61,62
Several studies support an association between dopamine dysregulation and increased aggression. Most animal models of self-injurious behavior suggest a similar mechanism of dopamine dysregulation.65 Such findings have led to the use of antipsychotic agents that act as dopaminergic antagonists to address aggression and self-injurious behavior. There is a significant body of evidence to support the idea that the typical or conventional antipsychotic agents are not well tolerated by many persons with autism, due primarily to medication-induced extrapyramidal movements, including akathisia.66–68 The atypical antipsychotic agents are less likely to elicit extrapyramidal side effects than the typicals due to more potency at serotonin-2A receptors versus dopamine receptors. However, the atypical antipsychotics have been implicated in weight gain and sedation in some persons with autism, and the potential serum prolactin elevation related to risperidone use is of concern. Elevated serum prolactin may lead to amenorrhea, galactorrhea, and osteoporosis in females and gynecomastia and sexual dysfunction in males. The minimum degree of prolactin elevation that is clinically relevant is uncertain; however, substantial and sustained increases in prolactin are most often associated with adverse effects.69 Of the atypical antipsychotic agents, risperidone currently has the most significant evidence-based support for use in treating behavioral problems associated with autism. Risperidone is FDA approved for treatment of aggression, self-injurious behavior, temper tantrums, and mood changes associated with autism in children and adolescents.56,59,68,70–76 Olanzapine has one published, double-blind, placebo controlled trial in children and adolescents with autism, as well as a number of open-label trials and case reports. There are generally positive results reported in global improvement scale assessment, however, the significant weight gain noted in these trials makes it a second-line agent to risperidone.70,71,77 Case reports and open-label trials generally support the use of quetiapine, ziprasidone, and clozapine as alternatives in attenuating behavioral symptoms such as self-injury, aggression, agitation, and repetitive behaviors. Concerns with the potential to cause agranulocytosis and the requirement for periodic complete blood count monitoring with clozapine and reports of consistent sedation with quetiapine should be considered as therapeutic decisions are made.65,78–85 Most published studies indicate atypical antipsychotic-associated improvements in nonspecific behavioral concerns such as aggression, self-injurious behavior, irritability, and anxiety.66,68 However, few studies with atypical antipsychotics show a sustained statistically significant impact on any core features of autism, such as deficits in language usage, impairments in social reciprocity, or behavioral rigidity.56
Serotonin modulates sensory perception, motor function, memory, and sleep. Many of these functions are dysregulated in persons with autism.86 One theory posits possible deficits in the serotonin transporter gene in persons with autism; however, data are conflicting in support of this.87 SSRIs have been used to treat stereotypies, ritualistic behavior, aggression to self and others, repetitive behaviors, and over-adherence to routines.86,88 The increase in the use of SSRIs to address symptoms of autism corresponds with the emerging efficacy data, as well as the perception of increased safety of SSRIs compared to conventional tricyclic antidepressants. Fluoxetine, sertraline, and fluvoxamine are the SSRIs with the most published evidence supporting their use; however, reports on the use of other SSRIs also exist.65,86 There have been reports of worsening of agitation with the use of SSRIs in the treatment of behavioral symptoms associated with autism, underscoring the need for caution and careful monitoring when using these agents in this population.65
Mood stabilizing antiepileptic drugs such as divalproex are commonly used to treat bipolar disorder, and there has been much interest in their use in persons with autism to modulate impulsivity and aggression.89 Well-designed controlled clinical trials of adequate dose and duration are needed to evaluate the efficacy of valproate as suggested in open trials and case reports. However persons with autism who have affective instability, impulsivity, and aggression, especially with a comorbid seizure disorder or EEG abnormality, may benefit from a valproate trial.89 Similarly, the use of lamotrigine and lithium have not been systematically evaluated in this population. Lithium may be effective in treating mood lability in some persons with autism, but it has a narrow therapeutic index and requires significant clinical monitoring. The mood-stabilizing antiepileptic drugs are considered to be a safer option.90
The use of psychostimulants should be limited to those with autism and comorbid attention-deficit disorder (ADD). Psychostimulants are dopamine agonists, whereas antipsychotic agents are dopamine antagonists. It is hypothesized that psychostimulants increase dopamine in the striatum, potentially enhancing prefrontal cortical function and improving inhibitory control in individuals with impulse control disorders.91 The psychostimulants methylphenidate and dextroamphetamine have been studied in persons with autism to address hyperactivity, impulsivity, and inattention. Dextroamphetamine has not been clearly demonstrated to provide clinical improvements, and the side effects of increased stereotypy and irritability in some patients were problematic.88 Methylphenidate has been demonstrated, in a small number of controlled trials, in doses of 0.3 to 0.6 mg/kg/day, to provide varying degrees of improvement in target symptoms.88,92–94 The side effects reported were similar to those reported with dextroamphetamine, as well as dysphoria, social withdrawal, and crying. Anorexia, increased aggression, and tic disorder have also been reported with methylphenidate.88
The α2-agonists, clonidine and guanfacine, have been widely used to treat hyperactivity and agitation in persons with autism. As with many psychoactive medications used in this population, there is a lack of supportive rigorous, methodologically sound studies.88 Sedation and hypotension were commonly observed adverse effects. β-Blockers (primarily propranolol) have also been used for symptoms of aggression and impulsivity. Similar to the α2-agonists, published literature is limited to case reports or methodologically unsound trials.65
Based on the hypothesis that a dysregulated endogenous opioid system can underlie the presentation of some behavioral symptoms of autism, naltrexone (an opioid antagonist) has been used to target extreme self-injurious behaviors in this population. In a meta-analysis of case reports and clinical trials of naltrexone use in autism (doses: 0.5–2 mg/kg/day), effectiveness was generally found, and the most predominant adverse effect was transient sedation.95 Limitations included small sample size, short duration of therapy, and inconsistent evaluative measures.
More well-controlled studies have shown some efficacy on self-injurious behaviors, minimal to no efficacy for other symptoms of autism (social deficits, stereotypies), and minimal adverse effects.88,96 Because of the apparent safety and specificity of naltrexone's therapeutic action, a trial may be merited in selected persons with autism who engage in clinically significant self-injurious behavior.95,96
Children with autism commonly have significant sleep disturbances which can adversely impact daytime functioning. Parents commonly rate sleep disturbance as a significant clinical issue. As with nonautistic individuals, it is important to determine the underlying etiology of the sleep problem. Behavioral intervention(s) (improved sleep hygiene, eliminating maladaptive sleep habits, parental education) should be undertaken prior to implementing pharmacotherapy. No medication has been FDA approved for pediatric insomnia. However, sedating properties of other medications in the existing regimen can be used to benefit sleep problems. This may entail administering an antiepileptic drug for seizures at bedtime or giving the larger dose of a sedating antipsychotic medication at bedtime.97,98
The current dearth of evidence-based psychopharmacologic and behavioral research in persons with autism is being addressed by a network of NIH funded research centers, such as the Research Units of Pediatric Psychopharmacology, Centers for Programs of Excellence in Autism, and Studies to Advance Autism Research and Treatment. The mission of these units is to foster well-controlled, multicenter, behavioral and psychopharmacologic intervention studies targeting behavioral symptoms in persons with autism.56
Among a number of scientifically unsupported treatments for autism is secretin, a polypeptide hormone promoted in the late 1990s as an efficacious therapy. Controlled trials found no reliable evidence of such efficacy.99 Elimination diets in which casein (from dairy products) or gluten (from wheat products) are excluded from the diet also have no scientific basis for efficacy. Additional scientifically unsupported therapies are facilitated communication, megadoses of vitamin B6 in combination with magnesium administration, dimethylglycine (an antioxidant) administration, nonspecific chelation therapy, and famotidine treatment.100,101
Sidebar: Clinical Controversy
While there is consensus that persons with autism require early detection and intervention with education to develop language skills as well as social and daily living skills, there is lack of consensus regarding how and when to use psychoactive medication for behavioral symptom management. In general, a careful clinical assessment should be performed establishing the severity and occurrence pattern of the behavioral problem. This assessment includes a thorough medication history, including doses utilized, response to treatment, and adverse effects experienced. In some cases the behavior in question is correctable with a change in environment, in demands, or identification and treatment of a medical problem. Psychoactive medications should be reserved for moderate or greater aggression/tantrums or self-injury in which other therapies, including behavioral ones, have been marginally effective or ineffective. Typically, combining behavioral and pharmacotherapy is most effective in severe behavioral problem management.106,141
Evaluation of Therapeutic Outcomes
Clinical investigators have used a variety of psychometric assessment instruments in attempts to measure changes in core symptoms secondary to treatment interventions.102 Monitoring the safety, efficacy, and tolerability of psychopharmacologic interventions in persons with autism are imperative to minimize adverse medication-related sequelae and optimize desired therapeutic outcomes.
A variety of instruments reported in the literature have been used in clinical trials to measure symptoms, such as communication impairment, restricted interests, and repetitive behavior. A review of many of these instruments is available elsewhere.103 Pharmacotherapy in autism is usually directed toward minimizing maladaptive behaviors, such as irritability, hyperactivity, compulsive, ritualistic, and perseverative behavior, and variants of self-injurious behavior. The Aberrant Behavior Checklist was designed for assessment of behavioral changes in institutionalized individuals in pharmacotherapy trials, however, a community-based version is also available.104,105 The Aberrant Behavior Checklist consists of 54 items divided into five domains: irritability, hyperactivity, stereotypic behavior, lethargy, and inappropriate speech. The lower the score in each domain, the greater the behavioral improvement. The Children's Yale-Brown Obsessive Compulsive Scale modified for Pervasive Developmental Disorders is a validated scale sensitive to changes in repetitive behavior severity pre- and post-treatment.106
Intensive medication-related side effects monitoring and assessment is important in this population, as self-reporting can be unreliable and can under-represent the presence of side effects. An instrument that is caregiver-rated such as the Monitoring of Side Effects Scale can be effective for this purpose. The Monitoring of Side Effects Scale is a multisystem, quantitative and qualitative caregiver assessment that rates the presence or absence and severity of a variety of potential medication-related adverse effects for clinician review.107 Signs and symptoms are written in layperson language and are listed by body area or system. As such, it is a broad-based screening tool that can be enhanced by side-effects specific scales such as those for akathisia (Barnes Akathisia Scale [BAS]), extrapyramidal effects (Simpson-Angus scale), or tardivedyskinesia (Dyskinesia Identification System: Condensed User Scale [DISCUS]).108–110
Use of atypical antipsychotic agents has been associated with increased risk of developing metabolic syndrome. Children receiving these agents should be monitored for hyperglycemia, dyslipidemia, and weight gain in a manner consistent with the consensus guidelines suggested by the American Diabetes Association and the American Psychiatric Association.111