The first fact we need to acknowledge when we talk about medications for autism, is that that there are no pharmacologic treatments that actually treat autism. There is no data to date to suggest that any medications actually teach skills or change the developmental course of the disorder. As such, medications are used to facilitate interventions shown to be efficacious in this population, and decrease the burden of care that is associated with some of the associated symptom domains of autism. In other words, it makes no sense to be prescribing medications for individuals with autism, especially for children, if no psychoeducational interventions are in place.
Having said that, there have been significant advances in the area of psychopharmacology of autism over the past two decades. Large, multisite randomized controlled trials have taken place by NIH-funded networks and we have a better understanding of the positive effects as well as the safety of commonly used medications in children with autism. A main limitation in our pharmacology research has been that until recently, we have had no molecular targets to develop medications for. As such, we looked at symptoms associated with autism (irritability, hyperactivity, repetitive behaviors) that have similarities with symptoms of other neurodevelopmental/neuropsychiatric disorders, hypothesized that similar symptoms across disorders may share neurobiology, and then “borrowed” such medications and tested them in children with autism. Although the approach has not always been productive, it has provided for a series of medications that are useful when considering pharmacotherapy for children with autism.
Irritability/Impulsive Aggression – Risperidone is the only medication that has an FDA indication for children with ASD, at the time of this article. There is now data from large randomized controlled trials to document efficacy of risperidone for irritability and impulsive aggression (RUPP 2002, Shea et al 2004). Discontinuation of risperidone in studies resulted in relapse rates between 25% and 62.5% (Troost et al 2005, RUPP 2005). Open label data, small randomized controlled data and larger emerging studies suggest that other atypical neuroleptics may be useful in this population. The main side effect of this class of medications has been weight gain. The reason we are concerned about this is that the mechanism of such weight gain seems to be insulin resistance, and as such, may predispose children to metabolic syndrome. In addition, concerns exist about the potential for developing movement disorders related to an extrapyramidal syndrome (EPS), and the potential of elevated prolactin levels seen with some of these medications to affect bone density.
Repetitive Behaviors – Previous data has suggested that SSRIs may be useful in the treatment of interfering repetitive behaviors. It is important to note, however, that not all repetitive/restrictive behaviors (RRB) that individuals with autism engage in are interfering. Further, some can be used to develop functional skills. Hypotheses were based on data from obsessive compulsive disorder suggesting that newer SSRIs, fluvoxamine and fluoxetine, may effectively treat compulsive behaviors. Hollander et al, (2005) found that fluoxetine may be useful in children with autism spectrum disorders (ASD) and McDougle et al. (1996) reported that fluvoxamine maybe be useful in adults with ASD. However, the largest clinical trial published in autism to date, found that citalopram had no effect on repetitive behaviors. Instead, the authors found that citalopram caused significantly more side effects than placebo, including agitation, insomnia and mood lability, as well as GI dysfunction. Hence, the effectiveness of SSRIs for repetitive behaviors is now questioned and studies with other SSRIs are needed before we could determine the usefulness of this class. On the other hand, studies of atypical neuroleptics have shown effectiveness for repetitive behaviors of autism (McDougle et al 2005).
Hyperactivity/Inattention – Currently, the DSM-IV does not allow individuals to be diagnosed with attention deficit hyperactivity disorder (ADHD) if they have a diagnosis of ASD. However, the data suggest that 40-59% of children with ASD meet criteria for ADHD (Gadow et al, 2006, Goldstein et al, 2004). Literature examining the effects of stimulants on individuals with ASD who express ADHD like symptoms have suggested that both stimulants (Rupp 2005, Posey 2007) and atomoxetine (Arnold et al 2006) are effective treatments for these symptoms. It is important to note, however, that the effect sizes are smaller and the prevalence of side effects is higher in the studies examining individuals with ASD than in the studies examining individuals with ADHD. Common side effects for both medications include irritability, insomnia, emotional lability, and anorexia. Further, like SSRIs, there is a black box warning about suicidal ideation for atomoxetine. The significance of this for individuals with ASD is not clear. There is also early, small randomized placebo-controlled data and open label data on Alpha 2 adrenergic agonists including clonidine and guanfacine. This data suggests that these compounds may play a role in the treatment of hyperactivity and inattention in individuals with ASD (Jaselskis et al 1992, Frankhauser et al 1992, Psey 2004, Scahill et al 2006).
Alternative Medications
Up to 70% of children with ASD are reporting complementary and alternative medicine (CAM) use (Hanson et al 2007). There are several types of CAM treatments that have been reported to be in use for children with autism. Most commonly, these include biological treatments that aim to reduce oxidative stress (e.g. B12, Vitamin C, hyperbaric oxygen treatments), change the GI flora (e.g. probiotics), reduce inflammation (e.g. omega 3 fatty acids), remove heave metals (e.g. chelation), and special diets with unclear biological mechanisms (e.g. GFCF). Other types of CAMs used in this population include Yoga, massage, brushing protocols and chiropractic therapies, as well as energy techniques such as Reiki. Unfortunately, there is very little data regarding the efficacy/safety of most of these interventions. Some early data is encouraging and justifies further study of some of these interventions, including omega-3 fatty acids, Vitamin C and possibly GFCF diet. Some, such as massage therapy and brushing protocols have a plausible biological mechanism, but no data exists to make a recommendation either way. Finally, a few likely have enough data to discourage their use given their safety profile (e.g. IV chelation) or lack of efficacy in randomized clinical trials (e.g. hyperbaric oxygen treatments). Given the widespread use of such interventions in the parent community, the research community has an obligation to produce well designed studies based on neurobiologically plausible hypotheses to establish both efficacy and safety of such approaches. At the same time, parents should consider that time lost experimenting with unproven and usually costly methods is invaluable as intervention with evidence-based treatments should be started as early as possible.
Where we go From Here
As previously mentioned, we have selected and tested medications based on the fact that they are effective in other disorders that have symptoms similar with common complaints in children with autism. There is now accumulating basic science work from genetics, neuropathology and animal models to suggest a few molecular targets that may be involved in the neurobiology of autism. Examples include the glutamate and GABA systems, neuropeptides such as oxytocin, and immune system dysfunction among others. These can be targeted with both novel traditional/western and alternative compounds. Furthermore, there are multiple complaints children with autism present with to a doctor’s office that are likely amenable to medication but have never been a target of proper intervention studies (e.g. anxiety), and randomized controlled trials in such areas are urgent. In addition, as children with autism grow up, we are realizing that we have tested very few compounds in adults with autism. At the time of this article, a PubMed search for clinical trials and autism revealed only 17 publications, including seven that were not related to medications. Given the developmental nature of this disorder, it is important that we start paying attention to research at all developmental stages, including adulthood. At the end of the day, we think that multidisciplinary approaches are beneficial for individuals with autism and therefore such models of care need to be tested. In other words, intervention research needs to mirror the way we practice, which usually means combining psychoeducational and pharmacological interventions.
Dr. Evdokia Anagnostou is a child neurologist who has subspecialized in neurodevelopmental disabilities. She completed medical school at McGill University in 1998 and her residency in child neurology in 2003. She subsequently completed a research fellowship in neurodevelopment disabilities at the Seaver Autism Center in New York. She is currently a Clinician Scientist at the Bloorview Research Institute and an assistant professor at the department of pediatrics at the University of Toronto. Her research program focuses on investigating developmental differences in the brains of children with autism using fMRI, MR spectroscopy, and DTI techniques and on psychopharmacology of autism. She is also the co-editor of the Manual for the Treatment of Autism published by APPI press in 2007.
