Some autism treatments high on promises, low on proof

National estimates show that one in every 110 children is diagnosed with autism, but limited research into treatment methods causes many physicians to struggle to identify the best care for patients.

Vanderbilt University researchers reviewed about 4,000 autism-related studies from January 2000 to May 2010 and found that although some treatments show promise, there is little evidence to support many medications and therapies now used. Three articles from their research for the Agency for Healthcare Research and Quality were published online April 4 in Pediatrics.

"We have several individual studies that demonstrate positive outcomes, but we need more duplication," said Melissa L. McPheeters, PhD, MPH, a co-author of the studies and co-director of the Vanderbilt Institute for Medicine and Public Health's Evidence-based Practice Center in Nashville, Tenn. "That does not mean that the interventions are ineffective. It means that we need more research."

Physicians typically like results to be proved in multiple studies before applying new treatments to patients, said Jeremy Veenstra-VanderWeele, MD, co-author of the studies and assistant professor of psychiatry, pediatrics and pharmacology at Vanderbilt.

There is no consensus on which interventions are most effective. Future research should aim to improve understanding about the causes of autism and ways to treat specific patients within a wide range of autism spectrum disorders, he said.

The three studies are part of a larger report on autism treatments AHRQ released on April 4.

"Autism spectrum disorders are frustrating and challenging for patients, their families and caregivers," said AHRQ Director Carolyn M. Clancy, MD. "This report will help parents and clinicians understand their options and design a course of treatment that is consistent with their goals and values."

But Scott Badesch, senior vice president of development and operations at the Autism Society of America, said the focus should be more on improving services for people with autism rather than research.

"We know if we could get a kid diagnosed by about age 3 and get them into treatment, that is the best chance of success," said Badesch, who has an adult son with autism. "The problem with that is that in a lot of areas, treatment is just not available."

Reviewing the literature

In their review of 34 studies involving at least 10 participants younger than 13, researchers found that early intensive behavioral and developmental approaches show particular potential. But research on these methods is difficult to do, Dr. Veenstra-VanderWeele said.

"Those studies are particularly time-consuming," he said. "You're talking about kids who are getting 30 hours of intervention a week."

Their review of medical treatments for children with autism spectrum disorders found that a few medications have been shown to benefit patients who have challenging or repetitive behaviors, but there is too little evidence to show clear benefit for most medications.

The antipsychotics risperidone and aripiprazole are the best-studied drugs to address challenging behaviors such as irritability, hyperactivity and agitation. But the drugs can cause adverse side effects, such as weight gain, sedation and extrapyramidal symptoms such as tremors, dyskinesia and rigidity. "There is very good evidence that side effects do occur and that they are substantial," McPheeters said. "That means we have to be very judicious in how we use them."

Positive results with the two drugs indicate a need to study other antipsychotic medications that don't have such dramatic side effects, she said.

A review of studies on the use of secretin found solid evidence that the drug isn't effective in improving patients' language, cognition, behavior, communication or socialization skills.

Dr. Veenstra-VanderWeele said all medications being used for autism have been adapted from other conditions. New research needs to find medications to target autism, he said.

Badesch's 23-year-old son, Evan, was diagnosed with Asperger syndrome in 1995. Getting the diagnosis took several years. "At that time there were not a lot of resources," he said. "We went to a number of doctors and therapists, and were told everything from there was nothing wrong to he was bipolar, he was schizophrenic -- everything under the sun."

Though there are more resources now, many families have trouble accessing them, Badesch said. Some insurance companies won't cover treatments, which can cost $70,000 to $80,000 per year.

With the lagging economy, Badesch said the priority should be providing more resources for families. "There needs to be a greater attention on those individuals than on research," he said.