Autism at the cusp: Better metrics and objective measures needed

Ruler-metric-shutterstock_70841518There’s been an explosion of scientific research in autism—from mouse models of genetic syndromes involving autism to culturing neurons from stem cells derived from patients’ skin to tracking EEG patterns in infants whose brothers or sisters have autism.

So I expected yesterday’s panel on Piecing Together the Autism Puzzle, part of Boston Children’s National Pediatric Innovation Summit, to be about the science. I changed my seat just before it started, so I could better view the slides.

Instead, the conversation turned to the insurance, public health and social justice aspects of autism. Take, for example, the rising incidence of autism, which the CDC places at 1 in 88 (and 1 in 54 in boys). Panelist Ami Klin, PhD, director of the Marcus Autism Center at Children’s Healthcare of Atlanta, noted that between the CDC’s 2002 and 2008 reports on autism, there was close to a 101 percent increase in autism prevalence in Hispanics and a 96 percent increase in blacks.

Thousands of children didn’t suddenly develop autism in a six-year span; rather, more were diagnosed with autism as awareness of the disease increased. Even so, diagnoses often don’t occur until a child is 3 to 5 year old, and only 2.5 percent of diagnostic assessments of autism are using the field’s best standardized tools. While multiple diagnostic tests are being researched—like EEGs or blood tests looking at gene expression—they’re still experimental.

“The vast majority of children are not being identified early,” said Klin. “Pediatricians are supposed to conduct screenings at 18 to 24 months. But uptake is slow—pediatricians feel they have very little to offer the parents. ”

Health economics of autism

Right now, that’s true. Drug trials—focusing on single-gene disorders causing autism—are in early days. Currently, the best treatment available is intensive early intervention programs, such as ABA or Floortime, which can help children develop socially and prevent some of the intellectual disabilities, language problems and behavioral problems associated with autism.

“If we manage to get children to speak by age 3, they could be in regular schools,” said Klin.

But insurance usually won’t cover early interventions. “We’ve fought state by state,” said panelist Robert Ring, chief scientific officer of Autism Speaks. “Thirty-eight states have mandated coverage, but it reaches only a fraction of individuals.”

In short, we need to rethink autism’s health economics. An earlier speaker at the summit, Jeffrey Leiden, MD, PhD, CEO of Vertex, said the company is working to demonstrate that the higher price tags of new targeted drugs for cystic fibrosis (CF) are outweighed by the costs of outcomes averted, including lung transplant.

Targeted drugs are where the autism field is headed, but if it’s hard to document the economic benefit of a CF drug, it’s exponentially harder to make a business case in autism, where there aren’t yet objective indicators of benefit, and where costs are harder to measure: lost parental time from work, emergency-room charges for self-inflicted injuries, not to mention years of psychiatric medications to take the edge off the aggression and anxiety common in autism.

Recent research funded by Autism Speaks indicates autism costs society a staggering $126 billion per year. But at a breakout session later in the day, John Gaitanis, MD, assistant professor of neurology at Brown University, noted that the costs are spread across insurers, schools, state agencies and families themselves. “The challenge with the cost issue is the multiple different people and groups that pay,” he said. “If medical insurance had to pay the full lifetime cost, there’d be more interest in biomedical treatments.”

Add to that is the fact that autism is really a collection of different pathologies that we’re just beginning to define. Each is likely to respond differently to therapeutic interventions. If we did a clinical trial that lumps together disparate patients, or includes some who may not even have autism, it is unlikely to show a clear benefit. At the same time, we might miss drugs that could be effective in a subgroup of patients.

Big data?

And such metrics are where the autism field is lagging behind the science.

“The research ecosystem is generating enormous amounts of data. The size of the data is going to grow, but right now, I don’t see a systematic approach for how to take on [the analytics],” said Ring.

Consider the task: “At least 400 to 500 genes underlie autism, and 2,000 genes make up a synapse,” said panelist Mriganka Sur, PhD, of the Massachusetts Institute of Technology. “There are trillions of synapses and many ways to mess up the parts of the brain that create social interaction.”

Add in the interaction between genetic and environmental factors, a territory that’s only beginning to be charted, and it’s clear no one research center will be able to mount enough data to make autism a treatable, reimbursable disease like CF or stroke. That’s going to take a national collaborative effort.