In one of the largest studies of its kind, researchers have discovered 24 new "high impact" gene changes missing or duplicated stretches of DNA linked to autism, and confirmed 31 known markers.
University of Utah scientists who collaborated on the study and have a financial stake in Lineagen Inc., which paid for it say the findings show the company's genetic screen for all these markers is ready for clinical use.
But that's a matter of debate in scientific circles, and experts warn against families placing too much stock in the proliferating number of genetic tests for autism. The risk of false negatives is high and positive results may not guide treatment, they say.
Lineagen markets its test, FirstStepDx PLUS, primarily to pediatricians, not geneticists or families. It sells for $5,200 and is covered by some insurance policies, said company CEO Michael Paul. The 15-person company also makes genetic counselors available to help physicians and patients interpret results.
"There are long waiting list to get into geneticists. We believe that performing this test sooner, rather than waiting, can lead to earlier diagnosis," he said.
Mark Leppert, a distinguished U. professor of human genetics who collaborated on the Lineagen study, stopped short of advocating use of the test. He characterized the study as "a major step forward in understanding the genetic markers associated with autism."
He is one of two U. researchers who worked on the study and have stock options in Lineagen, which paid to publish the results Monday in the open source journal, PLOS ONE. He is also an unpaid scientific adviser to Lineagen.
Predicting risk • Autism spectrum disorder is a group of conditions that vary in severity, from social awkwardness to significant developmental delays. It is estimated there may be as many as 400 genes or chromosomal regions linked with autism not to mention environmental factors thought to play a role.
It's a complicated disorder, acknowledges Leppert. "It's going to take a long time for the scientists and doctors to figure out how all these genes lead to autism."
Scientists have identified fewer than 20 percent of the genes linked to autism, said Lynn Jorde, chair of the U.'s department of human genetics, who was not involved in the study.
"Any genetic testing is only going to identify a small portion of cases," he said. "Families should be aware that even if the child has been diagnosed as autistic, genetic testing probably won't find a cause."
The advance of micro array DNA technology has accelerated the search for autism's missing genetic factors.
Using this technology, U. researchers in the Lineagen study analyzed DNA from 38 Utahns in families that have more than one member with autism. They discovered 153 new genetic changes, or variants.
Then they set out to find which of the new variants, along with 185 spotlighted in past studies, are most closely linked to autism. They set the bar at those that double autism risk.
For this they turned to Hakon Hakonarson, director of the Center for Applied Genomics at The Children's Hospital of Philadelphia.
Hakonarson, who has no financial ties to Lineagen, designed a micro array test to screen 9,000 DNA samples for these variants, including 3,000 from a registry of autism patients run by the advocacy group Autism Speaks. The rest, samples from children with normal development, came from the Philadelphia Children's Hospital.
"We found 55 variants that were meaningful," said Hakonarson, which included the 31 previously identified.
Testing for these variants predicts risk. It isn't going to give someone "a yes or no" diagnosis, Hakonarson said. But finding one or two variants could amplify a person's risk by 15 to 30 percent, he said.
Deciding to test • Guidelines for genetic testing vary.
Deborah Bilder, a psychiatrist and assistant professor at the U. who specializes in autism, follows guidelines from the American Academy of Pediatrics and refers patients to geneticists for a "proper work-up" when they have a "clear family history" of autism.
Such information is useful because it can change the way doctors treat patients thought to have inheritable developmental disabilities on the autism spectrum, such as Fragile X syndrome, Bilder said.
"The [PLOS ONE] study is interesting. They found a few new variants. But as a prescribing physician, it's not going to change the way I treat patients."
The American College of Medical Genetics and Genomics, recommends chromosomal microarray testing for all children with autism spectrum disorders. The test offers a detailed analysis of chromosomal abnormalities as it examines a patient's genome to detect extra or missing genetic materials, known as copy number variations.
Gail E. Herman, president-elect of the group and a pediatrician and geneticist at Nationwide Children's Hospital in Columbus, Ohio, said such tests on autistic patients typically come back normal. Occasionally maybe in 15 percent of cases the tests point to a contributing factor.
Rarely do the tests pinpoint the cause.
"It's important to do it because sometimes you find something that you can really say, 'This is the cause of the autism in that child,'" she said.
That can be a relief to parents who feel guilty, she said. And sometimes the testing can help predict whether a family will have another child with autism. For example, the risk may be lower in a family in which the parents don't have the variation that the affected child has.
Herman, who declined to comment on Linegan's particular test, read the study and said it wouldn't immediately change the way children are tested.
"These tests all have to be developed and validated. There's so many new things happening in autism that every week a company would have to totally redesign their test" to keep up.
However, she said the study could help interpret results of tests provided by other companies or labs, taking into account the new variations.
"There's lots and lots of new [genetic discoveries] coming out. What we don't know is what does it mean in terms of, are you going to treat the child differently? If you've got this particular CNV, does it mean that you're going to respond to one type of treatment better than another? â¦
"Finding the variant is the first step. What we really need is a lot more clinical information and then following over time to see if having a particular variant means you're going to do differently than a child with a different variant."
U. of Utah's stake in Lineagen's autism test
Announcing its genetic research published in PLOS ONE this week, the University of Utah heralded how the research will be used to "validate" a commercial test sold by Lineagen, Inc.
The U. has a financial stake in the private, for-profit Lineagen, which was launched in 2002 based on technology discovered at the university.
Bryan Ritchie, executive director of the U.'s Technology Commercialization Office, declined to discuss the details of Lineagen's licensing of U. research.
But he said in such deals, the U. typically earns 3 to 10 percent of profit in royalties. The school also usually owns an equity stake in companies founded on its research.