More sensitive prenatal tests detect autism-linked variants
A test commonly used to detect genetic abnormalities in babies with birth defects and children with developmental delay, mental retardation and other developmental issues is now starting to be used prenatally, raising questions among scientists, physicians and parents about how to deal with difficult-to-interpret findings.
Known as chromosomal microarray (CMA), the technology detects copy number variations (CNVs) — deletions or duplications of stretches of DNA that are too small to be seen using a microscope. CNVs have been increasingly linked to psychiatric and other disorders, including autism and schizophrenia, in the last few years.
New research shows the test detects more information than the traditional type of chromosomal analysis, known as karyotyping. Karyotyping is currently used prenatally to detect major genetic abnormalities in the fetus in pregnant women older than 35, those with a family history of genetic disorders, or when an ultrasound reveals potential problems with the fetus.
But CMA is much more sensitive, able to detect structural changes in fetal DNA smaller than 100,000 base pairs. Some of these changes are newly discovered and little-studied in medicine. In most cases, scientists don’t yet understand how CNVs, which can span dozens of genes, lead to different diseases.
“Interpreting that information prenatally is a very different discussion than when a parent brings in a child who already has an abnormality,” says Ronald Wapner, director of reproductive genetics at New York Presbyterian Hospital, Columbia University Medical Center, who recently completed a nationwide study of the technology. Prior to birth, “we don’t know if there are going to be phenotypic [or observable] consequences and that makes it more complex.”
Still, given its power, Wapner and others predict CMA, which is available through government-certified laboratories, will soon replace karyotyping.
“Among women who have high-risk pregnancies, obstetricians and patients are being much more aggressive about doing [CMA] testing,” says Wendy Chung, assistant professor of pediatrics at Columbia University in New York. Chung will follow babies from Wapner’s study who had an abnormal CMA result.
In Wapner’s study, DNA collected from samples of amniotic fluid or placental tissue from more than 4,000 pregnant women was analyzed via both karyotyping and CMA. CMA detected genetic abnormalities in 1 out of every 70 samples that had a normal karyotype. The results, presented at the Society for Maternal-Fetal Medicine annual conference in February, are under review for publication.
The technology can detect at least 150 known conditions with well-defined outcomes, such as Williams syndrome, which is linked to learning disabilities and heart problems, and DiGeorge syndrome, characterized by heart defects and other medical problems. Williams syndrome is caused by a deletion of roughly 25 genes on chromsome 7, while DiGeorge syndrome is caused by a deletion of roughly 30 genes in the middle of chromosome 22.
Both could previously only be diagnosed after birth using CMA or prenatally with expensive tests specific to each syndrome.
With hundreds of possible genetic disorders, it has not been practical to use traditional tests to screen for them all prenatally. So CMA, which assesses the whole genome, offers a clear advance.
But CMA can also detect genetic variations that can have much more variable results.
“I have hundreds of patients with DiGeorge syndrome, so I know how these kids grow up, and it’s easy to tell parents what to expect,” says Chung. “For new CNVs, we just don’t have nearly as many patients and not as much experience in it.”
One such variant is a duplication or deletion on chromosome 16 called 16p11.2. It is thought to explain about one percent of autism cases and is also linked to developmental delay, and speech and behavioral problems. But it is also sometimes found in people with no observable symptoms.
Given this variability, interpreting the risk of disease and explaining to expectant parents what detection of a 16p11.2 deletion is a challenge.
“There are guidelines for microarray interpretation, but they are limited to postnatal settings,” says Charles Lee, director of the Molecular Genetic Research Unit at the Harvard Cancer Center, who was not involved in the study. He says he hopes the experience of clinicians involved in Wapner’s study will help guide the development of new guidelines specifically for prenatal use.
“One concern some of us have is that obstetricians don’t understand this well enough,” says Chung. “If we don’t have a way of getting the correct information out to obstetricians and pregnant women, I have concerns the information will not be used in the most powerful way. We want to make sure women won’t be told their child will have a terrible condition when that’s not the case, or told something is nothing, when it’s not.”
Currently, physicians base their interpretations of genetic tests on a government-sponsored database that gathers data from clinical labs around the country, as well as scientific literature and data from ongoing projects. These sources have figures on how often a specific variant has been linked to different disorders, for example.
One of the variants that CMA can detect is 16p11.2.
Researchers and physicians are beginning to get a better sense of the spectrum of effects linked to the 16p11.2 variant as a result of projects like the Simons Variation in Individuals Project (VIP). This effort, supported by SFARI.org’s parent organization, has collected blood and tissue samples as well as clinical information from 200 carriers and their families.
According to preliminary data from the project, nearly 20 percent of those with the deletion meet the requirements for Autism Spectrum Disorders, and about 8 percent of those with the duplication. Approximately 80 percent have other disorders that sometimes co-occur with autism, such as anxiety or attention deficit hyperactivity disorder.
Though these statistics will aid obstetricians in informing parents of the potential outcomes of 16p variants, they may not tell the whole story.
“It’s unclear whether the people in the Simons VIP project reflect the full spectrum of people with this variation,” says Chung, principal investigator for the study. The cohort may be skewed toward people who have symptoms that are severe enough to have been picked up by parents or physicians. “We don’t know how many ‘average’ people walking the streets of New York have this same deletion,” she says.
As CMA testing enters clinical care, “physicians should discuss the potential outcomes before the test, so that parents can decide what their threshold for knowledge is,” says Wapner. “Those who have results linked to autism need to sit down with someone who can fully explain what that means.”
Chung also emphasizes that parents are likely to react very differently. “I think some families will be concerned enough to terminate the pregnancy, and others will say this will be useful to know once the baby is born,” she says. For example, parents who proceed with the pregnancy may recognize the signs of neurodevelopmental problems in their child much sooner as a result of the test.
How people will react to information about risk of disease is a major concern of personal genomics, even in adults. Will someone who carries genetic risk factors for type 2 diabetes, for example, exercise and diet to try to reduce his risk? Or will he give up because he believes the disease is inevitable?
This potential may be magnified in parents, whose behavior could be altered by their perceived implications of a genetic test result even before their child is born.
“Does it change in positive or negative ways how they rear the child?” Chung asks. "Some parents may be more vigilant, with the idea they can start therapy early, while for other parents this may cause anxiety and stress and overanalysis of normal differences in development in the child."