Inequity of genetic screening: DNA tests fail non-white families more often

DNA sequencing for metabolic diseases in newborns may eventually replace the traditional method of genetic screening used to check all new babies for a group of rare, inherited diseases that can cause severe neurologic damage and death if not detected quickly.

But currently, research is showing that those advanced methods of genetic tests aren't equally useful for everyone: They're less accurate for non-white families, raising concerns about how historical gaps in whose DNA gets studied produce inequities in medical care.

The problems with these genetic tests are detailed in a new Stanford Medicine study published recently in Genetics in Medicine. The authors describe how the inequity of the pool of information gathered affects families and call for better scientific approaches to filling the knowledge gap.

Traditional newborn screening, in use since the late 1960s, measures biochemicals in the baby's blood or urine -- most typically via a few drops of blood from a heel prick soon after birth. It offers clear-cut diagnoses without relying on genetic knowledge, and is not subject to the same biases as genetic tests.

Testing biochemicals in the blood is similar to tasting a cake after it's baked, whereas testing genes is more like reading the recipe to figure out if it will work.

Testing biochemicals in the blood is similar to tasting a cake after it's baked, whereas testing genes is more like reading the recipe to figure out if it will work.

That explains why advanced genetic tests have gained prominence the past decade. They can help families not only learn a baby's diagnosis after birth, but also get clues to the risk of disease during pregnancy, and help them understand the risk of having a subsequent child with the same condition later.

For now, the researchers conclude, that when genetic and biochemical tests give conflicting answers, families and doctors should rely on the biochemical results. Meanwhile, scientists must work to expand the knowledge base of genetic variation for minority populations so that their results aren't often declared to be a "variant of uncertain significance" when the biochemical test clearly shows a significant threat to the child.

"It's what happens when the database is full of white individuals and this is a gene variant in the Hispanic population," said the study's senior author, Christina Tise, MD, PhD assistant professor of pediatrics. "It feels very unfair."

Receiving ambiguous genetic test results can distress families at a time when they are overwhelmed with an extremely sick newborn, and can cause confusion among medical providers, Tise said.

"The disorders we're screening for are very severe," Tise said, adding that it is not unusual for newborn patients with metabolic disease to fall into a coma within days of birth. "If we can catch these patients before they become symptomatic, we can save lives and improve their outcomes."

Although biochemical screening is the universal standard, for a variety of reasons, genetic testing is increasingly being offered, either before or after a baby's birth. Some groups are starting to advocate for whole-genome sequencing of all newborns in neonatal intensive care units, but the study's results cast doubt on whether that approach will be equally helpful to all children and families because of those historical gaps in DNA collection, Tise said.

Her team decided to conduct the study to document the challenges they were seeing among their diverse patient population when genetic tests were used.

'Inborn errors' of metabolism

Since the 1966, when California began screening babies for a disease called phenylketonuria, U.S. newborns have been screened in early life for a growing list of genetic diseases that affect the machinery of metabolism. Most such diseases are inherited in an autosomal recessive pattern, meaning the baby inherits two non-functioning copies of the same gene, one from each (unaffected) parent.

Several metabolic diseases interfere with the body's ability to process amino acids, the building blocks of food proteins. During pregnancy, when the fetus' wastes filter into the mother's bloodstream, the mother's body compensates for the abnormal fetal metabolism. But soon after birth, as amino acids are not broken down, toxic molecules accumulate in the newborn's blood that may lead to seizures, coma or other severe neurological symptoms.

Newborn screening has been designed to catch these diseases early: During their first few days, infants have a few drops of blood collected from their heel onto a card, then analyzed for biochemicals that signal danger. Follow-up testing of biochemicals in blood and urine can confirm whether the infant has a metabolic disease. Babies with such diseases can be treated with medication, specialized diets, and other measures to keep their brains safe.

Rather than waiting until birth to test a baby, some medical organizations now suggest screening the parents during pregnancy for disease-linked genes. In the best-case scenario, such "expanded carrier screening" offers families an early warning that their baby may be born with a metabolic disease, allowing for immediate treatment at birth.

In the best-case scenario, such "expanded carrier screening" offers families an early warning that their baby may be born with a metabolic disease, allowing for immediate treatment at birth.

Genetic testing is also increasingly available for infants after birth. But it only works if the specific gene change has been previously connected to a disease in the genetic databases used by labs that conduct these tests. Because of historical inequities regarding access to genetic testing and research , the tests are more likely to be informative for white families than others, but few efforts have been made to measure the gap, which the new study sought to do.

Confusing test results

Among her newborn patients with metabolic disease at Lucile Packard Children's Hospital Stanford, Tise noticed that many infants of Hispanic and Middle Eastern origin were receiving ambiguous results from genetic tests, either for testing done prenatally on their parents or via gene tests performed on the newborns themselves.

"During pregnancy, the parents were not given the same heads-up that a white family, from a population whose gene variants are very well understood, would potentially have had," Tise said.

After birth, the situation could be even more discombobulating: Parents might have a sick baby in the NICU, whose biochemical tests clearly showed a metabolic disease, but the gene testing would come back with uncertain, confusing results. It could lead to unnecessary debate about the child's diagnosis.

In the most heartbreaking cases, families had previously lost newborns who had been born in places with fewer medical resources, likely each to the same genetic disease. Gene testing could have given the parents peace of mind -- or the ability to screen future pregnancies -- had it worked, Tise added.

When you have this 'uncertain significance' result hanging around in your head, it leaves you in a different place emotionally.
Christina Tise

"There is something powerful about saying, 'This gene defect is the answer, not only for what's happening with your new baby, but also likely the reason your other children passed away.' It gives families the potential to heal and process what happened," she said. "But when you have this 'uncertain significance' result hanging around in your head, it leaves you in a different place emotionally."

Quantifying the problem

To give a better sense of the scale of the problem, Tise's team compared results for different types of newborn testing -- biochemical vs. genetic -- on all the patients referred to their metabolic genetic service over an 18-month period in 2020 and 2021. A total of 136 patients were referred for suspicion of metabolic disease based on results from the first step of biochemical newborn screening, the test that uses heel-spot blood samples. This test is designed to err on the side of over-sensitivity, meaning most patients who "screen positive" turn out not to have metabolic disease.

A total of 19 infants from the original group of 136 were ultimately diagnosed with metabolic diseases based on follow-up biochemical testing. Of these, 18 also underwent genetic testing. From the 18 who received genetic tests, 10 infants had at least one "variant of uncertain significance" even though all patients actually had metabolic disease, exactly the type of confusing situation Tise has seen in the past. And of those 10 patients with confusing results, nine were of non-white ancestry.

Had expanded carrier screening -- a form of genetic screening that looks for clues in parents before or during pregnancy -- been performed on the parents in the study, 20 of 36 parents, 17 of whom were non-white, would have received inaccurate results, indicating no risk for the metabolic diseases their children ended up having.

What happens next?

In addition to bringing attention to the weaknesses of genetic testing -- and advocating for continued use of biochemical tests even though they aren't as "advanced" -- Tise is thinking about better ways to systematically update the genetic databases to reflect disease-causing gene variants in non-white populations. Right now, the process is very cumbersome, essentially happening one genetic variant at a time.

"Every institution with a geneticist who works with newborns has this problem, and together we could make a lot of difference," Tise said. But the time needed to report these disease-causing genetic variants in scientific papers isn't usually built into geneticists' jobs. In the context of the enormous human genome (with 3 billion sites for genetic changes to potentially happen), trying to raise awareness about one genetic change at a time seems daunting.

In addition, physicians need to overcome reluctance to try genetic screening for families from minority backgrounds, even though the results may be confusing, because their participation ultimately helps other parents in the future.

The reality is if you don't test people of nonwhite ancestry, it's this futile cycle: The lab's databases never diversify either.
Christina Tise

"In the past, I have heard people say 'I don't want to send genetic tests for this family because we're going to get back variants of uncertain significance and it's going to be confusing,'" Tise said. "But the reality is if you don't test people of nonwhite ancestry, it's this futile cycle: The lab's databases never diversify either."

To illustrate her point, Tise describes an unexpected success. A few years ago, she reported a gene variant responsible for an unusually severe case of a rare disease called Zellweger spectrum disorder, getting the variant changed in medical databases from "uncertain significance" to "disease-causing." Over the following 18 months, to the surprise of her team, several additional families whose infants had this rare genetic variant were identified at Packard Children's and nearby hospitals in California.

"You never know when you have a variant that is really going to matter for people in the future," Tise said. "Until the science is done, we just don't know how prevalent some of these variants are."

Image: Emily Moskal