This story is from The Pulse, a weekly health and science podcast.
Aurea toddles around in her pink sparkly sneakers, climbing up the steps that, to her, are nearly waist high. Her tiny t-shirt is the epitome of how adorable she is. It says “you + me + snuggles.” Aurea’s father, Rafal Smigrodzki, watches over his little girl. He is clearly proud of her.
“She’s very lively. I think she’s a pretty, pretty happy baby,” Smigrodzki said, “a very often smiley baby.”
Of course, Smigrodzki thinks his baby is special — most parents do. But Aurea is indeed unique. She was born almost two years ago and happens to be the first child born as the result of a new type of genetic screening, which carefully selected her embryo.
Smigrodzki and his girlfriend used in vitro fertilization and an advanced selection process from a startup called Genomic Prediction. The New Jersey startup offers genetic tests and promises to help prospective parents select embryos with the best possible genes. The company says its test can screen embryos for a variety of diseases and health conditions, like heart disease, diabetes, or breast cancer.
Smigrodzki, a neurologist with a PhD in genetics, stumbled across the company in 2017.
“I was always interested and reading about all kinds of new developments,” he said. “And just happened to read an article in the MIT Technology Review about Genomic Prediction.”
The article, however, was critical of this kind of testing. It was titled “Eugenics 2.0: We’re at the Dawn of Choosing Embryos by Health, Height, and More.”
Antonio Regalado, the journalist who wrote the article, has been covering genetics since the late ‘90s. Since then, scientists completed the Human Genome Project, decoding the billions of letters representing the whole genetic blueprint for a human being.
Sequencing has come a long way since then and is a lot faster now. Scientists have much more data to work with and millions of samples in genetic biobanks.
All of this data has made it possible for scientists to look for patterns — genetic variations that link to diseases like diabetes or cancer, or even non-disease traits like eye color. Sometimes it takes just one single gene to code for a disease. But most of the time, it takes many – hundreds or even thousands. That’s called a polygenic trait. And a polygenic risk score estimates the effect of many genes on an individual’s risk for a disease.
“These big databases created that possibility of doing vast computerized comparisons of everybody’s genes to everything about them,” said Regalado. “I have to say that this technology of polygenic risk scores, and genetic prediction is really something important and new.”
Regalado has compared the technology to the 1997 dystopian science fiction film called “Gattaca.” In the film, society is divided into the genetic haves and have-nots, where those born with chosen genes are the ruling class and people born naturally are destined to be nothing more than low-skilled laborers.
Where Regalado saw a threat, Rafal Smigrodzki, Aurea’s dad, saw an opportunity — to give his child the best possible start.
“Both I and my girlfriend, have children from previous relationships. But still, one child is not enough. It doesn’t feel like I got enough enjoyment with children. And I think it’s good to have children because somebody needs to take over in the future,” Smigrodzki said. “So, between my girlfriend and I, we decided to have a child.”
Due to his girlfriend’s age, one of the couple’s only options was IVF.
Smigrodzki was interested in also using the tests offered by Genomic Prediction. They met with an IVF doctor in Charlotte, North Carolina. Since the testing was new, the company sent a scientist to explain the technology to the doctors on the IVF team.
“During the next visit, our physician told us that they would not let us use the technology, because that [would be] unethical.”
Smigrodzki disagreed with the doctor’s decision.
“A physician who tells you that they will not use a life and health-saving technology because it’s unethical is a physician who is making a big mistake. There’s absolutely no credible reasoning that would justify this kind of a decision,” Smigrodzki said.
This wouldn’t be the last time Genomic Prediction caused tension. Controversy and resistance seem to follow the company.
How Genomic Prediction is different
Stephen Hsu, a co-founder of Genomic Prediction, said he became interested in computational genomics about 10 years ago.
“It became possible to build predictors using machine learning and artificial intelligence, which would look directly at the DNA of an individual and could tell whether someone was at unusually high risk, for example, for a specific disease condition. When we started seeing how powerful these techniques were, we realized one of the most obvious applications of it would be in embryo selection,” Hsu said.
IVF clinics have long tested embryos for genetic diseases. But what differentiates Genomic Prediction from routine genetic testing is the company’s promise to give an overall health impression of an embryo — not just testing to see if it has a specific disease or not.
This is much more complicated than screening, say, for Huntington’s disease. A single gene causes that disease, and tests for Huntington’s disease have a 99.9% accuracy rate. But test results for complex conditions, meaning they involve many, even thousands of genes, like heart disease or diabetes, are much less precise. So instead of giving definitive results, Genomic Prediction provides a risk score – an estimate of how likely an embryo will be to develop a disease.
“One very common case, which we’ve seen many times already, is a couple comes in and they know they have a family history of ‘x.’ And just to take a concrete example, ‘x’ could be breast cancer. So it could be that in the mother’s family tree, there are women that had breast cancer,” said Hsu.
“And they may just say, ‘Look, I saw my aunt die from breast cancer, and I just don’t want it to happen to my daughter. So, I just want to make sure that of the embryos we look at, I want the one we implant to be normal risk, not high risk.'”
Testing for breast cancer risk may not sound all that controversial. But you could imagine where this could go.
Remember that Smigrodzki’s first IVF doctors refused to administer the test, calling it “unethical.”
To start, the same technology that can test for complex diseases can also test for things like eye color, skin tone, or height.
“So, the accuracy with which one can predict height, from DNA alone, for people who have good family environments, you know, growing up, the error is less than three centimeters now,” Hsu said.
That means the scientists at Genomic Prediction say they’re able to predict adult height within a bit more than an inch before you are even born. But they are not doing that yet.
“We only provide risk scores for very impactful common diseases. Because it’s just too controversial, people just get too excited,” Hsu said. “We can make predictions about IQ scores of individuals. But we do not provide any of that information in our IVF service, just because it’s too controversial.
But what about traits that fall somewhere in the middle of eye color and disease risk?
“We can predict BMI [body mass index],” Hsu said. “And there’s a whole complex of health problems that come with being significantly overweight. Now, are we going to report BMI prediction scores? Would people call us fat shaming if we did that? As you know, it’s extremely delicate.”
Hsu and his company are pioneers with this type of genetic testing. They are making choices on what diseases and traits to test for, how accurate the tests should be, and if they should even offer the test. And the work they are doing will influence the direction this industry takes.
“There are guidelines from the American Society for Reproductive Medicine. And those guidelines support the use of this kind of technology to prevent disease, they do not explicitly support the use of this kind of technology for cosmetic purposes, and they don’t really talk about IQ,” Hsu said.
“So, it’s kind of a gray area. We’re certainly not forbidden from doing it, but we just chose not to do it. Because from our perspective, there’s so much good that you can do just on the purely medical health risk front, that we don’t want that to get drowned out by this other stuff.”
Hsu never explicitly said where he draws the line. But he did say that people should have a better understanding of these tests.
“People should become aware of how the technology works, how does heredity work, how the genes work — people should learn that. And then, through a kind of democratic process, they should decide this is the kind of society we want to be. This is what we should allow. We like gene editing, we don’t like gene editing, we like embryo selection, we don’t,” he said.
He says this technology, the ability to screen embryos for diseases and even non-disease traits, is becoming more sophisticated, and more importantly, it is becoming more available. There are other companies that plan to offer similar testing.
Even though Genomic Prediction doesn’t offer scores based on intelligence, a recent article in Bloomberg follows a couple’s journey with embryo selection. The couple used Genomic Prediction and wanted more data than the company offers, so they downloaded the raw embryo data. They analyzed it to weigh the odds of mental health traits, like stress management and chronically low mood. They told Bloomberg that what was “most important to them was mental health and performance.”
A closer look at the testing
As this genetic modeling continues to improve and become more wide-spread, we must start asking the hard questions: How will we use it?
Laura Hercher, director of research in human genetics at Sarah Lawrence College in Bronxville, New York, says almost everything to do with embryos and manipulation of embryos stirs up some controversy.
Hercher is a trained clinical genetic counselor, licensed to work with patients. But lately, she spends her time educating on ethical, legal, and social issues in genomic medicine. She says these test results are much more nuanced than they may seem.
“I’ve talked to a lot of genetic counselors about embryo testing. And mostly we have a good laugh. I always say if I was going to write a textbook on genetics — which, God forbid — I would entitle it, ‘It Turned Out to be More Complicated Than They Thought.'”
One of the biggest challenges with these risk scores is that scientists simply don’t know all of the genes involved with complex conditions like heart disease or cancer.
Another problem is that the information scientists do have comes from genetic databases with overwhelming white, European participants. Applying that data to other populations, some scientists say, is misleading.
“It’s population data, which means that you have to be very careful about how you use it for an individual,” Hercher said.
“These genetic tests say something about your risk, the way your zip code, when you’re born says something about your risk. If we calculated it over a population, and said, what’s your chance of going to college based on the zip code you were born in? We could trace a relationship, right? That’s intuitive. But it would not be causal — it would not be the end point.”
So, the tests aren’t perfect. But they will continue to improve. Already, parents who use IVF can select an embryo based on its biological sex, which is not a concerning medical condition. They can also screen embryos for rare, devastating disorders like Huntington’s disease. But soon, testing could include conditions — that may or may not be considered a serious medical concern.
Hercher said, take ADHD or autism, for example.
“People might fear these conditions, they might particularly fear them based on sort of worst-case scenario thinking, that it can get very bad. But also, these are medical labels we have applied in some cases, to things that were well within the spectrum of, really normal behavior,” said Hercher. “And I think globally, big picture, when people look at the use of these technologies with embryos, one of their fears has to have something to do about reducing our vision of what is normal and acceptable in the human population. And specifically doing that, for people who can pay for it.”
Stephen Hsu from Genomic Prediction said that his company was starting to see inquiries from couples interested in doing IVF just so they could do the embryo selection process. Couples tell him:
“’Hey, we heard about this, and it just seems like such a smart thing to do. We’re actually considering doing IVF. Because we can get a better health outcome for our children, even though we’re not necessarily having fertility issues,’” Hsu said.
IFV is already very expensive, and the additional genetic test itself is about $2,000 per cycle. Hercher warns that we must be careful about how this difference in access will impact society.
“So, how do you draw the actual lines there? I think the only way to do it is to try and bring the population along with you,” Hercher said, “to try and explain the limitations of the genetics and explain the negative effects of too much selection and too much trying to control what a child is like.”
With all of that in place, Hercher said she doesn’t, “see an option other than to trust parents to make decent decisions.”
“Fundamentally, one reason why we don’t want to get into selection of things that are less like diseases and more like traits,” she added, “is that what we owe our children above everything else, is to accept them and love them as they are.”
Selecting the embryo
Rafal Smigrodzki says it was his sense of fatherly responsibility that influenced his decision.
“You see, it’s all about the children. It’s not about us, right?” Smigrodzki said. He said he expects that children “would be grateful” if parents choose certain traits for them, like intelligence.
“Intelligence is a good,” Smigrodzki said “It’s not a weapon that we use against each other in a fight. It’s actually something that we use to understand and control the world. And it’s precisely our intelligence that gives us the ability to progress — to build civilization. So, there’s absolutely nothing wrong, in my opinion, in selecting an embryo based on intelligence.”
For Smigrodzki and his girlfriend, they ultimately found an IVF provider in Washington D.C. that was willing to work with Genomic Prediction. They started with 16 embryos and eventually had five surviving embryos genetically tested.
“Out of the remaining five, one had a chromosomal abnormality, so it was immediately rejected. And out of the four remaining ones Genomic Prediction provided a list of odds for different conditions, including cancer, diabetes, hypertension, and cardiovascular disease,” he said.
It’s up to the parents to decide how to weigh those odds. Smigrodzki decided to select the embryo with the lowest risk of heart disease. Later, Genomic Prediction came up with an overall, “embryonic health score,” taking into account each possible condition, the odds, and the potential for a higher quality of life. Smigrodzki later had those embryonic health scores calculated for the four embryos he had to choose from.
“The embryo that became Aurea [had] by far the best health score of the four of them. So, it means that I could congratulate myself because I selected the best embryo.”