In 2013, Susan Klugman, an obstetrician and geneticist who is currently president of the American College of Medical Genetics and Genomics, got back test results that were “really weird.” Her patient was a pregnant woman who had opted for noninvasive prenatal testing, or NIPT, a screen for Down syndrome and a handful of other conditions that had been on the market for only two years. Now Klugman was looking at a lab report that suggested the 13-week-old fetus had a chromosomal condition that should have been lethal. And yet here it was on ultrasound, with a beating heart, developing organs and no sign of a problem. After several rounds of diagnostic testing, Klugman reassured the woman that her baby would most likely be fine. Nine months later, she read by chance in her local newspaper that the new mother had died of renal cancer. Mysterious test results and an unexpected cancer—surely, Klugman thought, there must be a connection.
For decades, prenatal testing for Down syndrome and other chromosomal conditions had fallen into two categories: invasive biopsies using a needle to collect fetal cells from the amniotic fluid or the placenta, or biomarker testing via a simple blood draw from the pregnant person. The invasive test carries a small but real chance of miscarriage. Biomarker testing is easier, cheaper and safer, but because it looks at proxy measures and not the fetal DNA itself, it casts a wide and leaky net. Some people carrying a fetus with a chromosomal condition are missed, while the majority of those flagged as high risk turn out to be false positives.
Thus, the introduction of a more accurate and noninvasive blood test in 2011 was embraced enthusiastically by expectant parents. Today, NIPT is used in more than 60 countries, and sales of the test have attained a market value of more than $4.5 billion. In the U.S. alone, it is used in more than a million pregnancies a year. Multiple labs offer NIPT. And while they have their differences, they have this in common: all of them work by examining something called cell-free fetal DNA.
On supporting science journalism
If you’re enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
In living cells, DNA is contained in discrete structures known as chromosomes. As cells die and are replaced, chromosomes are deconstructed, and tiny snippets of DNA are released into the bloodstream. If sequenced, these blood-borne snippets can be traced back to their chromosome of origin like jigsaw puzzle pieces using our human genome map as if it were the picture on the box. A random sampling of cell-free DNA will consist of DNA from all chromosomes in proportion to their size—the largest chromosome being represented by the most segments, and so on—and therefore we can predict the number of segments we expect to see from each chromosome of origin.
In a pregnant person, some of this cell-free DNA is fetal DNA from the placenta. In the epiphany that spawned NIPT, Hong Kong–based researcher Dennis Lo realized that even though fetal DNA is only a small part of the whole, the fetal genome can be accurately assessed if the numbers are precise enough because any underrepresentation or overrepresentation of a given chromosome is clearly coming from the fetus. After all, the pregnant person’s chromosomal status is known, so their contribution to the cell-free DNA in the sample is entirely predictable.
That’s a safe assumption almost all the time. But from the earliest days of NIPT, there were rare instances of abnormal results that could not be explained by variations in fetal DNA. Sheetal Parmar, senior vice president of medical affairs for women’s health at Natera, who has been at the NIPT lab since it began testing in 2013, recalls that it was clinicians who first drew their attention to a possible link between certain oddball results and malignancies in pregnant people. “It started with people coming back to us and saying, ‘Hey, this particular patient had this finding and has been diagnosed with cancer,’” Parmar says.
But anecdotes aren’t evidence, and this left the labs in an awkward spot. Telling a person they have cancer—an unknown cancer of unknown origin, requiring unspecified follow-up, which insurance will be unlikely to cover —is not something to be done lightly, much less when the person involved is pregnant. NIPT tests showing multiple missing or added chromosomes, which labs had begun to suspect might be a potential indication of cancer, were officially labelled “nonreportable” results. Unofficially, genetic counselors and obstetricians have told stories of off-the-record phone calls from friends at labs whispering suggestions that a certain patient should “get checked out.”
In 2019, convinced that this cancer signal was real, Diana Bianchi, director of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, began a study intended to provide clearer guidance to labs and clinicians about how and when to follow up on these atypical results. Called IDENTIFY (Incidental Detection of Maternal Neoplasia through Non-invasive Cell-Free DNA Analysis), the study offered a thorough evaluation to any patient currently pregnant or less than two years postpartum who didn’t already have a cancer diagnosis and whose results were “discordant”—that is, cases in which the fetus seemed fine, yet the pregnant person’s NIPT results showed a pattern of extra and missing chromosomes that should have been impossible in a viable pregnancy.
Many were skeptical. “Our colleagues at the National Cancer Institute thought we were on a fishing expedition,” Bianchi recalled. “They thought we were out of our minds.” But if IDENTIFY was a fishing expedition, it was one that caught a whale. An article in the New England Journal of Medicine published in December 2024 reported on the first 107 IDENTIFY participants. All of them were pregnant or recently postpartum; their mean age was 33. “I met and talked with some of them,” Bianchi told me. Their physical exams were generally unremarkable. “They were pregnant, glowing.” Of the 107, some 52—almost 50 percent—had cancer. By the time the paper was written, seven of them were dead.
Even before their first report, IDENTIFY provided labs with a talking point and a resource they could offer to primary care providers, who, as Bianchi notes, typically “have minimal education in contemporary genetics” and may not understand the significance of these unusual results. Illumina, another market leader in NIPT, began adding a note in 2019 about the “association with maternal neoplasia, both benign and malignant” for any patient with nonreportable results. Susan Hancock, a staff genetic counselor at Illumina, says that Illumina has reinforced this with phone calls to practitioners to give them information on IDENTIFY. “It was important to let them know about the association,” Hancock explains, “because, realistically, many providers just were not aware.”
Nobody goes to a results session for prenatal testing expecting to be told that they might have cancer. On a Friday afternoon late in 2022 when Erica Lucca got a call to say that her NIPT results might indicate a problem, she assumed it was about the baby. Lucca and her husband spent “the weekend panicking, freaking out.” When the geneticist she saw on Monday suggested that it was most likely a maternal issue, she was inclined to be dismissive. Lucca was 33 years old and in the best shape of her life, just finishing up an uneventful first trimester. “I could easily have been, like, ‘No, nothing’s wrong with me, I feel fine.’”
But Lucca, who lives in Washington, D.C., decided to take the advice she was given and enroll in the IDENTIFY study, conveniently located in nearby Bethesda, Md., a decision she now credits with saving her life. An eight-centimeter mass in her chest, discovered via a full-body magnetic resonance imaging scan, turned out to be a particularly aggressive form of non-Hodgkin’s lymphoma. Less than two weeks later, she started chemotherapy.
Pursuing treatment during pregnancy was not an easy decision. “When you’re pregnant, you’re supposed to, like, avoid deli meats and avoid drinking alcohol,” she says. “But somehow, you’re okay to take chemotherapy and sit for an infusion every three weeks for eight hours? It’s kind of a difficult concept to accept.” Still, multiple doctors told Lucca that her baby would be fine—but she might not be if she delayed treatment. Her final infusion was on April 28, 2023. Two weeks later her son was born. Almost two years later, both mother and son are thriving.
It was not easy taking care of a newborn while recovering from chemotherapy. “The first two months or so,” Lucca says, “I don’t know how I survived that.” But her conversation with Scientific American is peppered with expressions of gratitude. It’s hard to feel lucky about having cancer at 33 in the middle of your first pregnancy, yet Lucca is aware that, in so many ways, it could have been worse. Without the test, she might not have known until many months later, when the symptoms arose that are associated with advanced disease.
Early and nonspecific signs of cancer might have been written off as pregnancy-related complaints, something Bianchi and her team observed frequently in the IDENTIFY cohort. Lucca’s obstetrician might not have known to send her to genetics; her geneticist might not have been familiar with the IDENTIFY study. If it had not been close at hand, she might not have been so easily persuaded to go get checked out. (IDENTIFY covers travel costs for eligible participants, but patients may still struggle with time off and childcare.)
Cancer during pregnancy is rare, though it is on the rise in many countries, along with the average age of childbearing. In the U.S., an alarming trend shows an increasing incidence of colon cancer in younger adults. In the IDENTIFY study, colorectal cancer was the second most common diagnosis after lymphoma.) The majority of people with cancer in the IDENTIFY cohort were entirely asymptomatic and would not have come to medical attention for months or even years, meaning that the test itself has the potential to increase the number of cancer cases diagnosed during pregnancy through earlier detection.. Bianchi estimates that hundreds of pregnant women per year receiving suspicious NIPT results would benefit from appropriate follow-up care.
But in the real world, it is easy for these opportunities to intervene to be lost. In the spring of 2023, a woman in Chicago, much like Lucca, was referred to genetics by her obstetrician for routine discussion of atypical NIPT results. Her report included a note that genetic counselor Divya Ramachandra had never seen before: “risk of maternal malignancy.” Still, it wasn’t bolded or marked in red, and the report mentioned other risks as well. “It was in this long line of possibilities,” Ramachandra explains.
Ramachandra and her patient discussed the mention of maternal malignancy and, absent any guidelines, she suggested that they order genetic testing for the pregnant woman herself. Hesitantly, the expectant mother agreed to submit a request for testing to her insurance. That request was subsequently denied. To move ahead, they would have to appeal. “The patient told me … she’d give us a call about whether she wanted to complete testing or not,” Ramachandra explained. “She didn’t give us a call, and I didn’t chase her down.”
Three months later, Ramachandra heard from a medical geneticist who worked in the hospital’s lab. The woman, now in her third trimester, had just been diagnosed with lymphoma. “Do you remember this patient?” she remembers the doctor asking. “How did she go so far undiagnosed when she had these results early on?”
“She wasn’t trying to accuse me of anything,” Ramachandra said, “but of course you feel like, ‘Oh my god! This is my fault.’” Distressed, Ramachandra called the lab that had done the NIPT. “And that’s when I found out that they had called the obstetrician’s office multiple times to tell them about their really strong concerns about malignancy.” Genetic counselors at the lab had sent over a flyer about the IDENTIFY study to the ordering physician, but it hadn’t made it to Ramachandra.
Every practitioner has cases that haunt them, and for Ramachandra, this is the one. If she had known, would she have pushed harder? Could they have gotten insurance to cover the testing? Would they have found an oncologist willing to do a workup on a pregnant woman with no sign of cancer except a strange result on a prenatal screening test? “People are a little hesitant when it comes to pregnant patients,” Ramchandra explains.
The professional discomfort is real. “I think obstetricians are not used to having patients who have cancer and oncologists are not used to having pregnant patients,” Lucca says. Luckily for her, she was able to find a tertiary care center where all her providers worked as a team. “Every time I had an infusion,” Lucca recalls, “someone from maternity would come over and take the fetal heartbeat, just for reassurance that everything was okay.”
But unfamiliarity and fear can affect the care that pregnant patients receive. “One of our challenges in the beginning of the study,” Bianchi says, “was to find interventional radiologists who were comfortable doing a diagnostic biopsy, wherever it was, if the woman was pregnant.” Eventually, the IDENTIFY team was able to establish a referral network for biopsies despite what Bianchi calls “a historical culture of ‘don’t touch a pregnant woman.’”
A part of the message of IDENTIFY, Bianchi tells me, is that cancer during pregnancy can be treated. As Lucca learned, chemotherapy after the first trimester can be done with minimal risk to the fetus. In Belgium, Bianchi adds, radiologists have even developed a work-around for the use of contrast dye in MRIs, which is contraindicated for pregnant people.. “They have women drink an enormous amount of pineapple juice,” she explains. “There’s something in pineapple juice that acts as a contrast agent.”
But in some cases, says Reshma Jagsi, a bioethicist and radiation oncologist at the Emory University School of Medicine, the treatment that provides the best odds of survival may necessitate a therapeutic abortion—if, for example, a patient with a pelvic malignancy requires radiation treatment. In states with strict antiabortion laws, a mere discussion of abortion could put the provider at risk of civil or even criminal sanctions. Though most states have exceptions for the life of the pregnant person, Jagsi notes that many hospitals’ legal counsels have taken a “very conservative, very restrictive” approach to interpreting those exemptions. “In some cases, it’s just having a chilling effect because physicians don’t know what it means. And understandably they are afraid of the risks.”
Even in the absence of legal jeopardy, appropriate follow-up care for individuals with “nonreportable” results may require some unfamiliar stepping out of silos. Klugman notes that insurance companies are likely to require that oncologists, not obstetricians, order follow-up testing. “Are the oncologists ready for this? No way,” she adds. What IDENTIFY proves, Bianchi observes, is that “frontline obstetric providers need to take these results seriously.” As Klugman and others suggest, getting them potentially life-saving care may require that obstetricians, insurance executives and perhaps even politicians take these results seriously as well.
Many medical tests have unanticipated findings, but few are quite as startling as this: a screening test for the fetus, a technique now integrated into routine prenatal care, that has unexpectedly manifested the power to identify cancer in a pregnant person long before it would become apparent on any physical exam. “It was a happenstance for me, and I hope it is a happenstance for other women, too,” Lucca says. “It just feels like magic.”
