A New Last Chance

By

Monica Halem calls it the “fertility train.” Every woman who embarks on a cycle of in vitro fertilization is familiar with the ride: the multiple cycles of hormonal stimulation, the pain of the injections, the discomfort and the bloating; then the delicate harvest of eggs to be fertilized outside the body, and the anxious wait for genetic testing on the embryos to make sure they have the right number of chromosomes before they are transferred back; and then, if all of the embryo tests come back abnormal, or the embryos don’t implant, or the pregnancy ends prematurely in miscarriage, the process starts all over again.

“It’s a lot of highs, right?” Halem says. “You’re getting excited, you’re ready. And then when it doesn’t work, which is more times than not, it’s a very low low. Such a depressing low. I mean, there’s-been-times-I-couldn’t-get-out-of-bed low.”

She is sitting at the desk of her office in the dermatology clinic she runs on Fifth Avenue, right across the street from the Metropolitan Museum of Art, with its endless rooms of Madonnas bouncing Renaissance bambini on their laps. It is a hot June afternoon, and Halem, 47, looks youthful and sleek in a black summer dress, her skin preternaturally smooth (what do you expect from a board-certified dermatologist?). But tears form in the doctor’s eyes as she relives her ride on the train.

“And to pick yourself up, you’re like, Okay, what’s the next step? And the next step is to try again. And you end up just going and going and going.” Halem first climbed on the train in 2009, and it was a rocky journey — over the next six years, she would visit five different fertility centers, go through dozens of IVF cycles, and endure seven miscarriages. “And it’s so … ,” she adds in a whisper, “so devastating when it doesn’t work.”

Then again, as the baby pictures and crayon drawings that decorate the walls of her office attest, it did work for Halem. In 2012, after an estimated 15 to 20 cycles of IVF and six miscarriages, this single mother became pregnant — at age 42 — and gave birth to a daughter in February 2013. A year later, she tried to get pregnant with a leftover embryo from an earlier cycle and suffered another miscarriage. Then, toward the end of 2014, at age 44, Halem decided she still wanted to try again.

But almost immediately she ran into a problem. She wasn’t able to produce a normal embryo. And so she faced a momentous choice: She could either quit trying to produce her own embryos, or, her doctor suggested, she could do the unthinkable and transfer into her womb an embryo that, according to genetic testing, was “aneuploid,” meaning it had an abnormal number of chromosomes. She was so intent on having another child, and so doubtful that an abnormal embryo would take, that she requested more than one aneuploid embryo, each with an irregular number of chromosomes, be implanted.

“I’m a very persistent person,” she explains almost needlessly. “Like, overpersistent.”

Watch: Why IVF Genetic Screening Is "Not Ready for Primetime"

Aneuploid has always been a dirty word in biology, the defining trait of cells that don’t obey rules and cause problems. It has historically been the hallmark of cancer cells, which are so genetically mutinous that they proliferate as tumors and attach to distant tissues as metastases. In the setting of reproduction, an aneuploid embryo — one with either too many or too few chromosomes — can sometimes result in a child with a genetic disability; the classic example is Down syndrome, which occurs in people who have three copies of chromosome 21. But more commonly, as one leading IVF expert put it not long ago, an abnormal embryo is simply “destined to fail” — either it won’t implant or, if by chance it does, the resulting pregnancy will almost certainly end in miscarriage. Mother Nature runs a ruthless but efficient quality-control program — up to two-thirds of fertilized human eggs (fertilized the old-fashioned way) do not result in a pregnancy. For women of “advanced maternal age,” as the medical literature describes them, the odds against pregnancy are even steeper.

Like many of her generation, Halem arrived at a fateful biological crossroads in her late 30s. After training in general surgery, she specialized in dermatology and came to New York when she was 38. “Figured I would meet somebody here, have a baby,” she says in her upbeat, rapid patter. When that didn’t happen, she decided to go it alone and consulted an IVF clinic when she was 40. She assumed she would quickly get pregnant. “And I did,” she recalls. “I got pregnant right away, but I had a miscarriage.”

That’s when she first learned that a woman’s egg cells begin to decline in quality after age 35. In all her years of medical training, she had never encountered some of the hard truths of female infertility, including the fact that for women older than 42, nearly 90 percent of the embryos they produce are aneuploid. Following the miscarriages, she also learned that immunological issues and other complications contributed to her infertility. All things considered, the one successful pregnancy was a miracle. Trying for a second child seemed like pushing her luck, especially after the seventh miscarriage in February 2014. While she was still rebounding from that setback, her acupuncturist suggested she go back and bank more of her embryos.

“How can I do that?” Halem protested. “I’m 44. What’s the chance of me having a normal embryo at 44?” But she kept rolling the idea over in her mind, and barely a month later, she began consecutive IVF cycles almost every month at New Hope Fertility Center, a Manhattan clinic with a philosophy of using low hormonal stimulation to produce a minimum number of eggs per cycle; they are immediately fertilized, allowed to grow five days, biopsied, and then “banked” — frozen — until the patient is ready to use them. (In other centers, higher doses of stimulation can produce up to two dozen eggs per cycle, but the process can be more physiologically and emotionally draining.) By the fall of 2014, she had banked 18 embryos, at which point all the biopsies were sent for analysis.

The test, a standard (and expensive) feature at many IVF clinics, is called “preimplantation genetic screening,” or PGS. Technicians typically use a laser to pry five or six cells from the outside of each day-five embryo and send them to a laboratory for chromosome analysis. Each cell of a normal human embryo has 23 pairs of chromosomes, 46 in all, but it is surprisingly common for one chromosome to be missing (known as “monosomy”) or for there to be an extra chromosome (known as “trisomy”). If even one of the cells turns out to be abnormal, the test registers as abnormal. The entire embryo is considered aneuploid, and it is not transferred. As Halem was about to discover.

“Every single one of them came back abnormal,” she recalls. “And I was devastated … I mean, all the money and time and hope.”

In the midst of that very low low, Halem consulted an expert in reproductive immunology named Jeffrey Braverman, who runs IVF clinics in Manhattan and Long Island. By the time she went to him, she was resigned to using donor eggs. Unbeknownst to Halem, though, Braverman belonged to a loose consortium of IVF doctors in New York who had doubts about the reliability of the PGS test. His concerns took root as far back as 2007, when three relatively young patients had unusually high rates of abnormal embryos, according to the test. After reviewing her results in late January 2015, Braverman suggested Halem transfer some of the abnormal embryos.

“If you have a good-looking embryo, forget the aneuploidy,” Braverman said. By “good-looking,” Braverman was referring to morphology, the traditional way of grading embryos before the advent of genetic testing. The embryo might not have the right number of chromosomes according to PGS, but viewed under a microscope, it still bore the signs of a potentially viable embryo — the cells along the outer wall were evenly spaced and symmetrical, and a little clutch of cells in the interior — the part destined to become the baby — appeared tightly packed and well organized.

Fourteen of Halem’s 18 embryos were deemed “complex abnormal” — what IVF doctors sometimes call “chaotic,” with more than one abnormal chromosome. But four of the abnormals were “good-looking” to Braverman’s eye. Three of the embryos, all male, were “monosomies” (missing one chromosome); the fourth embryo, a female, had an extra bit of chromosome 17, making it a partial trisomy. Otherwise, they looked healthy, so Braverman suggested transferring a couple of them. “Why don’t you try this?” he urged Halem.

Halem thought the idea seemed crazy. “Nobody puts abnormals back in.”

But Braverman said there was a precedent. In fact, he’d done it before. About a year earlier, as part of a complicated IVF situation with a same-sex couple, he had transferred a donor embryo missing one chromosome into one of the partners; she had gone on to have a normal pregnancy and bore a healthy child several months earlier, in the fall of 2014. So he knew it was possible, though he warned Halem that transferring an abnormal embryo wasn’t without risks: not only the obvious possibility of miscarriage and even a genetically abnormal child, but subtler issues of “emotional trauma,” of getting pregnant but not knowing if the pregnancy was normal until a week-ten prenatal test. If it was still abnormal at that point, the fetus was very likely to either not survive or be born with severe impairments. As with his previous patients, he also insisted that she go forward only if she was prepared to consider terminating the pregnancy if the ten-week fetus, upon testing, still revealed chromosomal abnormalities. In both cases, Braverman says, “I didn’t want to force them into a situation where they had no options.” His warnings barely registered with Halem. “I didn’t really absorb the risks emotionally,” she admits.

At most IVF clinics, all 18 embryos would have already been discarded as abnormal; New Hope, according to its director, Dr. John Zhang, likes to hang on to even abnormal embryos for a while. In June 2015, on Braverman’s recommendation, Halem went back to New Hope to have the procedure — but she requested that all four embryos be transferred. The recent trend in IVF medicine is single-embryo transfers to avoid the possibility of multiple pregnancies, but given Halem’s age and the presumed odds against an abnormal embryo’s successfully implanting, Zhang felt it was reasonable to transfer all four embryos.

Halem was remarkably fatalistic — at first — about the idea of carrying an aneuploid embryo, in part, because she really didn’t expect to become pregnant. “You know, I just went with it,” Halem says. “I told myself, ‘Let me get over this last hurdle and then move on.’ ” Two weeks later, she went to her doctor’s office for a pregnancy test. It was positive.

Oh my God, Halem thought to herself. It was an exclamation of joy, immediately followed by dread. If it was abnormal, she thought, what was I going to do? “It was … it was a lot of emotion.”

Photo-Illustration: Justin Metz

If you think deliberately transferring an embryo with an abnormal number of chromosomes seems ethically suspect (if not an invitation to a malpractice suit), you’re not alone — mostly out of public view, a vehement scientific argument has been raging in the field of reproductive medicine over whether, when, and how these embryos should be transferred. For a lot of older women like Halem, the inability to produce normal embryos according to genetic testing can mean the end of the fertility train — they aged out, couldn’t afford another $15,000 IVF cycle, simply gave up. For these women, transferring seemingly abnormal embryos can give them a chance at motherhood that appeared otherwise impossible. “Seemingly” is the key word here, because behind this daring medical decision was the growing conviction among a handful of IVF doctors, led by Norbert Gleicher, founder of the Center for Human Reproduction in New York, that PGS testing misdiagnosed too many embryos as abnormal and that many of those “abnormal” embryos might be more viable than anyone thought.

There are no hard numbers of how many women may have lost their chance to get pregnant as a result. But even educated guesses provide little comfort. Braverman estimates that “the lowest possible number of women that lost the opportunity to have a child due to the inaccuracy of PGS” would be 20,000 over the past two decades. Using a different calculation, Gleicher estimates a much higher number: about 13,000 per year. Behind these numbers, however soft, lies a hard new reality in reproductive medicine: Embryos that possess chromosomal abnormalities can nonetheless produce healthy babies.

Indeed, researchers have begun to report preliminary but stunning evidence that mammalian embryos with abnormal chromosomes have the remarkable ability in some cases to “self-correct” during early development, either by editing out cells that possess chromosomal irregularities or isolating them in the placenta, where they have no apparent effect on a developing fetus. As a result, these doctors believe that the widespread, long-standing conviction in medicine that abnormal embryos are destined to fail may be scientifically incorrect.

And a growing number of IVF clinics have the healthy babies — dozens of them — to prove it. Gleicher’s loose consortium claims seven healthy babies so far; a group based at the European Hospital in Rome reported 24 normal births at a meeting in Europe last July; and the Bernabeu Clinic in Alicante, Spain, reported 13 healthy infants earlier this year — all from embryos initially deemed abnormal by genetic testing.

The practice of transferring abnormal embryos, first initiated in 2012 at several IVF centers here and in Europe, is so new that only in May did a committee of the American Society for Reproductive Medicine declare these transfers to be “ethically permissible,” and many patients are unaware of the possibility. By the time of Halem’s pregnancy, Braverman and his colleagues had attempted at least eight transfers of supposedly abnormal embryos, with five successful live births. But there were no published successes and only a handful of anecdotal cases. And as stoic as Halem had been about implanting the abnormal embryos, her attitude changed the day she got the news she was pregnant. “I was a wreck,” she admits. “I was shocked, absolutely shocked, that it took.” At week eight, when she went in for an ultrasound at Mount Sinai, no one at the clinic — not doctors or nurses or counselors — had ever heard the story she told. “I remember the nurses saying, ‘You did what? You put an abnormal embryo back in? You’re crazy.’ ” But this time the ultrasound detected a heartbeat, after so many ultrasounds that didn’t.

And of course she wondered if she would have to terminate. Before taking the ten-week prenatal test Braverman had required, called the chorionic villus sampling, or CVS, she had had to consult a genetic counselor at Mount Sinai, where she went for her prenatal care. “I told her I’d put an abnormal embryo back in,” Halem recalls. “And she’s like, ‘Well, the likelihood that this is normal, you know — this is not going to be normal.’ ” The test took place in early August 2015, and she found out the gender before the rest of her results were available. Only one of the four embryos had taken: It was the girl, the one partial trisomy. Again, joy mixed with fear. Two nerve-racking weeks later, the complete results of the CVS test came back. “I was in the Hamptons on vacation with my daughter, sick as a dog, and they called me, and they’re like, ‘It’s a normal girl!’ I just sat there and said, ‘Omigod, omigod …’ ”

The rest of the pregnancy was not uneventful. At one point, Halem started bleeding so heavily that doctors feared she would lose the baby. At 33 weeks, in the wake of a blizzard that dumped 27 inches of snow on Central Park, her water broke, but she made it to the hospital, and on January 26, 2016, she gave birth to a baby girl. A normal, healthy baby girl.

A baby born to one of Dr. Braverman’s patients from an abnormal embryo. Photo: Courtesy

The first time many IVF doctors heard about what had been going on in New York clinics was three months earlier, on October 20, 2015, in Baltimore, when Gleicher took the podium at the annual meeting of the American Society for Reproductive Medicine, the largest professional society of IVF doctors in the U.S. He announced, for the first time in a public setting, that the transfer of abnormal embryos had produced normal human babies.

A stocky physician with a combination of old-world gravitas and the swashbuckling ethic of new-world fertility mavens (he was born in Krakow, grew up in Vienna, attended medical school in Tel Aviv, and opened his first clinic in the mid-’90s in Chicago), Gleicher has always been a bit of a crusader and provocateur: In the late ’80s, he co-authored a study in the New England Journal of Medicine describing a pilot project at a Chicago hospital to reduce the “excessive” number of caesarean deliveries in the United States; he has questioned the current trend of single-embryo transfers; and, starting about ten years ago, he began voicing — with “evangelical” fervor, according to one critic — concerns about the validity of PGS.

In Baltimore, he described how doctors at three IVF clinics in New York City, including Braverman, had initiated an experimental program to transfer abnormal embryos into older women who had run out of other options. The 11 women were all older, poor-prognosis patients who, like Halem, had undergone IVF and been unable to produce normal embryos, according to PGS testing. They had been offered the opportunity to try to get pregnant with their aneuploid embryos. Three couples declined, but eight of the women wanted to try. “We thought there would be a lot of hesitation, and I can’t say they didn’t ask questions,” Gleicher recalls one recent Friday afternoon, sitting in his third-floor Upper East Side office. Mostly, he says, the women were surprisingly enthusiastic.

Five of the eight women became pregnant, and all five delivered healthy children. There were no miscarriages; the other embryos simply didn’t take. Gleicher’s clinic first attempted such a transfer in June 2012; among the initial successes was Braverman’s same-sex couple. (Halem, five months pregnant at the time, would be the next successful birth in the experimental program.)

Gleicher was seen as such an outlier in the field that many viewed the report with skepticism. But it was not a blip. One month later, in November 2015, a group of reproductive-medicine experts in Rome published a brief letter in the New England Journal of Medicine. The Italian doctors reported transferring abnormal embryos into 18 women who were otherwise unable to get pregnant and had failed to produce any normal embryos (according to PGS). Six of the 18 women became pregnant, and all six women bore healthy children. Other clinics began to report similar cases. “We know today of over 50 healthy babies born, and that’s probably just a fraction of what is going on,” says Gleicher. “I’m sure there are two or three times as many.” In May, Santiago Munné, a leading figure in reproductive medicine, and his colleagues reported that “more than 100 such babies have been born” so far.

You won’t find many press reports about Gleicher’s presentation, but you will find people who believe it upended, directly and indirectly, the current landscape of reproductive medicine. “All I can say is, it stunned a lot of us,” says Eli Adashi, a longtime reproductive-medicine expert at Brown University who was at the meeting. It challenged a basic IVF practice — suddenly “abnormal” embryos had to be considered, and perhaps even offered, as an option to women without any normal embryos. The possibility meant that thousands of women who have been told they can’t have a biological child may find themselves with a reason to try again.

These successes, though small in number, also raised the alarming possibility that the PGS test, which purported to distinguish normal from abnormal and which had been used as the main diagnostic test to determine embryo quality over the last 20 years, may have been giving misleading results and that as a consequence tens of thousands of potentially viable human embryos have been unknowingly discarded. In 2015, according to the most recent CDC figures, there were approximately 186,000 IVF cycles in the United States, which some experts in reproductive medicine believe could have produced conservatively 558,000 embryos, around 20 percent of which end up undergoing PGS analysis, according to industry scientists. And although there is considerable disagreement on what percentage of those embryos contain chromosome abnormalities, mainstream doctors consider 40 percent to be a reasonable figure, suggesting that in the U.S. alone, nearly 45,000 embryos may have been classified as abnormal and ticketed for disposal in a single recent year, many of which may have produced a normal baby.

“You inevitably throw away normal embryos with the test,” says Richard Paulson, who heads the department of reproductive medicine at the University of Southern California and serves as the current president of the ASRM. “How many? Some people will say around 10 percent. I think it’s 40 percent. I think it’s a large number of embryos that we’re throwing away because of the inefficiency of PGS.”

All these estimates are fuzzy because no one keeps track of discards. But there’s little dispute about the amount of human heartache associated with those discarded embryos. “For me, the tragedy is that people who could have had a baby didn’t have a baby,” says bioethicist Josephine Johnston of the Hastings Center in Garrison, New York. “And that makes me weep.”

“What really makes this in my mind such an unprecedented scandal,” Gleicher was saying that afternoon in his office, “is the consideration that we as a profession have been disposing of thousands and thousands of completely normal embryos, with normal potential. And nobody — nobody — has had the guts to stand up and say we are sorry.”

In order to understand why the PGS test results could be inaccurate, it helps to think of an embryo as a classic soccer ball. In fact, when Paulson engaged in a debate on PGS at a medical meeting earlier this year, he stopped at a sporting-goods store and bought a soccer ball to make precisely this point. Many soccer balls have a mix of black and white facets on the outside of the ball; many embryos at day five similarly have a mix of white (euploid) and black (aneuploid) cells on their outer surface. If you biopsy from a white part, you’ll think the embryo is entirely normal; if you biopsy from a black part, you’ll think it’s entirely abnormal. If you happen to biopsy at a seam, you might get a mix of black and white cells. Embryos that display this mix are known as “mosaics,” but the version of PGS that entered use around 2010 still deems these embryos abnormal. “The fact that it looks like a soccer ball means that you’re going to get potentially a different biopsy [result] every time you do it,” says Paulson.

A few years ago, Gleicher and his colleagues, including former Tufts University embryologist David Albertini, devised a research project to gauge the validity of PGS: The researchers would send cells from the embryos that had already been deemed abnormal by a PGS lab in New Jersey to another PGS lab in Miami to see if the results matched. They withheld the earlier results, explaining to the second lab only that it was part of a research project. It was a small sample — too small to be statistically significant — but the findings, published in 2016, hardly inspired confidence in the accuracy of PGS. The results from the two labs agreed less than 20 percent of the time. Four of the 11 embryos initially deemed abnormal were found to be normal. In one instance, the two labs didn’t even agree on the sex of the embryo.

How could an embryo, one with abnormal cells, produce a normal child? Perhaps the strongest scientific evidence that mosaic embryos can tolerate — and overcome — chromosomal abnormalities came from research, reported last year, from the University of Cambridge laboratory of Magdalena Zernicka-Goetz — a ten-year project triggered by her own gut-wrenching personal experience with an aneuploid pregnancy. In 2006, when Zernicka-Goetz was 42, CVS testing suggested her fetus bore an extra chromosome. She decided not to terminate the pregnancy and ultimately gave birth to a healthy son, now 10 years old. But to deal with her anxiety as she awaited further test results, Zernicka-Goetz sketched out plans for a series of experiments to see exactly what happens to an embryo that possesses abnormal chromosomes.

She and her colleagues employed sophisticated molecular tools to induce “abnormal” cells to form in mouse embryos, in effect creating mosaic embryos. Then they tracked what happened to these abnormal cells, in real time, as the embryo developed. To their surprise, they discovered that mouse embryos have the capacity to eliminate these bad-apple cells in the part of the embryo that becomes the fetus as a normal part of development. The research, led by Helen Bolton and published in Nature Communications, demonstrated that if there are enough normal cells, even embryos marbled with up to 67 percent abnormal cells have the potential to grow into baby mice — and did so in the experiments.

Zernicka-Goetz’s study highlights several reasons why critics doubt the validity of the PGS test. First, the cells from the outer wall, the ones biopsied for the PGS test, aren’t even destined to become the baby; instead, they’ll go on to form the placenta. Zernicka-Goetz found that abnormal cells on the outer shell, the black spots of the soccer ball, can persist even as the part of the embryo destined to become the fetus weeds out abnormal cells. Moreover, if human embryos are anything like mouse embryos, the self-correcting changes that Zernicka-Goetz observed can occur later in embryonic development, after PGS testing is conducted in human embryos. According to her, “it is very obvious to scientists that this is not the way to go.”

Reproductive medicine is just beginning to respond to these startling findings. The American Society for Reproductive Medicine is poised to issue a guideline to physicians, according to a version of the document obtained by New York, stating that there is “insufficient evidence to recommend the routine use of” PGS “in all infertile patients.” The turnabout is the latest chapter in a decades-long battle over the effectiveness of genetic testing of embryos.

In the afterglow of the Human Genome Project, with its giddy promise of unprecedented molecular precision, preimplantation genetic screening became an irresistible addition to the fertility doctor’s tool kit. As far back as the early aughts, however, PGS has been dogged by what Joyce Harper, a leading embryologist at University College London, calls a “checkered history.” Harper and other critics complained that studies purporting to show the efficacy of the testing were small and underpowered; in addition, there was no government oversight, because the Food and Drug Administration decided preimplantation testing did not fall under its regulatory jurisdiction. The initial version of the test, used on day-three embryos, began to fall out of favor after a rigorous research report in the New England Journal of Medicine in 2007 showed that PGS actually decreased pregnancy rates in women over 35. In 2008, the ASRM issued a policy statement saying PGS was ineffective. Adding injury to insult, a 2013 paper by Richard T. Scott and colleagues at Reproductive Medicine Associates of New Jersey reported that the biopsy necessary for PGS testing itself significantly impaired the ability of day-three embryos to implant. The test, which was offered from the late ’90s to about 2010, actually did harm.

To PGS proponents, these were predictable growing pains for a technology that was evolving and whose shortcomings were often attributed to “very inexperienced hands.” With the advent of more powerful analytic technologies, the second version of the test became more precise, faster, more definitive. Around 2010, this newer version of PGS began using a biopsy on day-five embryos to assess all 23 chromosomes, delivering results in some cases within hours.

It was this second version, though, that Gleicher was questioning anew in the fall of 2014, just as Halem was recovering from her seventh miscarriage and banking embryos for yet another try. Within a two-week period, four relatively young couples came to his center for consultations. All the couples had undergone IVF cycles at other centers, and none had managed to produce a single normal embryo. One couple, in their early 30s, had produced more than 20 embryos; all of them, according to PGS, were abnormal. “We were reviewing those cases,” Gleicher recalls, “and we said, ‘This is statistically not possible.’ If we hadn’t seen four or five [cases] within a very short time period, we might not even have noticed. But this was for us the decisive moment.” Either “abnormality” was a lot more common than believed, or maybe “abnormal” wasn’t as abnormal as the test implied. After assessing those cases, Gleicher’s Center for Human Reproduction announced a new policy in the fall of 2014: It would begin an experimental program transferring aneuploid embryos in women who PGS testing indicated had no normal embryos. The program included the Center for Human Reproduction, Braverman IVF and Reproductive Immunology, and Andrea Vidali’s clinic, New York Reproductive. Monica Halem was one of the first patients.

The new abnormal has forced a major rethink in assisted reproduction — one that is still sifting down to IVF clinics. “We used to think of embryos as binary — they were either normal or abnormal,” says David Keefe, who heads the department of obstetrics and gynecology at NYU’s Langone Health. “But as in many things in medicine, it’s become more complicated than that.” And with the complications have come debates and controversies.

Of the 100 or so babies born to date following abnormal transfers, there have been no public reports of genetic abnormalities in any of the children. Some doctors warn that it’s too soon to tell whether there are any long-term health effects. “I’m sure that sooner or later there’s going to be one,” Gleicher admits, “but I don’t think the chance is higher than otherwise.” Braverman thinks the odds are actually very low: Since the vast majority of test-tube embryos never underwent genetic testing, IVF doctors have routinely been transferring mosaic embryos for decades. As one recent paper by Munné and Dagan Wells of Oxford University put it: “Given their high frequency, it is inevitable that millions of mosaic embryos must have been transferred, unknowingly, since the advent of IVF.”

Munné — one of the pioneers of PGS, who is now the chief scientific officer of CooperGenomics, a major testing company — says the recognition of mosaic embryos poses a new dilemma for doctors and patients. “What you do with them depends on your philosophy,” he says. “If, let’s say, you have lots of embryos and you want to maximize getting pregnant as soon as possible, and not miscarrying, then you transfer the normal embryos and you don’t transfer the mosaics. If you have the other philosophy, and you want to transfer anything that has any potential, even though it has risks, then you would transfer the mosaics.” Some clinics have already started to do this; Munné in fact just led a multi-center study reporting 58 successful pregnancies out of 143 mosaic transfers (a success rate of 41 percent). Other clinics still oppose the idea of mosaic transfers. In at least one instance, Gleicher says, a prospective mother who came to him after having undergone IVF at another clinic needed a lawyer to regain custody of her own “abnormal” embryos.

Another complication is the future role of PGS in sorting through this new reality. Some 200,000 embryos in the U.S. are screened using PGS, according to Munné, and some IVF clinics routinely submit every embryo for tests. Around 2014, the industry introduced a powerful DNA technology, known as next-generation sequencing, to PGS tests. (The latest iteration of PGS has also acquired a new name — “preimplantation genetic diagnosis-aneuploidy,” or PGD-A.) Keefe, a strong proponent of the test, says that this latest version is sensitive enough to distinguish between “a little abnormality and a lot of abnormality.” And Munné argues that three studies have already shown this new version of the test improves outcomes, with a fourth large study to be released this fall. “I think if you put them all together, there’s definitely enough evidence that this is working,” he says, dismissing the historical criticism that the previous PGS validation studies were too small and underpowered. He also rejects the notion that it may have been responsible for the discarding of potentially viable embryos. “The lab makes the diagnosis,” he says, “and then the couple decides whether they want to transfer.”

Instead of distinguishing solely between abnormal and normal, the new PGS creates a third category, identifying embryos as mosaic when the test detects a mix of normal and abnormal chromosome configurations in the biopsied cells. These mosaic embryos can range from a little abnormal (20 percent) to a lot abnormal (up to 79 percent), based on an analysis of five cells, and there is already argument about what is a safe percentage of mosaicism to transfer. This new category is controversial and has not yet been endorsed by professional societies like ASRM.

As Munné puts it, “It’s not as clear-cut as before. Before we had black or white, with some errors. But now we have black and white, with no errors, but a gray area.” At the same time, Munné downplays the impact that mosaic embryos will have on IVF practice — there is, he says, “a lot of noise, and probably we’ll find out that they are not so important. But we’ll see.” How doctors and patients make sense of that gray area will be a challenge. The test’s proponents like Keefe insist that the new version is 98 percent reliable and finally delivers on the promise of improving pregnancy outcomes and reducing miscarriages.

An increasingly vocal alternative camp suggests, however, that the new test is inaccurate and may not even be necessary, arguing that Mother Nature is the best arbiter of an embryo’s competence. “It’s not a zombie technology, because the technology is improving,” says Raoul Orvieto, who heads a large public IVF clinic in Israel. But he believes the test remains fundamentally flawed in part because the biopsy only plucks cells from outside the embryo, not the part that becomes the fetus. The test, he says, “is looking in the wrong place.” (Munné admits that the biopsy issue is “still under debate.”) The industry definition of mosaic embryos, Orvieto adds, is “bullshit on bullshit … I know how it works and I know how to do it. But I don’t do it because I don’t believe it gives the best outcomes for my patients.” There may even be a reason to avoid it, according to Paulson. He worries that, like the ’90s version of PGS, the newer tests may do damage to the embryo “as a result of the biopsy process” and says “there is ample indirect evidence, circumstantial evidence, that you in fact are doing harm.” Munné says the “biopsy is safe, provided that you do it right, like any procedure that involves manual skills.”

Meanwhile, you can go onto many IVF-clinic websites and see how PGS is currently pitched to patients: It increases pregnancy success rates, lowers the risk of miscarriage, and avoids the time (and expense) of multiple IVF cycles. But if you go to the website of Britain’s Human Fertilisation and Embryology Authority, which regulates IVF medicine in the U.K., you will read that there is “limited evidence” for those claims.

The dilemma posed by mosaic embryos has muddied the water so much that Pasquale Patrizio, head of fertility medicine at Yale University, and Sherman Silber of St. Luke’s Hospital in St. Louis recently suggested that “perhaps the best advance now for IVF is to take a step backward” and simply transfer day-five embryos without testing them first, predicting rates of live births and pregnancies “will quickly surge” as a result. Gleicher estimates that the increased transfer of mosaic embryos — embryos that only a few years ago would have been discarded — could yield thousands of additional IVF babies each year, especially in older women who don’t produce “normal” embryos.

For the time being, people who embark on the fertility train now have access to the most powerful genetic technologies and yet may still end up in destinations like Paradox, Uncertainty, and, as they say on the fertility blogs, BFN — “Big Fat Negative.” Monica Halem has been there. In her rounds of IVF, she transferred supposedly normal embryos that didn’t take and a supposedly abnormal one that did. In the middle of our long conversation about aneuploidy and mosaics, Halem pulls out her phone, calls up a photo, and slides the device almost accusingly across her desk. It delivers a picture of her 16-month-old daughter, lovely in lavender pajamas, grinning impishly. “Here’s this little abnormal embryo, okay?” she says. “This is what they would have thrown away.”

*This article appears in the September 18, 2017, issue of New York Magazine.

*This article has been corrected to show that BFN stands for “Big Fat Negative” not “Big Fat Nothing.”

There Is Now Hope for Women Told They Hit an IVF Dead End