AS I LAY about recovering from whooping cough in late 2009, reflecting on all I had learned in my journeys, I had a vision of how much easier social justice work around scientific research might become if it were consistently evidence-based. Scientists, the vast majority of whom I now understood to care deeply about social justice, would have to respect evidence-based activism. Maybe if everybody just agreed to discuss what we really knew—rather than imagining, assuming, and suspecting based on loyalties to particular theories or persons—disputes could be sorted out peaceably. Researchers and advocates could come together, look at the facts, and—as in the case of climate change, AIDS research, and intersex care—the great majority of researchers and activists would agree on what had happened and what needed to happen. The light of many minds would show a way forward. After carefully investigating and getting reasonable people to see what had actually happened in the Bailey and Chagnon cases, I felt downright optimistic that if we all simply agreed to talk through what we knew and didn’t know, true cases of injustice would be relatively easy to spot, expose, stop, and ultimately prevent.
As if by providence, at that moment—just as I was feeling tremendous clarity about what activism should and shouldn’t look like—a small group of my old intersex-rights allies, including a couple of clinicians with whom I’d remained close, called me to ask me to help with an ongoing travesty of justice. The crux of the matter was this: A major clinician-researcher in the intersex field (someone we all knew well) had been promoting a high-risk drug regimen to pregnant women who, along with their mates, were genetic carriers for a particular intersex condition, one caused by elevated levels of androgens in female children. The idea behind the intervention was this: If, through genetic screening, a clinician identified a woman as being at risk of having a daughter with this condition, then as soon as the woman got pregnant, a doctor would start dosing her with the steroid dexamethasone. If all went as planned, the dexamethasone would dampen the effect of the androgens, preventing intersex development in the daughters.
The major proponent of this fetal intervention, Dr. Maria New—a distinguished member of the National Academy of Sciences and a charismatic pediatric endocrinologist still actively working in her eighties—was going directly to the support groups for families with this condition to strongly recommend the intervention to them. She was selling parents on this “treatment” by claiming that “with nearly 20 years’ experience, the treatment has been found safe for mother and child.” This wording made it sound as though the Food and Drug Administration (FDA) had approved this use, when the FDA hadn’t. While doctors are allowed to prescribe drugs “off-label,” i.e., for uses the FDA has not approved, doing so means acting without the benefit of extensive FDA review of a use. In fact, the medical-scientific literature was utterly devoid of well-controlled studies of efficacy and long-term safety of prenatal dexamethasone for intersex prevention. Doctors didn’t know what the intervention really did in the long term, although many were very worried; dexamethasone was being used specifically because it had the power to cross the placenta and permanently alter the course of fetal development. Indeed, lab-based studies of prenatal dexamethasone exposure in animals, including nonhuman primates, suggested that dexamethasone exposure could shift fetal development in deeply unpredictable ways, including by changing brain development.
It got worse. Even while Dr. New was describing this use as safe and effective, openly taking credit for having gotten over six hundred pregnant women to follow this prenatal regimen, and boasting that she had at Mount Sinai School of Medicine in New York “the only clinic in the United States which routinely provides this service,” she was at the same time taking National Institutes of Health (NIH) funding to study, retrospectively, what had really happened to children and mothers who had been exposed. To get the NIH funding, Dr. New told the agency something very different from what she said in her public promotions of prenatal dex: “The long-term effects of dexamethasone on the children who received it as fetuses and on mothers who were exposed to it while they were pregnant have not been determined.” She told the NIH that her large “accumulated” clinical population of exposed women and children made her especially well positioned to do this research.
Telling pregnant women a fetus-altering drug is “safe and effective” and then using them to get grants to see if it really is safe or effective? The failure of ethics in this case seemed so appalling and so obvious to me that it seemed a perfect case study of how to spot and stop a true and significant case of abuse of research subjects. And it involved the very issue around which most of my work had revolved: the often-harmful, non-evidence-based medical quest to “normalize” children. Here again was a situation where doctors, motivated by not-unreasonable fears of social stigma, were driving forth an unscientific and even unethical system of treatment—in this case, risking the health of pregnant women and their babies, completely outside of modern scientific standards for pharmaceuticals. Indeed, after I did a little more reading and investigating, the situation seemed so clear that I thought we could and should use this as a demonstration of evidence-based social justice protest—a model of how you appropriately use evidence, the press, and the Internet to reign in abusive research. This case could serve to show scientists why activists were very reasonable to still be vigilant about ethics violations—because terrible abuse really was happening to some vulnerable subjects—and to show activists why and how we should work to push for better evidence. Moreover, because the researchers calling for my help were calling me in as an activist, I thought this would be a great example of collaborative work between the two camps.
Yes, it would be relatively easy! The source of our outrage, the reason for our collaboration across the researcher-activist divide—it would all be obvious to everyone. The press, the universities, and the government would do what they’re supposed to do. People weighing in on the Internet would behave well. Everyone would see that this is how you do it: You push together for the truth.
Looking back, I’m guessing I was high on cough medicine. That said, even if I hadn’t been, I would have taken on this cause. It was too important to say no. This was one quest for pediatric “normalization” run completely amok. And although this work—which I guessed at the outset would take only three months—ended up consuming the next three years of my life, nearly crushing my reputation and my spirit, it did end up teaching me the last things I needed to learn from my long journey: how badly most people want simple stories of male and female, nature and nurture, good and evil; how the Internet has gutted the Fourth Estate; how the government is made up of fallible and occasionally disappointing humans; and why, more than ever before, democracies must aggressively protect good research.
• • •
AT THIS POINT in my intellectual development, after witnessing the personal consequences to individuals attacked for controversial research, to say that I was reluctant to call out a single researcher would have constituted a substantial understatement. But after my colleagues brought this case to my attention, I poked around and could quickly see why they were focusing specifically on Maria New. Dr. New, I realized uncomfortably, had made herself the eight-hundred-pound gorilla of prenatal dexamethasone for intersex prevention. She really stood alone. There had been a number of major pediatric researchers, including prominent physicians with whom I had for years argued over intersex care, trying to put a stop to her misleading promotion of this intervention as having been found safe and effective. A quick search confirmed why they were a bit frantic: There remained a complete absence of any properly controlled scientific studies demonstrating efficacy or safety in humans. There hadn’t even been any animal modeling of this use published. The animal studies that had looked at prenatal dexamethasone were designed simply to find adverse effects of prenatal exposure to dexamethasone in general; no animal studies had been published specifically looking at whether you really could stop intersex development via prenatal dosing with steroids. The number-one rule of pharmaceutical research? Before you touch a human, see if the intervention works in animals. This was especially true for fetal experimentation. Yet such animal modeling hadn’t been done, and New’s retrospective sampling methodology suggested that the results of her NIH-funded study would be of low scientific quality. While what she was doing didn’t appear to be illegal according to FDA prescribing rules, ethically it was extremely problematic.
Meanwhile, alarming data were emerging from a Swedish research team that had been tracking children there carefully from the start of prenatal dexamethasone exposure straight through later childhood. The Swedish data suggested that prenatally exposed children were showing increased rates of problems with memory and with social anxiety—signs that the intervention might have adverse brain effects. And the rest of their bodies? The Swedish group had found cases of growth failure and of delayed psychomotor development among the children prenatally exposed, suggesting unintended effects on the rest of the body as well.
Dr. New had not been studying the effects of exposure nearly as carefully as the Swedish team, even though she had taken credit for exposing well over fifteen times as many mothers and children. I did the math to figure how many pregnant women per month might be getting sucked in to this unethical cycle of ill-informed “treatment” and federal follow-up research grants, knowing the frequency of the condition at issue (somewhere between 1 in 10,000 and 1 in 15,000 births) and knowing that New’s clinic drew patients from around the world. It could be a dozen per month. Then I mentally put myself in the position of these pregnant women, most of whom would know only what was told to them by the intervention’s reassuring promoter. Then I thought about what New was doing—telling the families the intervention had been found safe while using previously exposed families to get NIH funding to see if it was safe.
Then I went back to my friends’ central question: What should we do?
You’d think that if anybody was qualified to answer the question How do you stop a scientific researcher?, it would’ve been me at that moment. The problem was that my work had only uncovered all the wrong ways to do it. What the hell was the right way?
• • •
TO UNDERSTAND HOW my colleagues and I were processing this scene, one needs first to know a little biochemistry, a little developmental biology, and a little history.
In spite of what popular culture might lead you to believe, males and females usually make the same types of hormones. In addition to making the hormones usually associated with their sexes, male bodies also make the “female hormone” estrogen and female bodies also make the “male hormone” testosterone. What determines male and female biology is, in part, the different levels of sex hormones that usually occur in males and females during sex development. Being estrogen-heavy will usually make you more female-typical, while being testosterone-heavy will usually make you more male-typical.
Ovaries equal more estrogen, and testes equal more testosterone. That’s why most females are feminine and most males are masculine. But again contrary to pop culture’s representations of sex, the ovaries and testes aren’t the only organs in the human body that produce sex hormones. Some sex hormones are also made by our adrenals, a pair of glands that sit just above our kidneys. These glands make androgens, a group of hormones sometimes called masculinizing hormones: androgens can push the development of sex differentiated-tissues (like genital tissue, breast tissue, and some brain cells) in a more male-typical direction during sex development, even if you have ovaries. Perhaps you already see where this is going: There exists a condition called congenital adrenal hyperplasia (CAH) that results in higher-than-normal levels of androgens. If CAH occurs in a genetically female fetus, in spite of the fact that the fetus has XX chromosomes, ovaries, and a uterus, the higher-than-normal level of androgens can lead that fetus to be born masculinized in terms of genital appearance. CAH is, in fact, the most common cause of congenital ambiguous genitalia in genetic females.
The reason it’s impossible to know the efficacy of prenatal dexamethasone for CAH without a blinded, controlled drug treatment trial—something that has never been done—is because there is actually substantial natural variation in genital formation of CAH-affected females, ranging all the way from nearly typical female to nearly typical male. The natural variation among XX children with CAH means that a child with the condition may be born with a typical clitoris, a large clitoris, or in very extreme cases even one that has the urethra running part or all the way through it, just like a penis. The labia majora may be separated in the typical female fashion or may be joined at birth, looking much like a male scrotum. The XX child with CAH may have a normal vagina, may have a vagina that will not allow normal menstruation and vaginal intercourse, or in very rare cases may be essentially missing a vagina altogether. Sometimes a CAH-affected XX child will have something called a urogenital sinus, wherein the urethra and vagina develop in such a way that they meet, creating the potential for urine and menses backwash and subsequent infections.
Like genitals, brains are also shaped by prenatal androgens, so it should not surprise us that many CAH-affected girls’ and women’s behaviors and interests show what researchers call behavioral masculinization. Although there is a lot of variation in this population, CAH-affected girls are more likely than average girls to be interested in boy-typical toys, like cars and trucks, than in girl-typical toys, like baby dolls. They’re more likely to want to play with boys and are more interested in rough-and-tumble play. CAH-affected women are more likely to be bisexual or lesbian, and less likely to be interested in becoming mothers. A sizable percentage of genetic females with CAH—as much as 5 percent—ultimately identify as male in terms of their gender identities. The behavioral outcomes in this population strongly suggest that gendered behaviors and sexual orientation have a biological inborn component, and in CAH, females are skewed in a male-typical direction.
CAH can actually show up in both males and females, but only in the females can it cause intersex development. (Extra androgens in a developing male won’t make his sex ambiguous.) Notably, in both males and females, CAH can be a dangerous endocrine disease. Because the adrenal glands are responsible for the body’s response to physiological stress, having CAH can mean a poor response to infections and traumatic accidents. In the most extreme forms of CAH, a newborn baby who isn’t diagnosed and treated promptly with steroids can go into an adrenal crisis and die within a few days of birth. The potential seriousness of CAH is why all fifty states now employ newborn screening for CAH. If prenatal dexamethasone could prevent or cure CAH—not just prevent intersex development—that would require a recalculation of the ethical equation of using (and pushing) this intervention. Unfortunately, the endocrine disease that is CAH can’t be prevented, nor can it be cured. No matter what you do to a CAH-affected fetus prenatally, the medical dangers of CAH will have to be managed postnatally with monitoring and medication. This means that even if prenatal dex prevents intersex development, an exposed child will still face the serious health consequences of CAH.
CAH is a recessive genetic condition, which means that if both parents are carriers for the condition, each of their children will have a one-in-four chance of having CAH. Prenatal screening of a fetus in the second trimester has been used for many years to reveal to parents whether a fetus has CAH. Some couples have historically opted to abort. But in the early 1980s, in France—where, perhaps not coincidentally, by the time you could diagnose a fetus with CAH, an abortion would have been illegal—clinicians figured out that dampening down androgens in utero might prevent prenatal atypical sex development in CAH-affected females. A 1984 paper by these French clinicians published in the Journal of Pediatrics showed apparent success in one case using a prenatal course of dexamethasone, and after that documentation of assumed success in just one human case, “at-risk” mothers-to-be throughout the world started being offered prenatal dexamethasone.
Right around the time of the French group’s publication, Dr. Maria New began making the intervention available at her Cornell University Medical College–based clinic in New York City. (This was years before she landed at Mount Sinai School of Medicine, across town.) Dr. New had become internationally known in medicine and among the CAH-affected population for having improved the health and fertility of countless patients through her research and clinical care. She was the top specialist in CAH care. It was therefore easy for her to reach out and recommend this intervention to large numbers of CAH-affected families and their personal physicians. And she did so with gusto.
But why would anyone take such risks with fetuses? To understand this, you have to understand what a strong revulsion and/or pity some people feel toward a person who is born with a body that falls smack-dab between male and female. Maria New was not alone in feeling that intersex should be prevented if at all possible; many of her colleagues had long had the same general philosophy, knowing as they did that, historically, in most cultures, people have been expected to adhere to one of two sex types, in body and behavior. Although in the younger generation, clinicians have changed their thinking about intersex, homosexuality, and transgender, Maria New, born in 1928, had been working consistently from a very traditional, very conservative approach to sex and gender. Here, for example, is Dr. New speaking to a group of CAH-affected families at a support-group meeting in 2001 recorded on video. The audience sees, on a screen beside Dr. New, the masculinized genitals of a baby girl born with CAH—a large clitoris, fused labia. The genitals look pretty male. Dr. New tells the families:
The challenge here is . . . to see what could be done to restore this baby to the normal female appearance which would be compatible with her parents presenting her as a girl, with her eventually becoming somebody’s wife, and having normal sexual development, and becoming a mother. And [internally] she has all the machinery for motherhood, and therefore nothing should stop that, if we can repair her surgically and help her psychologically to continue to grow and develop as a girl.
The message is clear: Genetic females are supposed to have petite clitorises, grow up to see themselves as typical girls, and become wives and mothers. Since the baby shown in the slide was already born, Dr. New was suggesting to the audience that this particular child would have to have her femininity saved by normalizing genital surgeries. From Dr. New’s point of view, better still would be the clinical approach of ensuring a sex-typical girl right from the start of development—with prenatal dexamethasone. Dr. New encouraged the parents at the meeting to see this and to understand how personally motivated she felt to recommend prenatal dexamethasone:
Now one of the biggest breakthroughs, and I think one of the most satisfying things I’ve done in my scientific life, was to participate in the program of prenatal diagnosis and treatment of congenital adrenal hyperplasia. I didn’t start it. It was pioneered by a wonderful scientific investigator in Lyon, France, by the name of Maguelone Forest. She is the one who started the idea that it might be possible to treat the fetus by treating the mother. And what we did in New York is simply to refine it, advance it, and make it more feasible. . . . I’ll give you some slides to buttress what I’m saying, [but] the conclusion is, it’s effective and it is safe. Notwithstanding any people who criticize me for what I’m doing, I think it’s one of the nicest things I can do for my patients, which is to prevent ambiguity of the genitalia, so that these children, when they’re born, do not have confusion of sex assignment—nobody can call it a boy when it is really a girl, and nobody has to undergo very difficult surgical procedures, the surgery for which is only done well in very few centers.
In making this presentation, Dr. New revealed that she herself is a carrier for CAH, a fact that probably explained how she got into this field of research. The message to the parents would have been clear: I am one of you, and I want only what is best for your children. And good news: “The conclusion is, it’s effective and it is safe.”
What is dexamethasone? A potent synthetic steroid, dexamethasone is actually used for a wide variety of medical purposes. Doctors ordinarily prescribe it for children and adults to treat inflammation, like the kind you might get with some eye or skin diseases, and to treat various autoimmune diseases, like psoriasis and lupus. Dexamethasone has sometimes also been given in the second or third trimester to pregnant women at risk of giving birth prematurely; that’s because some of the steroids known as glucocorticoids—the drug class that includes dexamethasone—can cross the placenta and push a fetus’s lungs to mature faster, giving the fetus a better chance at survival if born too soon.
But the prenatal use of dexamethasone at issue in the intersex story aims not at second- or third-trimester fetuses who might otherwise die; instead it aims at first-trimester female fetuses who might otherwise develop in sexually atypical ways. In fact, it’d be more accurate to say the CAH-related use of prenatal dexamethasone is aimed at embryos. Humans start to differentiate into male and female types really early, by the seventh week of development. If you want to reengineer human genital development, you have to hit an embryo—through its pregnant mother—with dexamethasone pretty much as soon as you know the mother is pregnant. Usually the parents are identified before pregnancy because they’ve already had a child born with CAH, but additional “at-risk” parents are picked up because they’ve decided to undergo preconception genetic screening, and the screening picks up that both parents are carriers. Through the CAH care networks, CAH-affected families steadily learned over the years to start a woman on dexamethasone the moment a pregnancy with CAH risk was confirmed.
The prenatal use of dexamethasone for CAH has always been “off-label,” meaning that, while dexamethasone has been approved by the FDA for some specific medical uses (like treatment of inflammation), this has never been one of them. In fact, the FDA has for years classified dexamethasone in pregnancy category C, those drugs for which “animal reproduction studies have shown an adverse effect on the fetus and there are no adequate and well-controlled studies in humans.” Why then would doctors still be allowed to use it in pregnant women? Because, as the FDA notes for category C drugs, “potential benefits may warrant use of the drug in pregnant women despite potential risks.” Some doctors, like Maria New, decided that the “potential benefit” of intersex prevention justified potential risks.
But one would expect such doctors to be terribly careful with a first-trimester sex-engineering attempt, particularly since the DES disaster. Starting in the late 1930s, physicians sometimes gave pregnant women diethylstilbestrol (DES), a synthetic estrogen, generally to prevent miscarriage. A controlled, blinded study published in 1953 in the American Journal of Obstetrics and Gynecology actually showed that DES didn’t work for this use; women who were given DES did not end up with healthier pregnancies. Nevertheless, obstetricians kept pushing DES on pregnant women, perhaps unaware of the study, perhaps disbelieving it, perhaps thinking DES couldn’t hurt. With no good evidence of efficacy or safety, about 5 to 10 million fetuses were exposed to DES by the time this use finally came into serious question in 1971. It was then that clinicians in Boston realized that girls and young women who had been exposed to DES in utero were developing a rare and often fatal vaginal cancer at an alarmingly high rate. They realized the DES link to this strange cancer cluster only because a woman named Penny Stone, a mother of a teenage girl with this rare vaginal cancer, insisted that researchers consider as a possible cause the DES she had taken when pregnant with her daughter Sheila. A few months after the Boston group made its findings known, facing intense pressure from a Congressional investigation, the FDA issued an alert, and the intervention eventually tapered off among obstetricians. Since then, researchers have linked prenatal DES exposure to increased risk of many reproductive cancers, as well as to major fertility problems in prenatally exposed males and females.
The negative effects of DES did not present themselves in obvious ways at birth, which is part of the reason they took so long to be recognized. But as the medical profession learned the hard way, when a drug given during pregnancy is studied unscientifically, even dramatic birth defects may not be tied to the drug until thousands of babies are harmed. In the middle of the twentieth century, the widespread use of thalidomide, a sedative and antinausea drug for adults, led to babies exposed in utero being born without limbs or in some cases with flipperlike limbs. It was only in the early 1960s when Frances Oldham Kelsey, a scientist at the FDA, pushed very hard to collect and analyze data that the disaster of thalidomide came to be widely recognized. By that point, about ten thousand children in Europe had already been born with shocking birth defects because their mothers had taken thalidomide when pregnant with them, typically to deal with morning sickness. Yet, until a prominent Washington Post article about Kelsey’s scientific efforts emerged, the buzz had continued to be that thalidomide was a safe and effective sedative and antinausea medication.
Everybody working in medical research knows that there are lots of formal protection systems in place to prevent another DES or thalidomide. For example, researchers who want to experiment on pregnant women and fetuses are supposed to jump through extraordinary regulatory hoops before being allowed to proceed. There’s also supposed to be a kind of cultural protection in place from the conventional wisdom developed in response to DES and thalidomide, a conventional wisdom that says you don’t mess with fetuses pharmacologically unless you absolutely must, unless, for example, you think the fetus will otherwise die, as in the deployment of steroids for premature birth. Even then, you are supposed to proceed with profound trepidation and enormous scientific care if you know the drug crosses the placental barrier and affects the fetus. DES and thalidomide showed us how many cases of harm it can take, if you’re not scientifically careful, to start to learn what effects you’re really having when you’re exposing fetuses to pharmaceuticals.
So imagine how careful you should be when you’re exposing embryos to a drug with the intention of changing how their tissues will develop.
And how often does a pharmaceutical intervention have in all cases only the effect intended?
Hint: Never.
• • •
AS I BEGAN THINKING ABOUT the dex issue, one of the clinicians I was talking to was David Sandberg, a pediatric psychologist from the University of Michigan and Mott Children’s Hospital in Ann Arbor, with whom I’d long been collaborating on efforts to push back against risky attempts at medically “normalizing” healthy kids. Neither David nor I saw in theory any deep moral problem with preventing the development of ambiguous genitalia in utero. Neither of us saw intersex genitals as any more special or sacred than any other genitals, and all other things being equal, especially in our still-crazy medical system and in our culture at large, a child would almost certainly be better off born with typical genitals. Doing something to prevent “the need” for a surgeon to come at these children’s genitals with a knife? In theory, sounds good.
But as David talked me through what was going on, he made clear that the major concerns he and others had were not with theory but with practice. First off, there was the concern that the drug had to be given so early that it meant women were started on it as early as the fourth or fifth week of pregnancy. At that point, the embryo being targeted would be as small as a sesame seed, and the total dose of glucocorticoids reaching it would probably be sixty to a hundred times the naturally occurring level of glucocorticoids. Much more disturbing, it wasn’t at all clear whether the pregnant women understood that doctors had very little idea what this drug might do to them or their children in the long run or that many experts were plenty worried. Twenty-five years into use, we didn’t even have a placebo-controlled study to show it even did what was intended. All we had were reports from the doctors like Dr. New who had themselves recommended the drug saying they judged the intervention to be fairly effective in reducing genital masculinization in the females. Not terribly objective science, to say the least.
Now data coming out of Sweden were showing statistically significant unintended effects on brain development. There could be lots of other problems we didn’t know about yet—and wouldn’t know about anytime soon if this experimental intervention was conducted unscientifically and without mothers even realizing that it was experimental. Not knowing they were in what amounted to a big, sloppy experiment, the mothers wouldn’t necessarily consider the possibility that their children’s physical and cognitive problems might be side effects of prenatal dex. Even if, like Penny Stone, they did suspect prenatal drug exposure to be the cause of problems, few would know how formally to report their concerns. In the United States, there was no standardized registry of those exposed in utero, only Dr. New’s closed clinical records. Dr. New was supposed to be using her NIH grant to follow up with those treated years earlier, but many families were either missing or choosing not to participate when Dr. New’s research team called on them to be in the survey.
Because I teach the history of medicine, my friend David Sandberg didn’t have to tell me to think about all the parallels to DES and thalidomide. It seemed to us as if the American clinicians using prenatal dexamethasone for CAH had learned nothing from history. But it was also like they were learning nothing from the contemporary scientific literature. By the time I turned my attention to this, bench researchers studying the physiology of development in nonhuman animals had steadily been reporting that glucocorticoids like dexamethasone, when deployed during fetal development, appeared to lead to “fetal programming” of adult pathologies in various organ systems; in other words, prenatal glucocorticoids could permanently skew the developing body, reprogramming the tissues to develop major diseases like cancer in adulthood. Just like DES.
Listening to David rattle off his concerns, I was reminded that Dr. New had left Cornell’s medical school a few years back under the cloud of a major NIH fraud case and had then taken a job in another part of New York City, at Mount Sinai’s medical school. In 2005, the Wall Street Journal had reported that the NIH grant on which Maria New was the principal investigator—the grant that had included Dr. New’s research on prenatal dexamethasone—had been at the center of a fraud investigation that resulted in Cornell paying the NIH a $4.4 million settlement. According to the whistleblower, a Greek pediatric endocrinologist named Kyriakie Sarafoglou, who had been working in Dr. New’s unit, there appeared to be “phantom research projects” i.e., research projects that were being funded but were not actually happening. (Perhaps this helped explain why, although New kept promising the NIH and her colleagues extensive follow-up data on long-term outcomes of prenatal dexamethasone exposure, the data were strangely absent in the published literature.) According to Sarafoglou, at first Cornell had tried coming after her for raising an alarm about what she was seeing, even requiring her to submit to a psychiatric exam. Instead Sarafoglou went to the Feds.
As he talked me through what he knew, David also wanted to make sure I understood that it wasn’t just the females with CAH who were being exposed. Because the intervention for CAH had to start so early in embryonic life if it was going to work as planned, it had to begin before doctors could know if the woman was even carrying the kind of embryo they were worried about: a female with CAH. Remember that, on average, only one out of four offspring of at-risk couples will turn out to have CAH, and of those the issue of ambiguous genitalia affects only the females. Bottom line: Seven out of eight of the embryos exposed to dexamethasone in the first trimester could never gain any of the intended benefits from the intervention, because only one of eight “at-risk” pregnancies would produce a female with CAH—but all would bear the unknown risks of early dex exposure.
Actually, when I redid the math after some research, it turned out to be more like nine out of ten fetuses exposed to the risks with no chance of benefit—because not all girls with CAH develop ambiguous genitalia, and because even the proponents of dex estimated that it achieved its goal only about 80 percent of the time. Yet 100 percent of these couples’ developing offspring would be blasted, starting in the embryonic stage, in hopes of “saving” 10 percent of them from developing to be sexually atypical. The fetuses identified in utero as females with CAH potentially faced even higher adverse-event risk than the rest, because for them, the dexamethasone dosing continued all the way through the pregnancy, whereas for the rest, the intervention ended after a few weeks, once the obstetrician determined that the fetus being carried was most likely not a female with CAH.
Given all this, even some fairly old-guard doctors were speaking out against this whole intervention, essentially saying “this is just plain nuts.” One senior pediatric endocrinologist, Walter Miller of the University of California’s medical school in San Francisco, had been warning about the dangers for years. Miller had finally declared in 2008 in Endocrine News that “the accumulating evidence that prenatal exposure to dexamethasone has mild but deleterious effects on the developing brain of the 7 of 8 fetuses treated needlessly should clearly indicate that prenatal treatment of CAH is fraught with ethical (and possibly legal) problems. It is this author’s opinion that this experimental treatment is not warranted and should not be pursued, even in prospective clinical trials.”
But the intervention wasn’t even being pursued in prospective clinical trials, except in Sweden, where in 1999 researchers intentionally halted access to it through any other route because of their findings that children had been harmed. As far as I could ascertain, in the United States there had been no meaningful long-term oversight from medical-school institutional review boards (IRBs, or medical ethics committees), so there were no meaningful protections in place to assure that these pregnant women were fully informed before taking on these significant risks. Without IRB protocols in place for a prospective trial that would track the children continuously, without independent safety monitoring, there would be no meaningful assurance that adverse outcomes would be detected and reported promptly.
As I soon discovered, the truly amazing thing was that numerous medical societies had long been calling for the highest scientific and ethical precautions when using prenatal dexamethasone for CAH. Even a committee from the top American pediatrics group, the American Academy of Pediatrics, (AAP), had called for the greatest caution in an open (and ugly) debate with Maria New in 2001, the very year she was recommending the treatment at that support group meeting. The AAP representatives admonished New: “The maxim of ‘first do no harm’ requires a cautious, long-term approach, which is why the Academy Committee unanimously agrees that prenatal glucocorticoid therapy for CAH should be confined to centers doing controlled prospective, long-term studies. The memory of the tragedies associated with prenatal use of DES (diethylstilbestrol) and thalidomide demands no less.”
Yet almost a decade after that admonishment, nothing had changed.
• • •
RESOLVED TO WORK on this issue, my only question was how. This was different—really different from our old intersex fights over genital surgeries. For one thing, I was actually ideologically allied with most of the medical establishment types this time around. For another, intentional exposure of human fetuses to a development-altering steroid was something about which all kinds of protections were already supposed to be guaranteed. It seemed we just needed to pull some particular lever. But which one?
Before I did anything, based on what I had learned from talking to criticized researchers, I thought I had best first write to Maria New and ask her whether she had been obtaining informed consent of the parents. Maybe I was wrong. Maybe there was an explanation for what I thought I was seeing. When we had met years earlier at a meeting of intersex specialists, I had walked in to find Maria New recommending my first book to her colleagues, even though it contained sharp ethical criticisms of them all. She surprised me then. Maybe she would surprise me now. So on December 8, 2009, I wrote to her: “Dear Maria, I hope this finds you well. I have two questions for you with regard to prenatal dex treatments for CAH: Do you have any forthcoming publications on the long-term outcomes of these treatments? May I see the consent form you’re using for this experimental treatment?”
There came no reply. But it was safe to guess that New saw my message; in early January, at a conference of pediatric endocrinologists in Miami, the clinicians who had called me found her even more defensive than usual. In a session on long-term follow-up of prenatal dex for CAH, New raced through the slides, failing to let anyone really see the data. From what those calling to report to me could see, more than half of those exposed appeared to be completely unaccounted for, and the scientific quality of what data New did purport to have was very poor. One of my colleagues present at the meeting, pediatric geneticist Eric Vilain of UCLA, had boldly pressed New during the Q&A: What did the consent for this look like? What was she telling the mothers? She responded that his question was out of line. She wasn’t going to answer him. Others told me that her response to Eric had caused the room to go still. Then she’d gone on to assure the audience that she wasn’t accidentally making the prenatally exposed boys gay. Perhaps she said this because she knew that in the Swedish cohort at least one boy exposed in utero had been born with genitals that suggested incomplete masculinization, a sign that perhaps the boys were also becoming collateral damage in the sex-normalization quest.
I couldn’t let this continue without trying to help bring it to light. After pulling all the published information I could find and looking for evidence of proper ethics and scientific oversight, what I was seeing just seemed to confirm our worst fears. Besides promoting the intervention via the support group for CAH and her own foundation Web site, when writing about CAH for various textbooks, Dr. New had made a point of plugging prenatal dex to other doctors, writing as if it simply was the standard of care among clinicians in regular practice. As a result, all over the country, obstetricians and genetic counselors were using prenatal dex believing it to be safe and effective.
Still unsure exactly what to do, I talked to Aron’s boss, Marsha Rappley, who just happened to be recently retired from chairing an FDA panel on pediatric therapeutics. Marsha seemed skeptical that people would really be giving a fetus-altering drug outside of prospective long-term studies, but she helpfully suggested that, in addition to writing the specific medical schools involved, we might also formally register our concerns with the FDA Office of Pediatric Therapeutics and the federal Office for Human Research Protections (OHRP). This was the big idea I needed—to call specifically on the federal agencies charged with protecting people in just these kinds of situations. The same day I talked to Marsha, I also purchased the domain name FetalDex.org as a way to provide information for everyone, and I let the editors at the Hastings Center, the largest independent bioethics center, know I was planning to write a blog post for them about prenatal dex. The article would be used to gather up names of people in bioethics and allied fields willing to sign letters of concern to the FDA and the OHRP.
Still feeling more than a bit queasy, I decided I’d also better write another e-mail message to Maria New, letting her know I was about to start criticizing her publicly and giving her another chance to show me evidence that she was protecting the rights of the families. The subject line I used this time was “IRB oversight,” to indicate clearly that I was worried about violations of regulations governing research on humans:
I am about to write and publish an article critical of your use of dexamethasone in the maternal-fetal period without IRB oversight. You have a major grant to study the long-term effects of this experimental treatment, which clearly indicates you understand that there are risks to the children (not to mention the mothers) subject to what you understand to be an experimental treatment, yet apparently you do not have IRB approval for the experiment that is the treatment itself. In the event you have IRB approval for the experimental, off-label use of dexamethasone in pregnant women “at risk” of carrying a child with 21-Hydroxylase Deficiency [CAH], please let me know that within 48 hours. In the event I am wrong, I will correct my work. Thanks.
Two days later, I got a response, not from Dr. New but from Jeffrey Silverstein, the director of the Program for the Protection of Human Subjects at Mount Sinai, i.e., the person who would be in charge of overseeing New’s human-subjects research.
Silverstein wrote to say, “While your email was unusual, I am responding in the spirit of academic collegiality. In the time frame given, I am only able to confirm that Dr. New has an IRB-approved protocol for the study of ‘Rare Diseases of Steroid Metabolism,’ that includes treatment with dexamethasone in pregnant women with 21-Hydroxylase Deficiency.”
I knew from searching ClinicalTrials.gov (the government Web site showing who has what federal clinical research funding) and Dr. New’s papers what that probably really meant: Mount Sinai had in all likelihood given permission for the low-risk retrospective follow-up studies on prenatal dex for prevention of sex atypicality, studies conducted years after the pregnancies, but not for the high-risk prenatal drug administration itself. If so, there were no ethical protections in place at the important moment—when the woman was offered the drug.
Nevertheless, I asked Dr. Silverstein to clarify: “I know Dr. New has IRB approval for the . . . ongoing study of children who were subjected to dexamethasone in the womb. I am asking whether she has IRB approval for the treatment of pregnant women themselves with dexamethasone, a drug use understood to be experimental and off-label.”
He didn’t answer.