Fertility clinics across the U.S. are prescribing a medication with a seriously concerning safety profile and no proven benefits.

Alessandro Bianchi/Reuters

When Susan Manning, a 39-year-old woman just a few weeks into her first pregnancy, wrote to tell me she had been put on the steroid dexamethasone to prevent a miscarriage--and to ask whether she should be worried about taking this drug--at first I could not even process what she was saying. Dexamethasone is known to cross the placental barrier and impact fetal development, so the very idea of first trimester exposure sets off warning bells. Besides, dexamethasone is not known to help in preventing miscarriage. Susan's story sounded too crazy to be true.

It also sounded too close to the history of DES (diethylstilbestrol). From the 1940s through the 1970s, some doctors gave pregnant women DES, a synthetic estrogen, to try to prevent miscarriage. In spite of clinical evidence that it didn't work as intended, millions of fetuses were exposed in utero before doctors discovered that prenatal DES exposure could lead to infertility and deadly cancers. Just last week, Eli Lilly & Co. settled a suit brought by four sisters who believe their breast cancers were caused by prenatal DES exposure.

But as it turns out, Susan (a pseudonym) had it right. Women like her, pregnant by virtue of in vitro fertilization (IVF), are today routinely put on dexamethasone for miscarriage prevention at some IVF clinics. Susan is being treated abroad at a high-profile clinic, but some American infertility clinics also advertise this off-label use of dexamethasone as if it is the standard of care.

I inquired about this with Dr. Geoffrey Sher, Executive Medical Director of the Sher Institutes for Reproductive Medicine, a high-profile infertility practice with offices across the country. He confirmed in an email that, "We recommend 0.5 mg - 1.0 mg. [of dexamethasone] orally daily (dosage varies based upon individual patient needs) from the time of initiating the [IVF] cycle through to the tenth week of pregnancy." He could not point to studies showing that dexamethasone helps prevent miscarriage, but argued, "Since there are so many other variables that are involved" in IVF pregnancies, studies "would virtually be impossible to do."

Lucas Jackson/Reuters

In essence, doctors using dexamethasone for miscarriage prevention are working from a physiological hunch. The hunch is that some women who seek out IVF have immune problems that cause their bodies to reject pregnancies. If dexamethasone suppresses a woman's immune system, maybe it will help her maintain a pregnancy. But, again, there's no scientific evidence that dexamethasone prevents miscarriage, and no evidence that this drug--a drug known from animal and human studies to have the potential to change fetal development--is safe to use in this way.

Surprisingly, it appears that even women like Susan, with no diagnosable immune disorder and no history of recurrent miscarriage, are being put on dexamethasone for miscarriage prevention by some IVF specialists.

I asked Dr. Ralph Kazer, Chief of the Division of Reproductive Endocrinology and Infertility at Northwestern University's Feinberg School of Medicine, to give me his thoughts on this use of dexamethasone. (We work at the same medical school, although Dr. Kazer and I have never communicated before this.) Dr. Kazer expressed concern, saying, "The extent to which early [pregnancy] losses are due to immunological problems is controversial, but it is almost certainly a relatively rare problem."

I asked Dr. Kazer if he knew of studies of efficacy or safety of this off-label (non-FDA-approved) use. The answer was no. "This is not complicated," Dr. Kazer wrote. "In the absence of good evidence for efficacy or a very specific medical condition, a drug like dexamethasone should not be given to pregnant women in the first trimester."

He explained why: "Dexamethasone is a Category C drug, which means that there are concerning animal data [about safety] but no good data in humans regarding teratogenicity [i.e., birth defects caused by a drug]. Such drugs should only be used when potential benefits outweigh potential risks. There is currently no good evidence that this criterion is fulfilled for typical IVF patients. Dexamethasone does cross the placenta, so at the very least, it reaches the fetus."

At Brown University's Alpert Medical School, Dr. Philip Gruppuso is a pediatric endocrinologist with research interests in the fetal origins of adult health and disease. When I told Dr. Gruppuso about this use of dexamethasone, a steroid in the class of drugs called glucocorticoids, he responded, "Regardless of the rationale, the first trimester use of glucocorticoids should be viewed as an experimental treatment that could have long-lasting untoward effects. Evidence from animal studies raises the possibility that glucocorticoids can alter the fetal epigenome." In other words, prenatal synthetic glucocorticoid exposure could permanently change the way a person's genetics will operate over his or her lifetime.

In fact, the reason Susan had found me was because I have been tracking the history of a case in which prenatal dexamethasone has been used with the specific intention of permanently altering the course of fetal development: some doctors have been using dexamethasone on pregnant women at risk for having a child with Congenital Adrenal Hyperplasia (CAH). An endocrine disorder, CAH can cause female fetuses to develop ambiguous genitals and other atypical sex anatomy. "Dex" is believed to be effective at reducing the odds that a female fetus with CAH will be born virilized.

The off-label use of dex for CAH has been very controversial--and not only because it may impact a developing child's body permanently in unpredictable and unintended ways. (Rarely does a drug do only what you want it to do.) Because the intervention has to start before fetal genitals differentiate into male or female types, doctors have had to expose fetuses before they could know whether the fetus really is a female with CAH. As a result, almost 90% of the fetuses exposed via their genetic-carrier mothers' ingestion of dex have not even been females with CAH.

Medical societies' consensus statements on the matter have consistently concluded that this use should be restricted to prospective, long-term, controlled trials, with full ethics board oversight. But this use of dexamethasone has occurred almost entirely outside the kinds of clinical trials that could establish whether it is safe and effective.

Based on an NIH report, thousands of pregnant women in the U.S. appear to have undergone dex intervention for CAH completely outside of prospective clinical trials, in spite of the American Academy of Pediatrics in 2001 having issued this warning: "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 and thalidomide demands no less."

John Sommers II/Reuters

Almost a decade after that warning, in 2010, a systematic review of the literature on prenatal dex for CAH still could find no high-quality evidence of efficacy, and no data at all on metabolic outcomes in the children exposed. A report on the systematic review noted, "because dexamethasone prenatal treatment is relatively new, no offspring have yet reached middle age where many problems can be expected to present." Everyone in the field understands that to be an uncomfortable nod to the history of DES, in which middle age has brought a new round of cancers to those exposed in utero.

The only place where prenatal dex for CAH has been studied the right way--in a prospective, controlled, long-term trial--is Sweden. In mid-2012, the Swedish research team reported that they had gone to their ethics board to tell them they were shutting down their trial use. The Swedes have so far found 8 "severe adverse events" among 43 children exposed in utero. Problems among those exposed include mental retardation (3 cases among the 43 children), memory problems, and growth disorders.

Not all those problems may have been caused by dexamethasone exposure, but the Swedish group concluded, "until larger and more conclusive studies are published, we do not consider it ethical to initiate further treatment." They added, "We find it unacceptable that, globally, fetuses at risk for CAH are still treated prenatally with DEX without follow-up." Meanwhile, the American Journal of Obstetrics & Gynecology has just published a review of the animal and human literature that concludes the intervention should not be used.

The doses of dex in IVF cases appear to be slightly lower and to start much earlier than those for CAH; the two interventions appear clinically different. But the CAH findings give plenty of reason to be concerned about any first trimester use, especially outside of prospective trials. It is worth remembering that the terrible effects of DES were found almost by accident, when doctors became aware of a cluster of young women with a very rare form of vaginal cancer. A mother of a girl with the cancer pushed doctors to consider the possibility that the cancer came from her daughter's prenatal exposure to DES, and that mother turned out to be right.

When they were giving DES to pregnant women, doctors didn't understand how DES could affect fetal development. They found out the hard way. By contrast, obstetricians use dexamethasone for CAH--and also sometimes use dex on fetuses at risk of being born prematurely--specifically because they know it can impact fetal development.

What about Dr. Sher's claim that IVF treatment is just too complicated to allow a study of dexamethasone for miscarriage prevention? The infertility specialist Dr. Kazer replies, "Such a study could certainly be designed, although it might be difficult to convince an IRB [i.e., ethics board] to approve it, given the lack of biological plausibility regarding potential efficacy."

What did Susan's doctor tell her about efficacy and safety when he put her on dexamethasone? Nothing until she pressed him, several weeks into the "treatment." At that point, she says, "The doctor had the nerve to tell me that they think the benefits outweighed the risks, but maybe I don't! I've never been party to the discussion. How dare they make that decision on my behalf."

She wrote to me, "I can't understand why nobody is even questioning the use of this drug; in some respects I wonder whether women who go through IVF end up having such low self-esteem that they don't have the confidence to question what the doctors say. They are 'grateful' that they are being treated and therefore don't dare put a foot out of line. I know for a fact that is the situation in my clinic; you never question [the doctor] and if you do create a fuss, then he won't treat you again." She added, "It's a very disempowering relationship."


Kari Christianson, Program Director of DES Action USA, was astounded when I wrote to tell her that dexamethasone is being used for miscarriage prevention. She wrote, "what strikes me most about the horror of dex use is the vulnerability of not only a developing fetus in pregnancy, but also of a mother (and a father) wanting to make the best decisions about a hoped-for child. But again and again, unproven and unsafe drugs are available, offered and given to pregnant women without fully informed consent or understanding at a most vulnerable time. It is unconscionable. To think that we have learned little or nothing in the over 40 years since DES health harm was brought to light is frightening beyond all reason."

I asked Dr. Sher what he tells women he is putting on dexamethasone. He answered, "We tell our IVF patients that in our opinion it is by and large safe to take as prescribed, that there are no proven developmental risks to the baby [and] that side effects to the woman are infrequent, temporary in nature and reversible on proper withdrawal." He went on, "We do not go into detail describing everything in the literature on what we consider to be safe treatments. That would be overly time consuming, impossible to accomplish thoroughly and comprehensively[,] and in my opinion [would] create unnecessary patient consternation."