Galileo's Middle Finger

by Alice Dreger

  This, I think, may explain why I can’t find anything, in all the records, indicating actual enrollment numbers for pregnant women in IRB-approved trials of prenatal dex. Although she got IRB approval for fetal research as required to obtain NIH funding, in practice, New seems to have treated dex as “a standard procedure [performed] at the patient’s request.” Although OHRP would later tell us that “written informed consent (which included disclosure of risks) was obtained from subjects for participations in these studies,” I can’t find any consent forms that appropriately disclosed risks, and I can’t find any direct confirmation of enrollment of pregnant women into consented research for intersex prevention. All we have is Cornell’s word to OHRP that they looked into it and nothing had gone wrong. Recall that Cornell’s internal investigation in 2002 also said nothing had gone wrong, and when OHRP came and did a real site investigation, they found much evidence to the contrary.

  If Cornell’s IRB administrators had audited their pediatric researchers’ records prior to Sarafoglou’s complaints, they might have picked up that the pregnant women being named in New’s publications might not have given appropriately informed consent to research. But it appears that no such auditing occurred at Cornell’s Children’s Clinical Research Center (CCRC), a unit for which New herself served as director. Via FOIA, I have obtained extensive records and correspondence on IRB review of pediatric research at Cornell during this period, and I can find no reference to any audits—the kinds of inquiries that would have checked to see if people were actually signing consent forms—until Sarafoglou’s complaints. The extensive OHRP correspondence resulting from the 2003-2005 investigation also makes no mention of regular internal audits, as one would expect if such audits had occurred. And the previously mentioned angry whistleblowing memo specifically decried the complete absence of regular audits, seeming to confirm that no pediatric research at Cornell was being audited during this period to ensure appropriate consent practices:

  [New’s IRB protocol] like all the protocols performed in the CCRC has never been through a SAC [scientific advisory committee] review or an IRB audit . . . The Dean’s Office and IRB may indicate to regulatory agencies that Cornell has made “numerous corrective actions” or are “developing new procedures for the IRB” or “initiating an audit program” . . . But these are EMPTY WORDS that do not reflect the reality at Cornell. If Cornell really bothered to perform all the things they say they will when denying charges to regulatory agencies, don’t you think they would have at least properly audited a few protocols in the CHILDREN’S clinical research center, especially protocols begun before 1998.

  Whatever she told the mothers, New consistently led the NIH to believe prenatal dex was experimental. Odds are NIH wouldn’t have funded it if she’d told them it was standard of care. The portrayal of dex as experimental for grant-getting purposes didn’t change when, shortly after leaving Cornell, Maria New landed herself a position as a pediatric endocrinologist at Mount Sinai School of Medicine, across New York City. In 2006, in a research progress report to the NIH from her position at Mount Sinai, New assured the NIH that, even though she had left Cornell, her research on fetuses continued unabated: “The prenatal diagnosis and treatment program has galloped ahead so that we have now diagnosed and[/]or treated 768 fetuses.” She specifically referred to fetuses being “treated” with prenatal dex at Mount Sinai as part of her research progress. Because NIH funding requires proof of IRB approval for grant-funded experiments, New got a letter from the head of Mount Sinai’s IRB to give to the NIH, indicating Mount Sinai’s IRB had approved the study of “prenatal diagnosis and treatment.” This letter was signed by the same administrator who told the OHRP in response to our complaints that there was no fetal experimentation of this kind occurring at Mount Sinai.

  It is possible that what happened at Mount Sinai is what I think happened at Cornell: New obtained IRB approval for a study of prenatal dex to show to the NIH, as the NIH would have required to fund her prenatal dex studies, but pregnant women given dex for CAH virilization prevention were, at Mount Sinai, probably treated like patients, not asked to sign consent to fetal research. This seems to be confirmed by what that Mount Sinai administrator told OHRP in response to our complaints: “the decision as to whether to treat prenatally with dexamethasone or not was made by the patient and her physician,” not under the auspices of an IRB-approved trial. So even though New very specifically represented to the NIH that these pregnant women and fetuses were human subjects of her research at Mount Sinai, they were probably afforded no protection by an IRB, because they were “just” patients when they were offered dex.

  Some might wonder, how did New even manage to get hired at Mount Sinai after the NIH and OHRP investigations of her work at Cornell? Just after she had to relinquish her leadership positions and income at Cornell, New had written to the NIH asking for money to support herself under the NIH Merit Award System. New pleaded to her friends at the NIH that, “As I am no longer Chairman of Pediatrics, Chief of Pediatric Endocrinology, and Program Director of the CCRC, I have much more time to devote to the research,” and added, “This is a hard time for me and I am deeply appreciative of your consideration.” Noting that “her circumstances changed abruptly this year when she had to relinquish” her leadership positions “and the salaries entailed in these positions,” her colleagues on the NIH staff “enthusiastically support[ed]” giving her the award and writing her a big check. The NIH kept funding her. With a big active NIH grant, and with a brand new Merit Award to her name, Mount Sinai probably saw Maria New as the goose that would keep laying golden eggs. The NIH-Cornell fraud settlement was only announced after New was safely ensconced at Mount Sinai, and by then, OHRP had dropped its questions about her research ethics. So, without any official blemish on her name, she just moved over to Mount Sinai and “galloped ahead” on prenatal dex. Continuing to advertise the intervention as “having been found safe for mother and child,” she drew in more and more families to her relocated clinic. Because she had so many more CAH-affected research subjects than anyone else, the NIH kept funding her. According to one medical school administrator who looked at what I found (one to whom I happen to be married), Maria New seems to have invented a perpetual motion machine of NIH funding.

  I think it is fair to summarize what happened this way: The pregnant women given dexamethasone under Maria New’s guidance may never have understood they were in a high-risk experiment, and the NIH may never have understood they weren’t. Most of the pregnant women, like those who told their stories to the Time reporter, probably thought they were taking a drug that was safe and effective, not experimental. And the NIH, being shown by New the official IRB approval for her studies and being told about the “progress” of fetal treatment research, probably assumed that all the women were being actively enrolled in careful experiments that would be meticulously overseen by the medical schools’ IRBs. Because the medical schools’ IRBs appear never to have audited New’s practices—at least until the investigations at Cornell—they do not seem to have been keeping careful track of what was going on.

  It is quite possible that the problem of blurring the line between clinical care and human experimentation in Cornell’s pediatric unit went beyond Maria New. Internal minutes from Cornell’s IRB show that, after Sarafoglou raised the alarm, a member of the IRB who had simply been signing off on New’s annual renewals said this: “one of the issues the Program Review Committee is reviewing is the blurring of patient care vs. research and whether patients are being told which procedures are part of their care and which are part of the research.”

  • • •

  THAT’S WHAT I NOW BELIEVE happened with New’s work on prenatal dex. Given that I was able to figure all this out—given that I learned all this chiefly through obtaining government records through the Freedom of Information Act (FOIA)—why didn’t the government do something when we made our complaint in 2010? That is a tale in and of itself.


  Recall that it was late 2009 when my allies called me to help. In February 2010—not realizing there had ever been an OHRP investigation into New’s work on prenatal dex—Ellen Feder, I, and thirty of our colleagues appealed to the FDA and OHRP.

  About three months later, AJOB released the target-article manuscript by Larry McCullough and Frank Chervenak attacking us as “unethical” and “transgressive” for our calls for a federal investigation. FOIA confirmed that, as I had suspected, Larry McCullough had sent the AJOB manuscript to OHRP while that agency was deciding what to do with our complaints. In his cover letter, McCullough told OHRP, “We hope that your [sic] will take our critique into account in your deliberations.” None of the authors’ conflicts of interest were disclosed to OHRP. For the target article, McCullough listed Baylor College of Medicine as his only affiliation. As it turns out, he also had two other active, paid faculty positions: at the medical schools of Cornell and Mount Sinai. The Cornell chair of Obstetrics, Frank Chervenak, coauthor of the target article, did not mention to OHRP that he had served as “key personnel” on Maria New’s NIH grant.

  After receiving our letters of concern, the FDA gave the job of the prenatal dex investigation to a physician-ethicist named Robert “Skip” Nelson. As I figured out much too late to do anything about it, not only was Nelson serving on the AJOB editorial board (with Larry McCullough), but he was also actively negotiating a new AJOB journal editorship in chief for himself, for a new offshoot journal called AJOB-Primary Research. Just to be clear, he was in negotiations with AJOB while he was running the federal investigation that AJOB was being used to undermine. Nelson’s new editorship came with a fancy title and ten thousand dollars a year, to hire an editorial assistant. In the e-mail exchange in which he informed his ethics supervisors that he was planning to accept this position with AJOB, Nelson said he hoped it would allow him to get an adjunct faculty appointment at Georgetown University’s ethics unit, but never mentioned what role the journal was playing in the FDA investigation he had been given to run. Remember the “coincidence” of AJOB officially publishing the McCullough et al. target article just as the federal findings were made public? E-mails discovered via FOIA suggest that Nelson, by then working for both FDA and AJOB, was keeping track of when the federal findings would be released.

  Nelson’s behaviors would be less concerning if what he had represented in his official findings had been factually accurate. In one of the most important revelations of the whole federal investigation, in the name of the FDA, Nelson had told everybody that Maria New sought and obtained an IND (investigational new drug) exemption from the FDA in 1996 for a study using prenatal dexamethasone to prevent virilization in female fetuses. This would have meant that way back in 1996, a high-ranking physician at the FDA had reviewed the intervention and had formally decided that it was not risky enough to require significant FDA review before proceeding.

  This seemed almost impossible—the FDA giving a pass on an experiment meant to try to change fetal development, with seven of eight fetuses deriving no possible benefit? When Nelson made the claim in 2010, we all believed it. When I obtained a copy of the 1996 exemption letter, however, I found that it actually refers to a very different thing. It refers to a proposal from Maria New “to utilize dexamethasone to treat pregnant women with a [sic] congenital adrenal hyperplasia.” Based on this wording, it looks quite possible that what New was proposing to the FDA in 1996 was to use dexamethasone on pregnant women who themselves had CAH; women with CAH often need to have their disease treated during pregnancy, to maintain their own health, and dexamethasone can be used to treat the disease. In other words, the exemption was probably not, as Nelson claimed, for a study of “the administration of dexamethasone during pregnancy for the purpose of preventing virilization in females with congenital adrenal hyperplasia.” It appears to have been meant for a study of using dexamethasone during pregnancy for the purposes of maintaining the health of pregnant women who themselves had CAH. That’s presumably why the physician at the FDA didn’t flip out at what New had proposed to do without full FDA review. What was most likely being proposed wasn’t a fetal engineering experiment!

  When I contacted the physician who had signed the letter back in 1996, he had no recollection of the matter. But for his part, in 2010, Nelson knew full well there was no documentation to support his public misrepresentation of the letter. In internal communications, Nelson told OHRP staff that the single vague FDA letter was the only record of all this, so they could know no more than what the letter said. He explained to OHRP that any related documentation (which would have explained exactly what New had proposed) had been shredded during an FDA move. But in his 2010 memo to OHRP and the public, rather than providing what the IND exemption letter actually said—that New had been given an exemption to study “dexamethasone to treat pregnant women with a [sic] congenital adrenal hyperplasia”—he made it look as though the FDA had already reviewed prenatal dexamethasone for virilization prevention in 1996 and had found it to be nothing especially concerning.

  If OHRP had done a full investigation in 2010—if the agency had bothered to look up its own 2003 investigation at Cornell, as I later did—I think it would have found as much reason to be concerned as we had. Instead, an e-mail obtained via FOIA of the 2010 investigation shows the head of OHRP thanking Nelson for suggesting that OHRP rely on his work. That is what OHRP did—relied largely on Nelson for OHRP’s conclusions.

  OHRP was not alone in heavily relying on Nelson’s review. A couple of months after the federal findings emerged, AJOB published a piece by Maria New extensively quoting Nelson’s FDA memo and concluding, “The recent reports by the Office of [sic] Human Research Protections and the FDA therefore make crystal clear that my research on prenatal treatment of CAH is and always has been both legally and ethically proper at every level.” The article included no disclosure of the FDA official’s (Nelson’s) relationship with AJOB. When we asked AJOB to make this disclosure and to correct New’s article’s title to make clear that the federal government had not “vindicated” the intervention, the editors refused. Thanks to Nelson and the rest of the AJOB editorial team, the medical literature continues to include a supposedly peer-reviewed article that says that the FDA has vindicated prenatal dexamethasone for CAH.

  When we set out to petition the FDA, I felt sure that the agency would at least stop Maria New from advertising prenatal dexamethasone for CAH as having “been found safe for mother and child,” because the language seemed to be the kind you’re only allowed to use for FDA-approved drug indications. New is subject to no such limitation. Thanks to a loophole New is still driving a truck through, only two classes of people—employees of a drug’s maker and researchers with active FDA investigator status—are prohibited from advertising an off-label use as “safe and effective.” Because New doesn’t have appropriate FDA investigator status for this intervention, she isn’t subject to the prohibition. Under current regulations, if you unethically study a drug use, you can also unethically advertise it. As a Hail Mary play, I tried appealing to the “bad-ad” division of the FDA. The staff of that division never even acknowledged my letter.

  • • •

  SWEDEN REMAINS THE only place where prenatal dexamethasone for CAH has been studied using a prospective, long-term tracking approach with full ethics board oversight and informed consent throughout. Just two months after the U.S. agencies decided our letters of concern were unfounded, the Swedish clinical researchers studying prenatal dex were so alarmed by their results that they went to their ethics committee to tell them they would no longer provide the intervention in Sweden even if a woman was willing to sign up for a prospective clinical trial. They shut it down.

  The Swedish group continues to track those already dosed. In a study of forty-three children given prenatal dex for CAH, results show that in the quest to sex-normalize the females, boys who are “accidentally” exposed may be genitally and be
haviorally feminized. This study also confirmed earlier findings of impaired verbal working memory in exposed children who turned out not to have CAH. In terms of major side effects among the forty-three children tracked, researchers documented a total of eight “severe adverse events.” This is an astonishingly high rate, albeit one that requires a larger study to establish whether these problems were caused by the drug. The “events” included growth disorders like the one described by the American mother in the Time magazine article, a mood disorder requiring hospitalization, and three cases of mental retardation. (Just to be clear: 7 percent of the prenatally exposed children in this prospective cohort have mental retardation. That’s about ten times the normal rate.) In response to these findings, the Swedish team has declared it “unacceptable that, globally, fetuses at risk for CAH are still treated prenatally with DEX without follow-up.”

  In 2012, two years after we made our complaints to the feds, New’s research group, led by Heino Meyer-Bahlburg, published a retrospective convenience-sample study of prenatal dexamethasone for CAH. The group looked at cognitive function in sixty-seven children prenatally exposed, including eight CAH-affected girls and fifty-nine boys and CAH-unaffected girls, a tiny portion of the number of children New told the NIH she has had in her prenatal treatment studies. While the new study’s finding “contributes to concerns about potentially adverse cognitive after effects of such exposure,” in sharp contrast to the Swedish data, this study by New’s group found some positive outcomes in terms of brain function—i.e., the exposed children appeared to do better on some measures than children never exposed! The paper noted that the result was confusing, and considered what it might mean. Among those considerations was not the most obvious possibility: the sample was highly skewed.