Editorial Note: This post is by Johanna Ryan. As with all posts by Jo, it unearths angles on current stories that everyone else seems to have missed.
A column here last month followed the legacy of Study 329 into the present. By taking apart one 2015 study of Vraylar, a new antipsychotic, I tried to show that clinical research in 2015 is even more ghost-written, and more tightly controlled by industry, than that infamous 1990’s study. I had planned to follow up with a look at what the official authors and contract researchers involved in that Vraylar study were up to now.
Then Robert Califf was nominated to head the U.S. Food and Drug Administration, and I recognized the real Ghost of Research Future. He was right there on Capitol Hill in a white coat and a winning smile, holding forth to politicians and the media at his Senate confirmation hearing.
Two facts about Robert Califf are beyond question. He is an expert on clinical trials, who is already seen as a leading architect of the future of medical research. And as the New York Times put it, he has “deeper ties to the pharmaceutical industry than any FDA commissioner in recent memory”. A lot of senior figures in medicine support Califf in spite of his ties to Pharma. The guy is just so bright, and understands the nuts and bolts of drug research so well! Surely a person like this is more useful than some outsider who offers only a squeaky-clean resume, they argue.
Other researchers and healthcare activists are not so sure. They point to clinical trials run by Califf and his Duke colleagues for Xarelto, a new anti-clotting drug, that were flagged by FDA reviewers as being biased in Xarelto’s favor. The studies were funded by Johnson & Johnson, and the concerns were serious enough that two senior reviewers voted against Xarelto’s approval. Former FDA cardiovascular analyst Thomas Marciniak called Califf “one of the architects” of a broken clinical trial system run by drug companies, a feat for which he should be “held accountable, not appointed to run the FDA.”
Most of the senators at his confirmation hearing shared a rare bipartisan enthusiasm for Califf, but a few had their doubts. Vermont’s independent senator Bernie Sanders, currently running for president as a Democrat, wants the FDA to take on prescription drug price-gouging. He doubts Califf is the man for the job. Senator Elizabeth Warren thinks the FDA must insure new blockbuster drugs like Xarelto are actually effective and safe. She has asked Califf for a look at the terms of his contracts with the drug companies that funded his research.
The Duke Clinical Research Institute (DCRI) will likely be Senator Warren’s first stop. Califf cofounded the DCRI and was its first director. Since then he’s helped launch a dizzying array of Institutes, think tanks and public-private partnerships: the Clinical Trials Transformation Initiative, the Clinical Research Forum, the Duke Translational Medicine Institute, which he headed just prior to joining the FDA, and more. But the DCRI is his signature achievement, his baby, so to speak. And if Duke University was the mother and Califf the midwife, the father was Research Triangle Park, the Silicon Valley of corporate medical R&D just down the road from Duke.
Back in the 1990’s, the problem posed by Study 329 was medical school researchers letting their work be funded and co-opted by industry. By 2015, things were worse. It was common to see these same researchers signing on to clinical trials designed and controlled by industry from the start, with minimal input from their universities. The Vraylar studies were an example of med school faculty lending their prestige and credibility to work that was actually done in dozens of commercial Contract Research Organizations (CROs) far from the campus, and written up by medical communications agencies hired by the drugmakers.
Robert Califf’s DCRI is a new hybrid: a CRO built right into the university, but whose mission is to solve whatever research questions matter to its paying customers, like any other CRO. DCRI describes itself as:
“… the only one of its kind that can offer all the services of a commercial contract research organization (CRO) with the academic credibility and expertise of an academic research institute. From planning to execution to publishing results, the DCRI excels at every facet required for a speedy, effective research project. Our unique operational model ensures that all aspects of a project account for both our dedication to patients and the business needs of our sponsors.”
Some of this research is done at Duke. Most, however, takes place at outside research sites recruited and managed by the trial experts at DCRI. In cardiology, DCRI touts its ability to run global commercial megatrials in dozens of countries.
DCRI also provides “strategic communications”: everything from crafting effective journal articles and conference presentations, to designing CME programs that make sure prescribing doctors absorb the key results of the sponsors’ research. It connects companies with “thought leaders” who “are uniquely positioned to understand the operational, financial, and regulatory implications of numerous project designs, to the great benefit of our clients.” It even offers clinical-research classes delivered directly to corporations through DCRI Learn: “Let DCRI customize an experienced-based learning program for your employees. Proposals are confidential and programs individualized.”
It’s hard to imagine an industry-sponsored project Duke faculty members could sign up for that would not “fit within the mission of the university”. Yet in 2005 “Faculty Connection, LLC” was formed to help professors do just that, with Robert Califf on its board of directors. FC negotiates contracts (and aims to limit confidentiality clauses to “five years or less”), takes care of required conflict-of-interest reports, and “shields consultants from personal liability” through its LLC structure, as this 2011 video explains. In return it charges a 25% overhead fee which it says is often paid by industry; if not, it takes 20% of the faculty member’s earnings.
Most of Califf’s personal consulting fees for work with industry appear to have come through FC. If so, the “nonprofit” to which he famously donated his fees would usually have been DCRI itself, through its fellowship programs for promising young Duke researchers.
“Bias in choosing the question is a much bigger problem than lying about the data.”
– Robert Califf, February 12, 2013
Xarelto is one of several new anti-coagulant drugs approved in recent years to prevent strokes and other clot-related complications, especially in people with heart-rhythm disorders such as atrial fibrillation. All aim to replace a much older, cheaper and more well-known drug called warfarin.
Anti-coagulation or “blood thinning” is a trick that needs to be done just right. It must work well enough to prevent formation of clots that can lead to strokes or heart attacks. But if it works too well, even minor internal bleeding can become life-threatening. That’s why warfarin requires careful dosing, limiting certain foods and alcohol, and periodic tests to make sure your blood level is in a therapeutic range. It’s also why people with milder cardiac conditions are often advised to take aspirin instead, which is less effective but simpler to use.
What drugs like Xarelto and Pradaxa promised was anti-clotting action as good as or better than warfarin, but in a convenient once-daily dose with no need for monitoring. That was Dr. Califf’s message when he presented the ROCKET-AF trial to the American Heart Association in 2011. Even then, he faced pointed questions about conduct of the trial, especially in the non-US trial centers. Warfarin doses for subjects in the control group were so poorly controlled that their blood levels were often outside therapeutic range. That set them up for a higher risk of strokes – and also made Xarelto look better than it was.
Cardiology expert Steven Nissan considered it a “fatal flaw” – and added that those more cynical than he might think the trial committee had planned it that way. The FDA reviewers agreed, pointing out that warfarin treatment in the ROCKET-AF trial was worse than either the Pradaxa trials or older studies. Patients in the U.S. who had access to good quality warfarin treatment might actually be in more danger, not less, on Xarelto.
ROCKET-AF had been billed by Califf and Duke as more convincing because it was a “real-world” trial, in which warfarin doses would be left to clinicians’ judgment rather than subject to a strict protocol. This could also lead to trouble if local trial sites had an incentive for sloppy use of warfarin, or trial designers had an incentive to choose second-rate clinicians. Xarelto would be “safe and effective” – but compared to what? Bias in choosing the question would have the same effect as rigging.
The reviewers’ next worry was a problem Xarelto shared with Pradaxa: the convenience of once-daily dosing with no blood tests, which just might be too loose to guarantee safety. Prior to the trial, the FDA had recommended that Xarelto be tested in a twice daily dose, due to its short half-life, but the ROCKET-AF group used the once-daily dose anyway – just like the trials of competitor Pradaxa. Critics charged that the makers of both drugs had sacrificed safety for the marketing value of “convenience” as an incentive to put more patients on the expensive new products.
There were two more potential safety problems: Stopping Xarelto had led to 22 strokes in 30 days, compared to just six in the control group – even though the Xarelto group were transitioned onto warfarin. It seemed quite possible that stopping Xarelto put patients in a “prothrombic state”. This could leave them in a dangerous bind if they needed to stop Xarelto, in order to have surgery for example.
Finally, unlike warfarin, whose effects can be reversed by vitamin K or plasma, there was no antidote for drugs like Xarelto and Pradaxa. In an emergency, patients could bleed uncontrollably and doctors might be unable to stop it.
To your family doctor, the lack of an antidote might sound scary – as it did to the FDA reviewers. However, developers of Xarelto insisted there was no cause for concern. A 2014 paper co-authored by Robert Califf pointed out that “all-cause death after bleeding was similar in the rivaroxaban arm compared to warfarin” in their study. It also noted that “although physicians may express a desire for specific ‘reversal agents’, these interventions are not used in the vast majority of bleed events,” and the reversal agents for warfarin were “sub-optimal” anyway. Reading that, physicians might think their worries were overblown after all.
So they might have been surprised to hear that DCRI was already working on an antidote to both Xarelto and Eliquis, made by a company called Portola Pharmaceuticals – with Robert Califf on the board of directors. In 2014 Portola announced that FDA had granted Fast-Track status to its antidote, andexanet-alfa, opening the way for faster approval with less evidence. Fast Track is reserved for drugs that address a life-threatening condition for which no approved treatments exist – and uncontrolled bleeding in Xarelto or Eliquis users qualified.
It wasn’t until January 26, 2015, when Califf joined the FDA as interim deputy commissioner, that he announced he would be retiring from Portola’s board and selling his shares. Fortunately this was after Portola’s January 9 announcement of the success of its Phase III trials, which sent its stock price soaring. (Califf also sold off a “significant” investment in N30 Pharma, a startup developing drugs for cystic fibrosis.) Duke scientists continue to work with Portola on another anticoagulant called betrixaban, and with Perosphere Pharma on PER977, an antidote to edoxaban which may reverse other anticoagulants as well. (Meanwhile, Boehringer-Ingelheim gained fast-tracked FDA approval for Praxbind, an antidote for the “life-threatening” effects of its own drug, Pradaxa. A single Praxbind injection will cost somewhere in excess of $3,000 wholesale.)
This wasn’t Califf’s only foray into anticoagulants and antidotes, or the business end of drug development. In March 2014 a similar fanfare greeted the announcement of fast-track status for Regado Bioscience’s anti-clotting drug and antidote combination, dubbed REG1 and REG2, for use in stenting and other cardiac procedures. A small number of serious allergic reactions in the Phase 2 trial three years earlier were not enough to slow the progress of Regado’s potentially life-saving therapy. Regado was known as a Duke spinoff, and Robert Califf had served on its its scientific advisory board since 2009, joined by DCRI colleagues Thomas Povsic and Robert Harrington.
A few months later, however, those allergic reactions were judged serious enough to shut down Regado’s Phase III study. By early 2015 the company had ceased to exist. Both Povsic and Harrington joined former Regado CEO David Mazza at Caladrius Biosciences, whose cell-based treatment for ischemic cardiac damage may well come up for approval during Califf’s tenure. Portola’s and Perosphere’s antidotes are sure to do so, and perhaps Portola’s betrixaban as well.
Names like Portola, Regado, Caladrius and Perosphere have seldom come up in discussions of Califf or DCRI. These companies don’t turn up in a search of the Sunshine Act database, which only tracks payments on behalf of drugs that are already FDA-approved. Often these days it’s established Pharma giants that will market the drugs and deal with any problems they cause, while the startups that handled the clinical trials may fade from view.
It’s unlikely the Senate would consider a seat on a pharma board to be a black mark for Califf, and certainly there is no shame in working on a drug like REG-1 that didn’t pan out. However, the results of such “failed” trials are just as important as successful ones, and sometimes more so. In recent years Califf has spoken and written extensively on the need for transparency, data sharing and the importance of trial registries like ClinicalTrials.gov. He’s also been asked to serve on both government and private task forces to fix a troubled clinical trial system.
So it’s worth asking: What about the trials for Regado’s REG-1 and REG-2? The initial good news about the drugs was discussed in several journal articles, but no results were ever posted on ClinicalTrials.gov. The results on the 3232 subjects in the terminated trial were never posted, and the registry says only that the trial is on “clinical hold.” The outcome was reported at a March 2015 cardiology conference, but not in any articles to date. A similar shortage of results seems to plague Portola, as well as DCRI studies of an Alzheimer’s drug and its work with Perosphere.
The stakes could not be higher. In the coming year the US Senate will complete work on a total overhaul of the drug approval process, known as the “21st Century Cures Act.” Faster trials, lighter regulation and a bigger role for industry, all Califf watchwords, are central to the bill. The FDA has also approved two expensive new biologic drugs to manage cholesterol (PCSK-9 inhibitors), whose safety and real-world impact on health need to be closely watched. Califf served as advisor to the makers of both.
The fight for open data access, in the wake of Study 329 and countless other scandals from Vioxx to Avandia, may be even more important. Here again, the news is strange. Bristol-Myers Squibb has announced its own “transparency” initiative in which an Independent Review Committee will oversee sharing of clinical trial data with selected researchers. The “independent” committee will be set up by the DCRI, and consist entirely of Duke faculty.
It’s all part of a Strategic Partnership between BMS and Duke which was publicly celebrated at this 2014 event featuring Califf. The brochure has all the feel of a wedding announcement. Should any doctor, even one as wise as Robert Califf, be asked to monitor and report on his own spouse?Share this: