Tuesday, July 11, 2006

Conflicts of Interest at Stanford: the Mifespristone Studies

Paul Jacobs' series on conflicts of interest at Stanford University also included a detailed case study.

Its subject was Dr Alan F Schatzberg, Chair of the Psychiatry Department at the Medical School. Dr Schatzberg is currently running a federal grant on mifepristone as a treatment of depression. Dr Schatzberg has previously been the senior author of two related articles:
- Belanoff JK, Rothschild AJ, Cassidy F, DeBattista C, Baulieu E, Schold C, Schatzberg AF. An open label trial of C-1073 (mifepristone) for psychotic major depression. Biol Psychiatr 2002; 52: 386-392. (Link here.)
- Flores BH, Kenna H, Keller J, Solvason HB, Schatzberg AF. Clinical and biological effects of mifepristone treatment for psychotic depression. Neuropsychpharmacol 2006; 31: 628-636. (Link here.)

The first article disclosed that the study was supported by a grant from Corcept Therapeutics, and that Schatzberg has "a financial interest" in the company, as did other authors.

The second article disclosed support from the National Institutes of Health. It further noted, "Dr Alan Schatzberg is cofounder of Corcept Therapeutics and is the only author involved in this resubmission who has any financial interest in the company. However, Dr. Schatzberg continues to be full-time faculty at Stanford University. Dr. Schatzberg played no direct role in the recruitment, assessment, or follow-up of subjects enrolled in this study. Dr. Schatzberg was not directly involved in the analysis of data stemming from this research."

Nonetheless, Jacobs' article showed that Schatzberg's ties to Corcept Therapeutics were extensive:
With the backing of Silicon Valley biotech investors, Schatzberg and [Dr Joseph K] Belanoff [a research fellow in psychiatry] in 1998 founded Corcept Therapeutics.

Belanoff became the company's full-time chief executive officer. Schatzberg took a seat on the board of directors and a part-time post as chairman of the company's scientific advisory board, a job that now pays him $60,000 a year. He and his family were granted 3 million Corcept shares for $1,000 -- today worth nearly $12 million.
After the company was formed,
Schatzberg decided he could no longer play a direct role in RU-486 studies in depression. But he remained principal investigator with overall responsibility for NIH grants that have paid for ongoing Stanford research on the drug Corcept hopes to market.

Schatzberg says he helped design the RU-486 studies being conducted on campus but leaves it to others to recruit patients, administer treatments and analyze results. Monitoring committees at NIH and Stanford review the studies to ensure patient safety.

However, Schatzberg acknowledges that he has considerable influence over the junior faculty members doing the studies. As chairman of psychiatry, he helps set their salaries and can affect their career advancement. And he continues as a co-author of the resulting papers.
After the study reported by the first paper above was performed,
on the strength of these early findings, Corcept had raised $29 million from private sources. Four months later, the results of the five-patient pilot study were published in a scientific journal, claiming 'a rapid reversal of psychotic depression.'

Just before Christmas that year, Corcept filed plans with the Securities and Exchange Commission to raise as much as $90 million in an initial public stock offering. While the company waited to issue stock, Schatzberg, Belanoff and others reported in the journal Biological Psychiatry the results of the larger Corcept-sponsored study of RU-486 in 30 patients at six academic centers, including Stanford.
However, some strongly criticized this study.

Bernard Carroll recalls first reading the Biological Psychiatry study in his home-office in Carmel, where he runs his non-profit institute, the Pacific Behavioral Research Foundation.

'The spin was unmistakable,'' Carroll said. His reaction, he said, was 'sadness and dismay at the low standards evident in the report and in the editorial commentary.'' That commentary, written by the journal's editors, described the work as a possible 'paradigm shift in the treatment of depression.'

Later that afternoon, he got a call from his friend Robert Rubin, then a professor of psychiatry at Drexel University and Allegheny General Hospital in Pittsburgh. The two had become self-appointed guardians of scientific rigor in psychiatric research -- gadflies who periodically fire off salvos to journals to complain about papers that don't measure up to their standards.

They sent such a letter to Biological Psychiatry. They criticized Schatzberg for claiming the drug 'may be the equivalent of shock treatments in a pill.' They accused the authors of 'obfuscation' and the editors of ignoring the study's 'fatal flaws.'

Schatzberg and Belanoff responded, saying the letter was an 'ad hominem attack,' and that Carroll and Rubin had 'cut and pasted snips of quotes . . . and mis-characterized what we said.'

Weighing what to do with this unusually heated exchange, the journal's editors sent the correspondence to four outside, anonymous reviewers, who split, two and two, on whether the letter and response should be published. In the end, the editors decided not to print the exchange, setting the stage for Carroll and Rubin to go public in Puerto Rico [at the 2004 meeting of the American College of Neuropsychopharmacology].

[There] Schatzberg's critics publicly unveiled their critique of his work in the form of a poster -- really a scientific paper in outline form, pinned to a board in an exhibit hall.

The contents were explosive.

The poster, by Drs. Bernard J. Carroll and Robert T. Rubin, systematically picked apart the conclusions in three published studies of RU-486 in depression, two by Schatzberg and his colleagues, one by an independent group.

But in a departure from the usual give and take of scientific debate, the poster quoted positive public statements about the drug by Schatzberg and other researchers and juxtaposed those statements against the individual's financial interest in Corcept.

There, for example, was Schatzberg saying RU-486 may be 'the equivalent of shock treatments in a pill' and a statement pointing out that he owned 3 million shares of Corcept stock. The point was hard to miss: Researchers with a financial interest were expressing 'considerable enthusiasm' for a treatment of questionable effectiveness.

The poster was seen by many of the top brain scientists in the country, people whose opinions matter deeply to Schatzberg as well as his critics. Carroll, former head of psychiatry at Duke University and now semi-retired in Carmel, recalls that people came up to him during the session, slapped him on the back and said, 'It's about time somebody said this.'

But he and Rubin soon learned that other members, Schatzberg among them, were furious. A lawyer for Corcept says the ACNP admonished Carroll and Rubin for their conduct. Both Carroll and Rubin say they were faulted for making their criticisms public, not for the content of the presentation.

Even now, Schatzberg can hardly contain his anger. 'Those that are critical,' he said, 'ought to be careful about impugning others.'

Two independent experts asked by the Mercury News to review the three key RU-486 studies co-written by Schatzberg side with Carroll and Rubin.

University of California-San Francisco Professor of Medicine Stanton A. Glantz, the author of 'Primer of Biostatistics,' found several statistical errors in the papers he was asked to look at.

'These are elementary statistical methods,' he said. 'They have applied them incorrectly.' The result was to 'bias their results toward reporting an effect when the data doesn't justify that. Scientists shouldn't make these dumb mistakes.'

The other expert is Steven G. Self, a professor at the University of Washington, who heads the program in biostatistics and biomathematics at the Fred Hutchinson Cancer Research Center in Seattle.

Commenting on the 2002 paper in Biological Psychiatry, he said, 'there is no evidence at all' for a meaningful difference in the response of patients given high doses of RU-486 and those given a very low dose presumed to have no effect at all. All you have to do is look at the data, he said, to see this. 'No formal statistics are required.'

Schatzberg said that applying statistical analyses to these small, preliminary case studies is 'absolutely silly.' And Corcept CEO Belanoff agrees.
I will add a few technical comments. The 2002 Biological Psychiatry paper was an "open-label trial" in which all 30 patients got some dose of mifepristone. It had no control which received placebo or another drug. Given the unpredictable course of depression, it is impossible to tell whether any differences seen in groups receiving different doses of the drug were due to effects of the different doses, or the natural variability in the course of disease across patients. Many authorities on evidence-based medicine would hugely discount studies of drug or other treatments that did not randomize patients to the treatment of interest or some comparison treatment. (See, for example, Guyatt GH, Sackett DL and Cook DJ. Users' guides to the medical literature. II. How to use an article about therapy or prevention. A. Are the results of the study valid? JAMA 1993; 270 2598-2601. )

The study is so small, moreover, that any differences could simply have been due to random noise. Thus, the study's conclusions that "mifepristone appeared to be effective" seem to go beyond the evidence provided.

Furthermore, the disclosure in this article seems not to convey the degree of financial involvement of Dr Schatzberg in Corcept Therapeutics.

The important investigative reporting by Paul Jacobs in the Mercury News gives us an important example again of how pervasive are conflicts of interest among leaders of academic medicine. Unfortunately, the pervasiveness of such conflicts means that everyone now ought to be skeptical about in whose interests academic medicine now speaks and acts.


Anonymous said...

The July 11, 2006 Wall Street Journal has a front page article outlining unreported financial ties in a JAMA article on pregnant women and depression, along with chilling incidents of birth defects associated with drug therapy. This article is to long to summarize, but worth the read.

The most chilling statement was that the doctors involved saw no reason to mention their financial ties to the companies since they did not view this as an issue.

Complicity theory? Scary, definitely.

Steve Lucas

Anonymous said...

Great post. The back-slapping ACNP attendee had it right: it's about time people start doing this.

I really must admit to some surprise, however, that this fellow has managed to raise so much money for his company on the basis of these piffling reports. An uncontrolled open label study of 30 odd patients? Good grief. In the world of EBM that ranks slightly higher than toilet paper.