Friday, July 25, 2008

MANAGING CONFLICTS of INTEREST at STANFORD

MANAGING CONFLICTS OF INTEREST AT STANFORD

The case of Stanford University and Dr. Alan Schatzberg, chairman of Stanford’s department of psychiatry, continues to raise questions. You can see previous discussions here. The questions concern transparency at the academic-corporate boundary, reporting conflicts of interest to NIH, and Stanford’s “management” of a faculty member known to have a significant conflict. The conflict involves a company Dr. Schatzberg founded (Corcept Therapeutics), a drug called mifepristone that Corcept has in clinical trials for depression, and NIH-supported studies of the same drug at Stanford.

Stanford’s position is that Dr. Schatzberg “has not been involved in managing or conducting any human subjects research involving Mifepristone …” Dr. Schatzberg’s 2006 published disclaimer stated “…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.” Stanford represented that this disclaimer applies also to earlier publications with Dr. Schatzberg as co-author.

This disclaimer is hardly credible, considering the responsibilities of NIH-funded Principal Investigators. I pointed out many of the inconsistencies before. Now there is new evidence that Dr. Schatzberg failed to maintain an arm’s-length relationship to the projects at Stanford.

In a 2008 review article, Dr. Schatzberg discussed the Stanford projects in ways that contradict the claim of an arm’s-length relationship. This article acknowledged Dr. Schatzberg’s NIH grant support at Stanford. Corcept Therapeutics was not acknowledged as a source of funding. The first concern is that, if Dr. Schatzberg’s relationship to the Stanford studies is as Stanford claimed, then he has no business publishing a NIH-supported review article that portrays his drug’s prospects in a favorable light. Hello! Is there a conflict of interest here? Review articles that assess a field and synthesize data form a crucial part of science that has to be off-limits to Dr. Schatzberg just as much as assessing patients in one of his clinical trials would be. His many favorable, even exaggerated, articles, reviews and commentaries since he founded Corcept should have come under this proscription. So much for Stanford’s “management” of the conflict. Dr. Schatzberg certainly had a role in managing the research supported by NIH at Stanford – he managed the climate of scientific opinion for his drug and he managed the tone of the NIH-supported publications from Stanford.

Second, Dr. Schatzberg made a claim of efficacy for his drug that differed from what was originally reported in an NIH-supported Stanford study. He claimed a 31% decrease of symptom ratings with a scale called the BPRS. The original report does not confirm that claim. From the published data tabulated in the report, the reduction of symptom severity was 20%. Readers can easily check that for themselves. So, in a current scientific review article Dr. Schatzberg deviated from the published record. He also inflated by half the efficacy estimate for his drug. Hello! Is there a conflict of interest here?

Third, this false claim indicates that Dr. Schatzberg performed and published his own reanalysis of the primary data from an NIH-funded Stanford study. He didn’t get the number 31% from the published article. Yet Stanford says he had no part in managing or conducting the research or in analyzing any data. So here we have an NIH-funded Principal Investigator, with a clear conflict of interest, who supposedly remains at arm’s length from the project, accessing the primary data files, running a new analysis himself, and publishing an exaggerated new efficacy result for his drug that does not match what he published previously as a co-author. That is inconsistent with Stanford’s defense of Dr. Schatzberg. Hello! Is there a conflict of interest here?

Fourth, in this review article Dr. Schatzberg presented the first data on mifepristone blood levels in a peer reviewed journal, along with an elaborate scientific argument for the importance of the blood level as a moderator of response. These blood level data came from Corcept’s clinical trials, but the target blood level stated by Dr. Schatzberg did not correspond to SEC filings and press releases from Corcept. The administrative issue here is how Stanford justifies the presentation and discussion of original scientific data by a Principal Investigator who is supposedly insulated from the scientific work of the project in order to avoid bias. Dr. Schatzberg’s discussion of the new data and his scientific arguments about blood levels form a crucial part of the scientific platform for his drug’s current prospects, and as such must be off limits, just as assessing patients in one of his clinical trials would be off limits under Stanford’s policy. Hello! Is there a conflict of interest here?

This current example undercuts the assertions by Stanford and Dr. Schatzberg that his conflicts of interest have been “managed” by the University. Dr. Schatzberg may not have assessed any patients in Stanford’s trials of mifepristone, but he has had the lead role in responding to scientific critiques, where he clearly was the manager. He also has had the lead role in selling the mifepristone story to the scientific community and in shaping the tone of the NIH-supported Stanford publications that Corcept relied on to raise capital. Hello! Is there a conflict of interest here? How does Stanford justify these academic-commercial boundary violations, and why does NIH not act on the known conflicts of interest? That was Senator Grassley’s question to Dr. Zerhouni today.