Monday, October 09, 2006

The Conflicted Defending the Conflicted: Infectious Disease Specialists' Attack on the LA Times Goes Awry

We have posted frequently, (e.g., here and here), about conflicts of interest affecting top US National Institutes of Health (NIH) scientists and leaders. Two months ago, we posted about the case of Dr Thomas J Walsh, Head, Immunocompromised Host Section, Pediatric Oncology Branch, Center for Cancer Research and the National Cancer Institute (NCI). A Los Angeles Times investigation by David Willman had suggested that Dr Walsh, while working full-time for the NIH, had received money from several pharmaceutical companies, and also had had spoken for these companies' products at several US Food and Drug Administration (FDA) meetings.

The Los Angeles Times article (no longer available from the Times, but online here) elicited an extraordinary response. Last month, 109 (one hundred nine) authors wrote a rebuttal in Clinical Infectious Disease. (Full citation: Anaissie EJ et al. Clinical research in the lay press irresponsible journalism raises a huge dose of doubt. Clin Infect Dis 2006; 43: 1031-9.) "To provide a case study for how inaccurate and dangerous the mainstream press can be if articles are not carefully written, as well as to correct inaccuracies and defend honesty in research, we offer our counterpoint to a recent article that questions the various systems of checks and balances that govern the conduct of clinical trials and implicitly accuses one of our infectious diseases colleagues of unethical conduct in 2 clinical trials."

The article by Anaissie et al focussed on two concerns, "selective enrollment of patients and picking doses of antifungals to bias outcomes" of two clinical trials Walsh lead, published in 1999: [Walsh TJ et al. Liposomal amphotericin B for empirical therapy in patietns with persistent fever and neutropenia. N Engl J Med 1999; 340: 764-71]; and in2004: [Walsh TJ et al. Caspofungin versus liposomal amphotericin B for empirical antifungal therapy in patients with persistent fever and neutropenia. N Engl J Med 2004; 351: 1391-1402.]

According to Anaissie et al, "The first layer of purported deceit implicitly alleged by Willman's article involved the selection of the type of trial that would have the highest likelihood of securing FDA approval." Walsh had originally advocated testing the drugs on patients with fever and suspected, but not proven fungal infections. Willman suggested that one reason for this was that such patients are more numerous than those with proven fungal infections. Thus, studying the former rather than the latter patient group would speed up completion of trials, and hence possibly marketing of drugs. Anaissie and colleagues made an apparently reasonable clinical argument for empiric anti-fungal treatment in patients with persistent fever and low white blood cell counts despite conventional antibiotic therapy. Curiously, the corroboration they cited came from two guidelines both published after both of Walsh's trials were designed.

Again, according to Anaissie et al, "The second and more serious of Willman's implicit accusations is that Walsh deliberately chose to administer lower, less-effective dosages of comparator drugs in the 2 trials." Their argument in support of the doses of the comparison drugs used by Walsh et al was based on two points. First, the dosage of the comparison drug used in Walsh and colleagues' earlier study was based on "prior studies of empirical therapy that established the safety of this approach." Second, "no evidence of superior outcomes associated with higher dosages .. has ever been published."

Having demolished these straw men, Anaissie et al concluded in a devastating manner,

The Los Angeles Times has failed its readership on all counts. The destructive nature of Willman's implicit allegations and the strong rebuttals made by several investigators and by Walsh's superiors several months before the publication of Willman's article should have prompted Willman's editors to scrutinize carefully the quality of his 'evidence.' Their failure to do so calls into question the credibility of the Los Angeles Times as a serious newspaper. Accordingly, the editors bear responsibility, together with Willman, for this publication. One might ask at what point reporters and editors cross the line of ethical reporting. If anything, this sad chapter should give the public a huge dose of doubt about this newspaper.

Sensational attacks on a respected academic and government employee (and, implicitly, on the entire drug-approval process) and fear-mongering addressed to the lay reader (implying that individuals should enroll in clinical trials at their own risk) may be attention-grabbing ways to sell more newspapers; for this purpose, Walsh served as a convenient scapegoat.

The greatest danger of articles such as Willman's is that members of the lay public do not read medical journals. By contrast, the Los Angeles Times is widely read, is disseminated online, and is perceived as an authoritative news source. Accordingly, there is good reason to fear that the public will conclude, on the basis of Willman's article, that the entire process of drug development in the United States and abroad is corrupt and that they should refrain from participating in clinical trials. We question whether Willman and the Los Angeles Times considered the possibility that future patients might suffer as a result of Willman's irresponsible report.

Now who is fear mongering? I would readily concede that the methodologic issues raised by Willman about the two trials lead by Walsh are complex and debatable. In fact, when writing our first post on the subject, I decided not to dwell on them for those reasons. However, Willman's article raised other issues that Anaissie et al virtually ignored.

These were the questions about Walsh's conflicts of interest. Willman had reported:
  • Walsh made favorable comments about an anti-fungal drug made by Merck & Co, Inc to an FDA panel, and in an editorial in the New England Journal of Medicine, while he was consulting for Merck and had served on an expert panel for the company. Walsh did not disclose his relationship with Merck to the the NEJM.
  • Walsh also appeared in front of an FDA panel in support of an anti-fungal drug made by Fujisawa USA (now Astellas Pharma Inc) while he was a consultant to Fujisawa.
  • Walsh presented information to yet another FDA panel in support of another anti-fungal drug made by Pfizer Inc while he was a consultant to Pfizer.
Willman noted that, "U.S. conflict of interest law generally prohibits a federal employee from representing anyone before a government agency, regardless of whether outside compensation is paid."

Yet Anaissie et al attempted to finesse this issue thus, "Willman further attacks Walsh on the inappropriateness of his advice to regulatory agencies." They argued,

It is entirely appropriate for a principal investigator and data review committee chair to provide advice to both the pharmaceutical industry and the FDA, particularly when he happens to be, like Walsh, an accomplished investigator with almost 600 peer-reviewed scientific publications and service on numerous scientific advisory boards. Being a federal employee does not disqualify Walsh from providing such advice; he is as qualified to do so as any other academic investigator with similar expertise.

The issue, of course, was not whether Walsh should have provided just advice to pharmaceutical companies and the FDA. The issues were, instead, whether it was ethical, or lawful, for Walsh to provide advice to FDA advisory panels while he was working both for the NIH and for the drug companies who made products that those panels were considering, and whether it was ethical for Walsh to be working for drug companies as a consultant while he continued to work for the NIH on projects related to their products. Anaissie et al did not even acknowledge these central concerns raised by Willman's reporting, much less attempt to refute them.

In fact, a few days after the article by Anaissie et al was published, Willman wrote another Los Angeles Times article about the results of the NIH internal investigation of Walsh. This noted that Walsh had received over $100,000 from multiple drug and biotechnology companies, and that Walsh had been getting consulting fees from Merck at the same time he took the lead for the National Cancer Institute (NCI) on a formal collaboration between the NCI and Merck. The report dismissed Walsh's assertions "that his reputation is sufficient to dismiss any questions about his impartiality." The report noted "at least 38 separate instances where he chose not to follow agency procedures. He actively chose not to adhere to policies because it was inconvenient or time-consuming; he knew it was likely his participation [with the drug companies] would have been disapproved. His actions reflected negatively upon the agency." Finally, it asserted, "Dr. Walsh has engaged in serious misconduct, in violation of the Department's Standards of Conduct Regulations … and federal law and regulation."

Although Anaissie et al may have not been aware of these findings when they wrote their article, they were left in the position of defending someone now accused by the NIH of violating federal laws and regulations.

Anaissie and colleagues' article just received a deserved scolding by Ivan Oransky, deputy editor of The Scientist, in the Boston Globe,
Unfortunately, the Clinical Infectious Diseases screed's omission of Walsh's financial relationship with Merck and other companies is consistent with how clinical researchers seem to feel about disclosure.
Message received: Researchers know best. Please stop bothering us with these silly rules and your sensationalist investigations. Sorry, neither Willman's irresponsible report nor the government's alleged heavy-handedness in trying to protect the public are going to hurt patients. What is going to hurt clinical research and patients is the continued lack of transparency by a number of researchers. Like most Americans, I'm a strong believer in the importance of research, clinical trials, and the role of drug companies in improving the public health. But I'm also a strong believer in disclosure. When someone hides something, I want to know why. If there's nothing wrong with financial relationships between researchers and drug companies, why not disclose them?
Olansky makes good points. The NIH internal report certainly suggested that some researchers may believe that their scientific or clinical abilities exempt them from having to deal with "silly rules," including those about conflicts of interest. However, the attempted defense of Walsh by Anaissie et al makes a larger point.

A full one and one half pages of the print version of the Anaissie et al article are taken up by disclosure of its authors' "potential conflicts of interest." Of the 109, approximately 54 disclosed that they worked as consultants to and/or for speakers' bureaus of pharmaceutical and biotechnology companies. Approximately 28 worked for Astellas Pharma US, 35 for Merck & Co, and 44 for Pfizer Inc. (Many worked for more than one company; some worked for more than 10 at a time.)

Thus, isn't it clear why this group had a problem grappling with, even talking about, the allegations of conflicts of interest affecting Walsh's work for the federal government, conflicts arising out of Walsh's paid work for Fujisawa (later Astellas Pharma US), Merck & Co, and Pfizer Inc?

What Anaissie et al succeeded in doing was reminding us all how pervasive conflicts of interest are in academic medicine. They also reminded us that the problem with conflicts of interest is that the people who have them are, well, conflicted. As Joe Collier recently wrote in the British Medical Journal, "people who have conflicts of interest often find giving clear advice (or opinions) particularly difficult." And it is unclear whose interests conflicted peoples' opinions serve.

I think Anaissie et al have demonstrated the glaring need for tough regulations requiring disclosure of all financial (and probably political) conflicts of interest affecting anyone who can make decisions, for individuals or organizations, in health care. Basically similar regulations should apply not only to government agencies and academic medical institutions, but also all other health care organizations, not-for-profit and for-profit included.


If you work in a health care organization, push for internal regulation. Write your congressperson about the need to regulate and discl0se conflicts of interest in health care.

No comments: