The Nissen and Wolski meta-analysis [Nissen SE, Wolski K. Effects of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med 2007; 356, online here] was to be the first published article to combine data from all relevant clinical trials of rosiglitazone then available. Although two major trials of Avandia had been published, its manufacturer, GlaxoSmithKline, had performed many other smaller trials of the drug which remained unpublished. These results eventually appeared on a web-site run by GSK. However, this web-site was relatively obscure. It had not been created voluntarily, but in response to a settlement of legal action that alleged GSK had suppressed clinical research about its antidepresant paroxetine (Paxil). (See Steinbrook R. Registration of clinical trials - voluntary of mandatory. N Engl J Med 2004; 351: 1820-1822, link here and our post here).
Nissen and Wolski found the site, compiled the results of trials on Avandia it contained, and combined their results with those of the few published trials in their meta-analysis. It is to the credit of Nissen and Wolski that they were able to figure out how to do this. It is not to the credit of GSK that they sat on the data from these trials, only put it on this web-site when compelled to do so, did not make any effort to publicize the web-site, and did not publish a meta-analysis done by company scientists that showed qualitatively similar results to that eventually done by Nissen and Wolski (see post here).
I originally thought the case raised two questions: 1 - what are the benefits and harms of rosiglitazone as a treatment of Type 2 diabetes, and therefore for which patients under what circumstances should this drug be used? 2 - what barriers have prevented physicians and patients from getting the best possible answer to the first question, and what can be done about them?
But the immediate response to the Nissen and Wolski meta-analysis was a spin cycle that seemed to obfuscate these questions, and the answers to him (see posts here, here, here, and here)
In the last few weeks, several articles about Avandia have appeared in the media, and in medical journals, that reconsider the issues, and in particular, the "Avandia Spin Cycle." First was a commentary in the European Heart Journal by Dr Steven E Nissen, the first author of the meta-analysis [Nissen SE. The rise and fall of rosiglitazone. Eur Heart J 2010: published online here.] Dr Nissen provided a narrative history of the Avandia case from his viewpoint, and summarized it thus:
What were the key mistakes and lessons learned from the rosiglitazone affair?
1. The FDA rushed to approve rosiglitazone because of hepatotoxicity concerns about troglitazone, resulting in failure to consider the ‘signals’ suggesting cardiovascular toxicity.
2. An early critic of rosiglitazone was intimidated by company representatives and effectively silenced, a process antithetical to the principals of open scientific discourse.
3. Although early warnings were issued for the risk of heart failure, these warnings went largely unheeded in the face of aggressive marketing and promotion suggesting cardiovascular benefits.
4. No well-designed cardiovascular outcome trials were ever conducted for rosiglitazone, despite evidence suggesting increased cardiovascular risks. The only cardiovascular outcome trial was an open label study driven by a soft endpoint (hospitalization) with low adherence to randomized medications, seriously underpowered, and not completed until 10 years following launch.
5. Although both the FDA and the company were aware of evidence of an increased risk of adverse cardiovascular outcomes, certainly by 2005, neither warned physicians nor the public.
6. When a meta-analysis of rosiglitazone was eventually submitted for publication, the company subverted the editorial review process by stealing a copy of the manuscript and used this advance knowledge inappropriately to unblind an ongoing randomized trial.
7. Approval of diabetes drugs based exclusively upon their glycaemic effects has been short-sighted and scientifically unwise. Drugs that lower blood sugar may have other adverse effects that overcome any inherent benefits.
8. The failure of 50 other PPARs, many for adverse cardiovascular effects, went largely unreported because of negative publication bias. Such knowledge might have warned the medical community about the potential risk of these agents.
Meanwhile, the New York Times published conclusions of some internal US Food and Drug Administration (FDA) reports, and of an investigation by the US Senate Finance Committee. FDA safety officials suggested that Avandia should be taken off the market. Some key points from the Senate investigation were:
The bipartisan multiyear Senate investigation — whose results are expected to be released publicly on Monday but which were also obtained by The Times — sharply criticizes GlaxoSmithKline, saying it failed to warn patients years earlier that Avandia was potentially deadly.In combination, the Nissen commentary, and the Senate report nicely summarized what I called the "Avandia spin cycle" above. It appears that research was suppressed, marketing was deceptive, and critics and whistleblowers were intimidated.
'Instead, G.S.K. executives attempted to intimidate independent physicians, focused on strategies to minimize or misrepresent findings that Avandia may increase cardiovascular risk, and sought ways to downplay findings that a competing drug might reduce cardiovascular risk,' ....
Finally, Dr Harlan Krumholz published a commentary about the thiazolidinedione drugs (a group that includes rosiglirazone) in Circulation Cardiovascular Quality and Outcomes [Krumholz HM. A perspective on the American Heart Association Presidential Commission Advisory on thiazolidenedione drugs. Circ Cardiovasc Qual Outcomes 2010; 3 available online, link here], and an op-ed in Forbes. The latter sums it all up nicely:
I want to believe in America's pharmaceutical companies. I want to believe that people in these companies believe that the best strategy for success is to do what is best for patients. I want to believe that they are interested in scientific truth and eager to know of any safety issues and ready to share that information with the public.
This week I was disappointed again.
Over the years GlaxoSmithKline ( GSK - news - people ) has repeatedly reassured the public about the safety of its blockbuster diabetes drug Avandia. But this weekend the Senate Finance Committee released a report revealing that inside the company Glaxo's own experts and advisors were raising concerns about whether the drug could cause heart problems all along.
The report, based on more than 250,000 internal documents, provides a rare and unsettling glimpse into the decision by company executives to deflect safety issues--even as their own experts agreed with conclusions of outside researchers who were warning the public about possible harms.
The documents reveal that company researchers were deeply concerned about the cardiovascular safety of the drug as far back as 2003. The pages of the Senate report read like a spy novel: Glaxo receiving confidential documents leaked by a sympathetic academic who consulted for the company; the company embarking on a campaign to intimidate critics who warned about potential safety issues with the drug; and executives pulling strings to release data early from a scientific study that was supposedly controlled by an 'independent' committee of researchers.
The story here is less about the drug--the Senate report breaks no new ground about Avandia's safety issues (even among experts there remains some controversy)--and more about the ethical behavior of a company. What is clear: Glaxo failed to disclose its own concerns even as it sought to discredit outside researchers who were raising questions about the drug.Policy Implications
This type of behavior is eroding the public trust in the pharmaceutical industry.
Finally, Dr Krumholz concludes with some solutions with which I heartily agree:
The fix is simple: Once a drug is approved, all data relevant to drug safety should be placed in the public domain and independent investigators across the country should be able to use it. There should be big financial penalties for withholding relevant information. Drug studies sponsored by industry must be truly independent--outside of company control. Companies should give outside investigators independence over every aspect of the study. There are too many examples of companies wresting control of clinical studies from their consultant investigators for reasons that seem more related to product promotion than clinical science.Here in the US, the debate over health care reform has reached a new phase. Today the President is hosting an attempt to get bipartisan discussion of reform going again. Now that there is a new opportunity for discussion, I submit that would be health care reformers ought to think about the underlying causes of our continuing problems with rising costs, declining access, stagnant quality, and demoralized health care professionals.
And on all sides there should be a commitment to protect against the intimidation of academics who are willing to raise questions about the safety and effectiveness of company products. The free flow of information about the effects of drugs and medical devices will best serve the public's interest.
One set of causes that is rarely discussed has to do with the distortion of clinical research by commercial research sponsors with vested interests in the results turning out in favor of their products, and the further distortion of clinical discourse and decision making by the deceptive marketing practices of these same organizations. If we started introducing steps like those suggested by Dr Krumholz into health care reform, maybe we really could decrease costs, increase access, improve quality, and renew professionalism.
*Bring Back the DSI?
To lighten things up a bit, and to explain the title...
When I was in school in the 1960s, I first became a fan of the Hardy Boys mysteries, then the slightly more mature Tod Moran series. While on the Neiuw Amsterdam (the 1938 version) one day out of Kingston, Jamaica, I viewed Dr No, the first James Bond movie, a sophisticated spy thriller, not a spoof like some later entries in the series. I later read all the classics, like those by Agatha Christie, Earl Stanely Gardner, Rex Stout, Ellery Queen, etc. So it should be no surprise that I, like many other schoolboys and girls, collaborated to form an amateur detective and intelligence agency. But this was the sophisticated 1960s, so we called it the Department of Scientific Investigation (DSI). We skulked around New York City and its suburbs looking for spies, saboteurs, and common criminals. Luckily, we never really found any. As more realities intruded, I gave up any ideas about becoming an investigator or intelligence agent (although I still read mysteries, spy novels, and thrillers to this day, but mainly on airplanes.) Once I became a physicians, those days in the DSI seemed very remote (although the title does now seem like it ought to be part of the NIH ;-) ).
However, after reading Dr Krumholz's article above, maybe we need to resurrect the DSI. At least, maybe we health professionals ought to consider how we can become better and more organized watchdogs for the kind of problems revealed in the Avandia case, and in many other cases discussed on Health Care Renewal.
But if any of my old DSI colleagues want to set up a reunion, it would be fine with me.
ADDENDUM (25 February, 2010) - see also comments by Dr Howard Brody on the Hooked: Ethics, Medicine and Pharma blog.