Intended for healthcare professionals

Feature Drug Regulation

Rosiglitazone: what went wrong?

BMJ 2010; 341 doi: (Published 06 September 2010) Cite this as: BMJ 2010;341:c4848
  1. Deborah Cohen, features editor, BMJ
  1. Correspondence to: dcohen{at}

    Over 10 years after the diabetes drug rosiglitazone was approved by regulators, and despite studies on tens of thousands of people, questions remain about its cardiovascular safety. An investigation by the BMJ looks at why this happened.

    It was, as one Food and Drug Administration (FDA) adviser put it, a “perfect regulatory storm”—a combination of problematic data, uncertain clinical need, politics, and poor drug company behaviour.

    Now, 10 years after its approval by regulators in the United States and Europe, the widely prescribed blockbuster diabetes drug rosiglitazone may be about to fold. Two months ago, in July 2010, the FDA convened a 33 member expert advisory panel to decide whether it should be withdrawn from the market in the light of evidence that it may increase the risk of myocardial infarction. Earlier this year a US Senate finance committee report had detailed concerns about the paucity of evidence to support the use of rosiglitazone and about the way in which the drug was evaluated and licensed.1

    At the advisory meeting, members of the public heard a damning analysis of the RECORD trial, commissioned by the European Medicines Agency (EMA) when it approved the drug in 2000 in order to determine its safety. Millions of prescriptions later and with the drug still on the market around the world, this trial and other post-marketing surveillance have failed to resolve the concerns.

    To date, the FDA and the EMA have decided that the drug is safe enough to stay on the market. But the story reflects badly on almost everyone involved: the regulators, the manufacturer, GlaxoSmithKline, and the clinical community. It has also raised a host of questions. Why did the regulators accept such poor evidence on benefit and safety for rosiglitazone? Did GlaxoSmithKline mislead the regulators? Should the drug have been licensed in the first place and should it now be withdrawn? Why haven’t patients in the UK and Europe been made aware of the concerns about rosiglitazone’s effects? And is the current drug regulatory system up to the job?

    The FDA meeting was held in the open, in front of a packed audience including the world’s media. Ahead of the meeting, the FDA published the 765 page report circulated to panel members.

    This is far removed from the secrecy shrouding proceedings at Europe’s regulator, the EMA. The BMJ has talked to a range of experts close to the European regulatory process and submitted a series of Freedom of Information requests to the EMA, but we still have no clear picture of why, after initial rejection in October 1999, the EMA gave market authorisation to rosiglitazone in July 2000 in the absence of new evidence. Neither have doctors and patients been told that in July the UK’s Commission on Human Medicines— in an unanimous vote— advised the Medicines and Healthcare products Regulatory Agency (MHRA) to withdraw the drug. In a statement, the MHRA has confirmed that the evidence now suggests that the risks associated with rosiglitazone outweigh the benefits and “that it no longer has a place on the UK market.” But a “dear doctor” letter sent to UK doctors in July advised doctors to “consider alternative treatments where appropriate.”2 The MHRA said that it used the information provided by the Commission on Human Medicines to push for a UK withdrawal as part of the Europe-wide review by the European Medicines Agency.

    Clinical implications

    According to John Yudkin, emeritus professor of medicine at University College London and endocrinologist; “No new patients should be started on rosiglitazone, and patients already taking it should be reviewed and alternative treatments considered. Those at higher risk of heart disease should be advised to stop taking the drug.”

    Rosiglitazone is one of two available glitazones known to reduce blood sugar and was heralded as a much needed new approach to improving outcomes and long term complications, including cardiovascular disease, for people with type 2 diabetes. Of the other glitazones, troglitazone was withdrawn in 1997 in the UK and in 2000 in the USA because of hepatotoxicity; pioglitazone remains on the market as a competitor to rosiglitazone.3

    A paucity of evidence

    Concerns were expressed early on about the paucity of evidence to support rosiglitazone’s use. According to documents obtained under the Freedom of Information Act, advisers to the EMA noted the lack of good evidence during its approval process.

    In comments sent to the EMA approval meeting in 1999, one expert adviser noted that without a long term study with hard primary endpoints it was not clear whether rosiglitazone would have any beneficial impact on cardiovascular disease. This adviser also questioned whether you could put a drug on the market without these long term data and was unconvinced that rosiglitazone in combination therapy offered any advantages over what was already available—metformin and sulphonurea combined, or insulin.

    Another adviser pointed out that safety problems were evident in the data presented by GlaxoSmithKline (then SmithKline Beecham) and asked the panel whether they should postpone approval until better data were available.

    Silvio Garattini, a member of the EMA panel in 2000 that approved the drug and director of the Mario Negri Institute for Pharmacological Research, told the BMJ that the documentation presented for approval was initially poor and that the studies were of a relatively short duration. The initial decision to reject the drug was overturned despite there being no new evidence, he says.

    When rosiglitazone was approved, even clinicians who were nominally supportive of the drug remarked about the poor evidence base and lack of long term clinical trials.4

    After approval, three EMA panel members—whose names were redacted from the minutes sent to the BMJ—remained concerned that “the long term risk/benefit of rosiglitazone is still unknown and that there are several safety concerns.”

    The EMA told the BMJ that the principal safety concerns at the time were that rosiglitazone induced weight gain, with possible serious cardiovascular effects; the induction of anaemia; and that rosiglitazone raises blood lipids, although this effect was of unknown clinical relevance.

    Why did the EMA let it through?

    Given these concerns and the lack of good evidence, why did the EMA approve rosiglitazone?

    Garattini said that rosiglitazone is “an example of what happens for drugs that have large commercial interest such as the antidiabetic drugs.”

    When appealing against a decision not to approve their drug, Garattini says pharmaceutical companies bring forward opinion leaders who are obviously favourable. These paid advisers give presentations to the regulators and companies turn to them whenever an oral presentation is required.

    On receiving the negative opinion Jan Leschly, chief executive officer of SmithKline Beecham, told the press it was “a temporary setback,” adding that “in the coming months we will be working with the committee to address their concerns. We are confident that by the end of March we will have demonstrated Avandia’s unique benefits in the treatment of type 2 diabetes to the CPMP [Committee for Proprietary Medicinal Products].”5

    According to Edwin Gale, a diabetologist and adviser to European regulators, in the years before rosiglitazone’s approval diabetologists were also putting pressure on the regulators and clamouring to use this new class of drug. Some of this clamour was fuelled by pharmaceutical analysts touting its blockbuster potential, which at the time they said was crucial to SmithKline Beecham’s future growth.6 7

    “There were tremendous expectations about Avandia—partly because scientifically it was extremely interesting. It was a whole new model of the way a drug could act. It affected the body and its energy metabolism in totally new ways that were very interesting and fascinating,” said Gale.

    One anonymous member of the EMA committee that approved rosiglitazone in 2000 told the BMJ that he had been contacted by respected members of the diabetes community to urge him to approve the “wonder” drug, something he had not experienced before.

    Speaking at the European Association for the Study of Diabetes (EASD) annual conference in September 1999, diabetologists urged the use of rosiglitazone as a first line treatment for type 2 diabetes.

    “Unlike most traditional drugs for type 2 diabetes, rosiglitazone works in a novel way to reduce insulin resistance, helping the body’s own insulin work more effectively and offering patients improved glycaemic control, as measured by fasting plasma glucose. The hope is that this will slow long-term deterioration,” Dr David Matthews of the Oxford Diabetes and Endocrinology Centre told the meeting.

    But it was this new way of working—the stimulation of genes that acted on more than blood glucose—that would account for its adverse effects. This should have led to greater caution in the regulatory process, says Gale.

    In 1999 in the United States, pressure was on the regulators to fast track rosiglitazone to provide a safer alternative in the same class of drugs as troglitazone.8

    For example, in the weeks leading up to the FDA’s initial approval meeting in 1999, the American Diabetic Association affirmed the importance of glitazones on its website. “These drugs have mechanisms of action—working directly on insulin resistance—that are not shared by other drugs in other classes for treating type 2 diabetes. These drugs offer new options to health care professionals who treat people with type 2 diabetes and represent important advances in drug therapy,” it said.9

    Was there a need?

    Before the final approval, the EMA committee discussed whether there was an unmet need among the treatment options currently available for diabetes and whether a niche indication could be appropriate. Minutes of the meetings show that this suggestion won out. The committee decided that rosiglitazone could be used in combination with other oral antidiabetics as a second line treatment in certain circumstances.

    However, Garattini, who was on the panel, was not convinced. “There was no need for another antidiabetic drug—there are so many already that are more or less the same,” he told the BMJ. And as one adviser from the FDA said at an advisory meeting, it’s for the regulators to protect public health and not to equip physicians with a broader array of medicines for clinical choice.

    The RECORD trial

    Confronted with weak evidence and an appeal from GlaxoSmithKline to reconsider its decision to reject the drug, the EMA’s committee discussed whether additional clinical trials should be required before or after marketing authorisation. In the event the committee approved the drug with the requirement for two additional studies once the drug was on the market. The first was study 211, a double blind trial of the effect of rosiglitazone on cardiovascular structure and function in patients with type 2 diabetes and chronic heart failure. And the second was a randomised trial of six years’ duration with the composite outcomes of cardiovascular hospitalisation and cardiovascular mortality—the RECORD trial.10 11

    According to Bo Odlind, EMA rapporteur at the time of the SmithKline Beecham’s first attempt to get Avandia approved in 1999, the EMA committee believed it was important to do a cardiovascular outcome study rather than one looking only at surrogate endpoints such as haemoglobin A1c—although he is critical of the RECORD trial’s open label unblinded design.

    And Odlind isn’t the only person to be critical. In an internal memo, Thomas Marciniak, a drug approver for the FDA, wrote: “We did not review the protocol [of the RECORD trial] prior to study implementation. If we had, we would have judged it to be unacceptable”11

    It seems GlaxoSmithKline knew this. A company slide show cited in a US Senate finance committee report noted that the RECORD trial did not provide sufficient data to test for cardiovascular safety. It also noted that GlaxoSmithKline was trying to create studies to counter the PROactive study on rival drug, pioglitazone (Actos) that Takeda planned to release.1

    In a statement to the BMJ, GlaxoSmithKline said that this was not the case. RECORD met its primary endpoint. “The study confirmed its primary hypothesis. It showed that cardiovascular hospitalisation or cardiovascular death (which includes heart attack, congestive heart failure, and stroke) was not statistically different between the two groups after an average of 5.5 years of therapy.”

    However, the EMA told the BMJ that it acknowledged weaknesses in the trial, including a low event rate in a high risk population of patients with diabetes, a high loss to follow-up, and the open label design of the study.

    “The RECORD study was designed some 10 years ago. Since then the design of post marketing studies has evolved,” a spokesperson for the EMA said.

    Problems with post-marketing surveillance

    But even without the flawed design, was commissioning a trial after approval to resolve safety concerns the right approach? Garattini says that regulators request additional trials after approval to overcome a stalemate when concerns about toxicity exist. “It is not the best way because you need a long time to do the study and meanwhile the drug remains on the market. By the time the study is finished the drug patent is finished so there is no inconvenience to the company. This is what happened with sibutramine as well, which was eventually withdrawn,” he says.

    Garattini is concerned more broadly about the current reliance on drug companies to perform post-marketing surveillance. “The EMA has never produced a document indicating the percentage of fulfilment of such commitments,” he says. “For the FDA, it’s about 30%.” The BMJ has asked the EMA to say how many companies carry out their post-marketing surveillance commitments to be told that it has never published a comprehensive report on post-marketing commitments. But a study on the number, type, and status of post-authorisation studies requested by Committee for Proprietary Medicinal Products (CPMP) for centrally authorised products in the year 2007 until 2010 is pending.

    Agencies swamped

    Another flaw in post-marketing surveillance is that regulators sometimes fail to act on safety information appropriately when they are given it. In 2004, with increasing numbers of people taking rosiglitazone, signals of adverse events were picked up by WHO. They sent GlaxoSmithKline an alert about cardiac disease. GlaxoSmithKline conducted a meta-analysis and confirmed an increase in cardiac events to the FDA and the EMA in 2006.

    The FDA has been accused of sitting on the reports and not sufficiently alerting the public.12

    “These data were kept from the public because of acceptance of the proprietary nature of companies’ trial results, even when they concern the safety of marketed drugs—a proposition that badly needs public debate and reconsideration,” Jerry Avorn, professor of medicine at Harvard University, told the BMJ.

    Labelling in Europe by the EMA was updated shortly afterwards to reflect the results of the GlaxoSmithKline meta-analysis, with cautions about cardiovascular risk.13

    Sales still booming

    Safety concerns didn’t seem to hit sales of the drug, and in early 2006 GlaxoSmithKline won approval for a combined product of rosiglitazone and glimepiride in both the US and Europe.14 Rosiglitazone remained in wide clinical use. At the end of 2006 and beginning of 2007, the sales of Avandia were up.15 16 It was GlaxoSmithKline’s second biggest drug, making an estimated $3billion per year.17 And it was outselling its rival, pioglitazone.18

    Beginning of the end?

    But in 2007, the fortunes of rosiglitazone began to change with the publication of a meta-analysis of GlaxoSmithKline’s study reports by Steve Nissen and Kathy Wolski in the New England Journal of Medicine.19 It claimed that rosiglitazone was “associated with a significant increase in the risk of myocardial infarction” compared with placebo or other antidiabetic regimens. When the adverse effects of the drug became widely known in the wider medical community, sales halved.17 20

    Nissen’s ability to access study reports arose out of a court case in New York. As part of a settlement with the state over GlaxoSmithKline’s nondisclosure of possible heightened suicide risk among teenagers taking antidepressant paroxetine (Paxil) the company had to put all its recent clinical studies on a website.

    Whatever the criticisms of this particular meta-analysis were, it allowed academics to scrutinise the study summaries.

    “It’s important to realise what an important role publicly available trial results data played in the rosiglitazone story. Having this information posted on the GlaxoSmithKline website made it possible for Steve Nissen to perform his critical meta-analysis published in NEJM in May 2007, which really ‘broke the case wide open’ on this matter,” says Jerry Avorn, professor of medicine at Harvard Medical School.

    “Requiring the posting of clinical trial results on should help provide a warning system for other drug risk issues in the future, as will the growth of FDA’s new Sentinel system for post-marketing surveillance. The Avandia case provides compelling evidence of the vital necessity and importance of both of these developments,” he added.

    FDA rules allowed greater scrutiny

    For those with access only to the medical literature, unpicking the evidence behind rosiglitazone has not been easy. Bristol University diabetologist, Edwin Gale—who was chair of the EMA’s scientific advisory group on diabetes—complained in 2001 in the Lancet how little data on rosiglitazone had been placed in the public domain. Company documentation at that time cited five confidential files, 13 abstracts, and four papers (two of which were clinical) for rosiglitazone.3

    The regulators have access to more information. Since the 1950s, FDA rules have required drug companies to turn over all individual patient case reports from their clinical studies, not just the statistical summaries but the reports that permit re-analysis of how each case was coded.

    It was the availability of these case reports that allowed Thomas Marciniak, an FDA medical officer, to scrutinise the RECORD trial. He was asked in October 2009 to review the cardiovascular events in the trial.

    GlaxoSmithKline’s own analysis showed that 321 (14.5%) of patients treated with rosiglitazone compared with 323 (14.5%) controls experienced either cardiovascular death or hospitalisation. In essence, rosiglitazone did not cause more cardiovascular problems than metformin or sulphonylureas.

    The EMA accepted these findings when it received the findings of the completed trial in 2009. “No difference in the number of adjudicated primary endpoint events for rosiglitazone (321/2220) versus active control (323/2227) (HR 0.99, CI 0.85-1.16) was observed” the EMA’s scientific summary said.13 It took the FDA to thoroughly analyse the trial, which raises questions about the EMA’s ability to oversee post-marketing trials.

    The FDA had the individual case reports to do a more thorough job, but not the resources. In house statisticians could not wade through the voluminous RECORD trial dataset—running to 1438 pages for one patient, and several hundred pages for most of the other 4500 patients. But Marciniak analysed 549 patient case report forms, including 278 from the rosiglitazone group and 271 controls. Of these 549, 100 were a random sample and the rest had been the subject of adjudication disputes.

    Marciniak found problems that cast doubt on GlaxoSmithKline’s analysis, which he detailed his damning report for the July FDA advisory committee.

    His concerns were over study design, confounding variables that could have biased the overall findings, and the conduct of the study. He considered that these limit “any reassurances that RECORD can provide regarding the CV safety of rosiglitazone.”1

    Much of his concern hinged on data ascertainment in patients who stopped the trial or were lost to follow-up. Marciniak detailed 11 different conduct problems including failure to refer events to the board for adjudication, missed endpoints, insufficient collection of information, and issues with data handling.

    Specifically, he detailed eight failures to refer patients for adjudication—all in those taking rosiglitazone. “The eight cases weren’t the only problem cases I found. I classified the problem cases into categories and the eight were ones of ‘failure to refer for adjudication.’ Of 549 patient case report forms reviewed I found 70 serious problems, four to one favouring rosiglitazone,” he said in an interview.

    GlaxoSmithKline says there was no wrongdoing. In a statement to the BMJ, a GlaxoSmithKline spokesperson said “An inspection of the RECORD study by the FDA concluded that there was no evidence of systemic or pervasive findings that would undermine the reliability of the RECORD data.”

    Nevertheless, in his report Marciniak said: “While these numbers may seem small compared to the size of the trial, note that about 15 more [myocardial infarctions, MI] in the rosiglitazone arms are needed to change the GlaxoSmithKline MI results to a relative risk of 1.4 and a p value of 0.042”—which would make an increase in myocardial infarction statistically significant.

    Marciniak said it was “a huge challenge to try to find those few needles in the haystack,” and it’s certainly too much for the under-resourced European regulator. But it’s a job that’s needed, he says. According to Odlind, unlike the FDA, the EMA takes a top down approach—it takes the study summaries and asks the drug companies for more data if it sees fit.

    Marciniak is a keen advocate of accessing raw data: “You will not find the truth in drug review unless you dig,” he told the BMJ. “I believe the FDA approach is better or potentially more thorough than the EMA’s, but it also needs more complete implementation. One public suggestion has been to release the raw data to academic organisations. That would be an advance, but I believe that most of the academic organisations don’t realise that you need not only raw computer data files but also the case report forms and a variety of other source documents to understand completely a study,” he said.

    GlaxoSmithKline had employed an independent statistician, as is a prerequisite for the publication of pharmaceutical sponsored trials in some medical journals, notably JAMA. Marciniak does not blame the statisticians for not picking up the flaws in the case reports, as they can work only with what is given to them.

    Shining a spotlight on GlaxoSmithKline

    Neither the Nissen meta-analysis nor Marciniak’s digging show GlaxoSmithKline in the best light. In a Senate committee finance report published in February 2010, GlaxoSmithKline executives stood accused of focusing “on strategies to minimize or misrepresent findings that Avandia may increase cardiovascular risk.” Internal documents show that GlaxoSmithKline quickly published an interim report of the RECORD trial to counter the negative effect of Nissen’s meta-analysis.1

    The company, however, said in a statement that the interim analysis of the RECORD study was conducted urgently to “gather additional information about the potential risk for patients.”21

    To add to the company’s woes, on the day of the committee meeting this year the New York Times splashed with allegations that GlaxoSmithKline—then known as SmithKlineBeecham—had started a secret trial to see if rosiglitazone was safer for the heart than pioglitazone.22 “Not only was Avandia no better than Actos, but the study also provided clear signs that it was riskier to the heart. The company did not post the results on its Web site or submit them to federal drug regulators, as is required in most cases by law,” the article alleged.

    To date, there are no head to head trials of the two drugs in the public domain—companies do not have to demonstrate added therapeutic value under European Union or United States law.

    The news story went on to quote from an internal company memo: “‘Per Sr. Mgmt request, these data should not see the light of day to anyone outside of GSK,’ the corporate successor to SmithKline.”

    GlaxoSmithKline, however, say that the emails were “selectively disclosed by lawyers seeking damages” and that “other documents [were] taken out of context, which therefore are incomplete and misleading.” They also said that the “assertion that this study informed GSK’s views about heart attacks and Avandia is completely unfounded”.

    Decision time

    As Gerald Van Belle, director of the Clinical Trials Center at Washington University and FDA advisory panel member, put it—it was a “perfect regulatory storm.”

    At the hearing in July, the FDA panel voted that the available data supported a conclusion that rosiglitazone increases cardiac ischaemic risk in type 2 diabetes patients.

    But the question of what to do raised mixed answers. There were five options to choose from, ranging from the removal of the black box warning to withdrawal—and the breakdown of the votes would mean that it could be subject to interpretation. A majority of votes recommended keeping the drug on the market, but with more warnings or restrictions. But viewed in a different way, a majority recommended either removal or severely restricting access to rosiglitazone.

    However, they voted to continue the TIDE trial—a study commissioned by the FDA to assess rosiglitazone’s cardiovascular safety—which some argue the FDA should have asked for at the outset. This trial has since been put on “clinical hold” by the FDA.

    Janet Woodcock, director of the FDA’s Center for Evaluation and Research, admitted the advisory meeting featured conflicting opinions between two branches of the FDA.

    Marciniak agrees: “The Avandia advisory committee meeting was a battle between the “diggers” [Marciniak and a few others] and the “deliberators,” [meaning] the rest of the FDA and most of the committee and, I believe, the EMA.”

    Trying to gain an overall perspective of deliberations within the EMA has been far trickier. The BMJ attempted to speak to people who had sat on panels for the MHRA and the EMA. But they were bound by confidentiality clauses. The EMA would not release the names of the members of the scientific advisory group discussing rosiglitazone under the Freedom of Information Act. Secrecy also shrouds the UK’s regulatory agency.

    Where next for diabetes drugs?

    While the focus has been firmly on rosiglitazone, what about pioglitazone? Its manufacturer Takeda had benefited from the controversy of rosiglitazone. But like rosiglitazone, pioglitazone is associated with an increased risk of oedema, heart failure, and bone fracture.23

    As Professor Van Belle said, he doesn’t want to be sitting at an FDA advisory meeting in three years’ time discussing pioglitazone. Professor Gale is also concerned. In a letter to the UK regulator in 2008 seen by the BMJ, he wrote that “there is an urgent need to determine the safety of pioglitazone”.

    “Pioglitazone may or may not prove to be safer than rosiglitazone. There is an urgent need for more and better data addressing this issue. On present evidence, its safety cannot and should not be assumed,” he wrote. Takeda say that they continuously monitor the safety and efficacy of their compounds.

    Meanwhile other anti-diabetic drugs using a similar pathway are in development: the chequered history of the glitazones not having deterred manufacturers. According to reports, Dr Reddy’s Laboratories and Nordic Bioscience’s partial PPAR-gamma agonist balaglitazone met its primary endpoint in its first phase III trial in patients with type 2 diabetes (reduction in glycated haemoglobin). It was claimed to be “non-inferior” to pioglitazone. The companies are currently in discussions with regulators and hope to eventually file the drug in the European Union and the United States.24 And both Roche and Metabolex have drugs in phase two trials.25

    Will regulators, industry, and the clinical community do a better job for patients next time?


    • Troglitazone withdrawn from UK market after six weeks because of hepatotoxicity


    • April: American Diabetes Association declares that the drug’s properties are not shared by any others, offering new options to healthcare professionals

    • May: Rosiglitazone approved as monotherapy by FDA with label precautions for use in patients with heart failure

    • September: European Association for the study of Diabetes heard that rosiglitazone would be useful as a first line therapy

    • October: Rosiglitazone turned down by EMA by 14 of 25 votes


    • March: Troglitazone withdrawn in United States

    • July: Rosiglitazone given market authorisation in Europe with restrictions and with warnings on heart failure. GlaxoSmithKline (GSK, then SmithKline Beecham) is asked to conduct two post-marketing trials: one study to look at effect on cardiovascular structure; the other to assess cardiovascular safety—the RECORD trial.

    • October: Pioglitazone approved in Europe


    • February: FDA approve new warnings on potential for heart failure


    • With an increasing number of people taking rosiglitazone, World Health Organization picks up safety signals and alerts GSK

    • June: GSK ordered to publish summaries of results of all its clinical trials on its website once a product has been launched in a settlement in New York


    • September: Internal GSK meta-analysis finds 29% non-significant increased risk of ischaemic cardiovascular events


    • February: FDA approves Avandaryl (rosiglitazone maleate and glimepiride)

    • April: FDA approves new warnings on risks of cardiovascular events

    • May: Internal GSK meta-analysis finds 31% increase in ischaemic events

    • June: EMA approves Avaglim (rosiglitazone maleate and glimepiride)


    • May: New England Journal of Medicine publishes meta-analysis reporting 43% increased risk of myocardial infarction

    • June: NEJM publishes interim analysis of the RECORD trial

    • July: FDA advisory committee finds increased cardiac ischaemic risk but votes to keep drug on market

    • October: European Medicines Evaluation Agency asserts positive benefit-risk profile, recommends new warnings for patients with ischaemic heart disease

    • November: FDA approves new boxed warnings that drug may increase myocardial ischaemic events, including myocardial infarction, though evidence “inconclusive”

    • December: UK Medicine and Healthcare products Regulatory Agency warns drug might be associated with small increased risk of cardiac ischaemia


    • Updated internal GSK analysis finds no risk of myocardial infarction or other major cardiovascular events


    • March: International Journal of Cardiology meta-analysis finds no risk of myocardial infarction

    • June: RECORD trial published in the Lancet. EMA add a statement to their scientific information document saying there was no difference in the number of adjudicated primary endpoints between the arms of the study.


    • February: US Senate finance committee releases report that includes internal FDA safety report calling for drug to be withdrawn

    • February: GSK responds with 30 page document

    • June: David Graham’s study leaked to the Pharmalot blog. It is published in JAMA regardless. At the same time, another JAMA journal, Archives of Internal Medicine, publishes an updated meta-analysis by Steve Nissen

    • 13-14 July: FDA advisory committee meeting held. FDA drug approver gives damning verdict on the RECORD trial. Majority of committee vote either to withdraw the drug or restrict it severely

    • 15 July MHRA meet. Commission on Human Medicines vote to withdraw rosiglitazone

    • TIDE trial suspended by the FDA

    • 19-22 July: EMA meet to discuss rosiglitazone

    • 26 July: the MHRA send out “dear doctor” letter advising doctors to consider alternative treatments where appropriate

    • September: EMA will finalise its review


    Cite this as: BMJ 2010;341:c4848


    • doi:10.1136/bmj.c4812
    • doi:10.1136/bmj.c4805
    • Competing interests: All authors have completed the Unified Competing Interest form at (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous 3 years; no other relationships or activities that could appear to have influenced the submitted work.

    • Provenance and peer review: Commissioned, externally peer reviewed.


    View Abstract