Hailed ‘METEOR’ statin trial results not as stellar as we are led to believe

Back in December, I wrote a blog post which focused on the habit drug companies have of causing the medical profession and the public at large to forget what medicine is for [1]. The point I made is that much of medical research is not focused on health itself, but what are known as ‘surrogate’ outcomes or ‘end-points’ such as blood pressure, cholesterol levels or blood sugar levels. In this blog, I gave examples of some conventional treatments that might help surrogate end-points, but don’t seem to do much for us in real terms. Sometimes, the treatment might be downright dangerous.

This week saw the publication of a study which is another example of why it’s important to look more deeply into the reported results of a trial. It concerns research published in the Journal of the American Medical Association reporting on the results of what is known as the ‘METEOR’ trial. In this study, the effects of the cholesterol-lowering drug rosuvastatin (Crestor) in healthy middle aged individuals, some of whom had evidence of fatty build-up (atherosclerosis) in one or both carotid arteries (the main vessels supplying blood to the brain).

Compared with placebo (inactive medication), treatment with rosuvastatin over two years reduced cholesterol levels, and also halted the progression of atherosclerosis. This all looks good on the surface of it. And not surprisingly, the publication of these results got a lot of press around the World and I’m sure made rosuvastitin’s manufacturer, AstraZeneca, very happy.

But let us not lose sight of the fact that the main outcomes this study was designed to assess (cholesterol levels and atherosclerosis) are surrogate end-points. They do not measure the effect of rosuvastatin on so-called ‘cardiovascular’ events such as heart attacks. The whole point of reducing cholesterol, we are told, is to prevent such events, after all.

What this study showed was that in the placebo group, no individuals suffered a serious adverse cardiovascular event. On the other hand, 6 individuals in the rosuvastatin group suffered a total of 8 serious adverse cardiovascular events (including heart attack and angina). Admittedly, the serious adverse effect rates in the treated group were low. But there’s no getting away from the fact that when we look at these outcomes (rather than the surrogate end-points the study was designed to examine), then the results of the METEOR trial don’t look as stellar as the press reports make them out to be.


How drug companies can cause us to forget what ‘medicine’ is really for

2. Crouse JR, et al. Effect of rosuvastatin on progression of carotid intima-media thickness in low-risk individuals with subclinical atherosclerosis: The METEOR Trial.JAMA 2007; March 25 [Epub ahead of print]

15 Responses to Hailed ‘METEOR’ statin trial results not as stellar as we are led to believe

  1. Brian Donovan 30 March 2007 at 10:58 am #

    I was on statins for roughly four years and in that time although I did not make the connection I experienced increasing pain in my legs. It got so bad that I could only walk 25 metres or so before the acute pain made me rest. I had numerous tests at the hospital to try and find out what was causing the pain but no reason could be discovered.

    It was only in a telephone conversation with a relative when I chanced to mention the leg pain I was suffering when she told me she knew of two other people on statins who were suffering varying degrees of leg pain. I closely checked the leaflet that accompanied the statins and found that leg pain was a fairly common side effect in in extreme cases it could actually be fatal. This should be highlighted more strongly to people being prescribed statins. I am no longer on statins and
    my leg pain has disappeared.

  2. Neil 30 March 2007 at 1:24 pm #

    According to the editorial in the Journal of the American Medical Association,

    “The METEOR trial was not powered to assess the effect of rosuvastatin therapy on clinical events. Among 984 patients, there were only 6 ischemic events, which curiously all occurred in the rosuvastatin group. Although the ASTEROID trial demonstrated that rosuvastatin induced regression of coronary atherosclerosis,24 to date there have been no randomized trials showing this particular statin to reduce risk of clinical events.”


    “Should low-risk individuals undergo routine arterial imaging followed by statin therapy when evidence of asymptomatic disease is discovered? On the basis of current evidence, including the METEOR trial, the answer is clearly no.”

    It was hoped that there would be regression of plaque as well as halting of the growth of existing plaque, but this did not occur.

    Interesting but hardly compelling me to take a high dose of a powerful drug for the rest of my life. Ultimately, reduction in clinical end points, particularly death, is the real goal. METEOR was not designed to look at this aspect, so I would question the point of the whole exercise.

  3. Tiggy 31 March 2007 at 2:52 am #

    My dad, who is on statins, has had serious sudden and debilitating pains in his legs twice and had to call the doctor in. Each time there has been no real explanation and certainly his being on statins was not referred to even by his regular GP. One doctor suggested it was athritus, but it wasn’t like athritus at all. His walking has also got increasingly painful.

    Dyou think beta blockers are a good idea? They seem to make people so dopey!

  4. ben goldacre 2 April 2007 at 6:17 pm #

    hi john,

    someone just sent me this, wondering why i didnt cover the meteor media over-reporting scandal.

    if these trials really were overstated by credulous journalists in the newspapers as you suggest, i’d lap up the opportunity to criticise them.

    what media reports have overstated the results of the Meteor trial?

    there are no positive reports in the media at all, that i can find, and no news stories for consumers at all.

    the only mention i can find of Crestor, Meteor or Rosuvastatin in the archives for this week are short pieces on the business pages of the Daily Mail and the Times, essentially slating it, saying that the trial had mixed results, and were a disappointment. i have pasted these for you below

    other than that, i can’t find a single trace of it being mentioned anywhere in mainstream media, anywhere.

    seriously, where have you seen the Meteor trial “hailed”, and “getting a lot of press”, in “stellar” reports? i’d love to write about it if they exist.

    27 Mar 2007: The Times – Page 46 – (483 words)
    AstraZeneca shares dip as Crestor falls short of trial hopes Factbox
    By: Robin Pagnamenta Health Industries Correspondent

    Shares in AstraZeneca, Britain’s second-largest drug company, fell to their
    lowest level this year as results from a trial into its blockbuster
    anti-cholesterol drug failed to live up to expectations.

    A two-year trial into the long-term benefits of Crestor showed mixed
    results on patients with atherosclerosis, or hardening of the arteries.

    The trial indicated that, when compared with a placebo, Crestor
    significantly reduced the rate at which plaque built up in the arteries of
    patients with early-stage atherosclerosis who were at low risk for heart
    disease. Prescribing Crestor , however, failed to make any substantial
    reductions in existing levels of plaque in the same patient group.

    Atherosclerosis is the most common cause of heart disease and
    stroke-related death in the world. AZ earned more than $2 billion in annual
    sales from Crestor for the first time last year.

    AZ shares fell 1.2 per cent, down 37p to Pounds 27.62, a new low for 2007,
    wiping Pounds 560 million from its market value as investors reacted to the
    latest in a string of setbacks to AZ’s stock of drugs . The shares have
    fallen from a high of Pounds 35.29 last year after investors endured a
    series of failures in the company’s drugs pipeline or delays in getting
    drugs to market.

    Last week the company said that an experimental drug for heart disease, AGI
    1067, had failed crucial tests. AZ has already pulled three drugs before
    they got to market, all under the watch of David Brennan, the chief

    In January, an experimental stroke treatment being developed with partner
    Renovis failed in a late-stage trial. The company’s Galida for diabetes and
    Exanta for blood-clot prevention have also failed.

    Last November, the firm said it would have to carry out another mid-stage
    trial to Cytofab, its septic shock drug , after consultation with US and
    European medicine regulators.

    The latest mixed results for Crestor make it more difficult for AZ to
    differentiate the drug -also known as rosuvastatin -from rival statins such
    as Schering-Plough’s Vytorin.

    Crestor is already widely prescribed for high-risk patients, but AZ is keen
    to market it more widely to patients in the earlier stages of the disease.

    “Overall, the data supports the current use of Crestor , but does not make
    a strong case for ‘preventive’ use of the drug in early-stage patients,
    which would have been the best outcome,” Peter Cartwright, an analyst at
    Evolution Securities, told clients in a research note.

    The results of the study on the drug were unveiled on Sunday by the
    American College of Cardiology. Crestor has been approved by regulators in
    90 countries and has been prescribed to nine million patients.


    * Britain’s second-biggest drugs group

    * Sales in 2006 totalled $26.5 billion

    * Operating profit in 2006 was $8.2 billion

    * Employs more than 66,000 people

    * Crestor generated sales of more than $2 billion in 2006

    27 Mar 2007: Daily Mail – Page 70 – (291 words)
    Analysis Uniq gives food for thought

    AZ looks like a safe haven

    VITAL STATISTICS (drugs trial)

    One-year share performance: -5pc

    ASTRAZENECA’S cholesterol lowering drug Crestor showed mixed results in
    trials on almost 1,000 patients.

    While the so-called Meteor study revealed a 40mg tablet significantly
    slowed the progression of atherosclerosis n furring of the arteries n it
    did not reverse the process as many had hoped.

    It must, however, be pointed out that these were people at low risk from
    heart problems.

    And in fact an earlier trial indicated it reduced plaque build-up where
    disease was in its advanced stages.

    More relevant to the future of Crestor is whether it hits projected peak
    sales of pounds 2.3bn, which looks more likely as safety fears fade.

    With the shares close to a year low, AZ may be a good bet for those looking
    for a safe haven for their cash n particularly if the market cuts up rough

    VERDICT: Buy on sustained weakness

    20 Mar 2007: The Times – Page 44 – (489 words)
    AstraZeneca stung by late-stage trial failure of heart drug Factbox UK
    By: Robin Pagnamenta Healthcare Industries Correspondent

    AstraZeneca announced yesterday the failure of one of its few remaining
    drugs in late-stage development, raising further the pressure on the
    company to rebuild its weak pipeline of new pharmaceuticals.

    Britain’s second-largest drugs company said that a trial of AGI-1067, its
    experimental drug for heart disease, had failed to reduce deaths or
    complications when compared with a placebo. A phase-III clinical trial
    involving 6,000 people indicated that the drug had failed to “meet its
    primary endpoint of a statistically significant relative risk reduction”.

    AstraZeneca said that development of the drug had not necessarily been
    abandoned because niche applications could be found. The trial did identify
    positive findings for a few patients, such as those with diabetes.

    The company said that it would work with its collaboration partner
    AtheroGenics, the American biotechnology company, “to fully analyse the
    data” and then take 45 days to plot a course of action.

    In 2005 AstraZeneca paid $50 million to AtheroGenics for licensing rights
    to AGI-1067. The companies had hoped that, if successful, the drug could
    have generated billions of dollars in sales a year.

    However, David Brennan, the chief executive of AstraZeneca, admitted to The
    Times in December that it was a “very high-risk project” for which there
    was “a low probability of success”.

    The disappointment of AGI-1067 represents the fourth failure of a
    late-stage drug for AstraZeneca since the start of last year. Exanta, a
    blood-clot medicine, Galida, a diabetes drug , and Cerovive, an anti-stroke
    drug , were abandoned last year. Furthermore, patents on some of the
    company’s key products, such as Nexium, a stomach-acid drug , and Crestor ,
    an anti-cholesterol agent, will expire in the next few years.

    Shawn Manning, of Bridgewell Group, said that the failure of AGI-1067 would
    increase the pressure on AstraZeneca to buy new technology to bolster its

    “They need to bring in at least $1 billion to $1.5 billion in revenue by
    2009 to replace products that are coming off patent,” he said. “If they
    can’t do it by buying in products, then they will have to consider
    large-scale mergers and acquisitions.”

    However, Peter Cartwright, an analyst with Evolution Securities, said that
    AstraZeneca had “sufficient momentum from currently marketed products plus
    line extensions” to sustain top-line growth rates until 2010.

    AstraZeneca and AtheroGenics will publish the full results of the AGI-1067
    study at the American College of Cardiology annual meeting in New Orleans
    this weekend.

    Shares in AstraZeneca slipped 20p to Pounds 28.61 after the news.
    AtheroGenics, which is listed on Nasdaq, closed down $4.74, or 61 per cent,
    at $3.09.


    * AstraZeneca is Britain’s second-largest drugs group, after
    GlaxoSmithKline Sales in 2006 totalled $26.5 billion, with an operating
    profit of $8.2 billion

    * The company spends more than $16 million every working day on the
    research and development of medicines

    * It employs more than 66,000 people worldwide

    * It operates 27 manufacturing sites in 19 countries

    * Its corporate headquarters are in London and its R&D headquarters are in
    Sodertalje, Sweden

  5. Neil 2 April 2007 at 7:21 pm #

    I googled ‘meteor trial’ There were cautiously positive web press articles e.g. on medscape http://www.medscape.com/viewarticle/554104

    but nothing in the mainstream press appeared in a shortish google.

    I guess equivocal news is no news for them, they do like the polar extremes.

  6. Dr John Briffa 2 April 2007 at 8:35 pm #

    Hello Ben

    In my blog (above) I claimed that the METEOR trial results ‘got a lot of press around the World’ and that the ‘results of the METEOR trial don’t look as stellar as the press reports make them out to be’.

    By way of example, let’s consider AstraZeneca’s press release itself. It is entitled: “First Study To Show Positive Benefit On Atherosclerosis For People With Early Signs Of Diseased Arteries”, and is subtitled “METEOR Trial Shows CRESTOR Slowed Progression Of Atherosclerosis In People At Low-Risk Of Coronary Heart Disease.” The press release also includes a quote from the study’s lead author which reads: “It’s exciting to see that by using rosuvastatin we can potentially slow or even stop the disease progression in people with relatively modest atherosclerosis.”

    Not surprisingly, AstraZeneca’s press release ‘hails’ the METEOR trial results and gives a strong positive message about the use of rosuvastatin in individuals at low-risk of heart disease. The entire press release can be found here: http://www.astrazeneca.com/pressrelease/5317.aspx

    The reason that I mention the press release specifically is because both it and its very positive message have had very wide coverage on the web. Typing the title of the press release (“First Study To Show Positive Benefit On Atherosclerosis For People With Early Signs Of Diseased Arteries”) into google returns more than 19,000 results.

    Also, typing ‘rosuvastatin’, ‘carotid’ and ‘atherosclerosis’ into google (in an effort to identify reports of the METEOR trial) returns almost 42,000 results, and some random clicking on these reveals them to be generally positive reports of the trial results.

    To my mind, this amounts to a helluva lot of positive press for the results of the METEOR trial. And because this trial measured surrogate end-points, I maintain that its results cautious and conservative interpretation.

  7. ben goldacre 2 April 2007 at 9:28 pm #

    hi john,

    you say: “And because this trial measured surrogate end-points, I maintain that its results cautious and conservative interpretation.”

    seriously dude, i very heartily encourage you to “maintain” that, nobody is questioning that for one moment, in fact, you don’t even need to “maintain” it, everybody agrees with you, in fact, as i have shown, contrary to your claim… even the MEDIA know the results of the meteor trial are uncompelling.

    amazing news about a company press release being positive by the way. who’d have thought? just joshing.



  8. ben goldacre 2 April 2007 at 9:32 pm #

    do let me know if you ever come across an example of a drug company press release getting inappropriate positive coverage by the way, i’d love to get a story like that in, luckily the media are a lot more critical of that kind of stuff than they are of the CAM & food supplement chaps.

  9. Dr John Briffa 2 April 2007 at 9:59 pm #

    Hi Ben

    “everybody agrees with you, in fact, as i have shown, contrary to your claim… even the MEDIA know the results of the meteor trial are uncompelling.”

    I’m not sure how you can conclude that the media know the results of the METEOR trial are uncompelling, when there are literally tens of thousands of posts on the web that report the ‘positive’ results of the METEOR trial quite unquestioningly.

    “amazing news about a company press release being positive by the way. who’d have thought? just joshing.”

    The point I was making (in case this was not clear to you already) was not that the press release was positive. It was that it had been reported very widely indeed, with little in the way of critical appraisal.

    “do let me know if you ever come across an example of a drug company press release getting inappropriate positive coverage by the way, i’d love to get a story like that in”

    Try the reporting of the METEOR trial for starters 😉

  10. ben goldacre 2 April 2007 at 10:19 pm #

    john: can you actually give me an example of the press positively reporting the meteor trial? not 10,000 google results with the search terms you typed in. an example of the press positively reporting the meteor trial.

  11. Dr John Briffa 3 April 2007 at 6:46 am #

    Hi Ben

    Can I suggest you:

    1. Read these definitions of ‘press’ extracted from dictionary.reference.com:

    • The collecting and publishing or broadcasting of news; journalism in general.
    • The entirety of media and agencies that collect, publish, transmit, or broadcast the news.


    2. Look at this (just one of many similar reports to be found on the web)


  12. Pat 3 April 2007 at 9:41 am #

    hi john

    that is not an example of the press positively reporting the trial.

    that is the press release from eurekalert, the science and health press release service. it even say so on the page.


    a press release is not press coverage, it is the press release.

    can you actually give me an example of the press positively reporting the meteor trial?

    should be easy, from all those tens of thousands you’ve found with your google search.

  13. ben goldacre 3 April 2007 at 7:05 pm #

    thanks john,

    the webpage which you rather triumphantly refer to as evidence of positive reporting in the press is in fact the press release.

    they make that quite clear, and they even source it from eurekalert, the industry standard health and science press release service,

    here is the very same press release on eurekalert:


    it was very kind of you to give me those definitions of “press” from dictionary.com but i don’t think a press release satisfies any one of them.

    a press release is a press release.

    please do let me know if you come across any examples of the media reporting the meteor trial positively, or indeed reporting any pharma press release inappropriately positively, i’m always on the lookout for opportunities to write about them.

  14. Dr John Briffa 4 April 2007 at 6:43 am #

    Dear Pat and Ben

    Yes, of course the link I provided was to a press release. That’s partly my point ” that the press has reported the METEOR results positively and unquestioningly. I can’t imagine a more blatant example of that than simply posting the drug company press release (or a version of it), can you?

    Now that we’ve each had are say I will not be posting any further comments regarding this particular topic. I suggest we let the readers decide whether tens of thousands of postings of a drug company’s ‘positive’ findings by press release (or version of it) on the web constitutes ‘positive press’ or not (if they care at all, that is).

    Ben, I don’t have any hard and fast rules about the length of comments, but for any future correspondence you might want to bear in mind the advice you offer on your own site which states that “…if your post is more than one thousand words long then you are officially a loser.” (Your first comment above ran to almost 1400 words).

  15. Neville Wilson MD. 1 December 2007 at 1:20 am #

    Dear John,
    I’m pleased to note the declared truce and that Ben has been gently nudged into silence on the issue of “positive press reporting”of METEOR.
    There is undisputed evidence of “positive” reports which emanate largely from the lead researchers and their press releases.These have limited value considering the obvious conflict of interest on behalf of the trial spokesmen.
    What is more relevant, and disconcerting, is the recent FDA granting of “an additional indication for Crestor to slow the progression of atherosclerosis in patients with elevated cholesterol”.
    It must be noted in METEOR that, a)Crestor was shown to slow the rate of progresion of atherosclerosis, but did not produce disease regression, nor did the drug halt or reverse progression of subclinical atherosclerosis, despite some press proclamations about its “benefits” in terms of “halting early changes in the blood vessels that lead to atherosclerosis”.
    b)improving a surrogate endpoint,such as CIMT, does not equate with clinical benefit, and in this trial failed to yield a clinical benefit.In fact the opposite ocurred as 6 patients in the Crestor group suffered 8 events and one death.
    c) Measuring mean CIMT is irrelevent without a corresponding measurement of vessel lumen diameter, since it does not take account of compensatory vasodilatation within the vessel.And no clear relationship has been demonstrated between CIMT and the development of atherosclerosis.
    The METEOR TRIAL has been touted by pharmaceutical sponsored spokesmen as being of great “BENEFIT” despite its clear absence of clinical benefit, and by so doing adds the dangers of discouraging non-pharmaceutical approaches to primary prevention of cardiovascular disease, and encouraging routine arterial imaging of low risk patients with a view to pharmaceutical intervention

Leave a Reply