New evidence shows the drugmakers conducted several studies of the popular cholesterol med that raise questions about its risks to the liver, but the results were never published,
The New York Times reports. Partial results, alluded to in documents on the FDA web site, raise questions about whether Zetia can cause liver damage when used long term with statins. Most people who use Zetia take it along with Lipitor, Crestor or Zocor, or in a single pill, Vytorin, that combines Zetia with Zocor.
The discovery of the unpublished research comes as Merck and Schering are already under criticism for not yet releasing data from an important Zetia study, called Enhance, that they completed early last year, the Times notes. The Enhance data may also contain important info about Zetia’s liver risks. At least some patients were dropped from the Enhance study after testing revealed that they had elevated liver enzymes, a Schering-Plough spokesman confirmed this week. But a full report on that trial, including the number of patients who had liver problems, will not be available until March.
Doctors say that by failing to disclose promptly all their research, Merck and Schering-Plough may be leaving the public with a misleadingly favorable view of Zetia’s safety and benefits. “You don’t want to have data missing,†Bruce Psaty, a professor of medicine and epidemiology at the University of Washington, tells the paper. “When there have been adverse effects, when the benefits don’t look impressive, those are the trials that historically don’t make it to press.â€
Before the drug was approved in 2002, one FDA reviewer said it should not be cleared for use with statins because the combo had caused liver damage in animals. And in the last two years, scattered case reports of severe liver damage in patients taking Zetia in combination with statins have appeared in medical journals. In the US, the product label for Zetia contains only mild warnings about the drug’s potential for liver damage. But since 2005 in Australia and Canada, regulators have issued a series of warnings about Zetia’s potential to cause hepatitis, pancreatitis and depression.
“We keep telling people we want to practice evidence-based medicine, and what we keep finding out is that much of the evidence is obscured,†Harlan Krumholz, a cardiologist at Yale, tells the paper. “There is important evidence, but it’s not in public view. It’s hidden from investigators.â€
UPDATE: Here is a Public Citizen newsletter from December 2004 warning against the use of Zetia.
15 Comments
No doubt that Merck and SP should be more transparent. And, key opinion leaders in the field should be able to review the data via published reports. However, based on this report, it looks as though the FDA did have access to the data and did consider it in its approval of Zetia and Vytorin. Not the best situation, but it could be worse.
I'm a bit confused. As I read the NYT article, FDA has the relevant studies. But there is reference in the article to the new sanctions the FDA could use (in the new FDAAA bill) for for studies not made known soon enough.
Understanding that those sanctions are not retroactive to the time the Zetia studies were done (and assuming the FDA is unlikely to use them anyway), do they (theoretically) apply to studies not that _are_, in fact, released to FDA but not more generally? Would appreciate clarification on this.
Also striking to me that Health Canada and Australian regulators took at least some action. Should we be relying on _their_ labels? (Not that that would have told us anything about the scale of the potential issue).
Meanwhile, it seems to me the key - besides scale - may well turn on the velocity of liver changes, assuming they are diproportionate. Do we have Rezulin Redux, in which "monitoring liver functions" turned out to be meaningless (and was certainly suspected as such by many)? Or are these problems that, at least for many, can be caught?
Finally, anyone know if Public Citizen or others have filed a FOIA to get the studies?
One of the changes in the FDA 2007 amendment should make this sort of behavior (if it did occur) illegal. Currently, companies must disclose the existence of, and results from a trial on clinicaltrials.gov if they want to publish the results in a top medical journal (which they do). Sometime after 12/26/07 and next year (depending on the seriousness of the disease - serious diseases will be held to the early 12/26 date, other trials at a time towards the end of 2008 that escapes me now) all comparative studies (and that includes placebo) other than phase 1 studies (which usually test kinetics in healthy patients) must be posted to clinicaltrials.gov.
Hi Hank,
Don't know the answers to all your questions, but there was a difference in the US and Canadian labeling for Pfizer's Chantix concerning suicide seen in studies. If you look at the post 'Up in Smoke?' you can compare the labeling through the links provided - if you're looking for an example of a contrasting situation, anyway.
Cheers ed at Pharmalot
No doubt the results are important and we need them, but did anyone doubt that there would be liver issues? Scattered reports for a statin in any situation are expected, aren't they? Did I miss something
I see enough typos in my message that I should clarify the main question:
Even though they would not apply in this case, do potential FDAAA sanctions apply to published/unpublished studies in general, in contrast with submission of studies to FDA? Are they tied into what is posted on relevant websites?
Otherwise, agree with Atlex in principle, but it also comes down to what FDA actually got - raw data, interepretations, summaries, etc.. As we learned with Baycol, fen-phen, et. al., there are lots of ways data can be submitted that still leave OND/ODS with a good deal of "detective work."
Thanks to Jack - didn't see that message when posted my last.
Re: Todd's comment, we obviously have much to find out. NYT piece "suggests" interaction issue may be part of it - like statins, fibrates, and rhabdo. Or, indeed, could be independent issue of zetia alone, compared with statins, re: liver function.
Hank, please tell me you are not a clinician. If you are practicing medicine with this knowledge base, please consider retiring this year.
Hank: My understanding is basically all human phase II-IV trials will require at least a posting to clinicaltrials.gov. This was common practice before, since if things turned out well the company would want it to show up somewhere prestigious (eg JAMA), and the editors of these journals all agreed a few years ago not to publish trials that hadn't been posted on the website. Now the FDA requires it and can legally penalize (fines?) companies that don't comply with the strict deadlines. The NIH is also under certain legally required timelines for maintaining the site. I don't really know how detailed the postings will need to be, and haven't seen any hard guidelines. It's entirely possible the synopsis on the website could be short enough or vague enough to hide something like this example. But on the other hand, it's tough to legislate exactly what companies need to post, since trial design varies.
Anyway, it prevents companies (and universities - I think it applies to them too) from conducting a trial in a certain indication and, when the results don't turn out right, not telling anyone about it. I personally don't think this was common practice (but I do think companies might drag their feet a little about publishing failed efficacy results - and journals don't exactly like to publish it either), but I'm sure it happened.
However, please also keep in mind, previously, all data from registration trials got sent to the FDA, and companies needed to send all adverse event reports to the FDA as well.
Ed: I saw some stuff in my post before that looked unclear after I reread it too. How about an edit button?
Again thanks, Jack.
Todd - Rather than attack, why don't you tell us what you found misinformed, misguided, etc.. No doubt, many of us could learn from you.
In the case of Vioxx, Merck didn't hide studies per se, but they failed to report pre-specified analyses to the FDA and to the world. For example, in the key Alzheimer's studies, Merck never reported the so-called "intention-to-treat" (ITT) analysis (which was pre-specified). The ITT analysis shows a clear signal against placebo for thrombotic adverse events long before the withdrawal. Merck instead reported an "on-drug" analysis, that, for various reasons, looks much better. The FDA would need a staff of millions to catch this level of subterfuge.
So, its not just about hidden data. Its also about hidden analyses. To this day, Merck has never published an ITT analysis of the Vioxx Alzheimer's studies.
Great point, Emil. From what you say, I take it that there is no pre-specification of analysis type that a company files with FDA? Or, if they do, still easy for FDA to "miss" and/or let it be?
The FDA has all the data analysis plans but does not seem to have the wherewithal to ensure the plans are carried out.
Much thanks. Consistent with every report on FDA that's been issued, I guess (for the past forty years).
Has the NNT question re: Alz. studies been raised in context of Vioxx "wider discussion"? As I recall (could be wrong), Merck talked about unblinding those studies after VIGOR to see if the same pattern showed itself. Claiming not, used that as a reason to minimize response to VIGOR (along with theory that naproxen lowers MI rate, etc.)
All,
This finding by the NYT should not surprise anyone who has knowledge about how Big Pharma operates. If it's good news, shout it from the mountaintop. If it's bad news, hide it in the closet. As the FDA is woefully underfunded, they need to be able to trust the Big Pharma companies to be honest. Unfortunately, as long as money is involved, this trust is misplaced. The Big Pharma companies will never be trustworthy. They've demonstrated over and over again that they just can't do it. It's a crime when human lives are affected, but there's money to be made. With all the recent bad news on Schering-plough, they now top the list of the most mistrusted.