Cardiologist: Glaxo’s Diabetes Drug Raises Heart Risks
Health Blog
Posted by Jacob Goldstein
May 21, 2007, 1:37 pm
A leading cardiologist whose challenges to the safety of several prescription drugs has limited or halted their their use has turned his sights on Avandia, a popular diabetes drug from GlaxoSmithKline.
The Cleveland Clinic’s Steven Nissen (pictured, left) suggests that Avandia, known generically as rosiglitazone, raises the risk of heart attack, in an analysis of clinical data published online today by the New England Journal of Medicine. Nissen found the pill seemed to increase the risk of death from cardiovascular causes, although that finding narrowly missed statistical significance, a standard measure of scientific validity. Other data recently showed the drug could weaken bones.
The release of Nissen’s critique on Avandia sent Glaxo stock down more than 6% in midday trading on the Big Board. For its part, Glaxo said in a statement that it “strongly disagrees with the conclusions” because they “are based on incomplete evidence and a methodology that the author admits has significant limitations.” The company said it stands behind Avandia and believes “its significant benefits continue to outweigh any treatment risks.”
As early as last September, Nissen told Forbes.com that one Avandia trial’s “trend toward increased cardiovascular events is very troubling.”
Nissen pored over 42 previously released studies to draw his conclusions. Nissen and his co-author determined that the odds of having a heart attack increase by 43% for patients who take Avandia, compared with patients who take other drugs or a placebo. Because the approach re-hashes data collected from multiple studies, a technique called meta-analysis, the work is often faulted for smashing together clinical findings that don’t belong together. Still, the approach has its supporters because it marshals data on large numbers of patients, sometimes fleshing out risks in a way that even well-designed, individual clinical trials do not.
Nissen has a strong record as an early and ultimately accurate critic on drug safety.
Here are some other noteworthy safety calls by Nissen:
Merck’s Vioxx was pulled from the market because it raised the risk of heart attacks and strokes. Nissen and Eric Topol, then a colleague at the Cleveland Clinic, raised concerns about the drug in JAMA.
Bristol-Myers Squibb and Merck failed in their efforts to bring Pargluva, a diabetes drug similar to Avandia, to market, after Nissen published an analysis of data in JAMA in 2005 that showed increased cardiovascular risk from the drug. At the time, he said the Pargluva trials provide “pretty strong evidence” of potential danger.
Johnson & Johnson’s Natrecor, a heart-failure treatment, remains on the market but has seen sales slacken as Nissen and others have pointed to cardiovascular risks with the medicine. He told WSJ last year that without more clinical data about Natrecor and its safety, “there isn’t sufficient evidence to support its use.”
Drugs to treat Attention Deficit Hyperactivity Disorder were given strong warnings about cardiovascular risks after Nissen led the charge to have the labels changed.
Permalink | Trackback URL: blogs.wsj.com
Comments
Report offensive comments to healthblog@wsj.com
Ah the caped crusader Nissen is at it again. Since when is his word greater than the doctor’s who ran the original tests to begin with. At this rate Nissen will have concluded that every drug ever made has a CV risk to it.
Comment by Michael - May 21, 2007 at 1:52 pm
A bit of a conundrum this as an objective of controlling diabetes is to reduce cardiovascular risk. So it matters what the increase in CV risk is compared to.
Comment by DavidC - May 21, 2007 at 2:14 pm
I wonder why multiple prospective studies have not shown this trend? Also, why did Dr. Nissen include studies where patients had MIs in his meta-analysis and not all studies? This one fact would seem to me to be cherry picking studies to prove a point. It is unfortunate that the NEJM has decided to publish this data in the manner that it did … without fair balance. Without fair balance, the only people at risk here will be the patients who will stop taking their diabetes medicine and will suffer irreparable harm from that fact. I hope Dr. Nissen figured that into his analysis…probably not.
Comment by John - May 21, 2007 at 2:20 pm
Importantly, the editorial in the NEJM states: “A few events either way might have changed the findings for myocardial infarction or for death from cardiovascular causes. In this setting, the possibility that the findings were due to chance cannot be excluded. In their discussion, the authors properly emphasize the fragility of their findings.”
Comment by thoughful - May 21, 2007 at 2:22 pm
And yet the NEJM decided to publish this study. It would seem to me that someone has an ax to grind here.
Comment by John - May 21, 2007 at 2:36 pm
Meta analysis is not science; it is a statistical technique that has been in use since at least the 1970s and that many statisticians in my day of research looked at with scepticism.
If Nissen was able to get Vioxx off the market with a meta analysis, and if he succeeds in keeping GSK stock down, or–horrors–has it pulled off the market, then indeed the old saying that the market can remain irrational indefinitely is proven yet again.
Comment by Ernie Bernard - May 21, 2007 at 2:40 pm
Axe to grind? The NEJM will only be satisfied when nothing is approved, save of course for PBAbortion for which we have a Constitutional right! By these same sorts of analyses we could readily pull aspirin from the market.
Comment by jr - May 21, 2007 at 2:42 pm
It seems that everyday a once celebrated research finding is being invalidated (vis Vitamin C, multi-vitamin supplements etc). What is next - drinking 8 glasses of water will weaken your kidney?
Comment by kash - May 21, 2007 at 3:53 pm
Another case of the tail wagging the dog?
Comment by EJL - May 21, 2007 at 4:28 pm
AVANDIAMET HAS PROVEN TO ME TO HELP.NEJM SHOULD NOT JUMP THE GUN JUST TO FILL ITS PAGES WITH FALSE REPORTING
Comment by JOSEPH VECELLIO - May 21, 2007 at 4:41 pm
The number of genes affected with PPAR’s is enourmous and varied with each compound; subsequently the clinical differences in pio, mura, tro and rosi glitazone are great. The adverse lipid effects have been a question hanging over Avandia and it is no surprise to me that Nissen came to this conclusion. I agree with other posts that the meta analysis is weak but I also must point out that he disclosed the inability to get the raw data. If you were GSK would you give out this data if it was not favorable? The next part of this story will be how much did GSK know and when regarding cardio safety. The Nissen shot was over the bow, where will the next one land on the GSK ship?
Comment by Joseph F Zastrow MD FAAFP - May 21, 2007 at 5:33 pm
Unfortunately, drug studies are limited in both their size and scope due to their huge financial burden, thus we are just now learning about the longer-term side effects of many approved drugs. Ultimately, the key will lie in doctors being far more prudent in recommending drugs and giving greater focus on proper nutrition to help their patients.
Comment by Concerned Citizen - May 21, 2007 at 7:09 pm
There is obviously more here than meets the eye. It will be interesting to see how this all plays out. Why a meta-analysis..
that is about the lowest form of scientific data. Why did he not include the other available studies in his analysis? What about long term studies…what do they show??Why did the NEJM ever publish this?
Comment by PugMan - May 21, 2007 at 9:04 pm
Dr. Zastrow hit it right on the nose–GSK should have allowed Nissen access to the raw patient data. That they didn’t raises questions–Why?
Comment by William - May 21, 2007 at 10:19 pm
blogs.wsj.com |
