FDA Reviewer: Avandia Trial Was Full Of Errors
19 CommentsBy Ed Silverman // July 9th, 2010 // 1:35 pm
With more than a thousand pages of memos, charts and studies decipher, there are all sorts of hints, allegations and conclusions to be found in the vast treasure trove that is the set of Avandia briefing documents posted on the FDA web site this morning. For those of you with time on your hand or skin in the game, feel free to get started here and here and here. You can read about the panel members below.
The meeting, of course, is a litmus test for the FDA which, under commish Margaret Hamburg, is trying to brand itself as a public health agency. With that in mind, In a briefing for the media yesterday, FDA deputy commish Josh Sharfstein maintained “we’ll make decisions in the best interest of the public” when it comes to whether Avandia is withdrawn or allowed to remain on the market. Yet the committee meeting is being held due to internal bickering over Avandia’s fate. “There’s not complete unanimity within FDA about the meaning of this data,” CDER director Janet Woodcock acknowledged during the same briefing.
Thix explains why the briefing documents amount to dueling studies, and the questions the advisory committee members will have to answer appear structured to allow the agency some wiggle room. Not only are there eight questions, in fact, several questions contain an unusual feature. Instead of being able to choose just A or B - for instance, Avandia does or does not increase CV risk - committee members can choose choice C, which indicates a finding can’t be made. It’s a way of opting out.
AS for those studies, one particularly harsh finding that will not make Glaxo happy was written by Thomas Marciniak, medical team leader in the FDA’s Division of Cardiovascular and Renal Products. He was chartered with reviewing cardiovascular events in Glaxo’s Record trial and came to the conclusion that Glaxo committed several important errors that undermine the validity of the trial. His review starts on page 16; here is an excerpt:
“Our summary is that Record was inadequately designed and conducted to provide any reassurance about the CV safety of (Avandia). The results do confirm and extend the recognized concerns regarding increased heart failure (HF) and HF deaths with (Avandia). They also suggest that (Avandia) increasees the risk for myocardial infarction (MI), although the confidence intervals for estimated MI hazard ratios are wide and include no risk while biases in the study suggest that the true risk could be higher,” he writes, adding that Glaxo introduced bias into the study design and the criteria Glaxo used for CV safety analyses apparently have “substantial errors.”
“Note that the point estimates for the hazard ratios (HR) for MIs (1.38-1.42) are remarkably close to the odds ratio reported in the Nissen meta-analysis (1.43),” Marciniak continues. “One does not have to be a mathematician or to perform the calculations to estimate that incorporating our Record results into a meta-analysis similar to Nissen’s will produce a risk for MIs that is highly statistically significant.”
To illustrate the yin and yang to which Woodcock referred, however, we also note that Ellis Unger, the deputy director of the FDA’s Office of Drug Evaluation, came to an opposite conclusion: “With respect to cardiovascular safety, the strength of reassurance provided by RECORD can (and will) be debated, but, aside from the known risk of heart failure, the study does not appear to demonstrate harm.”
As for the panel members, which you can see here, there are actually experts from two advisory committees - the endocrinologic and metabolic committee and the drug safety and risk management committee.
Familar names include Cliff Rosen, who headed the last big Avandia panel session in 2007; Sanjay Kaul, a UCLA cardiologist who was excluded from an FDA advisory committee last year at Eli Lilly’s request; Marvin Konstam, a Tufts University professor who was once caught up in the Senate Finance Committee probe into academic conflicts and a Merck ghostwriting scandal; and Curt Furberg, a Wake University professor who co-authored a 2007 New England Journal of Medicine editorial that was critical of GlaxoSmithKline’s Record trial, which the drugmaker points to as proof that Avandia is safe.
One interesting choice is Enrico Veltri, a non-voting member and former Schering-Plough exec who once told a Merck counterpart to ‘f… off’ during the run-up to the Vytorin scandal.
Justice in MI
Small historical footnote…
The Veltri thread that Ed links at the end led to what I consider one of the most informative discussions I’ve ever seen on Pharmalot. 149 comments, which went well beyond the immediate question of Veltri.
A “classic.”
Anonymous
Ellis Unger. Appointed as Deputy Director to Bob Temple effective Aug 10. 2008 after the position had been vacant for 4 years. (Hmm wonder why Bob Temple needed a Deputy either before or at this time.) This was after Ellis reported to FDA senior management on who may be whistleblowing to Congress. (Documents that were made public by the FDA showing his reporting on possible whistleblowers are in the possession of the WSJ and others.) In the next year he only involved in 2 approvals one of which also involves claims of a company rigging their studies to avoid safety findings and appropriate labeling.
Rezulin
I’ve seen this movie before.
doc
Avandia needs to go. Interesting note on Sanjay Kaul M.D. of UCLA. I’m still amazed that he was not included in the review board for Lilly’s prasugrel at Lilly’s request.
Sounds like the old inmates running the asylum scenario to me.
Know It All
The whole thing (FDA) reminds me of taking a shower with your clothes on!!!!!! We need a complete revamping of the FDA, get rid of the unfriendly management, before further damage is made.
Anonymous
FDA obviously has its problems, especially with the old guard.
But in this instance it was an FDA reviewer who got the data from GSK and called them on it.
If the allegations are true, it’s not FDA that deserves contempt.
Anonymous I
Anonymous on July 11, 2010 @ 5:23 PM is not associated with previous anonymous.
While Drs. Graham, Johann-Liang, and Gelperin at the FDA are to be commended for standing up to pressure. Dr. Woodcock’s insinuating herself into the scientific discussion of Avandia last week is inappropriate and likely represents her ongoing emphasis on treating companies as clients and the creation of a culture that promotes repeated retaliation against reviewers who try to act ethically that has occurred repeatedly on her watch.
The FDA is not monolithic andwhile there are some bright spots there are definitely parts that are deserving of comtempt.
Anonymous I
Know It All
Anon,
I wonder if the FDA did their job correctly, examining more closely; they would have uncovered information that would be enough to raise eyebrows and deeper thought. I don’t think this would have happened if they were on the ball.
Sorry, the FDA is not worthy of any commendation. This government subsidiary needs a COMPLETE makeover.
pharmavet
JiM. Foul-mouth executives are common in Pharma. I may have uttered one or two expletives in 28 years. However, unlike Rick Veltri, most executives are smart enough not to put it in print or not to hit the “send” button.
My second pharma job was working for a joint venture operation like Merck-SP. The facts are that 80% of JV’s fail; in the other 20% the two companies can’t stand each others’ guts.
Condor
On the subject of Veltri — Matt Herper noticed (tonight) that New Merck cut him loose — right at merger time (November 2009). The Forbes columnist and blogger deduced this when Veltri listed himself as an “independent consultant” — to be added to the FDA agenda on — of all matters — Avantia(!).
As if that drug didn’t have enough trouble hanging around it.
Wow. Tomorrow should be interesting. Buckle up, and pop the popcorn!
Namaste
Ed Silverman
Hi Condor,
Yes, that revelation was contained in the FDA briefing documents and noted below, although admittedly, at the bottom of a post about the Marciniak memo. For what it’s worth, Veltri is a non-voting member, which means he can curse all he wants, but it may not count for much.
http://www.pharmalot.com/2010/07/an-fda-report-says-avandia-trial-was-full-of-errors/
Cheers
ed
Salmon
Nonvoting members are typically there to push the industry’s agenda and they tend to be very good at it.
pharmavet
Thanks, Ed. I think that the concept of a ‘non-voting member” is still a potential conflict of interest because it doesn’t mean “non-speaking member”. Even if Veltri is silent during the Panel meeting, he is free to speak with the voting members during breaks and other down time. Therefore, even though he himself may be non-voting he can influence the votes of others. Perhaps the best way to avoid this would be to empanel him in a sound-proof booth like on the quiz show “21″, allow him to hear comments of others, but only permit him to speak into a dead microphone. That would also insure that the proceedings are kept clean.
Ed Silverman
Thanks Folks,
I confess that I was being a bit tongue in cheek earlier. I realize a non-voting member can, indeed, influence discussions and, in this case, may well present an industry insider’s view. Presumably, Veltri will be on his best behavior, though, in hopes that the other committee members will care to hear what he has to say.
Regards
Ed
Tea Partyer
Let freedom ring! Down with big government and regulation! It only serves to stifle capitalism. Everyone knows this is bad for the economy. Companies have to be able to make money without Big brother looking over their shoulder.
pharmavet
Does anyone know if there will be a webcast of the meeting?
Ed Silverman
Hi Pharmavet,
As far as I know, it can be viewed from one of these sites, although I believe payment is required….
http://www.fdalive.com/
https://secure.iian.ibeam.com/events/fdcr003/20100/index.jsp?autoLogin=
Hope this helps,
Ed
Condor
Thanks Ed!
“Let my [En]Rico [Veltri] go!”
With friends like these. . . I think Avandia is toast. We’ll know — in two days’ time.
Namaste
pharmavet
Thanks, Ed. My last pay-per-view was Ali vs. Foreman, 1974. Maybe it’s time to try it again.