FDA Panel Votes To Keep Avandia On The Market
32 CommentsBy Ed Silverman // July 14th, 2010 // 3:38 pm
And so the vote is in - 20 FDA advisory committee members voted for GlaxoSmithKline’s Avandia diabetes pill to remain on the market, but with either restrictions or labeling changes. By comparison, 12 panel members voted to withdraw the drug. There was one abstention.
The FDA doesn’t have to follow the advice of its panels, but generally does. And so, FDA officials face a Solomon-like situation - instead of maintaining the status quo or withdrawing the drug, they have impetus to develop a package of caveats that will appease some doctors and patients who maintain Avandia is their best choice, despite the cardiovascular risks. This possibility suggests Avandia could remain available elsewhere, although European regulators hold their own review next week.
For Glaxo, the outcome is certainly not the worse-case scenario, but this is also far from perfect. A risk management program and strengthened black box warning is certain to diminish sales going forward. Since the 2007 meta-analysis in the New England Journal of Medicine trumpeted heart risks, Avandia sales fell by about two-thirds worldwide, according to Sanford Bernstein analyst Tim Bernstein, who now calls the drug a “keeper.” In any event, Avandia represnts only about 2 percent of Glaxo sales.
The FDA panel, by the way, also voted in favor of allowing the continuation of the controversial TIDE trial - which the FDA required Glaxo to run to compare Avandia heart risks with Actos - if the drug remains on the market. The vote was 20 in favor of continuing the trial and 10 against, with two abstentions. Earlier, 18 panelists voted that Avandia studies raised “significant safety” concerns compared with other diabetes drug, while six panelists voted no and nine couldn’t make a decision.
“I understand that there may be a lack of consensus on the effects of this drug at the FDA. But the result of the advisory committee vote is gravely disappointing and raises serious questions as to whether the science was presented in an unbiased manner to the committee members,” says Congresswoman Rosa DeLauro, a Connecticut Democrat, in a statement. “Given the compelling evidence revealed in the weeks leading up to this advisory committee meeting, it is incomprehensible that 20 committee members still would voted to keep Avandia on the market…I urge the FDA to fulfill its mission of protecting the American public and taking this drug off the market, especially considering there are less dangerous alternatives.”
“We will, of course, continue to work with the FDA in the best interest of diabetes patients who face this chronic and serious disease,” Ellen Strahlman, Glaxo’s chief medical officer, says in a statement. “GSK is dedicated to sharing data about its medicines transparently and in a timely manner. We remain fully committed to maintaining best practice disclosure of clinical data to serve the interests of regulators, physicians and patients.”
stay sign thx to sangrancito on flickr creative commons
natasha
what a dumb panel.they voted not to terminate a clinical trial so they can find out in six years if avandia kills more patients.
patient advocate?? is GSK paying her off. she does not speak on behalf of most patients.
FDA failed in its job to protect the public, by requiring the company to do a proper clinical trial and collect real results. the FDA is a farce.
My Take
Big Pharma wins again! Patients lose again! What else is new?
DrugLag
We can now go overseas in order _not_ to get the drug!
Justce in MI
Just watched PBS Newshour on this. Krumholz describes a recommendation of significant restrictions on use, including who can prescribe.
Does that sound right? What relevant precedents are there for this–DEA licenses? Does this make a difference?
Doc
A group of highly trained experts vote on the safety of a drug after a comprehensive meeting where numerous opinions on the data were presented. The data is inconclusive, and the conclusions drawn by either side (FDA & GSK) depend on how one massages (statistics!!) the number. The outcome of the vote supports keeping the medication available to patients with stricter labeling to follow.
This is how science works, data is presented and debated and then in this case voted on by a group of experts. Saying the FDA has not fulfilled its mission is ridiculous, this type of forum is exactly how these issues should be resolved. This isn’t politics where the loudest voice is the presumed winner, as the congresswoman would like. I’m sure the congresswoman did not watch a bit of the proceedings or ever look at a single piece of data or publication related to the drug. I guess its easier to slam the FDA & drug companies for a sound bite rather than pass some legislation which might help her constituents.
SteveM
I understand how complicated the trade-space can be for pharmaceuticals.
However the biggest process gap is in the failure to inform the patient about the nature of the risks. Tell the patient up front about the Avandia issue and let him decide. But…
A nice little systemic dance in off-loading the need to inform has been stylized. The doctor off-loads it to the pharmacist, and the pharmacist off-loads it to the PI sheet included with the prescription package.
And the PI sheet is either too dense and/or too cryptic for the patient to understand. Plus, he intrinsically trusts his care providers. And even if he doesn’t, he can’t even ask the right questions.
So it’s CYA all around. The docs are happy, the pharmacists are happy and the drug companies are really happy.
Can someone explain my assessment of the process defect if I have it wrong?
Ed Silverman
Hi Justice in MI,
That’s a good question. Avandia is not a controlled substance - and I didn’t see the PBS news program - but you give me the impression that the drug would somehow be elevated to that level. Certainly, a Risk Evaluation & Mitigation Strategy, informed consent and patient guide might be envisioned. As to who would be allowed to prescribe, that’s something else.
Regards
ed
Salmon
Largely agree with you Steve. I would add that the PPI sometimes is too dense and/or cryptic and other times it’s too little and just fluff. Otherwise I think you’ve largely got it right.
By the way this isn’t the first time I’ve seen GSK commission a flawed study to refute and muddy the waters on a significant safety issue.
Also didn’t we also see this happen with Ketek where the flawed safety study was run by PPD whose CEO Ernest Mario was also formerly CEO of Glaxo.
As for scientific debate at these AC meetings. I’ve been to enough of them to see typically several types of members: 1. In the Company’s pocket (of course this includes the nonvoting pharma member and typically the patient rep who comes from a front org with funding from pharma) 2. Scientific Lightweights who don’t even understand the basics of what the underlying science really is (Typically the vast majority of the committee. That’s why you have MDs making judgements on statistical issues they know nothing about) 3. Those who don’t want to say anything to jeopordize their careers. (Vote for the 3rd option) 4. Highly competant scientifically and will say something (typically only 1 or 2 at most if you’re lucky). (When I say lucky if you have 2 of these it’s because under some circumstances they can drive the committee.)
As for the presentations you had very smooth GSK people and high level FDAers intentionally introducing doubt where there really isn’t.
All in all the results were predictable.
Salmon
Salmon
Ed, the FDAAA 2007 required this type of AC meeting. It also requires that FDA try to keep the drug on the market in the least restrictive manner possible. So we may have a REMS and maybe a patient guide but I doubt restricted prescribing or informed consent.
Salmon
Justce in MI
I think Salmon’s last point is quite important. If both docs and pts truly had to “sign off” on something, then I think this takes us out of the “warnings mean nothing” category. But I still haven’t found what the AC is actually A’ing in this regard.
Also wondering how much the ghost of Vioxx played any role, at least for some. As we know, FDA itself was, at best, ambivalent about Merck’s decision to withdraw, and, as I understand it, more inclined to recommend some version of what has been recommended here.
In any event, this will probably help GSK re: lawsuits, but not in many other ways. I can imagine their wishing, down the line if not already, that it had turned out differently.
Paul
Agree with Doc.
There are people that are so blinded by their hard positions that they will never concede even when reason, albeit unpopular, has prevailed.
Nathan
Thanks Doc — my sentiments exactly. What are the congresswoman’s credentials? It’s easy to bash scientists when you don’t understand the issues they are grappling with. Everyone has an opinion. Not everyone has an INFORMED opinion.
harpy
“GSK is dedicated to sharing data about its medicines transparently and in a timely manner. We remain fully committed to maintaining best practice disclosure of clinical data to serve the interests of regulators, physicians and patients.”
this. I don’t care about the drug - they should be nailed for this.
Ex-FDAer
What troubles me about the decision is that according to FDA reviewers Marciniak and Johann-Liang in one case Glaxo concealed safety information and in another case intentionally hid deaths and other serious AEs with Avandia in the RCT. Since this later trial is the basis for Glaxo shedding doubt on the observational data I believe that the credibility of the study needs to be taken into account.
Having reviewed numerous submissions to FDA from companies I can say that misrepresentation of deaths and serious AEs or intentionally not conducting proper assessment and follow up is a common tactic and is invariably used by FDA management on a regular basis to dismiss safety issues and prevent safety information from being released.
Based on my INFORMED experiences I am inclined to believe that Glaxo’s safety study should be ignored and more weight should have been give to the observational data.
As to JIM’s points. I doubt that there would be any actions at that level. At best I think we might have a black box warning and a PPI which may or may not be dispensed and will likely be tossed anyway. We have data to show that these sort of warnings etc. do very little to alter prescribing. Especially with a drug that’s already on the market where prescribers are likely to say well I’ve never had a problem with it (how would they know based on an N of perhaps 100 for the typical prescriber).
I’d say Glaxo clearly played this extremely well and won. If FDA does not do something strong I would say that transparency under Hamburg is a joke. For what good is it to be transparent in data presentation and decision making when you’re basing the ultimate decision on fraudulant data.
The FD&CA used to be about precaution. Instead having gone through public FDA documents I know and can point to drugs that are being approved even when the studies relied on don’t show efficacy. Meanwhile safety issues are being held to an almost impossible standard and are otherwise dismissed.
The decision in this case will not only effect drugs on the market but will also send a message to the reviewers in OND.
DON’T raise any concerns and just approve the drug. You don’t have sufficient data to stop it or even to force us to label it and even if the supporting data eventually vindicates you we will manipulate things to minimize lost sales to the company. In the interim you know what happens to anyone who raises any concerns about a drug.
Patients are the losers.
Justce in MI
I’m struck by the range of interpretations across “pundits”–Nissen, Krumholz, and others.
Re: GSK, I suggested above that, while it may help in the short run re: lawsuits, a continuing focus on Avandia (how it should be labeled, restricted, etc.) increases the likelihood of more bad press and more stuff (whistleblowers, et. al.) coming out of closet.
My prediction is that they will let the current hooplah die down (they hope), and withdraw the drug themselves within a year. They’ll blame the insurmountable pressures coming from pols, media, greedy trial lawyers, and also credit their own concern that there should be any such questions about a medicine of theirs.
anon
hamburg has to yank the drug. graham and marciniak did a great job. the vote is not relevant. more than half of the d’s would have voted for E, if they knew FDA management would back them up. lets see if hamburg sets a new course for drug review.
Sam the Man
Agree with Doc. Once you adopt ex fdaer’s POV that anyone who disagrees with your position is either an idiot or corrupt, you’re no longer a meaningful participant in in the discussion. All you can do at that point is sit around and bitch with people who already agree with you.
Ex-FDAer
Sam nothing like putting words in other peoples mouths is there.
I have provided my INFORMED opinion and do not appreciate ad hominum attacks.
I did not come to my opinion lightly but rather through years and years of multiple and repeated experiences with a number of drugs where I was unfortunately ultimately vindicated in my concerns. Thus I am inclined to believe Marciniak. In my book credibility is an issue in the present case.
If I hadn’t had the years of experiences I had, I don’t know if I would have voted any different than the majority of the AC members. Not that they are idiots or even corrupt but they may simply not have the experiences to recognize very slick and practiced manuvuers.
I guess in Sam’s word I was an idiot for I know I was not corrupt before and since I clearly don’t agree with Sam now I must be an idiot now too.
pharmavet
I predict that FDA will not yank Avandia for the simple reason that it would only leave one glitazone on the market for diabetes. FDA is loathe to disrupt the marketplace by creating a de facto monopoly for Actos, and that is why Avandia will stay on the market for now.
Salmon
Pharmavet,
While I agree that FDA will not yank Avandia, I don’t believe it will be due to the reason you stated. As I mentioned before you simply need to look at the FDAAA 2007.
The following is from FDAAA 2007 and includes FDA’s legal options at this point short of limiting distribution.
“(e) Additional Potential Elements of Strategy.–
“(1) In general.–The Secretary, in consultation with the offices described in subsection (c)(2), may under such subsection require that the risk evaluation and mitigation strategy for a drug include 1 or more of the additional elements described in this subsection if the Secretary makes the determination required with respect to each element involved.
“(2) Medication guide; patient package insert.–The risk evaluation and mitigation strategy for a drug may require that, as applicable, the responsible person develop for distribution to each patient when the drug is dispensed–
“(A) a Medication Guide, as provided for under part 208 of title 21, Code of Federal Regulations (or any successor regulations); and
“(B) a patient package insert, if the Secretary determines that such insert may help mitigate a serious risk of the drug.
“(3) Communication plan.–The risk evaluation and mitigation strategy for a drug may require that the responsible person conduct a communication plan to health care providers, if, with respect to such drug, the Secretary determines that such plan may support implementation of an element of the strategy (including under this paragraph). Such plan may include–
“(A) sending letters to health care providers;
“(B) disseminating information about the elements of the risk evaluation and mitigation strategy to encourage implementation by health care providers of components that apply to such health care providers, or to explain certain safety protocols (such as medical monitoring by periodic laboratory tests); or
“(C) disseminating information to health care providers through professional societies about any serious risks of the drug and any protocol to assure safe use.
clarkkent
Plenty of precedence for restricted availability of drugs, see Acutane and Thalidomide.
The difference in lawsuit money paid out between Vioxx and Celebrex should point to the reason GSK will not withdraw Avandia.
Justce in MI
Yeah, I’ve heard the lawsuit argument that Mr. Kent makes.
But what the Daily Planet does not know is what Merck’s liability would have been if they had kept going.
I suspect they had plenty of smart lawyers involved in deciding when to bail. And, indeed, I’d say the end result bears them out. The Celebrex saga isn’t over, and is a very different case for other reasons.
As for Acutane and thalidomide, if Avandia is ever that restricted, I agree, there’d be no point in doing anything but let things take their course.
pharmavet
Salmon, I agree with you that the FDA will use the legal options that you state as the public and formal rationale for keeping Avandia alive. Of course they would never publically state that they might be trying to influence the marketplace, but behind the scenes I believe that it somehow factors into their decision making process. FDA always likes to see at least two competitors in the marketplace; thus they were ok with removing Rezulin because there were at least two other glitazones in the same class that were on the market or entering the market with proven track records of safety.
ex-FDAer
Pharmavet, I beg to differ. Based on personal knowledge I believe that Rezulin would have been removed even if only one other glitazone came on the market. I don’t believe it’s an issue of monopoly but rather having another option of a drug in the same class. Personally I don’t think that having a replacement should have mattered as there were many other therapeutic options available for diabetics.
vince
The FDA under Woodcock’s orchestration once again choose to be a handmaiden to a bad drug. The same eminent physicians ‘ who allowed this drug in the first place decided to allow it agin ’surprise ,surprise’. Oh well to quote a line from the ‘Godfather’ … There’s a lot of money in them white powder..”
pharmavet
Ex FDAer I concede your point, but I think that you would agree that it took an inordinately long time to remove Rezulin from the market.
Based on my own experience I think that FDA is surprisingly indifferent when it comes to safety signals from the liver. In the mid 1980’s I worked on a CNS product that was causing a high incidence of elevated LFT’s 3-4X ULN. When the drug was removed the LFT’s went back to normal and there were no clinical problems. FDA wasn’t particularly concerned either. Long and short it took 11 years for the regulators to convince themselves that hepatotoxicity was a real problem and the drug was withdrawn.
I’m not being facetious when I say that FDA doesn’t pay close attention to liver safety signals until a few patients have had to undergo liver transplants.
Justce in MI
Definitely agree with Vet re: “inordinately long time.” Mac Lumpkin still defends it; so do others. It took a virtual war, whistleblowers, et. al. for withdrawal to happen.
It is defended, retrospectively, as having to do with being the sole glitazone. But, at least in the U.S., there is only one cox-2 (although not entirely FDA’s wish). I’ve heard folks at Pfizer blame Glaxo’s undercutting trog so that rosi would fare better. Their “theory” is that that’s why trog was withdrawan much earlier in the UK.
ex-FDAer
Pharmavet,
I agree 100% with everything you said in your last post. I have observed first hand how FDA handles signals of hepatotoxicity from a number of different drugs. It’s not only that FDA doesn’t pay close attention but in my experience there’s intentional downplaying and dismissing of things internally as well as trying to minimize things externally and that includes misrepresenting the science in white papers and in scientific conferences designed for that purpose. I’ve also seen what happens after signals are ignored or dismissed and the drugs have gone onto the market and it takes a lot more than a few patients before anything is done. The situation you describe with a drug that causes elevations (acute toxicity) and then resolves when the toxin is no longer given is not unusual. Neither is it unusual for LFTs to go back to normal with continued therapy. These common patterns are used by FDA to dismiss concerns as they don’t fall under Hy’s Law however, more needs to be considered. What about sicker patients. Are patients likely to receive other hepatotoxic drugs or to be at higher risk due to kinetic interactions, etc., etc.. Also with stopping a drug are you preventing a Hy’s Law case from developing as it means that not sufficient liver has died off in a short enough period of time to cause someone to be sick enough to require a transplant. These cases are typically seen in phase I trials with high doses just beyond or at the top of the therapeutic dose range but aren’t picked up in phase III as very few get the highest doses neither may there be any monitoring. So what happens when the dose is pushed in practice especially in say a 85 lb greatgrandmother whose liver is already half gone. Also chemical structure and known relationships to hepatotoxicity are completely ignored. Rezulin being a case in point. Based on the chemistry adn metabolism it should have been obvious very early in development that this would be a possible issue.
Also mechanisms of hepatotoxicity other than acute necrosis are ignored. We definitely know about structures and biliary problems as well as chemical structures and steatotic liver disease. So when there’s added tox from EtOH it’s used to dismiss any drug induced steatosis that will compound the problem in the real world.
I know the FDA players and in every case the so called experts inside FDA made the wrong call (i.e. the drug was exceptionally hepatotoxic when they dismissed the signal.) I know of no case where the FDA’s ‘experts’ ever said a drug was a potential problem and only 1 case where they didn’t out and out dismiss the possibility and that was when they knew Congress had already been informed and was involved.
larry
Sanjay Kaul, MD who voted to restrict avandia and seemed very anti-graham, is affiliated with the TCT crowd New York Interventionalists known for dodgy research and even more dodgy finances. Marin Leon is part of this group. Interesting Pocock who is affiliated with TCT and CRF, did the stats for GSK meta analysis. Pocock is affiliated with industry and again dodgy research Woodcock has shown herself to be a supporter of this sort of research-business model. business comes first, then adapt the research to the business model this was demonstrated by the RECORD trial. No results, missing data etc. it was a post-marketing trial, one that is used to sell the drug rather than collect data.
SteveM
Re: pharmavet - “Based on my own experience I think that FDA is surprisingly indifferent when it comes to safety signals from the liver.”
Great point. My favorite FDA administrative pathology is psycho-pharm. Which is replete with me-too drugs offering little or no additional clinical value over legacy drugs.
So a garbage drug like Cymbalta is found to wreck livers (and induce suicide, and trash patients who discontinue and…and…) Yet there’s a panoply of less dangerous anti-depressants that can be prescribed before Cymbalta is ever considered.
But will the FDA ever make that recommendation? Heck no. Will the FDA ever challenge Eli-Lilly advertising that suggests Cymbalta is a first line treatment? Heck no again.
I imagine others can make the same argument about when and why Avandia is actually indicated.
FDA is messed up…
pharmavet
Salmon, I thought of another example of how FDA tries to influence the marketplace through the approval process. Feel free to disagree. Those in the Psychopharm area will remember a drug called Buspar (buspirone), which was the first non-benzodiazepene drug approved for General Anxiety Disorder. BMS’s NDA for Buspar was cleary substandard. One of the two “pivotal” efficacy studies was a 20 patient trial out of Germany, I believe. I also believe that FDA’s approval was influenced by its desire to go get the first non-BZ into market.
For students of history, FDA is also influenced by external events. Some may recall that approval of the early oral contraceptives was very controversial. When I was in college I read a new book by Stanford professor Dr. Paul Ehrich, entitled “The Population Bomb”. In the book Ehrlich predicts that the world’s population growth would eventually outstrip available resources, as per the original prediction of Thomas Malthus. Many, including those at FDA took Ehrlich’s book quite seriously, particulalrly his recommendation of what became known as Zero Population Growth (ZPG). Responding to this mantra the FDA accelerated approval of the early OC’s.
Salmon
Pharmavet,
Interesting comments. I can’t agree or disagree simply because I don’t have any info.
I took a look at the public info on Buspar at FDA’s website. It was approved in 1986. There are no reviews or SBA available and the oldest labeling is from 2000 and there is no information on efficacy studies in the labeling. If it were approved based on one of the efficacy studies enrolling only 20 subjects (presumably since it’s for GAD it would be a very limited duration study) that would be interesting info.
By the way where did you come by this info?
While I have no doubt that people are influenced by information and current views I doubt that FDA is monolithic and don’t know to what extent something like OC approval would have been influenced by politics. While I don’t doubt that it would have been controversial to some people Since the law says a drug has to be safe and effective (relative of course) I don’t know how much legal leeway FDA would have had not to approve such a drug.