File this under 'What was he thinking?' In an op-ed in The Wall Street Journal, former FDA commish Andy von Eschenbach proposes that the "FDA should approve drugs based on safety and leave efficacy testing for post-market studies." How would this work? Well, Andy suggests creating pilot programs in which patients would be entered in registries, and the FDA and drugmakers would later determine whether a medicine is effective.
"Take regenerative medicine," he writes. "If a company can grow cells that repair the retina in a lab, patients who've been blinded by macular degeneration shouldn't have to wait years while the FDA asks the company to complete laborious clinical trials proving efficacy. Instead, after proof of concept and safety testing, the product could be approved for marketing witn every eligible patient entered in a registry so the company and the FDA can establish efficacy through post-market studies."
In other words, this approach envisions substituting epidemilogical studies for double-blind, randomized clinical trials. This is ironic, because Andy laments that the FDA is losing its status as the gold standard among regulators, which was the rationale for writing his missive. "By empowering the FDA to create new paradigms for evaluating the most promising innovations, Congress can ensure that the FDA serves as a bridge-not a barrier-to cutting-edge technologies," he intones (here is the op-ed).
Of course, the pharmaceutical industry would benefit, since fewer trials would be required prior to approval, which would lower development costs. Meanwhile, payers - both private and public - would provide coverage for medicines, even though efficacy remains unknown until post-marketing study results are reported. Then again, the industry has a poor track record of initiating post-marketing studies promptly, or even at all. How long such a scenario would exist is unclear.
Imagine, after all, the reaction if studies are not completed on a timely basis or show a lack of efficacy. Meanwhile, registers are ringing and some patients are being treated with medications that do not serve the hoped-for goal. Andy, who is now chairman of the Manhattan Institute's Project FDA, believes that "breakthrough technologies deserve a breakthrough in the way the FDA evaluates them." Would this approach amount to a breakthrough? Or a setback? What do you think?
Should Efficacy Be Tested In Post-Marketing Studies?
- No (78%, 201 Votes)
- Yes (22%, 57 Votes)
Total Voters: 259
34 Comments
In areas of unmet need, something like this could work. Especially as we embark on the age of electronic medical records and data mining, which will make reporting / evaluating efficacy less centralized. It isn't a big leap from accelerated approval, or the recent guidance on the biosimilar approval process.
Excellent idea to speed approvals. This would be nothing more than the situation that existed pre 1962, when the Kefauver Harris Amendment was enacted into law, requiring that drugs be tested for efficacy as well as safety. The pilot efficacy/registry program that von Eschenbach mentions would be the framework to reintroduce the DESI system for classifying drugs as effective, ineffective, or needing further study. However, instead of using DESI retrospectively lets apply it prospectively to von Eschenbach's pilot efficacy studies.
This would streamline review time, cut the cost of R&D and put the burden of the real proof of efficacy back in the hands of those bast equipped to evaluate it: the physicians who prescribe these drugs.
How about the pharmaceutical industry not charge for the meds while they are in this efficacy testing mode?
In theory they are then putting their money where they mouth is. Their benefit would be that they would have an established and well informed market when they finally are able to formally launch and start charging. There would also be increased motivation to get the studies done quickly. And if they fail to quickly prove efficacy the companies would be sure to get them off of the market because they would not be getting paid.
Could be a win for everyone.
This would lead to consumers having many expensive new options and no way to know which are beneficial. Increased costs - definitely. Improved health - maybe.
More importantly, it would wildly exacerbate the phenomenon described in a recent Jama "Reversals of Established Medical Practices", which shows clearly that once in regular practice, stopping use of a treatment is very difficult, even if it clearly doesn't work. See stents for stable coronary disease, bevacizumab for breast cancer, vertebroplasty, and more.
oii has clearly summed up the historical background that I didn't have time to verify when I first saw this. This would be a "fine eample of turning back the clock" - except I would choose a diferent adjective ...
Given oii's enthusiasm for it, read after I first saw thsi and voted on it, I am truly glad I voted "NO.' Nice to have one's judgment validated, as it were.
This is tricky and may only make sense in areas where there is true unmet medical need. That is something, efficacy not clearly established, is better than nothing. Kind of like the push towards "compassionate care" during the Aids crisis.
On the otherhand, where there are other treatments available, with demonstrated efficacy, is it worth the risk to use the public at large as "guinea pigs" to prove or disprove the efficacy of a new drug? Will there be any formal monitoring and protection like in a clinical trial? Is it ethical to prescribe a drug of unknown efficacy when a known effective drug is available?
As I understand, prior to the Kefauver era, there were relatively few effective drugs at all. In that environment, the focus was rightfully on not making things worse with unsafe poisonous drugs. However, with the advent of "wonder drugs" this moral dilemma evolved.
As I also understand, at that time, it was recognized by regulators that proving efficacy could not be done by individual physicians acting alone or in groups. As I recall the AMA believed individual physicians could determine efficacy but were overuled.
So this issue should also be carefully considered. Forgive me but "Those who forget the past are doomed to repeat it"
Maybe with the internet, and some careful regulations, meaningful data could now be collected and analyzed quickly, post launch.
However, I hardly believe that drug companies would be in any hurry to get any drug off the market (e.g. Voixx). Hence, there is still a place for regulations.
In summary, I beleive this is tricky issue.
Good Night!! Sounds like stepping into it big time.
"Andy laments that the FDA is losing its status as the gold standard..." Just now realizing that?
By the way that very comment about being the gold standard was/is the foundation of the preemption argument. Huh
"Imagine, after all, the reaction if studies are not completed on a timely basis or show a lack of efficacy." - Best to just announce that this is finally the standard practice and get it over with!
Are we sliding uncontrollably down the slippery slope?
Observer, thanks for backing me on the historical info. Perhaps "pilot studies" could be thought of in a way other than registries. For in fact, what we do in Phase II are also called pilot studies or early hypothesis testing. I would submit that two well designed, well controlled Phase II pilot efficacy studies of 100-200 patients that strongly suggests efficacy could then lead to market authorization without going te "gold standard" Phase III. The MD could then be the determiner of efficay.
Observer, since you are familiar with DESI, you might agree with me that in our litigious society you could not do a modern day DESI retrospective review. When I was with Abbott, as in many companies we had to deal with lawsuiits from patients who had suffered adverse events from DESI drugs that had been newly classified as "not effective", whereas pre-DESI there there was no efficay statement required.
Of all the comments read, Bossart's seems the most appealing compromise. Of course, this idead won't work unless there was a rock solid system for tracking "experimental" patients, full disclosure to the patients that they are participating in a trial drug, and that effectiveness rating be consistent and fully reported.
The suggestion is nightmarish, in my view.
What the thalidomide disaster showed (and which was the reason Kefauver-Harris passed in the first place) is that docs are NOT able, on their own, to monitor either efficacy or safety.
Like a lot of extreme right positions, it is substituting faith-based policy for scientific evidence. Gingrich made the same suggestion in the 90s, btw, although in an even more misguided missile way.
There would be more injuries. If only by way of non-treatment. The huge cost of studies would shift to patients and insurers for ER visits, hospitalizations, increased labs, increased doctor visits, alternative and corrective products, lost wages, etc. Nothing is free.
This should put an end to preemption. If FDA is not checking for safety and effectiveness then they could not assume responsibility. We must then rethink what defines a frivolous suit.
Here's a thought: what if the same non-test for efficacy was applied to class-3 devices, where preemption reigns.
"Whoops, sorry, that pacemaker doesn't work. We'll have to...Mr. Brown, are you listening to me. Mr. Brown? Whoops...."
Are drugs that different. As Who's post above, some (probably very large) number of folks with serious illness will go without treatment at all. While the industry makes hay.
This is lunacy disguised as policy.
JiM, to get a device approved it is pretty much a non test for efficacy. You just have to show that your device is no worse than what's already out there.
If you look back to editorial in Science editorial 23 sept 2011 you will see the same idea - from Andrew Grove, former CEO at Intel
I think the post-market efficacy study approach ignores the fact that "safety" is a relative concept. All drugs have some side effects and unintended outcomes. Thus, we judge a drug to be safe and effective when its potential benefits outweigh its risks.
Without efficacy data, we can't determine whether a drug's risks are worth it.
Ah, words from Andy, a not-so-golden oldie...
Given that the FDA's prescription drug safety record is already crummy, what would they be left to "do" if testing for efficacy was saved for...later?
Here's hoping PDUFA reauthorization reflects some semblance of this reality. Hope springs eternal.
Boy! Andy really wants to lower the bar for drugs, because it's convenient to doctors and makes a lot of money for pharmaceutical companies. Oii is right, you just have to show that your device is no worse than what's already out there. It's never been shown that the use of PET results in cancer patients live longer (does the modality in question improve clinical outcomes), it only had to show that it was no worse than what's already out there (CT, MRI, Ultrasound, X-rays, etc.) or just manage patients on the basis of history and physical examination ("how are you feeling?"). All the CMS/ACRIN data showed in the NOPR registry is that doctors pay attention to the test results. And CMS' National Coverage Decision was based on "accuracy" of PET, not "efficacy."
Greg is right. Whereas a decent CAT scan may pick up enough metastatic cancer to stage it as IIIB, for example, a slightly more sensitive and much more excpensive PET scan may pick up enough extra tumor to call it a Stage IV. In the end the treatment is pretty much the same and you'll be dead in about the same amount of time.
This to concur with Health Scientist.
FDA safety and efficacy are relative matters, as we all know, and the clinical trial system is based on making the best guess after weighing both within the already limited data and review period.
Whether via phase IV, or through a much larger pool of pt. experiences, the original balancing act may well be rebalanced.
So the current system is already minimalist (yes, I know folks here who put trials together will not agree) relative to the longer history of a drug. Still, without even this first step, we back to the thalidomide era.
Re: device efficacy, thanks for comments. Still unclear how this varies from drugs. Isn't the drug efficacy criteria also "better than nothing" (i.e., placebo).
LOL,.. Andy must be having a Prozac moment.
Isn't this another example of a pendulum swinging? After Thalidomide we got Kefauver-Harris that requires efficacy studies. This worked and got increasingly complex until the 1980s with AIDS. Then Act-Up, among other,s argued that AIDS patients needed rapid access to experimental drugs as they were dying and there was no alternative. So they got early access and... some of those drugs didn't work.
Martin Delaney was one of the strong proponents of speeding up access but much later admitted that there might be some usefulness in efficacy studies so that people are not harmed by useless drugs or drugs with more side effects.
Of course, today is a new day. We have more tools to track all of those patients who are experimental naturalistic subjects. Implants with GPS are good.
Eschenbach's idea is dangerous. Half of drug candidates fail phase III for lack of efficacy of safety, and, from the half that reaches regulators, 40% do not get approved because they aren't good enough to bother. That's a 70% attrition rate from phase III to regulatory decision. Eschenbach's proposal would expose patients to drugs that are either ineffective or dangerous in 70% of the cases! And patients (or society) would have to pay for the privilege! I can't see how one can ethically justify this.
Bernard, I believe what I was taught: a good drug should show efficacy in Phase III trials IF they are designed properly. Likewise an otherwise efficacious drug will FAIL in Phase III trials if the trials are poorly designed. My experience has shown that there are good drugs that should have been approved but failed in Phase III due to poor design and/or poor execution. Why not give these drugs a second life postmarketing with Andy's proosal?
Re: OII above, aren't companies free--indeed, encouraged--to confer with FDA about trial design before initiating them? If failure is due to poor design, one wonders how likely it is that the drug itself is well designed? But I will take your word for it that you believe you have seen such.
In the meantime, you are suggesting that the correction to a poorly designed trial is not to design a better trial, but to test out efficacy on with no prior trial.
It would be interesting to have real data for this discussion--like the drug effiacacy study that followed Kefauver-Harris. As I understand that history, a lot of drugs that docs "in their experience" thought were useful did not pass when tested scientifically.
Re: Erica's point, Act Up changed its views over time. When the right tried to enlist them in the "deregulatory" efforts of the mid-90s, they rejected the invitation.
BTW, if anyone doubts the politics involved in A's proposal, do a google search with "Manhattan Institute Sourcewatch."
Usual suspects as supporters--Koch bros., big tobacco, right-wing family foundations, etc.
Not that there's anything wrong with that--just useful to understand the poltical soul brothers with whom Andy currently hangs.
Perhaps the biggest problem with this idea is the question of reimbursement. As a recent "Health Affairs" article pointed out, it is exceedingly difficult for an insurer to stop paying for a treatment once it has started to do so even if that treatment is later shown to be ineffective. It's awfully hard to explain the details of the situation to a patient who is sick or in pain and may well be grasping at straws that "...no, we used to cover this product, but now we have learned that it probably won't help you." Could any health plan in the country make that argument in a credible fashion, even if it were true?
OII, I think trial design is key, but am not sure poor design is at the heart of the problem. Drug companies are encouraged to consult with FDA about drug design. I would also think that, with many companies having been around for decades, they should know how to do it right. In any case, they can also avail themselves of the the best expertise, and you would think that, before, committing tens of million of $ to a trial, they would do just that.
I feel the root of the problem is rather 'pipeline stuffing'. CEOs need a pipeline to show Wall Street they are doing their job and their companies can withstand patent expirations. Unfortunately, innovation cannot be ordained. So the heads of R&D comply by moving to the the clinic poorly-researched and marginally innovative compounds that are doomed to fail, and indeed fail.
The innovation crisis is not a failure of science. It's a failure of leadership and governance. That's what we must fix.
The other issue I have with Eschenbach’s proposal is that, if companies under pressure can get to the market without demonstrating efficacy, they will flood it with compounds of dubious efficacy, and then create demand by using all the marketing tricks in their bag. It will be a race to the bottom that will not serve patients or innovation. Scholars such as Grabowski and Glenn have shown that demanding regulatory standards produce a world-class pharmaceutical industry, while permissive standards destroy innovation by removing the incentive to perform. Let's not dismiss the lessons of history.
(this is the continuation of my previous message. I pushed 'submit' too early.)
The innovation crisis is not a failure of science, it's a failure of leadership and governance. That's what we must fix.
The other issue I have with Eschenbach’s proposal is that if companies can reach the market without demonstrating efficacy, they will flood it with compounds of dubious value, and create demand by using all the marketing tricks in their bag. It will be a race to the bottom that will not serve patients or innovation. Scholars such as Grabowski and Glenn have shown that demanding regulatory standards create a world-class industry while permissive standard destroy innovation by removing the incentive to perform. Let's not dismiss the lessons of history.
Bernard, I agree with you about pipeline stuffing. To know this all you need to do is attend or listen to a few dog and pony shows where the CEO'S present the pipeline powerpoints to the investment community, attempting to impress them with their rich pipeline. Used to work, but the analysts have become much more sophisticated lately.
Andy was drunk, right? Had to be to come up with this Beavus and Butthead *idea*...
Billions of $$$$ invested in computers to do modeling and DNA research and we're back to "...it's the packaging that sells, not the ingredients..."
Make stuff up, note it only killed one subject (human/animal) and then launch it to see what it does...? Seriously? No target/goal to hit like *efficacy*?
Please, make the world a safer place and just go home if that's what you do at *work*.
A new low, if that was possible...
we can see how well that works from the Avastin/breast cancer debacle. and if the drug in question does turn out to be ineffective, or worse actively harmful, - who takes responsibility? I can just hear it now - "we didn't know!"
it is true that at the current time devices passed through the 510(k) process only have to show they're not much different from an approved device, but that is on the verge of change. or on the verge of the verge of change. it could tip back the other way. but the IOM strongly recommended radical reassessment of the 510(k) process and many people have been gravely wounded by products that passed because of it. most recently women treated with surgical meshes.
Even Europe, not long ago emulated for its quick and easy device approval standards, is facing some tough scrutiny and perhaps revisions of its CE Mark approval process.
he should be fired immediately. the idea of it. ridiculous.
This is a truly outrageous suggestion that supports the concept so beloved by many--public investment, private profit. The suggestion that efficacy be based on post-marketing studies means, after being sold studies. So these studies would be paid for by the public--the people buying medicine, which might or might not work. Or it would be paid for by insurance companies, which would cost all of us paying for it.
Who would benefit? The pharma companies, which would no longer have to pay for clinical trials--would in fact turn clinical trials into a profit center, since every drug would have to be bought, and paid for, to be "tested."
The only way this would work is if the pharma companies donated the drugs--see how that suggestion flies.
another example of how well this works with devices is DePuy's ASR hip replacements. from today's NYT: http://www.nytimes.com/2012/02/15/business/hip-implant-the-fda-rejected-was-marketed-abroad.html
Just more hype from von Eschanbach. Here's a blast from the past:
On February 9, 2007, the Food and Drug Administration and the National Cancer Institute held a groundbreaking seminar on immunotherapies: Bringing Therapeutic Cancer Vaccines and Immunotherapies Through Development to Licensure. Hosted by Andrew C. von Eschenbach, M.D., who was sworn in as the 20th Commissioner of the U.S. Food and Drug Administration on December 13, 2006, the theme of the seminar was summed up (search on 'February 9, 2007') by Dr. von Eschenbach as follows:
"Cancer vaccines and immunotherapy are ready to move beyond proof of concept studies, and we are ready to move them to the patients as safe and effective treatments for cancer. Therefore our primary focus over the next few days is to share our collective expertise and find solutions to the challenges of bringing cancer vaccines and immunotherapies through development to licensure."
The intent, simply put, was to find ways to develop therapies that stimulate a patient’s own immune system to fight cancers. Importantly, for an agency that frequently is viewed as a roadblock to the introduction of innovative, novel treatments for life-threatening diseases, Dr. von Eschenbach stressed that under his direction, the "FDA will be a bridge to the future, not a barrier."
As it turned out, Dr. von Eschenbach’s talk of the FDA being a “bridge to the future” was a cruel joke. Events almost exactly 3 months later would show his “bridge” to be nothing more than agency’s euphemism for Saint Benezet's bridge in Avignon, France…a bridge that only extends halfway across the Rhône.
This guy is just full of wonderful ideas. So...how did the one above work out for you?