"Kick the FDA" sweeping the nation as the favorite cocktail party game

from the author disclosure of first article:
"Dr. Wood reports receiving lecture fees from the Pharmaceutical Research and Manufacturers of America, serving on the board of directors of Antigenics, and serving on the scientific advisory board of Sapphire Therapeutics, in which he holds stock options."

Oddly enough, he doesn't mention financial conflicts of interest as an issue requiring reform at the FDA. While he is correct that the FDA's funding hasn't kept pace with its responsibilities (though the funding level of the FDA is almost directly determined by lobbying of members of Pharma), I think the things that need the most change are transparency and elimination of conflicts of interest.
howzabout:
1. ALL phase III clinical trial data have to be made public when the drug is submitted for approval; trials have to have their original endpoints analyzed even if the companies decided to switch endpoints during the trial or end the trial early.
2. No financial conflicts of interest for FDA employees and advisory committees involved in drug regulation, including agreements not to seek employment with companies(or their direct competitors) whose products they assessed for a full year after their regulation duties.
3. Compulsory monitoring of the first ~10,000 patients who take the drug for several years. Again, there has to be mandated public disclosure of the data - right now data like this is a patchwork, and the data is often suppressed for years if it's ugly.

1. Why analyze abandoned endpoints? That makes no sense.

2. Not to seek employment for a year in their field of expertise? How are you going to get competent people if you put that kind of restriction on things? A big part of the problem now is that most FDA personnel have a limited understanding of the industry and how it functions (and must function) in real life, and thus often fail to make comprehensive enough inspections.

3. And you think drugs are too expensive now? You're going to monitor every patient for every new drug for years? That would literally add tens of dollars to the cost of *each* dose of *every* drug approved, whether there was any reason to suspect a safety concern or not.

All this stuff is not cost-free, nor do we have infinite resources to do them. The FDA is already overstretched, and each of the measures you propose above would add greatly to that with little value added. Risk-benefit calculations need be made for any actions taken to determine if the action is worth taking; reflexive calls for more and more regulation without paying real attention to the whys and hows will not improve the situation.

PhilB

philB:
1. Why analyze abandoned endpoints? If they switch from mortality to a surrogate marker endpoint, they should still evaluate mortality. Easy enough. And if they terminate a trial because the results are just so darn good - would it hurt to continue monitoring the health of the subjects and report after the orignal end of the trial? Both of those issues have caused problems in the past few years with both efficacy and side reactions of new drugs that could have been prevented.
2. After working for the FDA The MDs could still practice medicine, the professors could still do academic research, and the administrators could still move to companies for which they didn't regulate products. Everyone can stay in their field of expertise, they just can't profit from their decisions at the FDA. I'll put it to you that ANYONE taking money from a company or negotiating to do so is ipso facto not fit to regulate that company. It's been proven again and again and again in this country and around the world.
If you work in pharma: how would you feel if the majority of the members of the advisory committee that will rule on your first-in-class product had financial relationships with your competitor, which happened to submit a very similar product six months after yours? What if they have been offered jobs at that company but haven't told anyone yet?

3. Please. Monitoring doesn't cost that much. Pharma has been doing postmarketing trials (It's called phase IV) on drugs in order to get accelerated approval for years, and often doing it voluntarily to look at getting approval for off-label usage. The difference would be in forcing them to start the trials immediately and then making them cough up the data in a reasonable time, instead of companies sitting on bad data for years while the execs sell off their stock. We'd end up with a more accurate list of contraindications, the company ends up selling to a smaller patient population. Sorry about that.

This is not a reflexive call for regulation; it's addressing the whys (conflict of interest, trial data hidden to protect the company's interest over that of its customers, trial endpoints fudged to put drugs in a better light) and the hows (transparency, better and more comprehensive data, uncompromised regulators).
And I think these measures would chalk up quite nicely on the risk and cost benefit ratios when you look at how much money has been spent on drugs that aren't effective, are dangerous, or are being misprescribed. If measures like this had been in place years ago Vioxx would have been marketed to a smaller population from the start and might still be on the market today, as opposed to being a ten billion dollar clusterf*ck.

All of your measures may make sense in some cases, but not in all, and should be mandated to be universal practice.

1. Sometimes analyzing the abandoned endpoint might be useful, but often not. And by saying "would it hurt to continue monitoring" you severely underestimate the cost and time it takes to do this stuff. The biggest part of development costs for a drug are spent in the clinical trials; it's not a trivial effort at all to extend them, continue to monitor, analyze extra data sets, etc. One of the major problems with our system now is that drugs are very expensive; you really are talking about making that problem significantly worse.

2. Agreed that any one who takes money from a company they regulate is doing wrong. That's corruption, and is already against the law. If someone is found to be doing this, he should be punished severely. But restricting their prospects after leaving the FDA is not going to change the regulatory behavior of either the corrupt or the honest. It's a pointless we-gotta-do-something add-on of no value.

3. Again you severely underestimate the cost of phase IV trials. They make sense in some cases, but mandating "compulsory monitoring of the first ~10,000 patients who take the drug for several years" for every drug as you suggested is a huge effort that will be *long ways* from "doesn't cost that much". Just think about any given patient; however much it will cost to monitor that patient for "several years" is exactly what would have to be added to the cost of *that patient's* prescription to pay for it. That's not going to be less than a couple hundred dollars *per patient*. That would hurt a lot of people. A better (and better enforced) system for reporting adverse events would accomplish nearly as much for far less cost. But compulsory monitoring of every patient for years is not even close to tenable.

Not to mention that such additional measures and costs will make many more drugs of marginal profitability not worth developing. Another problem we already have is that it is often not financially viable to develop drugs for conditions that are not common. The more difficulty and expense we add to the process, the more conditions and patients that will not have treatments made available at all. That's not cost-free either.

PhilB

Look. The FDA has been compromised in the past 7 years. As has the EPA and other government bodies. It's all politics.

Blame Bush.

Not the FDA.

Yeah. It's his fault.

I completely agree that Bush has been a crappy president. But he's not responsible for everything that is wrong with this country, and it's not all going to magically get better once he leaves office, no matter who replaces him.

Especially when the message is delivered as above, with the subtext that the Democratic Party is so much better. They're responsible in recent years for slightly less than half our problems as well. And they mainly aspire to be responsible for slightly more than half. Hooray.

PhilB

In essence, the FDA represents the people; we own it. So if we're unhappy with how it is performing, it is our - through our representatives - fault for not fixing it.

You have to admit it is pretty hard to swim upstream against a marginal majority swayed by propaganda, a President who flaunts the separation of powers, and sides exclusively with big business when it comes to government regulation, then add to that the first 6 years of that Presidency having a Congress that not only rubber stamped everything the President wanted, but blocked all Congressional oversight including going so far as to deny a room for an oversight committee to even meet in. Then consider that same President appointed hundreds of puppets in as many positions in regulatory agencies as possible, and let any action not favored by big business to either go unfunded or unstaffed.

And, yes, it was that bad.

PhilB:... it's not all going to magically get better once he leaves office, no matter who replaces him....

While I do think Bush and his insider crowd are responsible, I agree with this part of your comment. It is disconcerting that we've not heard the question asked or answered by any of the candidates, "Are you going to clean house of all the Bush appointed cronies and puppets, and are you going to see to it regulation enforcement actions left unfunded by Bush are refunded when you take office?"

The FDA has been politically compromised in the past 7 years by the present administration.

Place the blame where it belongs.

PhilB: Ahem:
1. Granted, it wouldn't be applicable in all cases. But trials where they swap endpoints or patient subgroups after looking at data midway have already lost a great deal of their predictive power; it's just how statistics work. And trials that end prematurely don't yield information on long term health effects.

2. Changing their job prospects WILL change the actions of the (potentially) corrupt. Not that long ago Boeing lost a multibillion dollar government contract because the government contract negotiator was also negotiating for a job at Boeing. Legislators and their staff are not allowed to lobby congress for 1-2 years after they leave their offices, and have to make public any job negotiations they make while in office. If someone approves a drug for Pfizer and then turns around and accepts a job there a few months later their judgement is suspect.
3.
Like I said, companies already monitor for adverse events after their drugs are approved:
http://www.pharmalot.com/2008/04/us-sues-bayer-over-baycol-fraud/
"As part of her job, Simpson gathered adverse events reports for Baycol and other statins and, in her report, noted that Baycol garnered much larger numbers, but she was ordered to remove that info before the report was distributed internally �in case of litigation.� Simpson also alleges Bayer orchestrated a ghostwritten article in the American Journal of Cardiology to allay fears over rhabdomyolysis and that Bayer reorganized its safety team to better control the flow of information about Baycol side effects. There was also a misleading press release and a Dear Doctor letter that downplayed the side effect."

It wouldn't be a new added cost, it's an existing cost. Drug companies already routinely gather and analyze the data. The difference would be in making the data public in a timely fashion as opposed to publishing it or burying it as the company sees fit.
How expensive? let's budget $200 (your number) a patient for a drug rep to gather reports of adverse events once a year for two years and have someone tabulate the data. Total cost for 10,000 patients: about 0.2% to 0.4% of the cost of getting the drug approved.

more on "why analyze abandoned endpoints?"

http://www.pharmalot.com/2008/04/what-negative-data-paxil-selective-rep…
Back to the new study. A key finding about 329: There was no significant difference between the paroxetine and placebo groups on any of the eight pre-specified outcome measures. However, by the time the data were analysed, many other new measures had been added to the list of secondary outcomes.

"For instance, along the way, four of the eight negative outcome measures specified in the protocol were replaced with positive ones. In fact, the study finds that in 329 disclosures of overdose and mania were edited out, and serious adverse events on (Paxil) were attributed to other causes. And the original authors of 329 didnt conduct any analysis to determine whether side effects occurred more commonly on Paxil compared with a placebo."

well, that's why.

1. (and "more on") Yes, the data should be well analysed where it exists, but again that's not the same as your initial proposal to always mandate that the studies be continued and analysed for the original endpoints regardless. It was that blanket proposal that I was objecting to as onerous and unrealistic. And of course when you're dealing with editing out adverse effects or other falsities as noted in your last post, you've gone past endpoint analysis into clear investiagort bias (if not actual fraud), which is a whole different problem than changed endpoints.

2. It's ALREADY illegal to negotiate for a job or otherwise take money or favors from companies you are regulating. If someone approves a drug for a company, and then "turns around and accepts a job there a few months later", that does not mean corruption at all if he was not looking into that at the time he was regulating the product. Conversely, someone who went over and approved the applciation for a drug might well be a very logical person to bring on board after he leaves the FDA; he's already familiar with the product and the regulatory situation, and might be quite valuable in applications for the same drug in other countries, for instance.

I'm not saying there's never any corrpution. I'm saying there are already adequate laws against corruption; enforce them, investigate allegations, and so on. Adding extra layers of restrictions on the honest doesn't advance the cause.

3. Yes, companies often monitor patients afterward. More often, the monitoring is passive; any "adverse events" are supposed to be reported, and the company is expected to followup on those, but that's different than actively monitoring thousands of patients who may or may not have adverse effects. And again, when you're talking about removed info and ghostwritten articles and such, you've moved off the subject; such shenanigans have nothing to do with the amount of monitoring; if someone is willing to mislead or commit fraud, that's (a) already illegal, and (b) a different problem.

And that $200 per patient for 10,000 patients come to another $2,000,000, for each new drug, that has to come from somewhere. Maybe that's less than 1% of the cost of bringing the drug to market, but that's not trivial.

skeptigirl wrote: "It is disconcerting that we've not heard the question asked or answered by any of the candidates, "Are you going to clean house of all the Bush appointed cronies and puppets...", ". Of course they will. They have to make room for their own cronies and puppets, don't they? As if cronies and puppets are a brand new phenomenon invented by Bush. (rolleyes)

PhilB

skeptigirl,

There are approximately 3,000 presidential appointees. When his term of office expires, so does theres (with the exception of an odd few, like the governors of the Federal Reserve Board, whose terms overlap presidential terms). The vast majority will be swept out by the new president, even if it's John McCain, as these are not party appointees, but essentially personal appointees.

Phil is right, the new president will appoint his (or her, although that now looks unlikely) own cronies. Some president's cronies have been less egregious than others, however, and it would be quite a shock if the next president's cronies were as bad as the present ones.

That said, however, I'm not implying they're the sole cause of the FDA's troubles. As with any bureaucracy, it faces a lot of perverse incentives and conflicting demands.

the fda cannot be taken seriously...not after they approved reclast for osteoporosis and tried to market the theoretical harms of cough syrups for kids as a major medical breakthrough.....their news bites are just so crappy.....

I really enjoy your blog, that I've read a few times, I've thought that a blog might correspond with your interests...
This blog is called Scientific Discovery and collects any important news regarding the global science world such as biology, health, medecine, technology, engineering, astronomy, environment ect.

Here is the adress: http://blog.elabspace.com/science/

Regards,

Pauline

I'm assuming that like other clinical studies that informed consent of the participants would be needed. What if people who would benefit from a hypothetical drug wanted treatment without being involved with the study? If it is unethical to continue a study that shows early on that a treatment is far more effective than placebo, in what circle of hell would it be ethical to deny patients a drug that has proven to be (at least enough to get FDA approval) safe and effective, just because they do not want to participate in a study?

Perhaps some are forgetting the actual purpose of these drugs we are debating- to help treat people who are ill, not just to produce data for some to drool over.

Please visit:

http://nostradamus-usa.netfirms.com

to see how we stopped James Randi's Million Dollar Paranormal Challenge and caused the President of American Atheists, Ellen Johnson, to quit her job...

Donde esta AT?

nice post thanks.

thnaks you. nice job.

thnaks you. nice job.

There are approximately 3,000 presidential appointees. When his term of office expires, so does theres (with the exception of an odd few, like the governors of the Federal Reserve Board, whose terms overlap presidential terms). The vast majority will be swept out by the new president, even if it's John McCain, as these are not party appointees, but essentially personal appointees.

Phil is right, the new president will appoint his (or her, although that now looks unlikely) own cronies. Some president's cronies have been less egregious than others, however, and it would be quite a shock if the next president's cronies were as bad as the present ones.

That said, however, I'm not implying they're the sole cause of the FDA's troubles. As with any bureaucracy, it faces a lot of perverse incentives and conflicting demands.

it is nice thanks

thanks you for post it is nice