1. SteveSC on April 19, 2005 9:51 AM writes...

OK, if the FDA allows drugs on the market after just safety trials some people will spend their life savings on a drug with unknown efficacy. But some of the drugs will work, and those patients will be better off. Some drugs won't work, but some of those patients would have spent their money on apricot pits or magnet therapy anyway. The issue is not whether a drug works, but where the decision is made. Why should a bureaucracy make the decisions for the 'average' person while forbidding the individuals and their doctors from applying their specific situation knowledge to the decision?

I think one problem is that the FDA decision is all or nothing. Once a drug is approved it is 'Katie bar the door!' and drug companies do everything possible to put the pills down people's throats with seemingly little regard on what is medically best. The FDA needs a few 'provisional' approval categories. For example, a lifesaving drug could be given a provisional approval good for only two years (or some other duration), not allowed to advertise, and the approval could automatically expire if certain efficacy measures were not made. Some drugs could leverage the controlled substances example and doctors could only prescribe them if they filled out (electronically hopefully) forms that included the data required to track efficacy and side-effects. Etc.

Fedex spends a lot more time knowing where its packages are than the FDA and pharmaceutical companies do tracking the effects of drugs on patients.

2. jsinger on April 19, 2005 10:58 AM writes...

OK, if the FDA allows drugs on the market after just safety trials some people will spend their life savings on a drug with unknown efficacy. But some of the drugs will work, and those patients will be better off. Some drugs won't work, but some of those patients would have spent their money on apricot pits or magnet therapy anyway.

Steve, if that seems like a good plan, why not have half the patients take the prospective drug, give the other half a placebo, do the whole thing in a rigorous way and have the pharma pick up the tab? And get the same outcome, but with some useful information about efficacy intead of worthless anecdote?

ie, the way we do things now...

3. jsinger on April 19, 2005 11:01 AM writes...

OK, if the FDA allows drugs on the market after just safety trials some people will spend their life savings on a drug with unknown efficacy. But some of the drugs will work, and those patients will be better off. Some drugs won't work, but some of those patients would have spent their money on apricot pits or magnet therapy anyway.

Steve, if that seems like a good plan, why not have half the patients take the prospective drug, give the other half a placebo, do the whole thing in a rigorous way and have the pharma pick up the tab? And get the same outcome, but with some useful information about efficacy intead of worthless anecdote?

i.e., the way we do things now...

4. Mike on April 19, 2005 2:09 PM writes...

I thought this was going to be Derek Lowe the pitcher's blog. Dammit.

5. SteveSC on April 20, 2005 9:57 AM writes...

Doing things the way we do now is not sufficient. Clinical trials still need to be performed, I am not arguing for dumping them. But entry into trials is very tightly controlled--not everyone can enter them--and for 'popular' drugs access can be difficult to get.

In any case, this misses my point. Why should the government make the decision once the drug is deemed safe? Are french fries safe and effective? They sure taste good, but as a nutritional source they have a lot of negatives, especially compared to other foods available on the market. So should the FDA ban french fries or just let the consumer decide whether to coat his/her arteries in fat?

6. qetzal on April 20, 2005 1:07 PM writes...

SteveSC,



One problem is that drugs are never really deemed "safe." Safe is really short-hand for "safe enough for the intended use, given the reasonably expected benefits."



Nevertheless, I'm sympathetic to your argument. It does seem hard to justify, from first principles, that the gov't should have to power to decide what you can use as a drug. Particularly for promising new drugs targeting life-threatening conditions, it seems reasonable to argue for early access.



I think there are some big practical issues, though. For example, Pharma's not going to invest in large scale manufacturing unless they're allowed to market the drug and sell it at a suitably profitable price. If you tell them they can't do that until after they've proven efficacy, they won't usually have enough incentive to make the drug generally available at an early stage.



Another issue is that making the drug available early can actually make it harder to prove efficacy in subsequent trials. Suppose PharmCo has a new drug, Cancerase, that people think is really promising for late stage lung cancer (based on animal studies, let's say). FDA tells them they can start selling Cancerase based on adequate human safety data, but if they don't prove efficacy in 4 years, using standard Phase III trials, it will get pulled from the market.



PharmCo agrees, starts selling the drug, and also starts trying to test it in Phase III. You have late stage lung cancer, and they want to recruit you for their trial. You have two options. You can participate in the trial, and have maybe a 50:50 chance of getting assigned to the placebo arm. Or, you can just buy and use the drug, since it's readily available anyway. If you're a desperate patient, which are you likely to choose?



Because they're already selling Cancerase, PharmCo may now find it difficult to recruit enough Phase III patients to prove efficacy in 4 years (or whatever time limit you want to set). Depending on the details, they might have been better off waiting to market Cancerase once they'd proven efficacy.



Then there's the issue of whether Medicare or insurance should pay for unproven drugs in this scheme. Of course, you could say "No," and argue that people who can afford it should still have the freedom to pay for it out of their own pockets if they wish. I would agree with that, too, but such an approach would be a political & social minefield.



Bottom line for me - I would like to see more options for making 'unproven' drugs available to folks with life-threatening diseases. I think there are ways to do that better than we do now. But barring a major revolution in our drug regulatory framework, I think this will continue to have a pretty limited scope.

7. Derek Lowe on April 20, 2005 1:56 PM writes...

And there's also the legal matter. Unless there's some sort of tort protection, it'll be a great temptation for the lawyers to sue us for unexpected effects of experimental drugs. Even when they're clearly marked "experimental."

8. SteveSC on April 21, 2005 10:02 AM writes...

qetzal,

I agree that safe is a relative concept, after all, even water and oxygen are not 'safe' in too high a volume or concentration. But to be used in people we need some data on where the lines of toxicity will be crossed.

Regarding manufacturing, by the time a company is doing Phase III studies it has already completed +90% of the work required for commercial scale manufacturing, e.g., all the formulation, analytic and process validation, component testing, etc. All that is left is tranfer to the commercial facility (which may be the same as the development facility). The cost of manufacturing itself is, except for some biologics, a very small portion of the costs, and I doubt this would be much of a factor in the investment return analysis.

On your points about clinical testing, you may be right that patients will avoid trials if they can afford the drug. But I wonder how much things will change once the FDA (and pharmacos) have to admit that we don't know whether some drugs work? Right now, the typical person (and many doctors) believe that if the FDA has approved it, it must work for anything the drug companies tell them. Also, real world use can add to the information on efficacy as long as there is some coordination of data collection. If all else fails, money can be thrown at the problem--pay enough an people will sign up for the trial.

Regarding Medicare payments, I think a 'provisional' status would help clarify the situation. Medicare doesn't pay for a lot of FDA-approved treatments now (mostly medical devices, since drugs have not until recently been generally reimbursed by Medicare). Not paying until the data is in makes sense, and patients can sign up for a trial if they can't afford the drug.

As for Derek's legal issue, this risk already exists for drugs in trials, and for drugs being used off-label.

Bottom line: yes it will take a 'revolution' in the regulatory framework to solve the problems, but we have to start somewhere.

9. SteveSC on April 21, 2005 10:08 AM writes...

Sorry about the long paragraph

I am trying this one with the HTML tag