Bioethics discussions are fun when you aren't about to die.
July 26, 2007 6:48 PM   Subscribe

Dying for Lifesaving Drugs: Will desperate patients destroy the pharmaceutical system that produces tomorrow's treatments?Reason Magazine
posted by BrotherCaine (43 comments total) 2 users marked this as a favorite
 
Medical News Today article. Sorry, I just realized this post needed more work. I've been posting on other blogs where the post quality is generally lower lately.
posted by BrotherCaine at 6:56 PM on July 26, 2007


"After last year's ruling in the alliance's favor, the FDA argued that the group no longer had legal standing to sue it, since none of the patients who had signed the original affidavits were still members. They were all dead."

*blink*

Holy crap!
posted by ZachsMind at 6:58 PM on July 26, 2007


That's a very loaded article, at least from the first few paragraphs I've read so far. Lots of people die from experimental drugs. There was an interesting article a while back about a doctor who was treating AIDS patients in the 80s. One guy couldn't get into a trial, and the doctor felt that if he had, he probably would have lived. Another patient got into a trial, and died like most of the people who tried the drugs.

The other thing is that there are laws that allow drug companies to give out 'compassion' doses of experimental drugs without affecting trials. As we go forward, though, taking these drugs will become more and more like Russian roulette. Maybe it'll save you, maybe it'll kill you. My feeling is that people should be able to take the risk, but who is going to want to be the control group?
posted by delmoi at 7:01 PM on July 26, 2007


The Libertarians at Reason Magazine don't believe in "control" groups, just "free market" groups... (LOLIBERTARIANS)
posted by wendell at 7:05 PM on July 26, 2007 [1 favorite]


Sound medicine interview (mp3)
Hosts: Barbara Lewis, Eric Meslin, PhD
Interview: Frank Burroughs, President
Abigail Alliance for Better Access to Developmental Drugs
posted by BrotherCaine at 7:10 PM on July 26, 2007


Two words: informed consent.

According to the articles, most of these drugs have already passed initial FDA approval for human testing; there is already some general idea of how the drugs will work. Are they dangerous? Yes, potentially lethal - but if there's no other reasonable alternative I don't see the problem in letting informed patients have access.

As long as the whole process has an open peer-reviewed system - the patient is informed of the potential risks and rewards and voluntarily opts into treatment, and the entire protocol (successful or not) is recorded - then the data obtained should be just as useful as if it were obtained through clinical trials.
posted by Benny Andajetz at 7:11 PM on July 26, 2007 [1 favorite]


Anyone who argues for the unencumbered right of patients to take developmental cancer drugs must grapple with the fact that 94 percent of them will not work.

Is it really that loaded? I tend to read with multiple sides of an issue in mind, so I don't always see bias as strongly as some readers.
posted by BrotherCaine at 7:14 PM on July 26, 2007


Benny, except that as they point out no one is going to want to volunteer for double blind studies where they might wind up in the control group if they have another option to get access to the drug.
posted by BrotherCaine at 7:16 PM on July 26, 2007


Benny, except that as they point out no one is going to want to volunteer for double blind studies where they might wind up in the control group if they have another option to get access to the drug.

Good point - I don't know how I missed that. Very thorny ethical question. I understand the scientific desire for double blind studies, but that's not the only way to get useful information.
posted by Benny Andajetz at 7:31 PM on July 26, 2007


94 percent of them will not work

I don't like that statement very much.

The problem is the definition of "working." The drug approval process typically requires very large number of subjects and typically must more efficacious than currently available drugs where the criteria for "efficacious" can be ephemeral or convoluted. The end-points for clinical trials are typically set out before the trial opens (to gain approval); the drug may do something really well but if the initial end-points that were set out for the study aren't met, then - no approval for the drug.

Also, drug trials only seek approval for very specific things. There could hypothetically be a drug that's approved for one type of cancer but (while it might be marvelous for treating another type of cancer) is not approved for the other kind.

Other issues: getting treated early in the development of the cancer is a good thing. The longer the cancer has developed, the more difficult it is to treat it.

Also, lots of drugs fail - but many more fail during phase I (toxicity - usually in healthy subjects) or phase II (doseage in patients) than in phase III (efficacy - treating much larger numbers of patients) trials. Drugs that make it to phase III have a very good chance of not killing the patient but is only hopeful for treating the patient.

I'd be perfectly comfortable with entering experimental treatments if I understood what that treatment entails and why it might work for my version of illness. For example; I'd be fine with adoptive transfer immunotherapy for cancer but I'd never touch (nor ever let anyone) that super monoclonal anti-CD28 treatment that was discussed in (ask)metafilter previously.
posted by porpoise at 7:43 PM on July 26, 2007


From the magazine that brought you this. The writers of Reason, as a matter of course, stand firmly against regulation of anything, and I'm sure they have someone on staff who would argue with a straight face that hospital sanitary regulation is a terrible evil.
posted by StrikeTheViol at 7:44 PM on July 26, 2007


crap - posted the wrong link, and can't find the one I was thinking of.

I need to find something that tests my blood alcohol content before letting me post anything to metafilter

posted by porpoise at 7:45 PM on July 26, 2007


From the magazine that brought you this.

Did you actually read it? Don't confuse good intentions with good lawmaking.
posted by Mr. President Dr. Steve Elvis America at 8:00 PM on July 26, 2007


Those of you writing this off as another Reason free-market screed don't appear to have read the article -- there's some really interesting stuff near the end about adaptive clinical trials, which my advisor works on. The piece raises a lot of questions generally about whether the current process is really the best in light of new medical and statistical methods that have emerged, and even offers some pretty scathing criticisms of the pharmaceutical industry's profit motives.
posted by transona5 at 8:02 PM on July 26, 2007


Even more than the regulatory barriers, this deep-seated fear of disorder works against the alliance's agenda. "I don't have a right to fly somebody's experimental airplane," reasoned Bruce Chabner, clinical director of the Cancer Center at Massachusetts General Hospital, Boston, in the August issue of the New England Journal of Medicine, "so why should I have the right to some drug that a company has dreamed up?"

Right. Because if Aunt Sally can't fly an SR-71, she'll be dead in three months.

*hopes that douche gets inoperable penis cancer*
posted by Avenger at 8:11 PM on July 26, 2007 [1 favorite]


well, my partner died because he didn't qualify for a protease inhibitor (AIDS) trial circa 1995. his read counts were fucked up. had he gotten into that trial i believe he would be alive today. protease inhibitors are now the life saving medicines keeping people with AIDS alive. i guess you know where i stand on the issue.
posted by brandz at 8:16 PM on July 26, 2007


Yes, I did. It's a flawed law, but without it, I wouldn't have been able to get inside the restaurant I had lunch in a few days ago. Reason would rather have private charity and the kindness of strangers do the job. As flawed as the system is, I'd rather wait on the behemoth to draw around its gaze than have to supplicate myself at parishes and synagogues to get things accomplished there.
posted by StrikeTheViol at 8:16 PM on July 26, 2007 [1 favorite]


To make my peeve clearer, I agree with the fundamental goals of the Abigail Alliance. I think that disbanding the FDA, for which I believe this was written as ammunition, is a bad idea.
posted by StrikeTheViol at 8:31 PM on July 26, 2007


Wow. Talk about a dishonest article. The article makes a big deal about how AIDS activists managed to push the FDA, but it doesn't mention that AZT is often used as an example of how fast-tracking and similar can fail and actually result in more patients dying because ineffective drugs or treatment regimens are detected later and better drugs or regimens aren't researched earlier.
posted by Ethereal Bligh at 8:32 PM on July 26, 2007


I was going to post a response to Mr President Thingy, but StrikeTheViol has it.
posted by spaceman_spiff at 8:33 PM on July 26, 2007


Yes, I did. It's a flawed law, but without it, I wouldn't have been able to get inside the restaurant I had lunch in a few days ago.

You like a bad law because it benefits you. Perfectly understandable, but of course not everyone is going to share your view.
posted by Mr. President Dr. Steve Elvis America at 8:34 PM on July 26, 2007


Naturally, but plenty do. A minority of those who don't wouldn't care if I lay dying in the street, much less are able to visit their shop, which makes me like this terrible law just that little bit more.
posted by StrikeTheViol at 8:43 PM on July 26, 2007


if the truth be known, AZT was the only drug available to those with AIDS in the mid-to-late 1980s. AZT certainly kept many, many of my friends alive, for years.
posted by brandz at 8:45 PM on July 26, 2007


This whole "control group" bullshit is monstrously stupid. We fucking well have incredible reams of statistical information about every fatal disease known to man -- why must we have half the people in a clinical trial die to reinforce those statistics?

I have a degree in math, sure the quality of the statistics will be impaired BUT if a drug is sufficiently powerful to be useful then the results will stand out like a sore thumb even if there is NO control group.

Get rid of the control group concept -- instead, spend a little extra money on data mining, you could do just as well and include everyone in these trials. Control groups make sense when there are "animals" who are "sacrificed," not when human lives are at stake.
posted by lupus_yonderboy at 8:55 PM on July 26, 2007


A minority of those who don't wouldn't care if I lay dying in the street, much less are able to visit their shop, which makes me like this terrible law just that little bit more.

That's a non sequitur. Anyway, you linked the article as an example of Reason taking an absurd position, but now you seem to admit that the ADA isn't a great piece of legislature, but it doesn't matter, because it benefits you. Apparently Reason should realize that it benefits you and not speak out against it for that reason.
posted by Mr. President Dr. Steve Elvis America at 8:55 PM on July 26, 2007


transona5 - adaptive clinical trials, which my advisor works on

Thank you for mentioning this. I'm less bitter about a system that kinda sorta works (hey, it's better than before the days of a FDA, but perhaps the FDA has gotten old and cancerous like Windows Vista) and slightly more optimistic that someone is working to reform the system.

In the current system, there are lots of patients who aren't good for the trials, and regardless of the system, trials that aren't good for the patients. Patients need to be educated - just because it worked or seems to work for "someone they know" doesn't mean that it's the right trial/treatment for them.

The "give me antibiotics, bitch!" vibe that a lot of people with illnesses that are (prima facie) being treated by highly hyped experimental treatments who aren't getting said treatment really bothers me.

Illnesses and diseases (and who gets them) are not homogeneous; a given disease/syndrome can have many different causes and the bodies that have them have many many different reactions to both the illness/disease and the treatment.

With that said, brand - yeah, in the current system, the drug trial in progress may very well have saved your partner, but your partner didn't meet inclusion criteria (or had exclusion criteria) - it's a damn damn damn shame that in most trials, anyone receiving the treatment has to be included into the statistics. The people who enrole patients into trials want the trials to be successful - putting patients who's illnesses that are too far advanced (even though they may benefit from the treatment) increases the odds that the treatment won't be approved because the patient will succumb to other problems not related to the specific drug in question.

Someone getting run over by a car has to be included in the mortality column even though it has nothing to do with the treatment (in all of the trials that I've been involved with).
posted by porpoise at 9:01 PM on July 26, 2007


What I meant to infer is that some private citizens have such limited regard for the disabled, that I would not expect them to treat me as a person unless law existed to state as much. More to the point, what would you replace the ADA with, MPDSEA? How 'bout the FDA? I think that both are flawed, but very much preferable to a solution involving placing both matters into the hands of whatever variegated private social services might exist at the state or local level.
posted by StrikeTheViol at 9:08 PM on July 26, 2007


More to the point, what would you replace the ADA with, MPDSEA?

It should be publicly funded, to begin with. If we think providing the handicapped with access to public accommodations and employment is a good thing, we should be willing to pay for it directly.

Particularly in the realm of employment, the ADA replacement would distinguish between two different issue that the current law conflates: 1) irrational prejudice against the disabled and 2) rational recognition that persons with certain disabilities cannot perform certain jobs as well as able bodies persons without incurring additional costs. If Congress wants #2, it should subsidize it.

Congress should give more thought to precisely who it is trying to help and draw up a clearer definition of "disabled."
posted by Mr. President Dr. Steve Elvis America at 9:26 PM on July 26, 2007


Interesting, MPDSEA. You'd tie it to political lobbying, then. In order to have a chance for consistent funding, advocacy organizations would have to push for appropriations yearly. It certainly could be done, it would simply mean the pace of adjustments would vary by how much the lobby managed to get in a given year. Given the growth of the AARP and disabled veterans organizations, it might even be possible to tax businesses to pay for it. What would you do with the FDA?
posted by StrikeTheViol at 10:09 PM on July 26, 2007


I'd just like to take a second to point out that this probably shouldn't be an argument for or against the existence of the FDA.

Most people on mefi can agree that #1) pure food and non-deadly drugs = good, while #2) waiting for life-prolonging medicines for no other reason than "the industry needs a body count" = bad.

We can hopefully have #1 without #2 even with the FDA still in existence, I think.
posted by Avenger at 11:40 PM on July 26, 2007


This whole "control group" bullshit is monstrously stupid. We fucking well have incredible reams of statistical information about every fatal disease known to man -- why must we have half the people in a clinical trial die to reinforce those statistics?

Hear that, everyone? We're done with medical research. That's it, it's over. We pretty much have all the information there is to get about every fatal disease known to man, so we can just throw away the idea of a control group, which is the very underpinning of that quaint concept called the scientific method. Hooray! I'm heading out to my local doctor and asking for the drugs that will make me super-fast, just like the Flash!

Thank you, lupus_yonderboy!

Speaking of which, is there's a cure for lupus? You know, the disease called lupus? Let me check with Wikipedia...

Lupus is treatable symptomatically, mainly with corticosteroids and immunosuppressants, though there is currently no cure.

Hmm, no cure. I guess they haven't gotten the message that we're done with medical research ... I'll go edit the article right now.
posted by Cool Papa Bell at 11:43 PM on July 26, 2007


Cool Papa Bell:

Methinks you're willfully missing lupus_yonderboy's point.
posted by Benny Andajetz at 5:42 AM on July 27, 2007


so we can just throw away the idea of a control group, which is the very underpinning of that quaint concept called the scientific method.

Not quite. It's a useful tool for implementing the scientific method when better statistical methods aren't known or applicable.
posted by transona5 at 6:35 AM on July 27, 2007


This whole "control group" bullshit is monstrously stupid.

Nope.

The purpose of the control group is not to develop statistics on the proportion of untreated subjects that have positive/negative results. It is to keep the researchers unaware of who is getting treatment and who isn't, so the researcher's bias does not affect the results. Double-blind studies (help) insure that the only difference between the treated and control groups is the treatment.
posted by hexatron at 7:19 AM on July 27, 2007 [2 favorites]


Hey, "Cool Papa Bell" -- you should be a little more polite, particularly when you are wrong.

The idea of the control group is not a part of the scientific method -- in fact, it's unclear to me exactly what a "control group" might mean in the context of physics, chemistry, astronomy, geology, most sciences.

A control group is (as transona5 says) a tool for getting better results in some sorts of statistical experiments, no more and no less.

For example, you have 200 cancer-prone rats -- 100 of them get exposed to a potential carcinogen, 100 are the control group and are left alone, then the results are compared. If statistically significantly more rats who are exposed get cancer than the control group, you can say that the substance is carcinogenic.

Now, if you *knew* exactly what the chances of the mice getting cancer on their ownsome was, you wouldn't need the control group. However, there are a lot of good reasons for doing it -- you don't really know what this batch of mice is like, you don't really know what your lab environment is like, it significantly reduces the uncertainty.

This is great for science. It isn't really so great for those extra 100 rats as they get "sacrificed" at the end of the experiment.

In the sorts of drug experiments above, you'd deliberately not treat half of your subjects -- in order to get a baseline to make your measurements from.

Trouble is that these aren't lab rats -- these are human beings who would probably rather not die. Moreover, we have almost 200 years of medical statistics about the mortalities of various diseases -- we could use these standardized statistics instead of the control group. Medical science does this all the time already, as an example when they rate the effectiveness of trauma centers (where it would be unethical to withhold treatment from badly-injured people just to get better statistics...)

If the medical establishment really was motivated, there are huge piles of data lying around that are not accessible to researchers, those being the treatment results of working hospitals. Of course, no one wants their personal medical data to be thrown around wantonly but I can't see any argument against releasing your data to be used only in statistical aggregate, particularly if it meant saving other people's lives.

To summarize:

1. the idea of the control group has nothing to do with the scientific method,

2. the control group is a useful tool to improve the quality of your statistics in certain classes of statistical experiments, but

3. it is perfectly possible to run rigourous medical experiments without a control group, and in many cases this is the only ethical decision, and

4. aggressive data mining might compensate at least partly for the lack of a control group in drug tests.

Are we on the same page now?
posted by lupus_yonderboy at 7:45 AM on July 27, 2007


What hexatron said.

There seems to be some confusion about what treatment the control group gets, as well. In the drug projects I've been involved with (for serious diseases like cancer), the control group does not get NO treatment, they get the best currently available treatment. The experimental drug is compared to the best available therapy, and has to work at least as well to get approval (a drug that produces fewer side effects, with the same cure rate, is likely to be approved because it improves quality of life for the patient even though it's no more likely to cure them).

And once a drug is approved, it can be prescribed "off-label" by physicians for just about anything. (Naturally, physicians are expected to use good medical judgement here, and many are reluctant to go out on a limb, but off-label usage is acceptable.)

Drug companies are well aware of this and sometimes take the strategy of seeking approval for whatever disease seems like the fastest route to market, regardless of how "important" or prevalent the disease is. The anecdotal data that flows in from off-label usage can be used to justify a clinical trial for other indications.
posted by Quietgal at 8:00 AM on July 27, 2007


Are we on the same page now?

No. You're apparently completely full of shit.

It's one thing to argue for improved practices that may speed research. It's another to start throwing around language that we don't need something as fundamental as research based on control groups. I mean, you're saying "aggressive data mining" is one of the answers. Except that data mining can't happen without the ability to compare data. In order to compare data, you must have good data. And good data comes from data integrity. And data integrity comes from ... control groups. You know, the data used to ensure that the data you're mining is correct in the first place.

You're 180 degrees from right. Ergo, full of shit. In your zeal to be so open-minded about new approaches, your brains have fallen out. Really ... I implore you ... just shut the fuck up. Because you're making MeFi look bad.
posted by Cool Papa Bell at 10:19 AM on July 27, 2007


BUT if a drug is sufficiently powerful to be useful then the results will stand out like a sore thumb even if there is NO control group.

When this happens in real life, they can easily unblind the study and shift everyone to the treatment group. It happens from time to time.
posted by ROU_Xenophobe at 12:32 PM on July 27, 2007


See also.
posted by myeviltwin at 4:13 PM on July 27, 2007 [1 favorite]


The control group is not specifically a part of the scientific method but it's used in medical research to satisfy one of the core components of the scientific method. Specifically, it's used in "the formulation and testing of hypotheses." Every clinical study begins with the hypothesis that the experimental treatment is significantly better than the current gold standard treatment. Without a control group, how else are we to determine what is truly an improvement vs "noise"? The presence of a control group with protocols is what separates medical research from the snake oil salesman hawking wonder cures for cancer based on the anecdotes of 5 people who were "cured."

Without the control group, you have no choice but to trust that the population receiving an experimental treatment is completely identical to the national population. The control group is your guarantee of true effectiveness in the presence of two identical populations.
posted by junesix at 4:28 PM on July 27, 2007


Does every study need a control group? If another clinical trial with the same type of patient population has just been attempted, can you juggle the numbers for it's control group to obviate the need for your own control group? It's not like your control group is ever a 100% match for your trial group anyway is it?

Also, it may be possible to tell the people that are treating the patients/subjects that you have a control group even if their isn't one to preserve the double-blind aspect of the study.
posted by BrotherCaine at 8:36 PM on July 27, 2007


Well if it's simply an observational study, then of course there's no need for a control group. But for any type of comparison to be made, you obviously need something to compare it to. Having a control group is simply the most ideal way to do it.

Theoretically you could piggyback a study on the patient population of another study but it's unrealistic and impractical for clinical studies of novel treatments. Clinical studies take years to develop and carry out. If you wait until another study publishes its findings before you start your own, by the time your own study is complete the patent will have nearly expired if not completely expired. And that's if you could even get the raw data of the other study and I've never heard of a company sharing its raw data.

It's not like your control group is ever a 100% match for your trial group anyway is it?
For most studies, it's not a match in terms of the obvious demographic factors like age, race, weight, height, income, etc. But the control and experimental groups must be 100% match for what's called inclusion and exclusion criteria. Before a study enrolls the first patient, it draws up a set of patient selection criteria. For a cancer treatment, they'll look for patients at a certain cancer stage depending on the treatment, with or without certain medical conditions and histories, and with or without certain prior treatments. Every study has a list of criteria and these will vary from study to study. This study criteria variation is also what makes it challenging to adapt the results of one study in comparison with another.

--
Back to the OP, I'm all for allowing patients access to experimental drugs. No, I don't think it will destroy the clinical trial system. As long as a firewall is maintained between the study populations and the patients who want access outside of the study loop, I don't see the harm. If anything, the companies should embrace it as a "public beta" of sorts. While they work on their own controlled studies, the public access groups can be asked to submit data. It would provide a lot of valuable information for what happens when their treatment is used in the real world where there is wide variety of patient types.
posted by junesix at 11:36 PM on July 27, 2007


Some commentators (both at MeFi and elsewhere) say that allowing terminally ill patients to acquire experimental medicines outside clinical trials would undermine the entire clinical trial system. They argue that patients would stop enrolling in clinical trials because they would no longer want to "risk" being allocated to the placebo arm of a trial.

But this argument is based on what I think is a false premise.

The Abigail Alliance does not in fact appear to be lobbying for unrestricted access to experimental medicines for terminally ill patients. You can read their citizen petition to the FDA (dated June 2003), and their subsequent response to an opposition letter (dated October 2003).

Based on these documents, the Abigail Alliance seems to support access to experimental medicines for those terminally ill patients who do not qualify for participation in relevant clinical trials.

Tyranny of the majority comes to mind re the current system...
posted by gbognar at 6:12 AM on July 28, 2007


« Older Don't Go Breakin' My Art   |   The Gays Are Getting Married Newer »


This thread has been archived and is closed to new comments