Professor Robert Rosenheck presented his blockbuster findings on long-acting risperidone at Stanford Psychiatry Grand Rounds this week. The injectable version of this anti-psychotic medication was heavily marketed as superior to its oral, less expensive cousins. The sticker price of $7,000/year was promised to be worth it because patients would need to be hospitalized far less frequently while on long-acting risperidone. The manufacturer raked in billions of dollars, much of it from the public purse.
But as Bob’s randomized trial in New England Journal of Medicine shows, none of the claims about this drug were true. Fortunately, Bob decided to launch his trial while federal research dollars were more plentiful than they are today. With the current withered state of federal medical research, pharmaceutical industry-funded studies are often the only source of information about whether a medication is effective and cost-effective.
And even when the government can support a study, it is today sometimes dependent on the manufacturer to donate the medication. In the case of the long-acting risperidone study, which was started 5 years ago, the donation was valuable but not critical (something like a half a million dollars within a ten million dollar federal budget). Today, the federal government budget available for such a study would be less, potentially leaving the researchers dependent on the manufacturer’s donation in order to conduct the trial. This is even moreso the case with recent biologic therapies that cost many times as much as long-acting risperidone.
Why does this matter? If the government approaches a manufacturer of an expensive drug about a study of whether an approved medication is no better than a generic drug or may be harmful to particular groups of patients, the manufacturer has economic veto power. The manufacturer can say, in effect, you can’t afford to do this study without a donation of our medication, and we won’t give it for any study that might harm our market share. On the other hand, if you want to compare the medication to one of our new alternative medications (all of which have more years left under patent protection), why then of course we’d make a public-spirited donation to the scientific endeavor.
This study was funded by both the VA Cooperative Study and by the drug manufacturer. It is not clear what the contribution was from each source. Is that made clear elsewhere?
The meaning of “unrestricted” grant from the drug maker is not crystal clear; do you know, Keith, whether this word has a legal definition?
Ed: That was what I referenced in the post. Bob said in his talk the VA committed $10 million, I roughed out the donation by multiply the number of experimental subjects times doses times cost and came with about on half million.
Unrestricted means that the investigator is free to publish whatever is found and to use the money without further monitoring by the donor.
OK, I got it from your post that the cost of the study was 10 million and the drug company gave one half million, which was apparently the cost of the study medication (retail? wholesale?) with no other direct money contribution; it also seems clear that your source for this information came from the author in the form of a grand rounds lecture, rather than being in a publically available online source. The clinicaltrials.gov website only gives the study design and enrollment information, and does not disclose dollar amounts of sponsorship. This was what I was wondering. If a reader of a study does not have personal communication from the author as to the cost and funding contributions, is there a way to get this information? It also seems that the sponsor’s contribution was limited to donating the medication, which you seem to have learned from the author rather than from a publically available source. Is this correct? Details of the sponsorship must be available somewhere; I just wondered where to look for this kind of information if the author is not giving grand rounds nearby.
What the federal government gives you can always find out, it’s public info (It might be on CRISP for NIH, VA publishes reports on this that go to Congress and can I assume be found on line).
I do not know if pharma is held to that rule, I suspect not on studies in which there is partial federal funding and for sure not on things they fund themselves. Indeed, there have been some recent scandals in which not only was it concealed that pharma funded a study but that they also did the data analysis and ghost wrote the journal article upon which an academic’s name appears. Many universities have been pushing back hard against this kind of thing, and I am proud to say that Stanford is one of them, but it still happens, and of all medical specialties, psychiatry has been the worst offender.
“If the government approaches a manufacturer of an expensive drug about a study of whether an approved medication is no better than a generic drug or may be harmful to particular groups of patients, the manufacturer has economic veto power.”
Not that it is likely to happen, but couldn’t one of the conditions of patent &/or FDA approval be that the mfr must make the drug available for FDA approved (or NIH approved) trials on a timely basis, and if it drags its feet or refuses, the patent is lifted?
Marcel: That is not the current law, but it could be.
This is horrifying but not shocking.
It’s way worse than you think. A little light reading for a Sunday evening:
http://www.theatlantic.com/magazine/print/2010/11/lies-damned-lies-and-medical-science/8269/
IIRC, one of the biggest promoters of statins as a prophylactic therapy has also taken large amounts of money from their manufacturer(s) to fund his research. I can’t remember if he also gets “consultant” payments, but that’s a good way to bet. Guess what his results show? Should you believe them? Maybe. Maybe not.
Just to add to the cheery mood:
http://www.nytimes.com/2008/06/08/us/08conflict.html
Has the NIH budget really been cut? This suggests funding has been flat rather than declining. It could well suffer real cuts in future under Republican-imposed austerity. Of course, using the regular GDP deflator could well understate the real trend; as in Keith’s example here of the cost of acquisition of drugs for testing.
John Ioannidis put this idea (that many medical “advances†are reversed as more information becomes available) on the map. Recently, a letter was published looking at the frequency of “medical reversal†during 2009 in the New England Journal of Medicine. “Medical reversal†was defined as a new trial contradicting current clinical practice. http://www.medpagetoday.com/PublicHealthPolicy/ClinicalTrials/27494 has a summary (details behind a paywall).
The authors found 124 articles published in 2009 which made some evaluation of the effectiveness of a medical practice. They found 16 articles which reported that a current medical practice was ineffective; these were classified as reversals. Many of the articles look at interventions other than pharmaceutical; some of them, however, look at drugs.
It gets interesting when you look at the studies which reported that a particular drug intervention is not effective in the population enrolled in the clinical trial. Some of the drugs are old and off-patent. However, there were 5 trials of current drugs in which their effectiveness was questioned by the trial results, and these were all sponsored by the manufacturers. This includes AstraZenica for studies of rosuvastatin and esomeprazole, Schering Plough for eptifibatide, Amgen for darbepoetin alfa, and Abbott for ezetimibe.
It would be instructive to conduct a larger study of medical reversal, searching for trials in other high-profile journals over a period of several years. Having only the results for one journal during one year is a start. But there are five examples of drug companies sponsoring and then publishing studies which run counter to their economic interests. I was expecting to find that these five studies would be sponsored by disinterested parties. There is insufficient information to allow conclusions to be drawn, but these do seem to represent counterexamples to the thesis that the drug manufacturer can control the course of scientific publication. That control almost certainly exists, but is very far from airtight.