I personally have no problems with drug approval in the absense of evidence of efficacy, so long as customers (doctors and patients) are not misled. Its a free world. I guess all patients have been told that a) the relationship between LDL lowering and actual clinical disease is not straightforward, b) that some drugs lower LDL but increase risk, c) that the relevant science prescribers have been permitted to see is a sack of potatoes.
The "Enhance" study of Ezetimibe was supposed to look at a surrogate endpoint that is one step closer to actual cardiovascular events - namely the development of atherosclerotic arterial plaques. So what happened?
- Oddly the reporting of "Enhance" was put off by a year, and it is now planned to remove the shroud of secrecy sometime in 2008 (possibly.....). Cardiologists expected to see results at a medical meeting in November 2006, then at another in March 2007, then at another this month. But none materialized .
- It turns out the companies have decided to change the study's primary endpoint in retrospect. To convince everyone that this is the right thing to do, they brought in an "outside" expert advisory "panel" who concluded that the wrong thing is the right thing (presumably for a fee). They then refused to state who these supposed experts were.
- It turns out that the company has had full control over all of the study data and that the lead "author" has not yet seen any of it.
- Lead investigator of the study, Dr John Kastelein (Academic Medical Center, Amsterdam, the Netherlands)stated that "The suggestion that the results are being suppressed because they are negative is simply wrong. People are assuming that anyone can take a peek at the data, but how can they do that if it hasn't even been unblinded and there are 40 000 images to analyze?"
- Firstly, we have no confidence (P=0.3) that the study has not in fact been secretly unblinded. This lack of confidence is the logical assumption following numerous odd events and data alterations which have taken place in the past (random examples here). Unblinding codes are presumably held by the same entity holding the raw data (I guess). That same entity has a huge financial stake in the outcome. That entity is not yourself Dr Kastelein. Nor is that entity a secure impartial honest third party (a complete guess).
- Secondly, as an experienced scientist you will be fully aware it is perfectly easy to fiddle the results of a randomized trial given a "blinded" study database even without the un-blinding codes. Ezetimibe has many side "effects" that distinguish it from placebo apart from it's intended clinical benefit (LDL lowering is itself such a "side effect"). It would be easy for an individual "exploring" the data to examine the relationship between LDL lowering and various "primary endpoints" without unblinding to get a pretty good idea of the "primary" endpoints to reject, exclusion criteria to be applied, or even the variables that might require a little "recoding".
Others  have pointed out some of the key problems with this research, and I have added a few more:
- Seek to get your drug on the market before we know whether it actually makes us live any longer or prevents real disease.
- Keep all the data under your strict control; don't even let the (so-called) lead investigator see it.
- Change the primary study endpoint in the middle of the trial.
- Pretend that this is science.
- Keep all the processes as secret as possible at each step of the way.
- Conduct studies that focus on patients who are not at all representative of the intended target customer for the drug.
- Pretend that you are a company based on science, but at the same time pretend out loud that altering data or endpoints in a still-blinded study cannot possibly be a means to induce bias.
- Hughes, Sue (2007-11-23). Concerns Raised on Delay of Ezetimibe Data. Medscape/Heartwire.
- Merck/Schering-Plough Update on ENHANCE Trial. Merck Website.
- Berenson, A (2007-11-21). After a trial, silence Page C1. New York Times.
- Brody, Howard (2007-11-24). Here We Go Again: Everything That's Wrong with Industry Sponsored Trials. Hooked: Ethics, Medicine, and Pharma.
- Blumsohn, Aubrey (2007-10-21). Vioxx and a quacking duck. Scientific Misconduct Blog.
- Blumsohn, Aubrey (2007-05-19). On the redefinition of research misconduct. Scientific Misconduct Blog.
- Berenson, Alex (2007-11-24). Cardiologists Question Delay of Data on 2 Drugs. New York Times.
- Herper, Matthew (2007-11-19). Drug Trials: The Vytorin Question. Forbes. Retrieved on 2007-11-27.