Michael O'Riordan, at the HeartWire, ran a story overnight that Marilyn Mann pointed out to us, last week -- O'Riordan has additional pull-quotes, and more retrospective background, though, so do go read it all:
. . . .The cancer signal first arose in the Simvastatin and Ezetimibe in Aortic Stenosis (SEAS) study, a trial presented at the European Society of Cardiology (ESC) in Munich, Germany last year and reported by heartwire at that time. Among those treated with the ezetimibe/simvastatin (Vytorin, Merck/Schering-Plough Pharmaceuticals) combination, there were significantly more cases of fatal or nonfatal cancer compared with those treated with placebo.
When the cancer findings became known, it led to an independent analysis of two ongoing ezetimibe studies, the IMPROVE-IT and SHARP trials, by the Oxford investigators, to determine whether the cancer risk was real or chance.
In IMPROVE-IT and SHARP, which provided more cancer data than the SEAS trial alone, including more data in patients with at least three years of follow-up, there was no increased risk of incident cancer or cancer mortality. When all three trials were combined, there remained an increased risk of death from cancer in the active-treatment arm.
As he also pointed out in his letter to the editor, Nissen, in speaking with heartwire, said the decision to analyze data from IMPROVE-IT and SHARP before those studies were completed sets a dangerous precedent, raises scientific and ethical issues, and is not reliable for the evaluation of drug safety.
"A decision to unblind a clinical trial is something that should be done under only the most extreme circumstances," said Nissen. "Once you unblind a trial, you forever alter the conduct of that study. Most individuals who think very hard about clinical trials will tell you that you better have an extremely good reason for unblinding an ongoing clinical trial, and I don't think there was a good reason here."
Moreover, Nissen said these two incomplete trials represent an incomplete experiment and notes that the mean exposure time to the drug in some patients is insufficient, and this dilutes the signal of risk. . . .
No comments:
Post a Comment