It is possible something unique to torcetrapib -- such as its tendency to raise blood pressure -- contributed to its failure, rather than a class effect.
a) The epidemiological data, on balance, says it is significantly harmful to have too little CETP (see for instance the trial last year with Pravachol where the patients with lowest CETP had MUCH higher event rates). But I'll agree it is noisy.
b) The blood pressure effect sure as heck isn't going to double the death rate from CVD (yes, the overall death rate went up 60%, but around 40% of the deaths in the overall deaths would be non-cardiovascular)
c) If ARISE (a CETP raising drug) results look strong in death/mi it puts more weight behind lowering CETP is a bad idea.
I find this quote "Dr. Nissen suggested that other compounds in the class should first undergo studies using imaging techniques, followed by trials to test whether the drugs reduce deaths, heart attacks and other events." particularly interesting. As if we really know what to image for? Clearly traditional IVUS isn't going to do the job because ILLUMINATE wouldn't indicate a huge increase in events. There are other options out there that hope to identify rupture prone plaque - but I believe it is not yet a well proven technology (but not yet read up on it in detail).
Clark
Market Data 
Markets 
AI
