I was wrong on hard events being stronger than 20% (with offsetting effects for revasc) but right on glycemic control.
Actually, it turns out I was right on offsetting effects for revasc. The p value of the primary endpoint was almost exactly 1.0 per Feurstein(sp?). I was just wrong on the magnitude of the hard endpoint efficacy. So they will definitely(!) have to run another big heart trial - my guess is they would want a >10,000 patient trial for 3 years but otherwise a similar protocol to ARISE. If I were they I would run it largely ex-US where there is less extraneous revasc for stable angina and I'd want to do some of the imaging that Nissen refered to to prove MOA inre rupture prone plaques.
BTW - I am disappointed in Feurstein. He is completely ignoring the diabetic data. His agenda seems to be to prove that he was right, vs an analysis of what is.
Market Data 
Markets 
AI
