Koo,

I understand the confusion now. In general, a greater relative risk reduction will require less patients to achieve statistical significance. Conversely, a lesser relative risk reduction will require more patients to achieve statistical significance. I went back and found Imtesty's "unofficial" transcript of the Sept 2015 Corporate Update regarding BETonMACE: 

"Where did we come to our sample size? Two-and-a-half thousand patients gives 80% power for a 30% relative risk reduction, which is much less than we saw in our Phase II studies. However, you always anticipate that your Phase III will be less than your Phase II. And the EXAMINE study tells us that we’ll have about a ten-and-a-half percent (10.5%) event rate in the placebo arm at 18 months. And our projections are for about a seven-and-a-half percent (7.47%) rate in the active arm. So we’re well powered for a 30% reduction."

So if the actual %RRR stays around 30% or greater at the 125 events futility analysis and 175 event sample size analysis, then most likely no more patients will need to be added because the study was already well statistically powered at that predicted %RRR. However, if the actual %RRR dips below 30% then there is an increased chance that BETonMACE may be underpowered with 2400 patients and that the DSMB will recommend adding more patients to ensure that BETonMACE is adequately statistically powered at the observed %RRR. 

Adding more patients should not be expected to change the actual %RRR though (i.e. improve from 25% to 30%). By the time the futility analysis and sample size analysis are performed, the DSMB should have a good idea of the actual event rate in the placebo and apabetalone groups. Any additional patients enrolled would still be recruited based upon the same criteria (diabetic, low-HDL, recent ACS event) and would be expected to experience similar event rates as observed in the original placebo and apabetalone groups.

BearDownAZ