Golf

Maybe you missed this?  It's pretty clear (bolding is mine).  

"The higher number of MACE increases the powering of BETonMACE (from the original 80% to approximately 85%), improving the likelihood of detecting a significant reduction of the trial’s primary endpoint using the original assumptions of a 0.7 Hazard Ratio with a p-value of <0.05. For clarification, the primary analysis as defined in the statistical analysis plan communicated with the Food and Drug Administration (FDA) will remain a time-to-event analysis."

The company knows people don't want to hear that success is a longer MACE-free survival (after all, the median dosing was only 27 months); people want to hear that you that you survive.  Facing up to mortality (we all have to die sometime) is not our strongest point!

Hence the emphasis on RRR.  The Hazard Ratio (0.7) reflects the initial assumption that MACE events would be 30% fewer in patients on apabetalone than those on placebo.  However, it is time to MACE that will be evaluated statistically. 

IMO it is very good news that a 50% increase in expected dosing time (from 18 to 27 months) was accompanied by only a modest 10% increase (from 250 to 275) in MACE events.  As Ionoclast has pointed out, this equates to a 41% or 61% reduction in MACE at a 7% or 8% placebo MACE rate.  This is based on the assumption that 2200 patients stayed in the trial (quite optimistic but the time in the trial counts for patients who drop out along the way so maybe we can go with it).

The increased power of the analysis is a huge benefit.  My opinion was that they cut the trial size and duration very very fine - a 20% chance that the results would not be significant even if the true MACE reduction was 30% was too high for my liking.  (I know, a bigger and longer trial means more money needed). Unlike Bear, safety issues were not top of the concern list for me, but the longer duration of patient dosing helps add data there too! 

All good - tick tock.

Jupe