Koo, 

Your link and Led's comment pretty much sum it up. The futility analysis is a more or less a very generous statistical test to determine if there is truly a difference between the control (placebo) and test (apabetalone) groups. At the halfway point in the trial (125 events) if the two groups are experiencing extremely similar event rates, or if the apabetalone group has a greater event rate, then the trial will likely be halted for futility. But if the apabetalone group is showing a strongly reduced event rate, then the trial won't be halted for futility. There may be a gray area of the apabetalone group only  showing a trend for reduced event rate, without clear separation from the controls. In this case, it might be a tough call for the DSMB. All just my opinion.... I'm no expert here. 

BearDownAZ