Got it.

I'm still struggling with the question:  does a reporting a successful topline in a CVOT trial mean that the parameter specified as the primary endpoint was signficantly different (in a specified one-way test) in the treatment and control groups (except when non-inferiority is the the primary endpoint)?

A quick use of google indicates:

For EMPA-REG: The primary composite outcome was death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke, as analyzed in the pooled empagliflozin group versus the placebo group.

For EXAMINE:  ..with a pre-specified non-inferiority margin of 1.3 for the primary end point of a composite of cardiovascular death, nonfatal myocardial infarction

Thus the endpoints here seem to be number of MACE events, not 'time-to' MACE events.

I am wondering if RVX chose 'time-to' MACE rather than occurrence of MACE as an easier bar to reach in a challenging trial environment (very sick patients with a high risk of having a MACE before ABL had a chance to work, small enrollment numbers and short time frame).  RVX would not want to risk another trial that 'failed' in spite of good results.  After all, the probability of achieving a successful BoM outcome even with a RRR of 29% was only 80%, good but not slam-dunk  They could easily (50% likelihood) have a good outcome (RRR of 25%) without achieving signficance.  Perhaps they envisaged the possibility of quite a few MACE events occurring quickly, before ABL had a chance to alter the outcome.  After the first spate of treatment-indifferent MACE events, the likelihood of MACE events taking longer in the ABL than in the placebo group might be higher than a large differential in the number of MACE events given that the trial ended at 250 MACE.  Just idel speculation...

Wouldn't mean that  other parameters weren't reprted of course...  For sure the %RRR is the eye-catching number and the NNT is the bottom-line economic consideration...

Thanks, Jupe