If PE will fail (miss the required p value) none of the SEs and TEs could be stat. significant.*
That's only true if the SEs are proportionate to the PE...It's possible and I believe will prove to be the case that V will be more effective for certain types of events than others...I believe the RRR in the specific event "reduction of fatal MI" will be higher than than overall MACE..considerably higher...Though I don't expect the PE to fail there could be a situation where the PE is close, but fails because it is less effective at lowering certain effects, but is very effective for certain specific effects.