One has to assume that normally fewer events would mean longer time to events.

But it seems at least conceivable that there could be fewer events but somehow enough of those that did occur happened more quickly to skew the result, since the endpoint was just time to events? 

Or that the difference between statins accounted for composite numbers but Rosuvastatin alone would have met.

Or that p values were only slightly too large

Or that all-cause mortality was lower even though time to 3 point MACE was somehow not (again, maybe unlikely but conceivable)

Or that, instead of MACE per se, other conditions saw clear improvement

Or that the data pointed clearly somehiw to a mechanism that would define and shorten a next trial, or suggest a different disease target.

Any of which would be second-choice to immediate success, but still quite positive and likely enough so to continue to attract funds.

Maybe?