They don't know the distribution of deaths between
Post# of 148179
However, as had been discussed here, even a spectacular 40-50% risk reduction at n=51 wasn't very likely to yield statistical significance.
It's prudent they decided to wait and conduct the analysis with a higher n...