As far as powering and statistical significance goes....I think its all relative.  Results that come from 2,500 patient years would be less powerful and statistically signficant than results of a trial with over 3,000 patient years. 

In simple terms (I'm a simple guy) the results of a trial with 3,000 patient years would be 20% more powerful (I think power and statistical signficance are synonomous) than the same trial with only 2,500 patient years.