What are the chances that two curves are different if they're identical at:
1) the beginning (day = 0)
2) the middle (median survival was same for both)
3) the end (24 month survival was same for both)
Even if you try and pull the "it works too slow and benefit is seen at 12+ months" angle, then you need to explain why all the benefit is lost at 24 months.
Unfortunately, this trial is now officially wasting money.