Try this instead. They had an info arm and expanded use arm they were monitoring. Perhaps they noticed some psPD issues. They didn’t know in 2015 if the trial failed on the old primary, but the external evidence demonstrated they should, at some point, switch from pfs to os as the primary. Further, they knew OS had crossover, so, out of an abundance of caution and in the interest of informative results, they compared the primary OS endpoint to ECA. However, in the interest of science and transparency, they kept the original placebo controlled OS endpoint for treatment as a secondary endpoint.
Market Data 
Markets 
AI
