RK, thx so much for the comprehensive re-count of the trial. I do not follow where you state there were "probably at least 190 patients before 2015". But, I follow the majority of the rest. The additional arms can cloud the critical path of the 348 patients in the base trial. You think that the company reached the 348 full enrollment or close to it around or shortly after the screening halt but decided to remain silent about enrollment due to regulatory communications I presume. The main assumption one has to make is what will the SOC PFS look like. Take a look at this overly simple set of numbers, using your enrollment numbers:
Active Placebo Total
Up to EOY 2013: 113 56 169
2014 48 24 72
2015 72 35 107
Totals 161 115 348
Now assume a median of active PFS of 24 mos and a placebo PFS of 12 mos and start with the assumption that 248 gets reached in Dec, allowing virtually a full 12 mos of 2016:
Evented 95 56 169
36 24 48
18 20 20
Totals 149 100 249
The most important assumption is the PFS for the placebo population. If the median of the active population is double the median of the placebo population then the target completion date is when 1/2 of the active population has evented, since the target number of events is approx. 2/3 of the entire trial population. But, if the active population median is less than double the placebo timing or greater than double this relationship does not hold. So, we can only provide an educated best guess for the placebo eventing and go from there. The higher the placebo eventing median the lower the active eventing median and vice versa. But, if we start with a 12 mos blended mgmt. methylation promoter population since that seems reasonable, things look good. If you assume a 15 mos placebo median it starts to tighten up considerably. But, isn't that quite high for a blended placebo population? Am I wrong here?
Again, I want to extend a deep appreciation for all your work here, wow.
AI
Market Data 
Markets 
