The most interesting part is that the Statistical Review makes it very clear that the exact form of the Cox Regression model was NOT prespecified. And when a Cox Regression was run using different methodologies the p value was substantially different. See Table 9 in the Statistical Review.
Note also that in adjudicating the 9902a CR the statistical reviewer bolded just one sentence:
The sponsor stated that Cox PHR model with covariates were using the same from Study D9901 before they looked at the data.
They then follow this up with:
Given the fact that data were available at the time when the sponsor developed an analysis plan for survival for Study D9902A, there was a chance that the analysis plan might be developed under the influence with the knowledge of the survival data intentionally or unintentionally.
Given that this is buried in the back of the Statistical Review there is a good chance this won't come up. But it is a key point and overall it does indicate PGS' level of cynicism is not unwarranted.
Market Data 
Markets 
AI
