I was wondering the same hopefully the presentation will explain further. Still is nice to be arguing between a 2.26 in a 100 chance of being due to chance versus and 8 in 1,000 chance of being due to chance. The latter is is clearly in the ITT population too which as investor pointed out over the last 3 months is the most the key analysis.
It could be that the p value for 0.008 in ADAS COG ITT was a pooled p value of 30mg and 50mg wheras the 0.026 p value from Sept 23 was just for either 30mg or 50mg considering it's roughly similar to the p values given for 30mg and 50mg each on their own in the AAIC slides. Obv the pooled p value was always going to be lower that the indivdual p values due to higher (n).
Market Data 
Markets 
AI
