Perhaps they are gathering additional information in the P2 part of the 2/3 trial and correlating it to those Pet scans. If this could be done it would save money and time in a larger P3 or P4.
It's all speculation, but other than the delay in late 2017 (which I believe was due to FDA issues not AVXL issues) Missling hasn't dropped a stitch so far when it comes to being prepared for, so I expect a pleasant surprise when we finally know exactly how Cognision works into the process.
Why did they only report ERP data from early in the study and not at 36, 48 or 52 weeks? Was the data positive or negative at these points? Was there any correlation to MMSE or ADS-ADL or concentration at later time points? Perhaps the answer to those questions relates to the absence of P300 as an endpoint in the 2b/3.