It should be no surprise that the dropouts in the Biogen trials did not rollover into the OLE just as the figures you quote supports.
How many dropouts were there in the Anavex AD P2b/3 trial?
Given any other potential inclusion/exclusion criteria, is a 90% or so OLE rollover rate uncommon in a CNS trial?
Is that standalone 90% figure a useful indication of how effective A2-73 was?
How effective did the placebo cohort patients feel A2-73 was for them? Apparently they must have felt it was quite good with such a high OLE rollover rate.