Np Hoskuld, even without ABCLEAR2/3 being pre-specified, when you add up all the factors - the extremely low p values, the breath of hitting on that many different endpoints, the fact that the effect remained and for ADL grew even wider from ITT in the subgroup at 192 weeks in the delayed start OLE analysis, the MRI added support for ABCLEAR 2, the fact that ABCLEAR 2/3 are quite large sub groups, the fact that it came from stat sig ITT data (in ADAS COG and CDR SB) and the biological plausibility there is a lot more chance than not that ABCLEAR 2/3 is genuinely adding extra efficacy on top of ITT rather than just a false positive.
It would seem crazy for me for regulators not to allow patients to try this drug now via conditional approval, especially when it is fairly safe.
There should hopefully be more data given on ABLEAR 2/3 in terms of how well original placebo patients did once they moved onto the drug - i.e original placebo ABCLEAR 2/3 vs non-ABCLEAR 2/3 from weeks 48 to 144/192. This could help validate even further that this sub group is outperforming the wider population.
Market Data 
Markets 
AI
