"In my opinion, the reduction in CHF readmissions is the most promising result in BETonMACE when it comes getting approval. The fact it was a secondary endpoint is not such a big deal. It was statistically significant and was in the total sample, not a subgroup."

I respectfully disagree here. The primary endpoint failed to acheive statistical significance. That means ALL other analyses, including both pre-specified secondary endpoints as well as pre-specified sub-group analyses of the primary endpoint, are not valid and are only considered exploratory.

CHF was at the bottom of the list in the hierarchal order of secondary endpoint analyses. That means even if the primary endpoint was achieved, all of the other secondary endpoints ranked above CHF would have needed to achieve statistical significance in order for the CHF p-value to be valid. But since the primary endpoint failed, none of the secondary endpoints are valid. CHF being at the bottom of the list hurts its case even more. 

As far as I know, the SGLT2i/DPP4i sub-group analyses were not even pre-specified. So there is a double whammy here. 1) Primary endpoint for total population failed; 2) SGLT2i/DPP4i is a post-hoc sub-group analysis.

In my opinion, the CKD sub-group is equally if not more promising than CHF. The CKD sub-group was 1) pre-specified and 2) had amazing reduction in all categories of MACE (excluding stroke of course, and including CHF). IF (big IF) the renal function data in the CKD sub-group shows improvement elicited by apabetalone, then hands down the diabetic stage 3 CKD population becomes the most exciting population for apabetalone. And even if apabetalone is neutral for renal function, the CKD population still seems to be a no brainer to go after.

BDAZ