"Do you know of any trials out there where they were successful for a 5 point MACE but not a 3 point MACE."

Not off the top of my head. I wouldn't be surprised if this has happened before though. In less than a week Amarin's REDUCE-IT trial will reveal more details on its top-line announcement of 25% RRR for 5-point MACE. Perhaps REDUCE-IT will be a contemporary example of what you describe.

"The phase 2 trials for apabetalone were all for 5 point and they got good RRR"

MACE reduction was not the primary outcome measure for any of the apabetalone Phase 2 trials. According to the ClinicalTrials.gov listings for Phase 2 ASSERT, SUSTAIN and ASSURE, only ASSURE included 5-point MACE incidence as a secondary outcome. Good RRR. But data mostly based on soft MACE events that are not the strict 3-point MACE (CVD death, non-fatal MI, non-fatal stroke), and based on very few MACE observations (only 35 total in ASSURE/SUSTAIN combined). See below.

"... and I think the company said reduce it by 10% and you should get the 3 point RRR,"

I don't recall the company ever saying that.

"... is that possible to say,... especially knowing the company has acces to all the blinded data,..?"

There are now two peer-reviewed published papers on the post-hoc analysis of ASSURE/SUSTAIN and ASSURE/SUSTAIN/ASSERT that provide itemized details on how many of each MACE event occured. It's not secret. I've broken down the ASSURE/SUSTAIN post-hoc analysis before in this post; however, it was based upon post-hoc MACE data presented at ESC 2014 that was just slightly different than what is in the Gilham et al. Atherosclerosis 2016 article. I don't have access to the Nicholls et al. Am J Cardiovasc Drugs. 2018 post-hoc that also included ASSERT. Probably mostly the same though. In post-hoc of ASSURE and SUSTAIN, there were a total of 35 5-point MACE events. Of these, only 8 were 3-point MACE events. Just keep these low numbers of observations in mind when trying to draw conclusions.

https://www.ncbi.nlm.nih.gov/pubmed/29027131

https://www.ncbi.nlm.nih.gov/pubmed/26868508

"I love the MOA of our drug and it goes a long way to having confidence that the top line data will be positive (above 25% RRR)."

It doesn't have to be above 25% RRR to be positive. The statistics are more important than the %RRR, as we discussed yesterday in the thread you started.

BDAZ