Spot on post @Pharmagang.No need to be an apologetic for the...

Spot on post @Pharmagang.

No need to be an apologetic for the sub-group analysis on the ARDS trial. It was one of the worst P3 trials I have seen in terms of controls of patients entering the trial. There will be such a huge variation that using the topline data would potentially be rejected by FDA anyway. Whilst the 'experts' will claim both arms were equally given the daily changes of the rapidly improving SOC, they easily dismiss that both arms were subjected to all kinds of completely unproven experimental treatments which likely damaged vital organs of a end of experimental line, high morbidity patient population either directly or indirectly due to delay in more effective treatments in time critical windows of the disease. These time critical windows were not equal across both arms as the MOA's are different between them. Additionally, given the huge variations in experimental treatments on these human guinea pigs, there was no way both arms ever received the same SOC. In such a small study, yes 300 is small for a P3, there would be no way to have those variances 'washed out'.

Maybe rushing straight to a P3 didn't allow enough data for an MOA/patient biomarkers and disease state to be postulated accurately to allow better screening of patients into the study. Yes, the desperate state of affairs had a lot to do with that as did the external funding source which removed some control of the study from Mesoblast. I won't try and go through the multitude of other reasons why this study, in hindsight now knowing the target PE, was doomed for failure short of Rem-L acting as a God-pill. But I would just reinforce that FDA would have picked the bones out of the poor controls, if they weren't going for such a crazy outperformance figure which again, in hindsight, was incredibly naïve and poorly risk managed. Again, no blame can be appropriately apportioned without a full knowledge of the who/what/when/how/why and not the point of my post.

My point is the data will only produce a Phase 2 level data set on PE as they dilute out the data into appropriate sub-groups to identify the signals and MOA. Unless, as you mentioned, they get a knock out secondary endpoint across the full dataset. And I note, a number of those secondary endpoints were used as PE's in other ARDS studies, including AtherSys, who didn't have the confidence/bravado/stupidity (take your pick) to go for morbidity as a PE.