Repros Therapeutics Inc (RPRX)

[iwfal]

Hard to know if you are referring to their blood levels at baseline - obviously the main characeristics were UF symptoms.

Agreed. But, for example, was the 'evaluable subgroup' substantially sicker than the rest of the ITT population? (Not an unreasonable question in the following sense - since perhaps it is because they were sicker that their anemia status was checked.)

The point here isn't to defend against every possible self selection issue - but to check if there were obvious baseline differences among the 'evaluable group' and the rest - if so then the ph iii enrollment criteria should perhaps change (although that itself is a risk that justifies a bigger trial).

The plain simple fact is the ph iii isn't much bigger than the ph iib (it is the smallest arm that really drives powering and I believe that was only about 15-17 'evaluable patients' per arm) - and the ph iib wasn't that stat sig for a subgroup at 0.002. So you need to really really check the subgroup or increase the size of your trial.


I believe the measure for Anemia is blood count (now I realize I dont know quite what that is - help - red cell count maybe) - but I beleive the endpoint is basically getting the level back up above the Anemia level (what was it, 11 decalitres or whatever ?)

I *suspect*, based upon the way that the PR was written, that the endpoint is improvement of 1.01 g hemoglobin/DL. But just speculation.

You could look at the presentation in the iBox - I would but I am not sure it is highly relevant.

They do not have that level of detail in any presentation I've seen.

The very fact that almost all the OTHER patients stopped bleeding makes it intuitive that both the sub-group and the rest count in the decision-making.

I agree that this is the most powerful indicator - but to go back to the subgroup equivalence: did the ITT population have the same baseline bleeding as the 'evaluable subgroup' and the same end-of-trial bleeding? That is probably the most interesting question of all. (The process of debating hones the questions - so thanks.)


Clark