biopharm, from what we heared there is no way this can be done. The correlation is with the statistical significant data (and hence of BIG VALUE for PPHM) but WORTHLESS for the FDA because on the PURE BLINDED DATA there was no statistical significance.
Hence the data they correlate with is only statistical significant for PPHM and presenting a correlation to the FDA with data that is NON STAT. SIG. is a futile exercise.
BUT...using the research of jbainseky's we know that in a NEXT trial they CAN (and as PPHM said will) include the biomarkers. They can then do BOTH at once, use the biomarkers as a selection and at the same time pre-define it for approval. They can only do this because they know the outcome of Sunrise is valid (and actually more then valid because the CTRL arm outperformed and that is ALSO embedded in the results that yield stat.sig. efficacy.
News 
Market Data 
Discover 
