I can understand separating blinded from unblended data/information in areas that are subject to interpretation, e.g. disease progression, but how could unblinding have any effect on something like expression of a biomarker and how long someone lives?