So here's an interesting question: What are some good examples of a failed trial where the company did a subgroup analysis that then was proven convincingly in a subsequent trial? Just trying to get a numerical sense of just how rare this phenomenon is. Peter