>Could in theory a phase IIb study (if extremely positive) serve as a basis for approval or is a phase III study always necessary. If for example a survival benefit were shown in spite of the original design that was not powered to do so.<
If positive, the ART-123 phase-2b trial can probably serve as one of the two pivotal trials needed for FDA approval. We know that with 800 patients (400 per arm) it is powered to show a survival benefit; so if ART-123 works in DIC, the trial ought to be successful. This is a big if, of course.