rrr,
In short, figures from studies a while back are generally used - partly because not enough real info is available for newer therapies.
For getting to market, they're roughly 10% PI, 20-40% PIIa-IIb, 70% PIII, 80% pre-approval (off the top of me noggin)
These are based on pharma products, so biotechs would likely be different
Furthermore, many 'easier to treat' indications are now gone, so making better, safer products raises the challenge.
Having looked at a number of drug pipelines, it surprises me how many products arent really very impressive or have inadequately designed trials - partly I think because many companies find themselves in a position where they need to at least have a go?!
As such, my feeling is that generalised success rates can be upped for impressive looking products from well thought out companies.
I'd say that OXB fits that bill...
Furthermore, I think widely accepted rates are 80% chance of getting through each stage (again from older studies of pharma product trials)