I don't know why there would be so much concern about the size of the trial given that RPRX management has the data it needs to make that decision, and they've decided that 15-17 in the smaller arm is sufficient to drive the results they're looking for.
Aka - you've been around enough biotech to know that companies routinely do not account for changed enrollment criteria, changed endpoints, changed SOC, ... and just plain good luck in the ph ii (aka Dew's Program Survival Bias) with adequate conservatism. So the lesson is that:
a) You really really really trust management
b) You look at the data itself (and note that perversely it is the companies with questionable management that don't share the data)
Therefore, common sense tells me that the decision of how many patients they decided upon was not made on economic, but rather statistical grounds, and isn't something that, if their biostats guys know what they're doing (and why wouldn't they)the trial is much more likely than not well powered.
Sadly, accounting for such things as I listed above isn't taught in mathematics. Seriously the people in biotech who design trials clearly do not learn from each other too well. If they did there wouldn't be the preponderance of failed trials on the right side of p=1.0. There wouldn't be a complete lack of randomized ph ii trials in cancer. ...
Market Data 
Markets 
AI
