Yes, so long as the efficacy data is still blinded to the company. In a Phase IIa they can do what they want--whereas they could it's problematic in a pivotal trial (though I have seen even pivotal trials add patients based on a blinded interim analysis--but they had to give results for the whole group).
RD2 was easily fixed, they simply altered the protocol for the group already defined by the higher dose, not yet run. It seems to me that they could have run more subjects for RD1 if they then treated the procedurally-challenged group as if they were dropouts, and not done an intent-to-treat analysis, but only analyzed 'completers.' It's not statistically kosher, but my guess is that it would have been sufficient for the purpose of showing POC to partners. They don't care if it's publishable (it wouldn't be).
More statistically sophisticated posters may well disagree.
NeuroInvestment
Market Data 
Markets 
AI
