XOMA - And I think the chance of stellar p value in the active trial is good - e.g. if the background rate of spontanteous remission is 20% and the treated arm gets 50% then the p value will be well below that threshold.
The Ozurdex Huron trial is probably the best comparison trial, as far as control arm expectations, to Eyeguard-A that I've found so far. The highest response rate for the primary endpoint of VH score 0 was about 21%, though that was at week 16 - at week 8 it was 12%. As you've mentioned before, I found treatment arm response rates in NIU trials that I've looked at, mostly TNF-alpha inhibitors, are generally 50% or better. However the response endpoint often consisted of a basket of metrics, e.g. visual acuity, different inflammation scores, tapering of current drugs, and meeting only one criteria qualified as a responder. What I have been trying to get my arms around is determining how well the various individual response rate metrics across trials correlate to an improvement in VH score.
Interestingly the Abbvie Humira trial in chronic uveitis is very specific about ramping down the corticosteroids to avoid confounding - to be entirely off them by 19 weeks. My guess is that Xoma (and Servier) did something similar - but I don't have Xoma management calibrated in this regard so you never know.
Along the lines of your point about background drug protocol, the Ozurdex trial specifically defined protocol for crossover to rescue medications, using LOCF as data of record (link below). I would think XOMA would want equivalent controls in the Eyeguard-A trial.
Market Data 
Markets 
AI
