INCY / MRK: I agree that, in general, data withholding is a bad sign, but I think it's overblown in this case.
Based on AF's article, in the NSCLC cohort INCY presented 35% ORR, 14 responders in 40 pts. If the additional 3 pts all responded, it would be a 39% ORR (17/43). If none of the add'l 3 pts responded, then it would be 32% ORR (14/43).
These numbers are not meaningfully different IMO, especially in terms of informing a go/no-go decision into Ph3. Given the competitive intensity, MRK/INCY clearly have a relatively low bar for making a "go" decision in order to beat others to market (not just in NSCLC). The ORRs from these small cohorts are inherently incredibly noisy, and shouldn't be taken as much more than signal-seeking. Is the combo safe? Is there some hint that the activity might be better than PD-1 alone? If yes to both, then they're pulling the trigger on a big Ph3 RCT.
Would it better if they showed the data? Yes, of course. But do I think this is a real red flag that undermines the conclusions of the data in the posters? Not really. In a world where companies are going straight into Ph3 studies based on handfuls of patients from uncontrolled Ph1 studies, bemoaning the absence of a few patients-worth of ORR data is tantamount to missing the forest for the trees.
News 
Market Data 
Discover 
