Correction:
This actual-checkpoint-progressor subset does not necessarily have better immu-conditions than the average level of the full set, since they also fail the biomarker assays. But as long as it is somewhat representative of the full set in terms of immu-conditions, we should expect to see the BORR for this subset to be somewhat close to that for the full set, given your theory is correct. This 33%<50% is a little bit weird.
Of course, the fact that 33%<50% may be just due to statistical errors and could be better purged in PISCES with larger cohorts to recover the true treatment effect of IL-12 on actual checkpoint progressors.
The true treatment effect on actual checkpoint progressors, in my viewpoint, is the ultimate variable of interest, for determining the value of the combination therapy, which would constitute a major share of Oncs total value based on current information.
Meanwhile, we will get a better sense of the value of the multigene platform during the SITC meeting.