Are you claiming that they shouldnt test further in high GPNMB patients in this setting or at all? I still want to see more mature data in the current trial before rushing to judgement.
If the latter, I'd say its presumptuous to kill it in the high GPNMB group without looking at the drug in a bit healthier group of patients.
Salvage setting also makes it hard to see the true potential of CDX-011.
I think it's not that meaningful to do that kind of statistical analysis due to really small sample size, including HG/TN group as well. I would conduct a randomized trial by enrolling patients with HG, but stratify by TN status.
n the HG without TN status the ORR comparison is 28% vs 20% with a p value near 1.0. I'd suggest it is a stretch to call that a trend. And the PFS data, being both more perceptive and of more clinical and regulatory importance, is HR is about 0.85 with a p value near 1.0. Again, hard to claim even a trend for HG without TN status. So unless they have some amazing OS data in that subgroup I would hope that they kill it.
The easiest regulatory path is clearly the HG+TN group. But it would be a mistake to dismiss the TN-HG and the HG-TN groups out of hand for a couple of reasons.
First looking at the TN-HG group: There was a single response here, and because of the small numbers (one more response doubles the response rate) and the presumed heterogeneity of this group (after all it is a diagnosis of exclusion) together with the tremendous unmet need, it would be a mistake to abandon this group.
Second, the HG-TN group: Here it is worth remembering that these were tremendously beat-up patients. It is quite plausible that you might see much better results in earlier stage patients with immune systems that are more functional. Obviously the regulatory bar is higher, but so is the potential.
The bottom line though is that this trial was just too small to be able to make sound decisions about what subgroups to pursue.