"There is little or no efficacy with this protocol in...TN by itself."
It's tricky because the same patients are being shown more than once in different groupings, but I think iwfal is likely correct.
Think of it as a Venn diagram - you have the TN circle and the High-Expression circle and they intersect. There is also the set of responders. If you look at the absolute number of responders, you can see that most of the responders (except perhaps 1) in the TN group are also in the HE group. So that implies there was not much in the way of response in the TN but not HE group.
Maybe the company would respond to the question "How many responders in the TN group were not also high expressors?