It's more complicated. I should have taken notes when I did the research, but actually there are various layers of psPD.
There are early, complete, partial, high methylated, low methylated, etc. Methylated MGMT versus psPD is not really a comparison since you can have both in either. Net Net the group of psPD in the info arm were almost certainly not expected to do as well as the group in the arm of 32. Those remaining in the 331 trial are on a spectrum for likely response somewhere in-between. I'd explain this further, but I'm fading. If the top ten in the indeterminate arm were primarily psPD, they still did very well for the type of psPD they were. I'm certain I've confused everyone, but it's true, imo.