>>why not enroll more than ~25% pf patients<<
Propobably most tumors had less than 25% cells stained positive for GPNMB. As explained by CMO, they used a highly non-selectively low GPNMB staining threshold (5% membrane/cell) that admitted more than 90% of the eligible and screened patients to the study. While there might be good strategic rationale (e.g. quicker enrollment, more TN negative patients, bigger market share should CDX-011 demonstrate efficacy in LG), in hind sight, a higher threshold is probably better, perhaps in phase III they may want to enroll either HG or TN.