While the Duke randomized study found similar rates of control with 1mm v. 3mm margins and higher radionecrosis with 3mm, it did not look at treating with no margin. Does margin size depend on modality (gamma knife v. linac or cyberknife) and dose fall-off?