Thanks for you help.
I did try a number of simple rotations like 3dWarp -deoblique and -oblique2card, but it just tilts the oddly-shaped brain (see attached).
The DICOM attributes I'm concerned with look like this:
(0018,1120) Gantry/Detector Tilt = 30.000000
(0018,1140) Rotation Direction = CW
(0020,0032) Image Position (Patient) = -120.000\-103.923\209.037
(0020,0037) Image Orientation (Patient) = 1.000000\0.000000\0.000000\0.000000\0.866025\-0.500000
Except for the z-value in the image position attribute, the values are the same for each slice in the volume. Based on what the image looks like after to3d, I'm thinking that I need to artificially push forward the slices as they move superior (based on the slice thickness and gantry tilt).
BTW, I have registered other CTs and MRIs with AFNI and it does quite well. Great tool.