How do you contour and dose definitive CRT target volumes for an unresectable NSCLC invading vertebral bodies?

This is both an excellent and a complicated question.

While I do not typically favor elective nodal radiation in NSCLC, I do think this is a case where some degree of elective coverage is warranted.

Different imaging modalities yield very different contours when it comes to bone invasion in NSCLC, and while I have not clearly come across data to document this, it is my experience that invasion on CT appears < invasion on PET which appears < invasion on MRI. Thus, for patients with vertebral body invasion, I would always obtain and fuse a high quality MRI with planning CT.

I would use IMRT with a single plan to 60 Gy in 30 fx, and attempt to take the gross disease to 60 Gy (tumor on CT + PET + MRI, plus appropriate small uniform expansion f...

While I agree with @Dr. First Last' thoughtful approach (in particular, always using spinal MRI to define the degree of invasion), I would allow a maximum spinal cord dose to reach 54 Gy in view of the potential devastating impact of any local failure, such as cord compression. In my mind, this is a situation similar to the Pancoast tumor, where I also advocate for maximizing local tumor control as the primary goal.

Regarding elective spine RT rather than only targeting the MRI-defined target, I agree again that there is likely no harm in bringing this elective volume to 45 Gy and hopefully addressing microscopic disease.