Prior to surgery, it can be helpful to obtain functional MR imaging to determine potential injury that could result from aggressive tumor resection, based on location. Intraoperative monitoring is often employed to further decrease the risk of adversely affecting function.

Following surgical intervention, immediate (within 48 h) postoperative MR imaging with contrast is strongly recommended. Thin slices (e.g., 1–2 mm) are preferred for enhanced image resolution and more accurate fusion with the planning CT images, which should ideally match the slice thickness of the MR images.

Radiographic imaging is helpful in treatment planning but almost universally underestimates the true extent of disease.

For patients with extensive edema and only a small volume of enhancing tumor, a diagnostic MR image at mid-treatment is recommended. In such cases, emergence of further areas of enhancement can be seen, requiring expansion of the boost (GTV2) volume to cover these areas. Figures 3, 4, and 5 include images that show multifocal edema and a small unifocal enhancing tumor. It is possible that enhancing disease may develop in other areas of edema over a span of weeks.

MR fluid-attenuated inversion recovery (FLAIR), T1, and T2 (if significantly different from FLAIR) sequences should be fused with the simulation CT scan for contouring guidance.

In situations in which cranial nerve abutment is a concern, a 3D fast imaging employing steady-state acquisition (FIESTA) MR series can be helpful in visualizing and sparing the nerves. A 3D spoiled gradient recalled (SPGR) MR image can facilitate contouring of the hippocampus.

For patients with recurrent GBM, the addition of PET/CT imaging should be considered. Because it is difficult to distinguish between postradiation changes and disease recurrence on MR imaging, PET occasionally can be helpful. In these instances, PET/CT should be fused to the tumor on simulation CT and to MR imaging. Amino acid PET imaging for this purpose is frequently used in Europe but is not approved for this use in the United States.

In the majority of situations, patients should undergo CT simulation supine with the head in a neutral position. Intravenous contrast is not recommended.

The head and neck must be immobilized, preferably up to the shoulders, using a custom mold or mask to increase setup accuracy. Bite blocks can be useful in those situations in which angular rotations are a problem (in our practice, these are used infrequently). One specific advantage of a bite block is the possibility of incorporating into it markers that can be tracked by cameras, so that intrafraction motion can be monitored. This would be useful in fractionated stereotactic approaches.

Thin-slice CT imaging is recommended (at 1–2 mm intervals) and, for ease of fusion, ideally should match the MR image slice thickness.

Multimodal MR fusion, using at least the contrast-enhanced T1 and FLAIR sequences, should be fused to the planning CT. Although the entire anatomic volume should be matched, the primary matching focus is the tumor. PET/CT fusion can be helpful in select situations, as mentioned above.

Standards for treatment of HGGs are directed by data from three seminal studies from the 1980s that highlighted the extent of microscopic disease well beyond enhancing tumors or radiographic changes (Hochberg and Pruitt 1980; Wallner et al. 1989; Kelly et al. 1987). This finding drives the principles of contour generation in these cases.