Vitrea software

All procedures for obtaining patient information and VS samples were approved by the Institutional Review Board. Patient demographics and case histories were obtained by retrospective chart review. 3D tumor volumes were calculated based on the axial post-contrast T1 magnetic resonance imaging (MRI) thin section images using Vitrea software [version 6.6.2, Vital images, Toshiba Medical Systems, Minnetonka, MN]. In all cases, the tumor was manually outlined on the axial post contrast images on all sections followed by computerized compilation of summated tumor volume and generation of a volumetric model. Segmented volumetric data was re-reviewed post analysis to ensure measurement accuracy. Tumor volumes were calculated from the MRI at the time of presentation, prior to radiation and prior to re-resection. For controls, we selected two patients with VSs who underwent subtotal resection of a VS with subsequent regrowth and re-resection of the VS without any IR therapy.
Samples from each specimen were fixed in 10% neutral buffered formalin and paraffin embedded. Five-micron thick sections of formalin fixed paraffin embedded tissue were stained with hematoxylin and eosin (H&E) and were also used for immunohistochemical assays.

Volumetric analysis of tumor size was calculated from postcontrast axial T1 sequences on preoperative MRI scans using Vitrea software (version 6.6.2, Vital Images, Toshiba Medical Systems, Minnetonka, MN). This protocol has been previously described and used by Bathla et al.^{15 (link)} Briefly, a 3-D volumetric model was constructed by completing manual segmentation of axial contrast enhanced images (slice thickness 3 mm, interslice gap 0.3 mm) on all sections followed by computerized summation. Post-volumetric data was reanalyzed to ensure accuracy of measurements. When present, fundal fluid size was determined by measuring the length of the CSF space from the lateral edge of tumor to the IAC fundus on axial, highly T2-weighted, thin cut images, such as Fast Imaging Employing Steady-State Acquisition (FIESTA) or Constructive Interface in Steady State (CISS) sequences. When fundal fluid was captured on multiple slices, the slice with the largest fundal fluid cap was used. Picture Archiving and Communication Systems (PACS) was used to calculate linear distances, and less than 1.5 mm was considered no fundal fluid.