Abstract
BACKGROUND: Preoperative advanced MR imaging techniques, like DWI/DTI and DSC, provide significant structural and functional information about tumor biology. Still, histopathological evaluation of biopsy samples remains the gold standard for the diagnosis of brain tumors. Here we present the use of standardized RCBV (sRCBV) in conjunction with intraoperative navigation for astrocytoma grading. This is the first comparison between tissue type and grade and measures of RCBV using spatially correlated imaging and tissue samples. METHODS: Histopathological findings of 107 tissue specimens from 32 adult patients were studied: non-tumor control tissue (n = 28), WHO Grade 2 Astrocytoma (n = 11), WHO Grade 3 Astrocytoma (n = 11), WHO Grade 4 Astrocytoma/GBM (new: n = 22; recurrent: n = 22; mixed: n = 13). Preoperative (≤ 3 months) DSC-MRI and T1 + C images were co-registered to MR Imaging uploaded for surgical navigation. Guided by intraoperative MRI snapshots of the tissue sampling site, a 3 mm spherical region of interest was drawn manually on pre-operative T1 + C images. Leakage-corrected sRCBV values were determined from raw DSC data. Statistical sample independence was confirmed before conducting statistical analyses. RESULTS: Standardized RCBV significantly distinguished tumor from non-tumor controls with robust sensitivity and specificity. Additionally, sRCBV differentiated between Grade 2 and newly diagnosed Grade 4 astrocytomas, as well as between newly diagnosed, recurring, and mixed GBMs with strong sensitivity and specificity. Furthermore, sRCBV revealed two statistically distinct patient populations in Grade 3 and newly diagnosed Grade 4 astrocytomas. Although not statistically significant, the median progression-free survival from initial diagnosis (surgery) for Grade 3 astrocytoma and newly diagnosed GBM patients with low sRCBV values was longer (73 days and 62 days, respectively) than for patients with high sRCBV values, regardless of treatment. CONCLUSION: The results of this study suggest that preoperative sRCBV has the potential to predict tumor characteristics beyond standard histopathology, making it an important supplementary indicator of tumor aggression.