Nimg-61. Patterns of Glioblastoma Recurrence in Low Field Intensity Regions During Ttfields Treatment

Abstract

Glioblastoma (GBM)-induced breakdown of the blood brain barrier is revealed by contrast enhancement (CE) on MRI and corresponds to the presence of bulk tumor tissue. However, vascular function in the CE area and the surrounding, non-enhancing (NE) area of infiltrating glioma cells is less defined. Resting-state BOLD (Blood Oxygen Level Dependent) fMRI has been shown to be sensitive to alterations in neurovascular coupling both within and outside the region of CE in GBM. However, the mechanism by which the tumor disrupts vascular control in these areas is not known. We hypothesize that the alterations in neurovascular coupling seen in GBM are due to the local effects of infiltrating glioma cells. To test this hypothesis, we took 25 radiographically-localized biopsies from CE (n=16) and NE (n=9) regions during open surgical resection of primary GBM in patients with preoperative resting-state BOLD fMRI. The BOLD scan was co-registered to the volumetric T1-weighted + gadolinium scan used for intraoperative stereotactic guidance and the BOLD signal intensity was calculated from a spherical region of interest (radius=2 mm) at each biopsy site. Cellularity of each sample was calculated by averaging cell counts from three high powered fields of highest cellularity on H&E stained sections. Linear robust regression was used to evaluate the relationship between mean signal intensity and cellularity. The BOLD signal intensity was positively correlated to tissue cellularity (correlation coefficient = 0.42, p<0.01), indicating that the neurovascular coupling mechanism is disrupted by local effects of glioma cells. Moreover, this result suggests that BOLD fMRI could serve as a clinically useful biomarker for glioma infiltration into non-contrast-enhancing brain regions.