NIMG-41. VARIATION IN PERFUSION AND PERMEABILITY MRI DEPENDING ON VASCULAR INPUT FUNCTION SELECTION. (9th November 2020)

• Record Type: Journal Article
• Authors: Abbasian, Parandoush; Ryner, Lawrence; Shankar, Jai; Essig, Marco; Pitz, Marshall
• Is Part Of: Neuro-oncology. Volume 22(2020)Supplement 2
• Page Start: ii157
• Page End: ii157
• Publication Date: 2020-11-09
• Subjects: Brain Neoplasms -- Periodicals; Brain -- Tumors -- Periodicals; Brain -- Cancer -- Periodicals; Nervous system -- Cancer -- Periodicals; 616.99481
• DOI: 10.1093/neuonc/noaa215.654
• Languages: English
• ISSNs: 1522-8517

Abstract

Abstract: BACKGROUND: According to the updated RANO criteria, MRI contrast enhancement and peritumoral edema during the initial 12 weeks within the radiation field following chemoradiation may be pseudoprogression (PP) or true progression (TP). This ambiguity results in diagnostic uncertainty and delays in effective therapies. MRI perfusion and permeability imaging markers, including relative blood volume (rBV), and volume transfer constant (Ktrans), may be useful in differentiating PP from TP. Both techniques rely on identification of a vascular input function (VIF) to produce reliable output maps. METHODS: Perfusion and permeability maps were acquired from 7 patients on a Siemens 3T Verio MRI as part of an ongoing prospective study on glioblastoma multiforme. Patients received standard surgical resection with concurrent chemo-radiotherapy treatment and MRI follow-up at one and every 2 months. Maps were based on the follow-up MRI. The 3D contrast enhancing lesion (CEL) was segmented from the T1-Post-Gd MRI. VIF pixels were identified in the internal carotid artery (ICA), the M2 segment of the middle cerebral artery (MCA2), the superior sagittal sinus (SSS), automatically using software (Olea Sphere 3.0.22), and auto-edited (removed pixels from the auto VIF definition that were outside the brain). RESULTS: For the different VIF pixel selections (ICA, MCA2, SSS, Auto, Auto-edited), mean Ktrans values from CEL were 0.32, 0.48, 0.05, 0.09, 0.08, respectively, showing a 10-fold variation. Mean rBV values from CEL were 2.70, 2.36, N/A, 3.10, 3.19, respectively, showing a 1.3-fold variation. CONCLUSIONS: VIF pixel selection is a critical step in generating reliable perfusion and permeability MRI maps. Variation of up to a factor of 10 in the Ktrans values, depending on VIF selection, was observed. SSS for the permeability VIF resulted in maps that most closely matched literature values, whereas perfusion imaging showed less sensitivity to VIF selection.