Xiaodan Li1, Shukun Liao1,2, Jun Hua3,4, Liuji Guo1, Danni Wang5, Xiang Xiao1, Jun Zhou6, Xiaomin Liu1, Yuefa Tan1, Lijun Lu7, Yikai Xu1, Yuankui Wu1,3,4. 1. Department of Medical Imaging, Nanfang Hospital, Southern Medical University, Guangzhou, P.R. China. 2. Division of CT & MR, Radiology Department, First Affiliated Hospital of Gannan Medical University, Ganzhou, P.R. China. 3. Neurosection, Division of MRI Research, Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA. 4. F.M. Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA. 5. School of Biomedical Engineering, Shanghai Jiao Tong University, Minhang District, Shanghai, P.R. China. 6. Department of Pathology, Nanfang Hospital, Southern Medical University, Guangzhou, P.R. China. 7. School of Biomedical Engineering and Guangdong Provincal Key Laboratory of Medical Image Processing, Southern Medical University, Baiyun District, Guangzhou, P.R. China.
Abstract
BACKGROUND: Inflow-based vascular-space-occupancy (iVASO) MRI is a noninvasive perfusion technique that does not require administration of exogenous contrast agents. Arteriolar cerebral blood volume (CBVa) obtained from iVASO MRI is hypothesized to be an indicator of tumor microvasculature. PURPOSE: To assess the diagnostic performance of iVASO MRI implemented at 3T in predicting histologic grades of cerebral gliomas. STUDY TYPE: Retrospective. SUBJECTS: Forty-five patients (31 males) consisting of 14 WHO grade IV glioblastoma multiformes, 14 grade III, and 17 grade II gliomas. FIELD STRENGTH/SEQUENCE: At 3T we acquired CBVa data using an iVASO sequence. ASSESSMENT: The maximum and mean CBVa (CBVa_max and CBVa_mean) values were calculated in the tumor and normalized to the contralateral thalamus (nCBVa_max and nCBVa_mean). STATISTICAL TESTS: Kruskal-Wallis test, Mann-Whitney test, and receiver operating characteristics (ROC) curve were used for statistical analysis. RESULTS: Both CBVa_max and nCBVa_max increased with tumor grade (P < 0.001). Grade II gliomas showed CBVa_max <0.78 ml / 100 ml in 10/17 cases and nCBVa_max <1.20 in 11/17 cases. Grade III gliomas showed both CBVa_max >0.78 ml / 100 ml and nCBVa_max >1.20 in 13/14 cases, and CBVa_max <2.06 ml / 100 ml in 13/14 cases and nCBVa_max <2.33 in 11/14 cases. Grade IV gliomas showed CBVa_max >2.06 ml / 100 ml in 9/14 cases and nCBVa_max >2.33 in 13/14 cases. The areas under the ROC curve, sensitivity, and specificity were 0.839 (P < 0.001), 92.9% (26/28), and 64.7% (11/17) for CBVa_max, and 0.883 (P < 0.001), 92.9% (26/28), and 70.6% (12/17) for nCBVa_max in the discrimination between grade II and high-grade (grade III and grade IV) tumors, respectively. DATA CONCLUSION: iVASO MRI might be used to help determine and predict glioma grade. LEVEL OF EVIDENCE: 4 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;50:1817-1823.
BACKGROUND: Inflow-based vascular-space-occupancy (iVASO) MRI is a noninvasive perfusion technique that does not require administration of exogenous contrast agents. Arteriolar cerebral blood volume (CBVa) obtained from iVASO MRI is hypothesized to be an indicator of tumor microvasculature. PURPOSE: To assess the diagnostic performance of iVASO MRI implemented at 3T in predicting histologic grades of cerebral gliomas. STUDY TYPE: Retrospective. SUBJECTS: Forty-five patients (31 males) consisting of 14 WHO grade IV glioblastoma multiformes, 14 grade III, and 17 grade II gliomas. FIELD STRENGTH/SEQUENCE: At 3T we acquired CBVa data using an iVASO sequence. ASSESSMENT: The maximum and mean CBVa (CBVa_max and CBVa_mean) values were calculated in the tumor and normalized to the contralateral thalamus (nCBVa_max and nCBVa_mean). STATISTICAL TESTS: Kruskal-Wallis test, Mann-Whitney test, and receiver operating characteristics (ROC) curve were used for statistical analysis. RESULTS: Both CBVa_max and nCBVa_max increased with tumor grade (P < 0.001). Grade II gliomas showed CBVa_max <0.78 ml / 100 ml in 10/17 cases and nCBVa_max <1.20 in 11/17 cases. Grade III gliomas showed both CBVa_max >0.78 ml / 100 ml and nCBVa_max >1.20 in 13/14 cases, and CBVa_max <2.06 ml / 100 ml in 13/14 cases and nCBVa_max <2.33 in 11/14 cases. Grade IV gliomas showed CBVa_max >2.06 ml / 100 ml in 9/14 cases and nCBVa_max >2.33 in 13/14 cases. The areas under the ROC curve, sensitivity, and specificity were 0.839 (P < 0.001), 92.9% (26/28), and 64.7% (11/17) for CBVa_max, and 0.883 (P < 0.001), 92.9% (26/28), and 70.6% (12/17) for nCBVa_max in the discrimination between grade II and high-grade (grade III and grade IV) tumors, respectively. DATA CONCLUSION: iVASO MRI might be used to help determine and predict glioma grade. LEVEL OF EVIDENCE: 4 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;50:1817-1823.