S Fang1,2, J Liang2, T Qian3, Y Wang5, X Liu2, X Fan2, S Li4, Y Wang5, T Jiang6,2,7. 1. From the Department of Neurosurgery (S.F., Y.W., T.J.), Beijing Tiantan Hospital, Capital Medical University, Beijing, China. 2. Beijing Neurosurgical Institute (S.F., J.L., Y.W., X.L., X.F., Y.W., T.J.), Capital Medical University, Beijing, China. 3. MR Collaborations NE Asia (T.Q.), Siemens Healthcare, Beijing, China. 4. Functional Neuroradiology Center (S.L.), Beijing Neurosurgical Institute, Beijing, China. 5. Beijing Neurosurgical Institute (S.F., J.L., Y.W., X.L., X.F., Y.W., T.J.), Capital Medical University, Beijing, China taojiang1964@163.com tiantanyinyan@126.com. 6. From the Department of Neurosurgery (S.F., Y.W., T.J.), Beijing Tiantan Hospital, Capital Medical University, Beijing, China taojiang1964@163.com tiantanyinyan@126.com. 7. Beijing Institute for Brain Disorders Brain Tumor Center (T.J.), Beijing, China.
Abstract
BACKGROUND AND PURPOSE: The accuracy of preoperative blood oxygen level-dependent fMRI remains controversial. This study assessed the association between the anatomic location of a tumor and the accuracy of fMRI-based motor function mapping in diffuse lower-grade gliomas. MATERIALS AND METHODS: Thirty-five patients with lower-grade gliomas involving motor areas underwent preoperative blood oxygen level-dependent fMRI scans with grasping tasks and received intraoperative direct cortical stimulation. Patients were classified into an overlapping group and a nonoverlapping group, depending on the extent to which blood oxygen level-dependent fMRI and direct cortical stimulation results concurred. Tumor location was quantitatively measured, including the shortest distance from the tumor to the hand knob and the deviation distance of the midpoint of the hand knob in the lesion hemisphere relative to the midline compared with the normal contralateral hemisphere. RESULTS: A 4-mm shortest distance from the tumor to the hand knob value was identified as optimal for differentiating the overlapping and nonoverlapping group with the receiver operating characteristic curve (sensitivity, 84.6%; specificity, 77.8%). The shortest distances from the tumor to the hand knob of ≤4 mm were associated with inaccurate fMRI-based localizations of the hand motor cortex. The shortest distances from the tumor to the hand knob were larger (P = .002), and the deviation distances for the midpoint of the hand knob in the lesion hemisphere were smaller (P = .003) in the overlapping group than in the nonoverlapping group. CONCLUSIONS: This study suggests that the shortest distance from the tumor to the hand knob and the deviation distance for the midpoint of the hand knob on the lesion hemisphere are predictive of the accuracy of blood oxygen level-dependent fMRI results. Smaller shortest distances from the tumor to the hand knob and larger deviation distances for the midpoint of hand knob on the lesion hemisphere are associated with less accuracy of motor cortex localization with blood oxygen level-dependent fMRI. Preoperative fMRI data for surgical planning should be used cautiously when the shortest distance from the tumor to the hand knob is ≤4 mm, especially for lower-grade gliomas anterior to the central sulcus.
BACKGROUND AND PURPOSE: The accuracy of preoperative blood oxygen level-dependent fMRI remains controversial. This study assessed the association between the anatomic location of a tumor and the accuracy of fMRI-based motor function mapping in diffuse lower-grade gliomas. MATERIALS AND METHODS: Thirty-five patients with lower-grade gliomas involving motor areas underwent preoperative blood oxygen level-dependent fMRI scans with grasping tasks and received intraoperative direct cortical stimulation. Patients were classified into an overlapping group and a nonoverlapping group, depending on the extent to which blood oxygen level-dependent fMRI and direct cortical stimulation results concurred. Tumor location was quantitatively measured, including the shortest distance from the tumor to the hand knob and the deviation distance of the midpoint of the hand knob in the lesion hemisphere relative to the midline compared with the normal contralateral hemisphere. RESULTS: A 4-mm shortest distance from the tumor to the hand knob value was identified as optimal for differentiating the overlapping and nonoverlapping group with the receiver operating characteristic curve (sensitivity, 84.6%; specificity, 77.8%). The shortest distances from the tumor to the hand knob of ≤4 mm were associated with inaccurate fMRI-based localizations of the hand motor cortex. The shortest distances from the tumor to the hand knob were larger (P = .002), and the deviation distances for the midpoint of the hand knob in the lesion hemisphere were smaller (P = .003) in the overlapping group than in the nonoverlapping group. CONCLUSIONS: This study suggests that the shortest distance from the tumor to the hand knob and the deviation distance for the midpoint of the hand knob on the lesion hemisphere are predictive of the accuracy of blood oxygen level-dependent fMRI results. Smaller shortest distances from the tumor to the hand knob and larger deviation distances for the midpoint of hand knob on the lesion hemisphere are associated with less accuracy of motor cortex localization with blood oxygen level-dependent fMRI. Preoperative fMRI data for surgical planning should be used cautiously when the shortest distance from the tumor to the hand knob is ≤4 mm, especially for lower-grade gliomas anterior to the central sulcus.
Authors: Marco Ciavarro; Eleonora Grande; Luigi Pavone; Giuseppina Bevacqua; Michelangelo De Angelis; Paolo di Russo; Roberta Morace; Giorgia Committeri; Giovanni Grillea; Marcello Bartolo; Sergio Paolini; Vincenzo Esposito Journal: Front Neurol Date: 2021-05-14 Impact factor: 4.003
Authors: András Folyovich; Viktória Varga; György Várallyay; Lajos Kozák; Mária Bakos; Erika Scheidl; Katalin Anna Béres-Molnár; Zita Kajdácsi; Dániel Bereczki Journal: BMC Cancer Date: 2018-10-03 Impact factor: 4.430