Abstract:
Background and purpose: Resting state functional magnetic resonance (RS-fMRI) correlation with pre-surgical functional status in patients with brain tumors is scarcely documented in the present literature. Aim of the present study was to investigate the validity of RS-fMRI as potential preoperative functional mapping tool in tumor brain surgery by exploring the association of motor and language RS-fMRI networks with subjects’ preoperative performance on motor and language clinical assessment respectively in patients with brain tumor.
Materials and methods: 85 patients presented with brain tumor entities and 27 healthy controls were prospectively recruited for the present study. Clinical sample was subdivided into two groups according to mass localization: patients with tumors in proximity to motor cortex (n=59) underwent clinical examination for gross (paresis- muscle weakness) and fine (finger tapping) motor deficits. Patients harboring tumors in proximity to the left inferior frontal gyrus (n=35) were clinically assessed for apparent (expressive aphasia) and subtle language function (phonological verbal fluency) disturbances. All patients and healthy subjects underwent RS-fMRI with motor and language resting networks being derived by Independent Component Analysis (ICA). Results: In the motor group, patients with paresis demonstrated significantly (p=<0.01) reduced resting state BOLD-signal intensity in ipsilesional motor cortex in comparison to the respective one in contralesional-intact motor cortex. Significantly (p<0.01) decreased BOLD-signal intensity was additionally noticed in ipsilesional motor cortex of patients with paresis in comparison to patients with normal muscle strength. Furthermore, in patients with intact muscle strength, a strong positive correlation (r=0.70, p<0.01) between ipsilesional pre-central gyrus BOLD-signal and performance on finger tapping task was demonstrated. Compared to the healthy group, clinical motor group showed reduced resting state network activity, with patients’ ipsilesional pre- central gyrus BOLD-signal intensity to be significantly (p<0.01) lower than normals’ left and right pre-central gyri BOLD-signal intensities. Concerning language group, patients presented with expressive aphasia exhibited significantly (p=<0.01) reduced RS-fMRI BOLD-signal intensity in left inferior frontal gyrus (Broadmann area 44) when compared with patients without aphasia. In non-aphasic patients, a strong positive correlation (r=0.70, P<0.01) between left inferior frontal gyrus’ BOLD-signal intensity and phonological fluency scoring was demonstrated. Similarly with the motor group, language group also showed significantly (p=<0.01) reduced left inferior gyrus RS- fMRI BOLD-signal when compared to healthy controls. Finally, RS-fMRI BOLD signal was not observed to have an association with demographic parameters (age, gender) for both clinical and healthy groups and with tumor histopathological grading for both motor and language clinical groups.
Conclusions: Our findings show a significant affection of motor and language RS-fMRI networks’ BOLD-signal intensity by the presence of a tumor and a correlation with clinical performance of patients providing thus evidence for the functional validity of RS-fMRI in brain tumor patients; our results indicate therefore, that RS-fMRI may be a valuable complementary tool for preoperative mapping of eloquent areas, at least in patients who cannot cooperate satisfactory in a traditional task-based motor and language fMRI.