ABSTRACT
Task-based functional MRI (tb-fMRI) represents an extremely valuable approach for the identification of language eloquent regions for presurgical mapping in patients with brain tumors. However, its routinely application is limited by patient-related factors, such as cognitive disability and difficulty in coping with long-time acquisitions, and by technical factors, such as lack of equipment availability for stimuli delivery. Resting-state fMRI (rs-fMRI) instead, allows the identification of distinct language networks in a 10-min acquisition without the need of performing active tasks and using specific equipment. Therefore, to test the feasibility of rs-fMRI as a preoperative mapping tool, we reconstructed a lexico-semantic intrinsic connectivity network (ICN) in healthy controls (HC) and in a case series of patients with gliomas and compared the organization of this language network with the one derived from tb-fMRI in the patient's group. We studied three patients with extra-frontal gliomas who underwent functional mapping with auditory verb-generation (AVG) task and rs-fMRI with a seed in the left inferior frontal gyrus (IFG). First, we identified the functional connected areas to the IFG in HC. We qualitatively compared these areas with those that showed functional activation in AVG task derived from Neurosynth meta-analysis. Last, in each patient we performed single-subject analyses both for rs- and tb-fMRI, and we evaluated the spatial overlap between the two approaches. In HC, the IFG-ICN network showed a predominant left fronto-temporal functional connectivity in regions overlapping with the AVG network derived from a meta-analysis. In two patients, rs- and tb-fMRI showed comparable patterns of activation in left fronto-temporal regions, with different levels of contralateral activations. The third patient could not accomplish the AVG task and thus it was not possible to make any comparison with the ICN. However, in this patient, task-free approach disclosed a consistent network of fronto-temporal regions as in HC, and additional parietal regions. Our preliminary findings support the value of rs-fMRI approach for presurgical mapping, particularly for identifying left fronto-temporal core language-related areas in glioma patients. In a preoperative setting, rs-fMRI approach could represent a powerful tool for the identification of eloquent language areas, especially in patients with language or cognitive impairments.
Subject(s)
Glioma , Magnetic Resonance Imaging , Humans , Language , Brain Mapping , Glioma/diagnostic imaging , Glioma/surgery , Prefrontal CortexABSTRACT
BACKGROUND: Few studies have evaluated the relationship between brain arteriovenous malformations (bAVMs) angioarchitecture and the response to Gamma Knife Stereotactic Radiosurgery (GKSR). METHODS: A prospectively enrolled single-center cohort of patients with bAVMs treated by GKSR has been studied to define independent predictors of obliteration with particular attention to angioarchitectural variables. Only patients older than 18 years old (y.o.), who underwent baseline digital subtraction angiography (DSA) and clinico-radiological follow-up of at least 36 months, were included in the study. RESULTS: Data of 191 patients were evaluated. After a mean follow-up of 80 months (range 37-173), total obliteration rate after first GKSR treatment was 66%. Mean dose higher than 22 Gy (P = .019, OR = 2.39, 95% CI 1.15-4.97) and flow rate dichotomized into high vs non-high (P < .001, OR = 0.23, 95% CI 0.11-0.51) resulted to be independent predictors of obliteration. Flow-surrogate angioarchitectural features did not emerge as independent outcome predictors. CONCLUSIONS: Flow rate seems to be associated in predicting outcome after GKSR conferring high-flow AVM a lower occlusion rate. Its role should be considered when planning radiosurgical treatment of bAVM, and it could be added to other parameters used in GKRS outcome predicting scales.
