الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 166 |  |  |
| | | from | 166 | | |
Abstract
Neurosurgery nearby functionally important areas posses a high risk for postoperative neurological deficits. fMRI is feasible for advanced MR-neuroimaging in the clinical setting and noninvasively provides an important diagnostic tool, which is otherwise unavailable(2).
Using the blood-oxygenation-level-dependent (Bold) signal allows measuring changes in neural activity in a non-invasive fashion. During task-based fMRI, Bold-signal changes are utilized to reflect the neural activity elicited by a predefined task. A typical experimental setting lies in recording fMRI data while a patient performs a specific action inside the magnet, e.g., moving pre-determined body parts or producing speech. Comparing the neural activation during these tasks with the level of Bold-activity while the patient lies restfully in the scanner results in statistical maps holding information on the individual functional anatomy of the examined patient(2, 9, 119).
We aimed to assess the role of fMRI in locating the eloquent brain areas and its impact on presurgical planning and evaluation of surgical outcome for the estimation of an as good as possible balance between maximal tumor resection and minimal loss of function.
This study was conducted on thirty patients, having different brain lesions and planning for surgical excision, we examined 60 different s-motor and language areas by tb-fMRI. Conducted using different motor and language paradigms.
Our study included 30 patients, 18 (60%) of them underwent fMRI on 3T magnet testing 41 (68.3%) different areas and 12 (40%) of them underwent fMRI on 1.5T magnet testing 19 (31.7%) different areas.
The majority of our patients (27 patient 90%) were right handed. They underwent fMRI testing 55 different areas (91.7% of the total areas tested).
Our study included two middle handed patients, underwent fMRI testing four different areas (6.6%) as well as a left handed patient, underwent fMRI testing receptive area of the language (WA), we considered the left sided hemisphere as a dominant hemisphere in the middle handed patients, as their laterality index was 20 and zero. The different paradigms were repeated to ensure reproducibility of 49 of the 60 stimulations (81.7%). Our studied patients suffered from lesions of different types and grades, including 13 patients suffering from high grade gliomas, eight suffering from low grade gliomas, two suffering from meningioma, four suffering form cavernoma, two suffering from AVM and one suffering from mesial temporal sclerosis.
The lesions of interest affected the examined related eloquent areas by infiltration, displacement and/or compression or did not affect them. The examined areas showed compromised signal intensity or did not show activation at all in 20 (34.5%) areas.
Nine (15%) of studied areas were associated with relevant (related to areas examined) clinical manifestations, while 51 (85%) areas were not associated with relevant clinical manifestations. There was a significant association between compromised activation signal of the examined areas and presence of clinically relevant manifestations.
The areas that showed compromised or absent signal with no clinically relevant manifestations or patient complaints (eleven areas) were language and language associated areas. They were replaced by contralateral areas signal activations.
Many of the examined s-motor and language areas showed simultaneous contralateral activation either in the dominant or non-dominant hemispheres, on 1.5 and 3T machines with no statistically significant differences between the two machines.
Contralateral areas activations during s-motor paradigms were significantly associated with compromised area of interest activation signal as well as with the presence of clinical manifestations.
Contralateral area activations during language paradigms were significantly associated with compromised or absent area of interest signal. Our studied patients did not include patients with manifestation related to language areas.
LMD between the areas activation signals and LOI measured in millimeters for areas showed activation ipsilateral to LOI. Twelve areas showed activation contralateral to the LOI, so we did not define LMD for them.
We considered false positive activation signal distance (FPAD). It is the maximum activation distance appearing outside cortical anatomical outlines. The mean false positive activation signal distance differed significantly between the 1.5T and 3T machines being more noted at 3T machines.
Some of the resulted areas activations (5 out of 60 =8.3%) were not matching the expected anatomical locations for them as considered by the ALM’s and decided from the conventional MRI images:
Our study included 54 examined areas for 26 patients underwent surgical excision of the lesions of interest. One of the other four patients refused the surgery due to presence of an operative risk on Broca’s area. Neurosurgery team offered non operative management for another one of them due to operative risk on hand s-motor, upper limb s-motor and Wernicke’s areas. The third one was an old lady, who had marked atherosclerotic changes resulted in neuro-vascular uncoupling. The last patient was suffering of a small cystic lesion likely neuroglial cyst and recommended for follow up to avoid general surgical risks.
Our fMRI results changed the management options (in about 51% of the examined areas) as well as intra operative procedures and decisions (regarding the dissection access and margins) (in about 70% of the examined areas).
The resulting language areas were coincident with the patients’ laterality in 23 (62.2%), while were not coincident with the patients’ laterality in 8 (21.6%) and were bilaterally represented in 6 (16.2%) of the examined areas.
Five of our study patients underwent surgery during awake craniotomy dealing with 13 areas, eight of them subjected to intraoperative electrophysiological monitoring (by direct cortical stimulation DCS). The results of the awake craniotomy and DCS were always supporting our results obtained from the preoperative non invasive fMRI.
Two of the examined areas (3.7) were associated with immediate post operative deficit, that was shortly post operatively resolved. None of the studied patients had recent post operative deficits.
There was no significant association between LMD and the immediate post operative status as well as the lesion excision either complete or partial, because we depended mainly upon the related sulcal anatomy in describing the lesions safety margins rather than distances measurement.
However distance below 10 mm in two patients underwent awake craniotomy without intraoperative monitoring was associated with development of minor weakness at the related motor areas during the dissection of the lesion’s related border in one patient and safety margin in the other patient. Both patients regained their full motor power at the recent post operative period.
Moreover nine of our studied patients underwent partial lesion excision because the lesions were sizable and high grade with risk of injury to other important intracranial structures.
Supplementary motor area activation during the s-motor and language paradigms occurred in 35 patients (60.3%) in association with the examined activated areas with no significant difference between the 1.5 and 3T machines.
However SMA activation was significantly more common during s-motor areas activations than during language areas activations.
Other areas signal activations occurred during language paradigms, namely, insula, middle frontal gyrus and anterior temporal areas. Middle frontal gyrus activation was the most common language-associated area activated during language paradigms.
Dorsal attention network showed signal activation in association with 31 activated areas during the s-motor and language paradigms. Spurious activations at the related vessels and even LOI were also noted.
However there was a significant association between lesion of interest spurious signal activation and the lesion type and grade.
Ten out of thirteen high grade gliomas showed spurious signal activation during the different paradigms. However the studied low grade gliomas, cavernomas, AVM and mesial temporal sclerosis did not show spurious signal activation.
Our study included two patients suffering from extra axial lesions (meningiomas), one of them showed spurious signal activation (it showed invasive signs) and the other did not show signal activation (meningiothelial meningioma).
Tb-fMRI has an established role in identifying eloquent brain regions essential for functions which can be at risk due to surgical approach, specially as there is considerable variability in function as well as anatomy between different patients and to preserve function during resection(98, 109).
Moreover, Anatomic landmarks may fail when tumor mass effect & edema effacing gyri and sulci, distorting the cortical anatomy(119). In addition intra axial neoplastic and vascular lesions or structurally evident epileptogenic focus may unexpectedly modify or shift functional areas(98).
Tb-fMRI provided localization of the s-motor, pre and supplementary motor areas, as well as lateralization, discrete focal localization and spatial mapping of language areas, such information cannot be obtained by any other techniques(5). It enables the neurosurgery team to pre operatively identify risks, to make preoperative therapeutic decisions, to wisely advise patients about the different estimated risks as well as benefits of the operative procedures and enables both to preoperatively explore other management options and treatment decisions. It helps neurosurgery team to preoperatively develop a plan, choose the safest entry zone and give a hope of maximum lesion resection with minimal post operative deficit facilitating post operative radiotherapy in patients with incomplete lesion excision(124). It helped decision making in choosing patient in need for awake craniotomy and/or intraoperative DCS, moreover guiding DCS reducing operative time(3).
The brain of each patient has its unique anatomy, and the resulting functional areas activations could present pathologically-induced unusual reorganization. Brain eloquent areas mapping is not standardized and must be done in a subject-specific pattern(115, 138).
Our study had some limitations as we did not include patients with high grade deficits as they will not be able to do the tbfMRI, moreover tb-fMRI failed in an atherosclerotic patient, but those can be evaluated by resting state fMRI. The limited number of patients underwent DCS to compare our results with DCS results, as DCS is considered the gold standard technique. Lack of standardized intra-operative radiological method for measuring the distance between the area activated and LOI after craniotomy, to be compared with DCS measured distance.
On the other hand our study offered some hope for future studies in need for dedicated work and support.