Article
Combined analysis of MR-spectroscopy and DTI data for surgical planning and image-guided resection of gliomas in eloquent brain areas
Search Medline for
Authors
-
Artem Rozumenko
- Department of Neuro-Oncology, Romodanov Institute of Neurosurgery, Kiev, Ukraine
-
Miriam H. A. Bauer
- Klinik für Neurochirurgie, Universitätsklinikum Marburg, Marburg, Germany; Klinik für Psychiatrie und Psychotherapie, Universitätsklinikum Marburg, Marburg, Germany; International Clinical Research Center, St. Anne’s University Hospital, Brno, Czech Republic
-
Barbara Carl
- Klinik für Neurochirurgie, Universitätsklinikum Marburg, Marburg, Germany
-
Christopher Nimsky
- Klinik für Neurochirurgie, Universitätsklinikum Marburg, Marburg, Germany
| Published: | June 2, 2015 |
|---|
Outline
Text
Objective: The surgical resection of tumors located within or in close vicinity to eloquent brain areas carries an increased risk of postoperative neurologic deficits that reduces quality of life in operated patients. Delineation of white matter tracts – using Diffusion Tensor Imaging (DTI) – in the vicinity of brain gliomas is therefore essential for surgical planning.
Method: A total of 5 patients (age range 50-77 years) with cerebral gliomas underwent preoperative 3T-MRI with Gd-enhancement (3D-T1W, T2W, FLAIR, DWI) and magnetic resonance spectroscopic imaging (MRSI). Brain tumors were located close to eloquent areas (Sawaya grade II) in all patients. The volumes of interest for MRSI covered the areas with decreased fractional anisotropy (FA) in projection of white matter tracts close to tumor borders. The ratios of N-acetylaspartate, choline and lactate-lipids were analyzed. The surgical 3D planning was performed with DTI-based fiber tacking and MRSI data. The surgical trajectory and volume of safe tumor resection were defined and calculated before operation. The tumor microsurgery resection was carried out using a multimodal navigation system and intraoperative ultrasound.
Results: MRSI allowed detection of spatial distribution of specific metabolite concentrations in pathological brain areas compared with the normal brain. The lactate-lipids concentration was associated with areas of anaerobic metabolism or necrosis. The lactate-lipids and choline ratios were used as a parameter sensitive to neural structure destruction and tumor cells active proliferation.
The combined analysis of MRSI with DTI helps to estimate the fiber bundles preservation by identification of their destruction due to tumor cell infiltration in areas with or without perifocal edema next to the tumor border. The multimodal 3D planning technology supported the ability of anatomic and metabolic data sets integration for verification of safe tumor resection limits.
Conclusions: The combined analysis of MRSI and DTI data seems to be a reliable method for the determination of structural integrity of the white matter tracts and may optimize the resection volume while preserving the surrounding functioning brain tissue.
