Introduction
The present study aimed to explore whether the integration of PET data into a cranial neuronavigation system in combination with cortical functional mapping would enable neurosurgeons to optimize the tumor resection without postoperative neurological deterioration.
Methods
Ten patients with intrinsic brain tumors including four oligodendrogliomas, two oligoastrocytomas, one pleomorphic xanthoastrocytoma, one glioblastoma, one gangliocytoma, and one fibrosarcoma were preoperatively examined with PET using C-11 methyl-L-methionine (Met) and F-18 fluorodeoxyglucose (FDG). Seven tumors located in the right frontal lobe or in the left parietal lobe, were close to motor or somatosensory cortex. The remaining three tumors were located in the temporal lobe. The tracer accumulation in the tumor was visually compared to the contralateral gray matter on PET images scanned 35 min and 50 min after intravenous injection of Met and FDG, respectively. The VectorVision2 (BrainLAB AG, Munich, Germany) incorporated with thin-slice, contrast enhanced T1-weighted, and T2-weighted MRI scans was used as a navigation system. Met and FDG PET data were converted and fused with MR images by PatXfar5.2 and VectorVision Cranial Planning1.3 (BrainLAB AG, Munich, Germany) using an automatic image-fusion algorithm. PET images and somatosensory evoked magnetic fields of magnetoencephalography (MEG) were also superimposed on corresponding MR images. Silicon tubes were implanted as an anchoring marker for a resection to the margins of the resectable tumors that was demonstrated as the area of increased Met accumulation on co-registered images of PET and MRI in the navigation system. The tubes were inserted before dural incision to prevent from the dislocation due to brain shifts by the leakage of cerebrospinal fluid. The correct localization of the precentral gyrus was intraoperatively verified by cortical somatosensory evoked potential monitoring.
Results
Met-PET clearly demonstrated tumor extent by increasing uptake of Met as a “hot” lesion in all patients. The areas of inhomogeneous FDG uptake in the tumors were completely surrounded by those of “hot” lesions on Met-PET images. The “hot” lesions on Met-PET were all larger in size than the Gd-enhancing lesions on MRI. The area of Met accumulation that is believed to be highly consistent with the distribution of tumor cells, extended sparsely over the T2-hyperintensity lesions. Postoperative MRI revealed that the tumors were totally resected in all eight patients, with silicon tubes implantation method. In contrast, the T2-hyperintensity lesions extending partly beyond the area of Met accumulation diminished on follow-up MRI. There was no additional morbidity caused by the surgical intervention in all patients. Follow-up MRI revealed no evidence of recurrent tumors in all patients 2 to 55 months (median, 31 months) after the tumor resection.
Conclusions
The integration of the preoperative metabolic (PET), functional (MEG), and anatomic (MRI) images provides important information to optimize the tumor resection without an operative neurological morbidity. The silicon tube implantation method under the monitor of this navigation system is simple and accurate to determine the border of the intrinsic brain tumors to be resected. The complete resection of tumors defined as high-Met-uptake would be associated with improvement of the survival with better quality of life.