Nicht-fluoreszenz-basierte Konfokale EndoMikroskopie in der Neurochirurgie: ein intraoperatives Hilfsmittel für die Tumorklassifikation und Resektionskontrolle?

Objective Extent of tumor resection and histopathological diagnosis influence survival time and outcome in patients with brain tumors. Recently a couple of studies were published that showed the potential of confocal microscopy using fluorescence to detect tumor in and ex vivo. To establish an intraoperative technique to differentiate between tumor and healthy tissue we used confocal laserendomicroscopy (CLEM; Storz: Tuttlingen, Germany) without fluorescence in tumor samples of patients. Methods Tumor samples of glioblastoma WHO°IV (n=11), meningioma WHO°I, II (n=12), metastases (n=3), oligoastrocytoma WHO°III (n=2), cavernom (n=2), oligoastrocytoma WHO°II (n=1), pleomorphic xanthoastrocytoma WHO°II (n=1), oligodendroglioma WHO°III (n=1), hemangiopericytoma WHO°II (n=1) and haemangioblastoma WHO°I-II (n=1) were examined. Before samples were analysed by conventional histopathology CLEM was used to examine tumor samples ex vivo but immediately after resection. At least 100 images of each sample were collected with a lateral resolution of 2 µm and a field of view of 300x300 µm. CLEM and histological images of each tumor were collated. In addition CLEM was also performed on fresh mouse brain. Images of cerebral cortex, white matter, corpus callosum and cerebellum were collected and compared with conventional histological images. Results In patients with high grade glioma and metastases normal brain tissue, infiltration zone, vital tumor cells and necrosis were detected reproducible using CLEM. WHO criteria such as cell density, microvascular proliferation and necrosis could be visualized. A differentiation of metastases vs. glioblastoma was problematic but the differentiation to normal brain tissue was obvious. Collagen patterns, calcifications and/or psammoma bodies were represented consistently in meningeoma and correlated to conventional histopathology. Images of murine cerebellum showed typical features such as molecular layer, purkinje cell layer and granule cell layer. Neural tracts in corpus callosum and white matter as well as typical aspects of cortex were identified. Discussion Non-fluorescence CLEM facilitates a distinction of brain structures as well as tumor tissues. Thus it is possible to detect infiltrative brain tumor margins and characteristics like cell density during surgery. Since no contrast media (fluorescence) is required using this technique in vivo is very easy. It may help to get intraoperative rapid section diagnosis more simply and to make a decision about operative strategies just as to make sure all pathological tissue is removed.