[Results and reliability of stereotactic and endoscopic biopsies in brain tumors].

Stereotactic neurosurgery was the first "minimal invasive method" in the field of neurosurgery, later followed and partly replaced by endoscopic techniques. One reason for such an approach is sampling of small tissue probes for diagnosis, e.g. in brain tumours not accessible to open surgery. The appropriate method in the hands of the experienced is the "Quetsch" or smear technique. Its reliability is limited by the fact, that the "architectural" or "tissular" components of tumours lack in those purely cytological preparations. Tissue architecture however is crucial for the assessment of different grades in glial tumour progression. The grade of a glial tumour is the most critical information for the patient and the therapist; grading of the supratentorialf gliomas of the adult by means of cellular and tissue pleomorphism therefore forms the basis of Zülchs system of classification and grading of all intracranial tumours by comparison of postoperative survival. The resulting four grade system--slightly modified--is part of the old and new issue of the WHO classification of brain tumours. In order to specify the possibility of correct diagnosis and grading in probes gained by the minimal invasive techniques, we present results of three diagnostic approaches: First: We report results of a study performed during the last twelve years in which a diagnosis of smear preparations had been made on neurosurgical specimens prior to conventional handling. The "blind" cytological diagnosis was then compared with the final diagnosis of the tumour using light and electron microscopy and immunohistochemistry. Second: We report results and estimates of tissue probes gained by the so called sandwich technique in which the removal of material for cytological analysis is done stepwise. By doing so, material representative for different compartments of the neoplasm is obtained. This implies collaboration between neurosurgeon and neuropathologist not only during the time of stereotactic action but also in the planing period; the correct interpretation of the different compartments delivered by imaging methods in this context is essential. Third: We present selected cases of probe sampling under direct visual control by endoscopy. This method is especially useful for tumours bordering the ventricular system. Surface structures and cyst linings can be visualized directly if the endoscopist is familiar with normal and pathological tissue appearance. The specimen for analysis may therefore be taken from the most relevant tumour region and the sandwich technic which means tissue damaging in multiple localisations can be partly or fully avoided.(ABSTRACT TRUNCATED AT 400 WORDS)