Material: 28 stereotactic biopsies of organic brain processes (brain tumours) were performed in the years 1997-2000 in the Department of Neurosurgery, Medical University of Warsaw. In this series the lesions were located in corpus callosum in 5 patients, in basal nuclei in 9, and deeply in the white matter of cerebral hemispheres in 14.
Method: The Baklund biopsy kit and Leksell's stereotactic frame were used, target localisation was based on the CT scan. Histological verification was based on hematoxillin--eosin staining, completed with histochemical evaluation if necessary. In 9 patients intraoperative smear evaluation was performed.
Results: Sensitivity of stereotactic biopsies was 86% (24/28), although the rate of conclusive biopsies was lower, being 60% (17/28). False negative results were observed in 14% of the patients (4/28). Analysis of the results revealed, that the sensitivity was not dependent on the size, neither on the location of the tumour, but was related to its morphology. The false negative results were obtained in the tumours with significant necrosis (as seen on CT scans). There were no surgical complications in this series.
Conclusions: 1. Difficulties in stereotactic biopsies of brain tumours are associated mainly with tumour morphology. In tumours with marked necrosis, other degenerative changes or cystic ones, higher risk of non-conclusive biopsy may be expected. 2. Size of the tumour and its location do not affect the diagnosis based on the stereotactic biopsies. 3. In the polymorphic tumours, the policy to take biopsy material from different tumour sites, should be a rule, as different parts of the lesion may represent different stages of malignancy and histological evaluation of separate parts of the tumour may lead to inadequate oncological treatment.