Reliability of simple cytological methods in brain tumour biopsy diagnosis

Stereotactic procedures in brain tumour diagnosis and therapy are used increasingly because small lesions are more frequently detected and methods of interstitial irradiation are gaining wide acceptance among neurosurgeons. The diagnosis of underlying pathological processes has to be performed from very small tissue samples. The method of smear and/or squash preparations therefore has long been used by neurosurgeons. Neuropathologists, formerly reluctant, are becoming more and more enthusiastic over this simple and elegant method. Since the diagnostic reliability in such small samples is questioned, we have undertaken to analyse 200 open tumour biopsies in a comparative investigation: the diagnosis from smear preparation on the one hand has been performed without knowledge of the lesion, and on the other hand the remaining material was submitted to the whole pattern of diagnostic methods including frozen sections, paraffin sections with special staining, immunohistochemistry, electron microscopy and tissue culture. Diagnostic reliability was estimated in a four step procedure: Is there tumour or not--is there neuroepithelial newgrowth or metastasis--what kind of tumour is it--has the degree of malignancy--according to the criteria of the WHO--been estimated correctly? The overall diagnostic reliability in our material is about 75%. This is less than the diagnostic reliability of other methods, mostly due to the lack of histologic "architecture" criteria in evaluating the cytological biopsies.