Cerebellar astrocytomas. Part I. Macroscopic and microscopic features.

Some of the pathological features of 112 cerebellar astrocytomas seen at The Radcliffe Infirmary between 1938 and 1984 have been described. These include the following: Macroscopic appearance: Cerebellar astrocytomas, when compared to other gliomas, have an unusual tendency to become cystic. Thus, if all ages are considered only 22% are total solid tumors while the others have either a large solitary cyst or many smaller ones. Although patients with cystic tumors remain free of recurrence far longer than those with totally solid astrocytomas, the proportion of cystic and solid elements does not significantly influence the overall survival period. Moreover, the degree of tumor demarcation is not an absolute indicator of the degree of malignancy. Extent of disease: Approximately 8% of cerebellar astrocytomas infiltrate the brainstem and these cases are associated with the poorest prognosis. Supratentorial invasion is very rare and CSF dissemination is very uncommon. Only one example of the latter was recorded in the present study. It is important to note that subarachnoid spread may also occur from histologically benign tumors as well as malignant cerebellar astrocytomas. Classification by growth pattern: The classification of cerebellar astrocytomas by "growth pattern" is clearly of limited use. This is due to the presence of substantial transitional forms in both the "juvenile" and "diffuse" as well as in the "fibrillary" and "protoplasmic" classifications. Confusion also arises because of the disregard for the quantity and distribution of certain features which, in turn, limits the usefulness of the "A" and "B" type classification system. Failure to define rigorously the cell type which predominates in the "cerebral" and "diffuse" forms also creates substantial difficulties in using Ringertz' nomenclature. Further problems are caused by Zülich's "polar spongioblastomas" since cerebellar astrocytomas are not composed of primitive spongioblasts, the "polar spongioblastomas" merely representing one subtype of cerebellar astrocytoma, namely the piloid or pilocytic variety. Attempts to classify cerebellar astrocytomas in a manner which does not rely upon any specific cell type but which depends instead upon the degree of homogeneity or heterogeneity of cell growth is of limited value. Grading schemes: Although the tumors studied in the present report were not graded, previous authors have shown that "grading schemes" are of little predictive value in assessing the behavior of cerebellar astrocytomas.(ABSTRACT TRUNCATED AT 400 WORDS)