Proliferating cell nuclear antigen immunoreactivity in human central nervous system neoplasms.

Abstract

Formalin-fixed, paraffin-embedded surgical specimens from 140 primary human central nervous system tumors, including 51 meningiomas, 26 astrocytomas, 26 anaplastic astrocytomas, 9 glioblastomas, 1 gliosarcoma, 8 oligodendrogliomas, 5 ependymomas, 2 subependymomas, 9 medulloblastomas, and 3 paragangliomas, were immunostained using a streptavidin/peroxidase method and the PC10 monoclonal antibody, which recognizes an epitope on the proliferating cell nuclear antigen (PCNA). The following PCNA labeling index (LI) mean values were found for the above neoplasms: meningiomas, 3.80 +/- 7.35%; astrocytomas, 0.65 +/- 1.03%; anaplastic astrocytomas, 8.46 +/- 7.95%; glioblastomas, 10.26 +/- 11.21; gliosarcoma, 46.34%; oligodendrogliomas, 2.31 +/- 3.59%; ependymomas, 1.12 +/- 2.10%; medulloblastomas, 23.91 +/- 11.95%; and paragangliomas, 2.07 +/- 1.86%. Collectively, our findings indicate that while benign central nervous system tumors generally have low PCNA LI values, consistent over-expression of PCNA epitopes was noted in some examples, especially in a number of meningiomas. Among the malignant neuroectodermal tumors, medulloblastomas were found to have the highest PCNA LI values, corresponding to their histological grade of malignancy, and malignant glial tumors generally displayed significantly higher PCNA LI values, than their benign counterparts. Although in our study mean PCNA LI values seemed to reflect histological grading, large discrepancies were noted in all tumor groups. Our data, therefore, suggest than PCNA immunoreactivity can not be considered reliable for predicting the prognosis of the disease in individual cases.