Journal of Neuro-Oncology 35: 93–100, 1997.  1997 Kluwer A cademic Publishers. Printed in the Netherlands. L aboratory Investigation Immunoreactivity of human MA b BT32/A 6 with neuroepithelial tumors Michael D. Dan, 1 Pradip K. Maiti, 4 Xuanmin He, 1 G. Yancey Gillespie, 2 William C. Halliday, 3 Ashok K. Prashar, 4 Albert D. Friesen 4 and Howard A. Kaplan 4 1 Department of Neurosurgery, L ouisiana State University Medical Center, New Orleans; 2 Division of Neuro- surgery, University of A labama at Birmingham, Birmingham, USA ; 3 Department of Pathology, University of Manitoba Health Sciences Centre, Winnipeg; and 4 Novopharm Biotech, Inc., Winnipeg, Canada Key words: glioma, tumor-associated antigen, immunohistochemistry, monoclonal antibody, IgM Summary The present study was undertaken to determine the pattern of immunoreactivity of BT32/A6, a human IgM monoclonal antibody (MAb), with the following histological panels: 1) 30 human and non-human cell lines, 2) 32 normal human tissues, and 3) 28 tumors of central neuroepithelial origin (16 astrocytic; 11 non-astro- cytic). Antibody BT32/A6 recognizes a surface and cytoplasmic antigen present on a variety of human tumor cell lines including gliomas, melanomas, neuroblastomas, and a few sarcomas. The antigen is present (at least focally) on 15/16 astrocytic tumor tissue sections (94%), and in some cases, on close to 100% of cells. All malignant cell types, including small anaplastic cells, giant cells, gemistocytic cells, and cells forming pseudo- palisades were labeled by MAb BT32/A6. Non-astrocytic neuroepithelial tumors did not stain appreciably with MAb BT32/A6. There was weak immunoreactivity in a small subset of normal human tissues of epithelial and lymphoid origin, with the exception of adrenal cortex, which exhibited weak to moderate staining. All normal tissues of neuroectodermal and mesenchymal origin were unreactive. In conclusion, MAb BT32/A6 appears to be unique in that it recognizes a highly-expressed astrocytic tumor-associated antigen that is pre- sent on both low and high grade tumors. This makes it a strong candidate for further studies aimed at establish- ing its usefulness in the treatment of human astrocytic tumors. Introduction For several years, investigators have been aware that some patients with cancer are capable of mounting an immune response to their own tumors. Malignant brain tumors appear to be no exception to this phenomenon, in spite of being regarded as arising in an immunologically privileged site. Thus, in an earlier study of patients with malignant brain tumors by Pfreundschuh et al. [1], sera from 30 out of 30 patients were found to contain antibodies that reacted with cell surface antigens of cultured auto- logous astrocytoma cells by an anti-C3 mixed he- madsorption assay. Aware of these results, we sought to develop and characterize fully human an- ti-tumor monoclonal antibodies (MAbs) from the B-lymphocytes of glioma patients. Since cancer is not generally regarded as an auto- immune phenomenon, we postulated that MAbs derived from the B-lymphocytes of actual tumor patients (as opposed to mice immunized with hu- man tumors) would lack significant cross-reactivity with normal cellular components (or at least cellu- lar components that are normally sequestrated from the immune system) as a result of clonal abor- tion of self-reactive subsets occurring early in de- velopment. We previously described [2, 3] 5 fully human