Monoclonal antibody immunophenotyping of solitary cerebral metastases with unknown primary sites

Various authors have reported the incidence of the solitary metastatic lesion to range from 25-85% of cases with known cerebral metastasis. In about 20% of these cases, no primary neoplasm is ever identified. The therapeutic alternatives considered depend to a great extent on the histopathology of the tumour. Unfortunately, the difficulty of making a correct diagnosis in this setting is compounded by all of the problems associated with tumour sampling, i.e. the representativeness of the sample, the often miniscule amounts of tissue obtained, mechanical deformation of the sample, etc. This study describes the use of a monoclonal antibody panel in the immunophenotyping of 45 tumours where the differential diagnosis was glioma versus metastatic tumour with no known primary site. The monoclonal antibodies used bound neuroectodermal antigens (UJ13A), cytokeratin (LE61), and epithelial membrane antigens (AUA1), as well as human milk fat globule antigens (HMFG1, HMFG2). The neuroectodermally-derived MAB UJ13A accurately and reproducibly differentiated gliomas and metastatic lesions with one exception. Oat cell carcinomas (n = 5) were positive for both UJ13A and cytokeratin (LE61). There were no cases in which metastatic tumours were positive for UJ13A and negative for cytokeratin. A combination of the LE61 and AUA1 MABs resulted in positive staining in 44/45 (98%) cases. Immunophenotyping made a substantial contribution in 13 cases (29%), and in three cases the MAB panel was responsible for the final diagnosis. The antibody panel described in this study enables primary CNS neoplasms to be accurately distinguished from nonlymphoid metastatic tumours. In addition, a positive diagnosis of carcinoma is made in 98% cases, although identification of the precise organ of origin is not generally possible.