H. Richard Beresford

Normal differentiation and malignant transformation of human cells are characterized by specific changes in surface antigen phenotype. In the present study, we have defined six cell-surface antigens of human sarcomas and normal mesenchymal cells, by using mixed hemadsorption assays and immunochemical methods for the analysis of cultured cells and immunohistochemical staining for the analysis of normal tissues and greater than 200 tumor specimens. Differential patterns of F19 (Mr, 120,000/95,000 glycoprotein), F24 (Mr, 95,000 glycoprotein), G171 (Mr, 75,000 glycoprotein), G253 (Mr, 90,000 glycoprotein), S5 (Mr, 120,000 glycoprotein), and Thy-1 (Mr, 25,000 glycoprotein) antigen expression were found to characterize (i) subsets of cultured sarcoma cell lines, (ii) cultured fibroblasts derived from various organs, (iii) normal resting and activated mesenchymal tissues, and (iv) sarcoma and nonmesenchymal tumor tissues. These results provide a basic surface antigenic map for cultured mesenchymal cells and mesenchymal tissues and permit the classification of human sarcomas according to their antigenic phenotypes.

We have characterized five distinct cell surface antigens of human astrocytomas and correlated their expression with the expression of glial fibrillary acidic protein (GFAP) and four previously defined cell surface markers of astrocytomas. One of the newly studied antigens, A4, which was originally detected on rat central nervous system (but not peripheral nervous system) neurons, is expressed on GFAP+ human astrocytoma cells, but not on GFAP- astrocytomas or a wide range of other neuroectodermal, epithelial, and hematopoietic cells. Antigens F19 (Mr 140,000/90,000 glycoprotein) and F24 (Mr 90,000 glycoprotein) also show restricted distribution and are expressed on subsets of neuroectodermal and mesenchymal cells. Antigens G253 (Mr 95,000 glycoprotein) and S5 (Mr 120,000 glycoprotein) are more widely distributed on the cultured cell panel. The distribution of these antigens was determined on a series of 22 astrocytoma cell lines and in normal brain tissue and the results were compared with the distribution of 5 additional glial cell markers: GFAP and cell surface antigens A010 (Mr 110,000 glycoprotein); AJ8 (Mr 100,000 glycoprotein); LK26 (Mr 35,000 glycoprotein); and Thy-1. Distinct patterns of expression on cultured astrocytomas and in neural tissues were identified for all antigenic systems studied, and cell surface expression of antigen A4 was found to correlate closely with GFAP phenotype of cultured astrocytomas. The antigens described in this study provide new markers to study normal glial differentiation and to correlate the phenotypes and biological behavior of distinct subsets of astrocytomas.

The neural cell adhesion molecule (NCAM) was discovered in a search for cell surface antigens of chicken neurons that contribute to cell adhesion and pattern formation during development. Homologous adhesion molecules have been identified in several species, including humans. In this immunohistochemical study, the authors examine the role of human NCAM in tumor diagnosis. The authors used a monoclonal antibody (MAb), 5.1H11, to examine NCAM immunoreactivity in frozen sections of more than 450 tumors, including more than 80 small round cell tumors (SRCT) of childhood and adolescence (neuroblastomas, Ewing's sarcomas [ES], peripheral neuroepitheliomas [PN], primitive neuroectodermal tumors [PNET], esthesioneuroblastomas, malignant ectomesenchymoma, medulloblastomas, small cell osteosarcomas, mesenchymal chondrosarcomas, embryonal rhabdomyosarcomas, and lymphomas). The authors show that 1) neuroblastomas and primary brain tumors are NCAM+; 2) ES, most PN/PNETs, and melanomas are NCAM-; 3) embryonal rhabdomyosarcomas and various other sarcomas are NCAM+; 4) neuroendocrine tumors are NCAM+; 5) subsets of carcinomas of kidney, ovary, lung and other organs are NCAM+; and 6) lymphoid tumors are NCAM-. Tests with normal fetal and adult tissues indicate that these findings reflect only in part the NCAM phenotypes of corresponding normal tissues. Notably the NCAM- phenotype of ES and PN/PNET is not explained by current histogenetic models for these tumors, which suggest a primitive neuroectodermal origin. Finally the authors show that NCAM expression among SRCT has an inverse relationship with the expression of p30/32MIC2, a cell surface antigen of ES and PN/PNET detected with MAb HBA71. These results suggest that immunohistochemical assays for NCAM and p30/32MIC2 expression may aid in the further characterization of SRCT of childhood and adolescence.

The surface antigens of cultured human malignant astrocytomas were analyzed by using mouse monoclonal antibodies. BALB/c mice were immunized repeatedly with either SK-MG-1 [a glial fibrillary acidic protein (GFA)-negative astrocytoma line] or SK-AO2 (a GFA-positive astrocytoma line). After fusion with NS/1 mouse myeloma cells, 12 antibody-producing clones were selected for detailed study. Serological analysis permitted the identification of nine distinct antigenic systems. Four monoclonal antibodies (Ab AJ225, Ab AO10, Ab AJ8, and Ab AO122) identified cell surface antigens preferentially expressed on tumors of neuroectodermal origin, and these antibodies subdivided the astrocytoma panel into distinguishable subsets. The determinant detected by Ab AO10 and Ab AJ8 showed mutually exclusive expression on the astrocytoma lines. The AO10 and AJ8 phenotypes appeared to reflect the differentiation state of the cultured cells; 4/7 AO10-positive astrocytomas expressed GFA, an intracellular astrocyte differentiation antigen, whereas all AJ8-positive astrocytoma (9/9) were GFA-negative. Five antibodies (Ab AJ10, Ab AJ9, Ab AJ17, Ab AJ425, and Ab AJ2) recognized determinants widely distributed on normal and malignant cells. Four antibodies defined in this study precipitated proteins from reduced preparations of radioisotope-labeled SK-MG-1 and SK-AO2 cells: Ab AJ225 (Mr 145,000); Ab AO122 (Mr265,000); Ab AJ10 (Mrs 195,000 and 165,000); and Ab AJ2 (Mrs 170,000, 140,000, 140,000, and 28,000).

In a prospective double-blind study, we compared dexamethasone and placebo in 33 subjects with lumbosacral radicular pain. Of subjects with resting pain, 7/21 improved on dexamethasone, and 4/12 improved on placebo. Of subjects with pain on straight-leg raising, 8/19 improved on dexamethasone and 1/6 on placebo. Of 27 subjects evaluated 1 to 4 years after treatment, 8/16 who had received dexamethasone were asymptomatic or had only occasional mild low-back pain, compared with 7/11 who had received placebo. Thus, dexamethasone is not superior to placebo for either early or long-term relief of lumbosacral radicular pain, but may reduce pain evoked by stretch of acutely inflamed spinal nerve roots.