Numerous retrospective immunohistological studies have shown a correlation between deletion of normal epithelial ABH """"""""blood group"""""""" antigens from noninvasive urinary bladder carcinomas and development of recurrent carcinomas with deep muscle invasion. We have recently shown that detection of the normally cryptic T (Thomsen-Friedenreich) blood group related antigen in tumors provides additional prognostic information. However, nothing is known of the biological basis for the correlations between the profile of blood group related antigens on tumor cell surfaces and malignant phenotype. Also, it is unclear whether purely qualitative antigen testing as currently performed will be clinically useful for assessing prognosis and therapy for the individual patient. Because of the inherent limitations of retrospective, immunohistologic studies, we have focused our attention on antigenically defined cell lines. Our overall objective is to test the hopothesis that the correlations observed in retrospective studies may reflect a quantitative link between expression of blood group related cell surface antigens and malignant phenotype (invasion, metastasis) at the level of defined tumor cells populations. We plan to establish by flow cytometry from continuous human urinary bladder carcinoma cell lines, sublines with defined profiles of blood group related cell surface antigens. We will then correlate antigen expression (quantitative cytometry, immunoelectron microscopy) with quantifiable parameters of the malignant phenotype of each subline; a) tumorigenicity; b) growth in semisolid media; c) hyperploidy; d) invasion (bone invasion assay); e) metastasis (ling colony assay).
