The ABO and Lewis blood group antigens are cell surface carbohydrate determinants formed by the sequential addition of saccharides to precursor backbone structures of membrane lipids and proteins. Suppression of normally active glycosyltransferases results in the absence of antigens that are normally expressed. ABH antigen deletion in malignant and premalignant urothelium has been extensively evaluated; it appears to correlate with significantly higher rates of tumor recurrence and disease progression. However, we have recently shown that the ABH blood group system is differentially expressed in the normal urothelium of secretors in contrast to nonsecretor individuals. The normal urothelium of nonsecretors does not express A, B or H determinants; therefore, deletion of ABH antigens can Q& be ascertained in secretor individuals. Although nonsecretors only comprise 2244% of the population, this observation mandates a reevaluation of earlier studies. Deletion of A, B or H antigens is noted in carcinoma in sifu (CIS), and in invasive and metastatic transitional cell carcinoma (TCC) of secretor individuals. Increased synthesis or activation of normally quiescent glycosyltransferases in bladder tumors can result in the expression of aberrant tumor-associated blood group antigens. lmmunohistochemical analysis has demonstrated that Lewis X (Lex) is not detected in normal adult urothelium except for occasional umbrella cells. However, papillomas, CIS and TCC expressed Lex in 84% of cases, regardless of grade, stage, blood type or secretor status of the individual. The presence of Lex-positive cells in bladder lavage specimens enhanced the detection of urothelial tumor cells, correctly identifying bladder tumors in 2531293 (86OA) cases compared to a 63% sensitivity for cytology alone. The specificity of Lex immunocytology was 87%; 57 of 65 control subjects were negative for the Lex antigen. Furthermore, the detection of the Lex antigen in exfoliated bladder cells is a useful marker for predicting bladder tumor relapse in high risk, disease-free patients. All 17 Lex-positive patients recurred with clinical evidence of disease between 2 and 33 months (mean 8.4 months), while only 8 of 39 Lex-negative patients recurred; 31 patients remain NED
( 2 4 0 months, mean 16.2 months).