Immunohistochemical evaluation of novel and traditional markers associated with urothelial differentiation in a spectrum of variants of urothelial carcinoma of the urinary bladder.

Data on immunohistochemical expression of novel and traditional urothelial markers in the wide range of urothelial carcinoma variants have so far been very limited. In this study, whole tissue sections from 130 bladder urothelial carcinoma and variants were stained with a panel of novel and traditional immunomarkers supportive of urothelial lineage. The positivity rates were as follows: (a) urothelial carcinomas with or without divergent differentiation: GATA3 (50%), S-100P (86%), uroplakin III (20%), thrombomodulin (40%), cytokeratin 7 (CK7) (80%), CK20 (55%), p63 (87%), and high molecular weight cytokeratin (HMCK) (89%); (b) urothelial carcinoma variants (micropapillary, plasmacytoid, nested, clear cell, and microcystic): GATA3 (88%), S-100P (96%), uroplakin III (33%), thrombomodulin (49%), CK7 (95%), CK20 (61%), p63 (69%), and HMCK (96%); and (c) undifferentiated carcinomas (lymphoepithelioma-like carcinoma, small cell carcinoma, sarcomatoid carcinoma and carcinoma with rhabdoid and giant cells): GATA3 (28%), S-100P (31%), uroplakin III (0%), thrombomodulin (22%), CK7 (50%), CK20 (3%), p63 (50%), and HMCK (49%). In urothelial carcinoma with squamous differentiation, GATA3 expression was lower (20%) in contrast to p63 and S-100P. In urothelial carcinoma with glandular differentiation, GATA3 (50%) and p63 (60%) expression was lower than S-100P (100%). p63 expression was relatively lower in micropapillary (54%) and plasmacytoid (50%) variants compared with the other urothelial carcinoma variants. This study provides comprehensive data for novel and traditionally used markers to support urothelial lineage in urothelial carcinoma variants. Our findings show that GATA3, S-100P, CK7, CK20, HMCK, and p63, in the appropriate differential diagnostic setting, are useful to support urothelial lineage of variant morphologies.

Immunohistochemical profile to distinguish urothelial from squamous differentiation in carcinomas of urothelial tract.

Urothelial neoplasms with squamous morphology raise the differential diagnosis between pure primary squamous cell carcinoma, urothelial carcinoma with squamous differentiation and secondary involvement by squamous cell carcinoma, for example, from uterine cervix. Accurate identification between these entities is critical due to differing prognosis and therapeutic strategies. We evaluated the utility of an immunohistochemical panel of 3 urothelial-associated antibodies (uroplakin III, S100P, and GATA3) and two squamous-associated antibodies (CK14 and desmoglein-3) in 50 primary urothelial neoplasms: 15 pure urothelial carcinomas, 12 pure squamous cell carcinomas and 23 urothelial carcinomas with squamous differentiation. Squamous differentiation was defined by intercellular bridges or evidence of keratinization. Pure squamous cell carcinomas were positive for CK14 (100%) and desmoglein-3 (75%), negative for GATA3 and uroplakin III; one case was S100P positive (9%). Pure urothelial carcinomas had an opposite pattern and were positive for S100P (93%), GATA3 (93%), and uroplakin III (67%) and were negative for desmoglein-3; CK 14 was positive in 27% of cases; 74% of urothelial carcinomas with squamous differentiation had expression of urothelial and squamous associated markers (S100P, 83%; GATA3, 35%; uroplakin III, 13%; CK14, 87%; and desmoglein-3, 70%), although reactivity for individual markers within some tumors did not always correspond with morphologic differentiation. Of the remaining 26%, 4 showed an overall "squamous" immunoprofile, whereas 2 cases showed a "urothelial" immunoprofile. Our study showed that a panel of five antibodies identifies squamous and urothelial differentiation in most instances suggesting potential diagnostic utility.