The histogenesis of primary nonurachal mucus-producing adenocarcinomas of the urinary bladder including signet ring cell carcinomas remains to be elucidated, since the normal bladder contains neither columnar nor mucus-secreting glandular epithelium. Based upon the assumption that adenocarcinomas may develop secondarily from pre-existent transitional cell carcinomas (TCC) by a metaplastic process, it was the purpose of the current immunohistochemical study to analyze whether urothelial carcinomas are capable of secreting MUC5AC apomucin, using the monoclonal antibody 45MI. This antibody has been initially demonstrated to strongly react with the mucus-producing columnar cells of the surface gastric epithelium, recognizing a specific epitope located on the peptide core of glycoproteins as major components of mucins. Nine of 64 uniformly differentiated papillary (14.1%) and 5 of 66 nonpapillary (solid) TCC with a uniform urothelial differentiation (7.6%) expressed the MUC5AC antigen, yielding an overall incidence of 10.8%. Transitional cell carcinomas with a focally altered cellular and structural differentiation (squamous cell, pseudoglandular, true glandular and mixed differentiation) stained positively in a substantially higher percentage of 43.8% (21 of 48 cases). A positive immunoreactivity was also observed in 3 of 19 mixed transitional cell and nonurothelial carcinomas. The tumor-associated resurgence of normally cryptic MUC5AC antigenic determinants in transitional cell carcinomas is considered as a re-expression of oncofetal antigenicity, probably as a result of the embryologic origin of the urinary bladder from the pluripotent tissues of the cloacal endoderm and the mesodermal wolffian ducts. Our findings may help to better understand the histogenetic development of mucus-secreting vesical adenocarcinomas from pre-existent urothelial carcinomas.

The histogenesis of nonurothelial carcinomas of the urinary bladder is difficult to understand, since the bladder is normally lined exclusively by transitional cell epithelium. To gain more insights into the pathogenesis of nonurothelial carcinomas, the morphology and immunohistochemistry of transitional cell carcinomas (TCC), mixed transitional cell and nonurothelial carcinomas, and pure nonurothelial carcinomas were comparatively studied. Of papillary and of nonpapillary (solid) TCC (overall incidence 6.8%), 4.8% and 15.4%, respectively, disclosed foci of altered celllular and architectural phenotypes, consisting of squamous epithelium, pseudoglandular formations, and true glands with or without mucus production. The diverse phenotypic variants develop obviously by a metaplastic process as a result of the well-known inherent potential of the urothelium to undergo several pathways of cellular differentiation. There is strong evidence that squamous cell carcinomas arise secondarily from a squamous metaplasia and adenocarcinomas from metaplastic glandular epithelium within pre-existing TCC following complete carcinogenic transformation of the initially bland-looking metaplastic tumor cells. The metaplastic origin of nonurothelial bladder carcinomas is supported by immunohistochemical findings. The high molecular weight cytokeratin 34betaE12 identifies tumor cells with squamous characteristics, helping to explain the development of squamous cell carcinomas. Secretion of MUC5AC apomucin is assumed to play a central role in the histogenesis of nonurachal mucus-producing adenocarcinomas, including signet ring cell carcinomas. Metaplastic phenotypic variants of TCC should be recognized as distinct tumor entities with the potential to transform into nonurothelial carcinomas and thus possibly implying a poorer clinical outcome than typical, uniform TCC.