Experimental models of bladder cancer: a critical review

We have reviewed critically the available models of bladder cancer, and have attempted to compare their strengths and weaknesses where appropriate data are available. Further direct comparisons of the applications of each model will be needed in order to rank them, and to identify areas of research where each model will predominate. Furthermore, more information will be needed to validate each model in the context of human disease. With the increasing range and sophistication of the tools of molecular biology, a very important future direction will be the characterisation of animal and human bladder cancer, and in particular the study of the changes from normal to neoplastic urothelium: tumor markers, chromosomal patterns and oncogenes that are associated with specific biological functions, such as invasion, metastasis and ultimate prognosis. The transfection of normal tissues by oncogenes to yield transformed or immortalised lines may be of critical importance in identifying the nature of neoplastic transformation. The use of animal tumors and xenografts, each with the availability of a physiological internal milieu (although different from human metabolic conditions) may yield useful systems for the testing of new therapeutic approaches. However, of the utmost importance is the continuing need to characterise and validate each model, to avoid multiple publications and nomenclatures pertaining to common lines, and to recognise the limitations of the heavily adapted long term cell lines in vivo and in vitro. The major deficiencies of the available lines continue to be found in the method of their application to basic research, rather than being inherent in themselves. Although there are many theoretical and practical applications of these models, it should not be forgotten that the direct study of human bladder cancer, including the appropriate processing of biopsy specimens, will remain integral to understanding the biology of this disease.