There is a steady search for procedure which could replace or at least reduce the frequency of the invasive cystoscopy in the surveillance of heterogeneous superficial transitional cell carcinoma (TCC) of the bladder. Recently, UroVysion FISH assay has been shown to provide with better sensitivity than the urine cytology except for the lowest stage pTa and grade I-II TCCs. Data indicate that this failure of the sensitive FISH might be due to mistargeting. Therefore, our aim was to elaborate a procedure enabling FISH analysis in phenotypically preselected urothelial cells, only. Cytokeratin 7 (CK-7) chromogenic immunolabeling was applied to various mixtures of negative and positive control cells as well as voided urine specimens. Cellular targets and CK-7 positive cells were identified by morphometric and pixel intensity indices using an automated microscope workstation. UroVysion FISH pattern was analyzed only in the subsequently relocalized CK-7 positive events. Automated phenotypical preselection of urothelial cells proved to have 97.3% sensitivity, 96.1% specificity, and 99.0% accuracy, whereas combined pheno-and genotyping revealed 93.3% sensitivity and 99.8% specificity, respectively. In clinical samples, the overall 20.4% FISH positivity gained by traditional target identification contrasted with the 55.6% positivity obtained by the combined method, by which the efficiency of identifying chromosomally aberrant cells proved to be two to threefold higher even in grade I lesions. FISH analysis of phenotypically preselected urothelial cells might represent a reliable asset in surveillance of low stage-low grade TCCs. ' International Society for Analytical CytologyKey terms transitional cell carcinoma; voided urine cell specimens; cytokeratin 7; UroVysion; fluorescence in situ hybridization AS transitional cell carcinomas (TCC) of the bladder are frequently multifocal, likely for this reason recur often (15-70% at 1 year) and may progress into higher stage and grade (7-40% at 5 year), thorough follow-up of patients is indispensable (1-4). However, the duration, the frequency, and the method(s) of choice for that is under discussion (5). It is because the superficial tumors accounting for the vast majority of TCC at presentation are not homogeneous either in clinical course or the underlying oncogenetic pathway (6). The noninvasive, papillary pTa lesions with frequent recurrence, but low tendency for progression to invasive tumor represent 80% of the superficial TCC and hallmarked by structural chromosomal abnormalities on 9q and 9p. The rest is composed from highly malignant, invasive tumors, mostly evolving from flat carcinoma in situ (Cis) driven by chromosome 17-17p13.1 (p53) numerical or structural abnormalities. From therapeutical point of view, the situation is even more complicated as in addition to low-and high-risk superficial TCCs, the interme-

Urovysion multitarget fluorescence in situ hybridization (FISH) assay is a promising tool for detection of bladder cancer, however, there is still no consensus regarding abnormal signal pattern and cut-off level, and the recommended targeting carries limitations similar to urine cytology. Aim of this study was to explore diagnostic benefits of a recently introduced method featuring target specific genotyping, as well as to investigate the feasibility of locally and statistically determined cut-off, compared with conventional evaluation scheme. Histology, cytology, and comparative FISH approaches were performed on 42 patients with high clinical suspicion for urothelial carcinoma (UC). FISH parallels were (1) Urovysion-alone (according to manufacturer's instruction); (2) Targeted-Urovysion (cytokeratin7 immunophenotyping followed by Urovysion), both of which evaluated by both conventional and statistical evaluation scheme. For statistical evaluation cut-offs and sufficient sample size were determined on controls and ratio of positive cells was recorded, whereas conventional evaluation relied on manufacturer's recommendations. The specificity of cytology, Urovysion-alone in general and targeted-Urovysion in general appeared 86%, 86%, and 100%, respectively. In the same comparison, overall sensitivity was 60%, 80%, and 93%, respectively. In superficial cases sensitivity was 48% for cytology, 72% for Urovysion-alone and 91% for targeted-Urovysion, while no prominent differences were seen in muscle invasive cases. The ratio of FISH positive cells was proportionate with both stage and grade, however, targeted genotyping could separate high grade/high stage cases more effectively. In conclusion, CK7 targeting raises diagnostic efficiency of Urovysion, and could be an ideal tool for identifying tumor cells in ambiguous cases or when other tumors are present. Statistical evaluation produces accuracy comparable with results of conventional evaluation, and with laboratories setting cut-offs individually but according harmonized protocol, it could aid method standardization. Furthermore, by providing additional quantitative information about tumor characteristics, is likely to have therapy relevant value in the future. ' 2011 International Society for Advancement of Cytometry Key terms bladder cancer; targeted FISH; diagnostic cut-off value; CK7 immunophenotyping; automation BLADDER cancer is the seventh most common cancer in women and fourth most common in men (1). Urothelial carcinoma (UC) is the most frequent form, standing as the second most common cause of death in urinary tract malignancies (1). After transurethral resection (TUR) recurrence rate is 30-70%, and the chance of progression at 5 years is 10-30% (2). Monitoring and thorough follow-up of patients is indispensable and early detection of high grade tumors could have a great prognostic

It is possible to automate signal enumeration of Urovysion using a user-trainable system, and achieve efficiency comparable to manual analysis. Previously introduced automated immunophenotypic targeting should further increase diagnostic sensitivity, while resulting in a comprehensively automated method. However, the problem of reduced detection accuracy in cases featured with low polysomy is likely to remain a great challenge of automated signal enumeration.

No abstract

All of these findings support that the age (which seemed the only independent variable) shows strong correlation with the presence of the fatty infiltration but obesity may also play important role in the development and the extent of this change. Because of its frequency in elderly, at least partly, the fatty infiltration might be responsible for the xerostomia. We also think that presence of fatty infiltration should be mentioned in the histopathological report of salivary gland biopsies.