Predictive biomarkers and signatures in urinary bladder cancer

There is a strong demand for reliable and cost-efficient tumor markers that are suitable for disease screening, prognostics and surveillance. Markers should fulfill certain criteria, such as high sensitivity and specificity and non- or low-invasive sampling techniques. Bioanalytical techniques allow the quantification of differences in expression profiles between tumor and healthy tissue at the transcript and protein level; deduced expression signatures and classifiers can then be used for differentiation of tumor types, tumor stages and tumor aggressivity. However, markers and signatures with sensitivities and specificities close to 100% are not yet available. Strategies that combine gene expression profiling in bladder tumor tissue and protein profiling in urine and serum are promising. Because signatures and classifiers identified in different laboratories can differ widely in their components, a strategy of harmonization and cross-validation is required. In this review, developments in tumor-marker identification with a focus on bladder cancer are described.