Prognostic biomarkers in non‐muscle‐invasive bladder cancer: biological basis and evidence

Objectives To review prognostic and treatment‐response biomarkers for non‐muscle‐invasive bladder cancer (NMIBC), with particular attention to biological function, specimen type, assay method, statistical evidence, and possible clinical use. Methods We conducted a focused narrative search of PubMed, Web of Science, and the Cochrane Library for English‐language human studies published from January 2015 to August 2026, supplemented by earlier pivotal studies and reference‐list screening. Priority was given to NMIBC cohorts reporting recurrence, progression, survival, or response to intravesical therapy. We considered cohort specificity, sample size, multivariable analysis, testing in an independent population, and assay standardisation. Results Tumour‐tissue markers such as stromal antigen 2 (STAG2), fibroblast growth factor receptor 3 (FGFR3), fibroblast activation protein‐α, the urokinase plasminogen activator system, and cadherins are linked to chromosomal stability, growth‐factor signalling, stromal remodelling, extracellular‐matrix invasion, and epithelial–mesenchymal transition. Urinary RNAs and plasma circulating tumour DNA permit non‐invasive sampling, although results depend on specimen handling and assay conditions. Blood‐derived inflammatory indices are inexpensive but non‐specific. Most evidence comes from retrospective single‐cohort studies; reported effects and cut‐offs are not always consistent, and few markers materially improve clinicopathological models. Conclusions None of the biomarkers reviewed can currently replace guideline‐based risk stratification or cystoscopic surveillance. Future studies should define assays and thresholds prospectively, test them in independent cohorts, and show that adding a biomarker improves prediction and patient management at an acceptable cost.