A Multi-Modal Approach for Decision Making in Bladder Cancer - Beyond the Abstract
In our Review, we synthesise evidence supporting a multimodal approach to decision-making in bladder cancer, highlighting how advances in artificial intelligence (AI), enhanced cystoscopic imaging, urinary biomarkers, multiparametric MRI (mpMRI), radiomics, and genomics can be integrated to improve diagnostic accuracy, risk stratification, and treatment selection. The clinical value of these technologies lies not in isolated adoption, but in coordinated use across the patient pathway.
At initial detection, enhanced cystoscopic modalities such as blue-light cystoscopy and narrow-band imaging improve sensitivity for carcinoma in situ and small papillary tumours, although uptake remains variable owing to cost and logistical considerations. AI-assisted white-light cystoscopy offers a complementary strategy, enabling real-time lesion detection and classification while reducing interobserver variability. Combined with structured video documentation and spatial mapping, these tools may transform cystoscopy into a more reproducible and data-driven diagnostic modality.
Non-invasive urinary biomarkers and cytology represent an important opportunity for early risk stratification and optimisation of surveillance. While no single assay can currently replace cystoscopy, multiplex transcriptomic and DNA-based tests, particularly when supported by AI-driven analysis, show promise in identifying patients at low risk of recurrence who may safely avoid frequent invasive procedures. Such approaches could reduce patient burden and health-care costs without compromising oncological outcomes.
Accurate staging remains central to treatment selection, particularly in distinguishing non-muscle-invasive from muscle-invasive disease. Multiparametric MRI, standardised through the VI-RADS framework, improves pre-operative assessment of muscle invasion and treatment planning. Notably, mpMRI is already embedded in prostate cancer guidelines, where its routine use has reshaped diagnostic pathways, potentially facilitating wider acceptance in bladder cancer. Radiomics and AI-based imaging models further enhance staging by extracting quantitative features beyond visual assessment, while AI-driven computational pathology is beginning to improve the reproducibility of tumour grading and staging.
Molecular profiling has revealed biologically distinct bladder cancer subtypes with important prognostic and therapeutic implications. Integrating genomic data with imaging and pathology may enable a shift from traditional clinicopathological risk groups towards biologically informed, patient-specific decision-making.
However, several barriers to clinical implementation remain. These include limited external validation, lack of standardisation, regulatory and cost-effectiveness challenges, and difficulties integrating new technologies into established workflows. Importantly, as bladder cancer care enters an era of biomarker-driven and precision treatment strategies, there is a risk that individual patient context may be overshadowed. Frailty, functional status, comorbidities, patient values, and tolerance for risk remain fundamental to truly personalised care and cannot be fully captured by algorithms or biomarkers alone. Multimodal frameworks must therefore support holistic clinical judgement and shared decision-making, rather than replace them.
In our Review, we propose integrated multimodal clinical pathways that apply emerging technologies at each stage of the disease course, from risk-adapted detection and staging to personalised treatment selection across non-muscle-invasive, muscle-invasive, and metastatic disease. Ultimately, the future of bladder cancer management lies in combining technological innovation with patient-centred care to improve outcomes while minimising unnecessary intervention.
Written by: Hasan Al-Sattar, School of Medicine and Biomedical Sciences, University of Oxford, Oxford, UK
Read the Abstract