NMIBC 5-year survival rate is about 90%, yet its 3-year recurrence rate is about 70%, and up to 20% progress to MIBC, which shows a dramatic decline in 5-year survival rate. International guidelines have stratified patients by their risk of recurrence and progression based on clinical and pathologic variables, yet issues such as heterogeneous risk groups, selection of appropriate adjuvant intravesical therapies, and decision-making on when to pursue radical surgery remain suboptimal. MIBC prognosis depends on the stage of the disease at the time of treatment. Neoadjuvant chemotherapy (nCHT) followed by radical cystectomy (RC) plus urinary diversion is the standard of care for cisplatin-fit MIBC patients, yet this strategy itself represents a challenge because of the proper selection of patients most likely to benefit, poor tolerability of platinum chemotherapies by a significant subset of patients, unknown response in variant histology, and the morbidity associated with RC. In mBC, early detection of metastasis, selection of systemic treatment regimens, and identification of patients at high risk of recurrence remain challenging.
In this context, the implementation of personalized medicine, which aims to tailor the right therapeutic strategy at an individual level (e.g., according to a patient’s genetic profiling, lifestyle data, and disease characteristics) for optimal outcomes with fewer toxicities and improved QoL, would be pivotal for BC optimized management. Artificial intelligence (AI), namely a machine’s ability that mimics human intelligence to perform tasks involving decision-making and problem-solving, has recently revolutionized the practice of medicine.2-4 By its emerging applications, mainly consisting of machine learning (ML) and deep learning (DL) methods, it holds potential in reshaping clinical practice, also in the field of BC.5
In this mini-review, we briefly summarize the role of AI applied at different steps of the current BC work-up (i.e., detection, grading, staging, risk stratification, treatment, and outcome prediction), which can all potentially promote a more individualized treatment for BC patients.6 To date, relevant studies exploring the application of ML and DL approaches have been carried out, and several AI tools have been shown to optimize the speed and accuracy of BC detection, provide more precise risk stratification and treatment strategies, and offer more powerful recommendations related to prognosis.6 However, several major obstacles still impede the broad and successful integration of AI in BC clinical practice. They include variability and quality of training datasets, different data sources, algorithm bias, and limited interpretability of results. Methodological issues, such as the lack of external validation sets, the use of the same dataset for both training and testing, as well as scalability, standardization, and the rapid evolution of treatment strategies, coupled with incomplete or outdated clinical data, further hinder the relevance and reliability of AI applications in this setting. Finally, ethical concerns (i.e., privacy, data security, transparency, accountability, and preservation of human autonomy) and economic implications must also be considered. Solving these challenges might mark a step closer to truly personalized BC therapy over the decade to come.
Written by: Martina Maggi, MD, PhD,1 Francesco Chierigo, MD,1 Giuseppe Fallara, MD,1 Letizia Maria Ippolita Jannello, MD,1 Marco Tozzi, MD,1 Francesco Pellegrino, MD,1 Felice Crocetto, MD, PhD,2 Daniela Terracciano MD,3 Roberto Bianchi, MD,1 Matteo Ferro, MD, PhD1
- Unit of Urology, Department of Health Science, ASST Santi Paolo and Carlo, University of Milan, Milan, Italy.
- Department of Neuroscience, Reproductive and Odontostomatological Sciences, University of Naples "Federico II", Naples, Italy.
- Department of Translational Medical Sciences, University of Naples "Federico II", Naples, Italy.
- Gontero P, Birtle A, Capoun O, et al. European Association of Urology Guidelines on Non-muscle-invasive Bladder Cancer (TaT1 and Carcinoma In Situ)-A Summary of the 2024 Guidelines Update. Eur Urol. 2024 Dec;86(6):531-549. doi: 10.1016/j.eururo.2024.07.027. Epub 2024 Aug 17. PMID: 39155194.
- Ferro M, Tataru OS, Carrieri G, et al. Artificial intelligence and radiomics applications in adrenal lesions: a systematic review. Ther Adv Urol. 2025 Aug 2;17:17562872251352553. doi: 10.1177/17562872251352553. PMID: 40761430; PMCID: PMC12319203.
- Ferro M, Crocetto F, Jannello LMI, et al. Emerging biomarkers in bladder cancer: Translating molecular advances into precision oncology. Crit Rev Oncol Hematol. 2025 Jul 23;215:104862. doi: 10.1016/j.critrevonc.2025.104862. Epub ahead of print. PMID: 40712901.
- Ferro M, Crocetto F, Barone B, et al. Artificial intelligence and radiomics in evaluation of kidney lesions: a comprehensive literature review. Ther Adv Urol. 2023 Apr 17;15:17562872231164803. doi: 10.1177/17562872231164803. PMID: 37113657; PMCID: PMC10126666.
- Ferro M, Falagario UG, Barone B, et al. Artificial Intelligence in the Advanced Diagnosis of Bladder Cancer-Comprehensive Literature Review and Future Advancement. Diagnostics (Basel). 2023 Jul 7;13(13):2308. doi: 10.3390/diagnostics13132308.
- Maggi M, Chierigo F, Fallara G, et al. Shaping the Future of Personalized Therapy in Bladder Cancer Using Artificial Intelligence. Eur Urol Focus. 2025 Aug 1:S2405-4569(25)00216-0. doi: 10.1016/j.euf.2025.07.011. Epub ahead of print. PMID: 40753031.