Abstract Background Apparent clinical complete response after neoadjuvant or multidisciplinary therapy in non-metastatic muscle-invasive bladder cancer has renewed interest in omitting or deferring radical cystectomy for selected patients. The central uncertainty is whether favorable endoscopic, cytologic, pathological sampling, and imaging findings can exclude both residual intravesical disease and occult systemic minimal residual disease with sufficient confidence to justify de-escalation. Main body We performed a decision-focused structured narrative review, updated through 5 June 2026, and an aggregate-data Bayesian calibration analysis. The review was designed to examine the post-treatment pre-omission decision point, not to establish a validated clinical pathway. Recent systematic evidence shows that clinical complete response definitions are heterogeneous and that concordance between clinical and pathological complete response is moderate. Using the pooled cCR-pCR concordance estimate as a pretest anchor, the residual probability of pathological disease after apparent cCR is approximately 49%. A negative molecular restaging layer would therefore need a negative likelihood ratio of about 0.12 to reduce residual disease probability below 10%, about 0.18 to reduce it below 15%, and about 0.26 to reduce it below 20%. Organ-preservation cohorts, trimodality therapy experience, and emerging immune-checkpoint-inhibitor strategies show feasibility, but they do not remove the need for stringent local reassessment, histology-aware selection, advanced imaging where appropriate, rapid salvage pathways, and prospective validation. Plasma circulating tumor DNA and urine tumor DNA should be interpreted as complementary, non-interchangeable readouts: plasma is more closely aligned with systemic molecular residual disease, whereas urine more directly samples intravesical tumor shedding. Conclusions Clinical complete response should be treated as a residual-risk state rather than proof of cure. Dual-negative plasma and urine tumor-DNA results may provide the strongest biological support for surveillance, but only within expert multidisciplinary programs or prospective protocols. Urine-positive-only, plasma-positive-only, dual-positive, and indeterminate states require prespecified actions before dual-compartment restaging is used routinely.
Drobot et al. (Wed,) studied this question.