Answers are grounded in this paper and cite their sources.
We read with great interest the recent article by Rubio-Briones et al. [1], which presents the largest multicentre evaluation of the Uromonitor®-version 2 (v2; U-monitor, Porto, Portugal) assay for molecular surveillance in non-muscle-invasive bladder cancer (NMIBC). The authors reported a sensitivity of 36% and a specificity of 93%, highlighting the ongoing challenge of implementing highly specific urinary biomarkers in routine clinical practice. Their robust real-world investigation prompted us to re-examine the cumulative diagnostic performance of Uromonitor across all published datasets to date. The only previously aggregated evidence, reported by Kravchuk et al. [2] in 2024, encompassed four studies with 1190 tests and yielded pooled estimates of 80% sensitivity and 95% specificity. Since then, four further independent cohorts have been published, expanding the evidence base to eight studies with 3196 evaluable test results [1, 3-5]. To provide an updated synthesis, we aggregated patient-level data from all available prospective and retrospective series employing this validated three-gene mutation panel (fibroblast growth factor receptor 3 [FGFR3], telomerase reverse transcriptase [TERT], Kirsten rat sarcoma viral oncogene homologue [KRAS]). Across these cohorts, histologically-confirmed urothelial carcinoma of the bladder (UCB) was present in 833 of 3196 tested patients (26.1%). The pooled diagnostic indices were as follows: sensitivity 0.55 (95% CI 0.52–0.58), specificity 0.95 (95% CI 0.94–0.96), positive predictive value (PPV) 0.79 (95% CI 0.76–0.82), negative predictive value 0.86 (95% CI 0.84–0.87), and overall accuracy 0.85 (95% CI 0.83–0.86). These results reaffirm the assay's hallmark specificity and predictive precision for confirming histologically-verified tumour presence, while underscoring that a negative outcome should be interpreted with clinical discretion during endoscopic surveillance. Advanced diagnostic metrics further delineate the test's discriminative capacity, including a Matthews correlation coefficient of 0.572 and an area under the receiver operating characteristic curve of 0.750 (95% CI 0.728–0.772). The likelihood ratios (LR+ = 10.6; LR− = 0.47) provide a clinically transparent perspective: a positive Uromonitor-v2 result increases the post-test probability of a histologically-confirmed UCB to ~79%, whereas a negative result reduces it to ~14%. Collectively, these findings demonstrate strong rule-in capability and acceptable rule-out potential under real-world surveillance conditions. Our synthesis extends and refines the conclusions of Kravchuk et al. [2], whose earlier meta-analysis is now supported by an evidence base nearly three times larger. The Uromonitor assay consistently maintains high specificity and PPV across diverse international cohorts, although sensitivity remains variable among studies. The observed heterogeneity, including that reported by Rubio-Briones et al. [1], appears to be driven primarily by pre-analytical variables such as urine handling, centrifugation, and DNA yield rather than by the assay design itself. Our group is currently conducting a multicentre validation of an optimised Uromonitor platform that retains the same three molecular targets but incorporates an enhanced urine-filtration system to improve DNA extraction efficiency. We anticipate that this refinement will further increase analytical sensitivity while preserving the excellent specificity already demonstrated. We commend Rubio-Briones et al. [1] for their pivotal contribution to the clinical validation of urinary molecular diagnostics in NMIBC. We hope that our updated synthesis will encourage further methodological harmonisation and foster continued confidence in molecular urine testing as a valuable adjunct to cystoscopy in NMIBC surveillance. The authors declare no conflict of interest related to this response. No funding was received for this study. This article is linked to https://doi.org/10.1111/bju.70114.
No takes yet. Share an insight, caveat, or question.
May et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: