Abstract
The emergence of genomic profiling has significantly deepened our understanding of urothelial carcinoma, signaling a promising shift toward more personalized, molecularly guided management strategies. While tissue-based analysis remains the clinical gold standard, it is limited by the need for sufficient tumor material, logistical sampling challenges, and clonal heterogeneity. 1 These limitations are compounded by the dynamic nature of tumor evolution, as shifting mutational landscapes make repeated biopsies both invasive and impractical in routine care.
This has led to growing interest in cell-free DNA (cfDNA), fragments of nucleic acids released into bodily fluids, primarily blood (circulating tumor DNA, ctDNA) and urine (urinary tumor DNA, utDNA). cfDNA carries genetic and epigenetic features that closely mirror the tumor of origin. 2 Its short half-life in circulation (16 min to 2.5 h), 3 coupled with its accessibility via blood or urine, makes it an ideal real-time, non-invasive biomarker. Recent advances in sequencing have markedly improved the sensitivity and specificity of cfDNA assays, enabling more widespread clinical application. 1
utDNA refers specifically to tumor-derived DNA shed into urine, either from the tumor surface or from exfoliated cells. It offers a unique window into real-time tumor biology. utDNA profiling captures both quantitative metrics – such as DNA concentration, variant allele frequencies, and methylation intensity – and qualitative features, including pathogenic mutations (e.g., FGFR3, TERT, TP53), gene-level copy number changes, and genome-wide aneuploidy.1,4 This rich molecular signature supports its use in diagnosis, prognosis, and monitoring of treatment response, especially in non–muscle-invasive bladder cancer (NMIBC), where invasive assessments are frequent and burdensome.
Patients with BCG-unresponsive NMIBC face limited effective bladder-sparing options, heightened treatment-related morbidity, and a persistent risk of progression. The recent FDA approval of novel intravesical and systemic therapies has created new opportunities to sequence treatments and delay or avoid radical cystectomy. However, this evolving landscape may also increase the risk of suboptimal treatment pathways, overlook progression of tumors, and missed windows for curative intervention. In this high-risk setting, the need for reliable biomarkers, particularly those that can predict treatment response, is increasingly urgent to optimize patient selection and preserve long-term oncologic outcomes.
The recent study by St-Laurent et al. (2025) 5 evaluated the clinical utility of utDNA analysis using the UroAmp assay, a next-generation sequencing-based platform developed by Convergent Genomics (South San Francisco, CA), in the SWOG S1605 trial. This was a single-arm, phase 2 trial investigating atezolizumab (a PD-L1 inhibitor) in patients with BCG-unresponsive high-risk NMIBC. 6 The trial demonstrated a 6-month complete response (CR) rate of 27% and a durable response in 56% of responders at 12 months. However, these results fell short of the predefined efficacy threshold, leaving pembrolizumab, approved based on the KEYNOTE-057 trial, as the only systemic agent currently FDA approved for BCG-unresponsive NMIBC.
The UroAmp assay
7
profiles urinary cfDNA across a 60-gene panel to detect:
Single nucleotide variants Small insertions and deletions Copy number alterations and copy-neutral loss of heterozygosity Microsatellite instability Genome-wide aneuploidy
The UroAmp minimal residual disease (MRD) report provides a binary result (positive/negative), a genomic disease burden score, and a predicted tumor grade. It is designed to support disease monitoring, recurrence detection, and therapeutic response assessment.
Of the 129 patients in the SWOG S1605 efficacy cohort, 98 had urine samples suitable for utDNA analysis. Baseline samples were collected from 89 patients, and 77 provided urine at the 3-month time point. Patients were stratified into two groups: those with CIS (n = 30 with CIS only; n = 26 with CIS + Ta/T1) and those with Ta/T1 disease only (n = 42).
The primary endpoint for CIS patients was complete response (CR) at 6 months, confirmed by mandatory biopsy, whereas event-free survival (EFS) at 18 months served as the outcome measure for patients with papillary lesions and overall study population.
So, what did they find?
Baseline utDNA positivity (after complete resection of all visible tumors) was observed in 61 of 89 patients (69%), including 73% of CIS cases and 62% of Ta/T1 cases. utDNA positivity, regardless of specific mutation, was strongly associated with inferior outcomes:
In CIS patients, the 6-month CR was 71% for utDNA-negative patients versus 13% for utDNA-positive patients (p < 0.001). Among Ta/T1 patients, the 18-month EFS was 71% (utDNA-negative) versus 43% (utDNA-positive). In the full cohort, the 18-month EFS was 51% in utDNA-negative patients versus 23% in utDNA-positive patients (HR, 2.8; 95% CI, 1.6–5.1; p < 0.001).
utDNA status at 3 months, after four of six atezolizumab cycles, was prognostic for subsequent event-free survival in patients without clinical evidence of recurrence.
In CIS patients who remained recurrence-free at 3 months, the 6-month CR was 47% (7/15) in utDNA-positive patients versus 100% (2/2) in utDNA-negative patients. In Ta/T1 patients, the 18-month EFS was 47% (utDNA-positive) versus 83% (utDNA-negative) patients. Across all patients, the EFS was 38% in utDNA-positive versus 86% in utDNA-negative (HR, 3.5; 95% CI, 1.3–9.1; p = 0.012).
Although the trial design precluded distinguishing predictive from purely prognostic value (since all patients received atezolizumab), the findings support utDNA as a prognostic biomarker of treatment failure and recurrence risk in patients with BCG-unresponsive NMIBC.
Similar findings have been reported in a cohort of patients with BCG-unresponsive NMIBC treated with nadofaragene firadenovec, in which 12-month recurrence-free survival (RFS) was 71% in utDNA-negative patients compared to 20% in utDNA-positive patients at baseline (p = 0.012). 8 Using 3-month urine samples, the RFS was 100% for utDNA-negative patients and 38% for utDNA-positive patients (p = 0.038), further supporting the prognostic value of utDNA testing in this setting.
Importantly, the supplementary analysis provides a compelling comparison of 3-month urine cytology and UroAmp performance metrics relative to clinical recurrence assessments at both 3 and 6 months.
At the 3-month evaluation, UroAmp demonstrated 100% sensitivity and negative predictive value (NPV), detecting all cases of high-grade urothelial carcinoma (HGUC) and reliably excluding recurrence in negative cases, although specificity was low (26%). In contrast, cytology showed high specificity (90%) but poor sensitivity (17%), failing to detect the majority of true positives.
When evaluating recurrence at 6 months, still based on UroAmp and cytology results obtained at 3 months, UroAmp maintained excellent sensitivity (97%) and a strong NPV (93%), again outperforming cytology, which continued to exhibit low sensitivity (18%) despite high specificity (94%). Notably, only 10 of 61 utDNA-positive cases were also cytology positive, highlighting the superior sensitivity of utDNA testing.
Previous studies have shown that utDNA assays, especially those using targeted mutation or methylation panels, often provide an “anticipatory diagnosis,” predicting recurrence before cystoscopic evidence appears, perhaps suggesting an explanation for their high false positive results. 9
These findings raise two important questions regarding the interplay between utDNA and cytology as prognostic biomarkers. First, is there added utility to urine cytology in utDNA-negative patients, given the latter's markedly high sensitivity and NPV? Second, in utDNA-positive cases, can cytology, with its high specificity, offer complementary value, or does it risk undermining the anticipatory advantage of utDNA by introducing delay or diagnostic uncertainty? As newer multi-omic biomarkers are increasingly utilized, we may gain a more comprehensive understanding of early recurrence patterns and refine our interpretation of apparent test discordances.
These diagnostic insights carry important implications for surveillance strategies. As highlighted by Ward and Bryan in the editorial accompanying this manuscript, 10 utDNA could also guide personalized surveillance. Patients who are both cystoscopically, and utDNA-negative (“double-negative”) may be safely monitored with reduced frequency. In contrast, molecular positivity, even in the absence of visible disease, may signal early recurrence, warranting closer follow-up. Given that frailty is common among the bladder cancer population, 11 de-intensifying surveillance protocols using noninvasive biomarkers such as utDNA has the potential to reduce the burden on patients and their caregivers, often without compromising safety. 12
Finally, an important step in clinical trials incorporating similar exploratory analyses is determining the optimal timing for utDNA testing. Baseline utDNA may help identify patients unlikely to benefit from further bladder-sparing therapy and prompt earlier consideration of radical cystectomy. Monitoring changes in utDNA over time, reflected in the genomic disease burden, may also provide a dynamic marker of treatment response or early recurrence, even in the absence of visible disease.
To conclude, this analysis supports utDNA as a noninvasive, biologically informative, and clinically actionable biomarker in BCG-unresponsive NMIBC. While further validation in randomized prospective trials is warranted, especially to define its predictive capabilities, utDNA has the potential to refine treatment selection, guide risk-adapted surveillance, and avoid delays in definitive therapy for those unlikely to respond.
As the field continues to advance toward individualized treatment paradigms, utDNA profiling represents a powerful tool, one that may finally surpass a long line of biomarkers that failed to gain clinical traction, leaving cytology as the sole standard for decades, and improve our understanding of tumor behavior and guiding more tailored management strategies.
Footnotes
Abbreviations
Author contributions
Both authors were involved in the conceptualization and writing of the manuscript
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Dr. Edward Messing is the Associate Editor for Paper Alerts for Bladder Cancer and did not participate in the editorial handling or peer review process of the article.