Urine-Based MRD Test Tracks Response to Bladder Cancer Surgery

Urine tumor DNA analysis offers a non-invasive way to track disease, but its accuracy is limited by mutations that arise in normal bladder tissue as people age. These background mutations can mimic cancer signals and reduce test specificity. Researchers have now shown that accounting for these field-effect mutations significantly improves urine-based MRD detection.

In the new study, researchers at Stanford Cancer Institute (Stanford, CA, USA) built on their previously developed urine Cancer Personalized Profiling by Deep Sequencing (uCAPP-seq), a tumor-naïve method for detecting cancer-associated DNA in urine. By analyzing urine samples from individuals without bladder cancer, the researchers identified an age-related accumulation of somatic mutations in normal urothelial cells, a phenomenon they termed clonal cystopoiesis.

To address this, they developed a tumor-informed uCAPP-seq strategy using a statistical framework called RePhyNERX, designed to separate true tumor-derived mutations from age-related field-effect mutations. The researchers first examined urine samples from 138 non–bladder cancer controls and observed a sharp rise in oncogenic somatic mutations with age, reaching a prevalence of about 20 percent in individuals older than 75. These mutations, originating from cancer-adjacent but non-malignant tissue, were shown to confound conventional urine DNA analyses.

The refined, field-effect-aware uCAPP-seq approach was then applied to 261 longitudinal urine samples from 61 patients with intermediate- or high-risk non–muscle-invasive bladder cancer treated with surgery followed by Bacillus Calmette–Guérin (BCG) immunotherapy. Using the improved method, the researchers identified three distinct molecular response groups. One group showed complete molecular response after surgery alone, with no detectable urine tumor DNA. A second group had residual disease after surgery but responded to BCG therapy, while a third group showed persistent MRD after both treatments.

Importantly, the findings, published in Cell, revealed that predictors of response differed by treatment type. High tumor mutational burden and pre-existing immune activation predicted response to BCG but not to surgery. These insights suggest urine-based MRD testing could guide more personalized treatment decisions and help identify patients who may benefit from intensified or alternative therapies.

Related Links:
Stanford Cancer Institute