A urine test developed at Stanford Medicine identified 95 of 100 bladder cancers in a proof-of-concept study published in Nature Medicine on October 2, 2026. The assay, called uRARE-seq, reads RNA that tumor cells shed into urine rather than searching for DNA mutations. Science News describes the technology as still in its infancy, and Stanford oncologist Maximilian Diehn, a senior author, says much work remains before it reaches clinics.
AI, explained every day: The Deep View is the go-to daily newsletter for 600,000+ founders, knowledge workers and professionals who want to stay up to date with artificial intelligence.
We may earn a commission if you sign up.
uRARE-seq reads tumor RNA instead of tumor DNA
Urine is a natural place to look. Diehn told Science News that it is “a very attractive fluid” for bladder cancer because it sits in direct contact with tumors. The method builds on a 2025 Nature paper led by M.C. Nesselbush that described the RNA detection approach, which had earlier picked up lung cancer signatures in blood.
The team’s argument for RNA over DNA is flexibility. Senior author Ash Alizadeh said, in a summary of the work, that RNA looks for a biological signature of disease rather than an exact genetic match, which sidesteps the problem of needing a tumor’s precise mutation. Lead author Kevin J. Liu, a graduate student, worked with Diehn, Alizadeh and urologist Joseph Liao on the paper.
515 people, 683 samples and a 90 percent specificity
Science News reported a study population of 515 cancer patients and healthy people, with the test catching 95 of 100 bladder cancers, early-stage cases among them. Stanford’s own news release counts 683 urine samples from patients and healthy volunteers, a difference that most plausibly reflects some people giving more than one sample, though neither account spells it out.
The release gives the fuller accuracy picture: 95 percent sensitivity in people with localized bladder cancer and 90 percent specificity, meaning nine in ten people without the disease were correctly cleared. It says the test outperformed standard urine cytology and a DNA-based urine test, but it supplies no figures for those comparators and no direct comparison with cystoscopy. Science News did not report specificity or false-positive rates.
A sensitivity figure alone says little about how a test behaves in a clinic. With 90 percent specificity, a simple illustration is that about 100 of every 1,000 people without bladder cancer would still be flagged and sent for follow-up, which is why specificity in a larger and more varied group will matter as much as the 95 percent top-line number. The figure also comes from a mix of known patients and healthy volunteers rather than from people who walked in with blood in their urine.
Cystoscopy today and what a urine RNA test could add
Diagnosis currently leans on looking inside the bladder. The National Cancer Institute’s bladder cancer diagnosis page describes cystoscopy, in which a thin lighted tube is passed through the urethra and fluid fills the bladder so the wall can be checked, with a biopsy usually taken during the same procedure. Stanford’s release notes that cystoscopy can miss up to 30 percent of cases.
Surveillance after treatment is where the gap looks widest, since patients with bladder cancer are typically watched for recurrence for years. Liao said there are currently no biomarkers at all that are recommended or approved for surveillance. In the study, the RNA test also flagged residual disease after surgery and after BCG immunotherapy, and pre-treatment RNA profiles fed a biomarker that predicted with high accuracy whether a tumor would respond better to immunotherapy or chemotherapy, though the release gives no numeric accuracy for that biomarker. The NCI’s overview notes that most bladder cancers are non-muscle-invasive, the kind where repeated checks matter.
The treatment-response work may prove as consequential as detection. Patients who responded to BCG had pre-treatment urine RNA rich in T-cell and immune-signaling genes, while non-responders showed signs of rapidly dividing tumor cells, according to the Stanford account. The team used those differences to build the biomarker for choosing between immunotherapy and chemotherapy.
Company ties and the road to multicenter trials
The work has commercial links, which readers weighing the 95 percent claim should factor in. Liu and several coauthors hold ownership interests in Resero Bio, a San Carlos, California, Stanford spinout formed to commercialize the technology, and two coauthors are company employees. Diehn, Liao and Alizadeh have filed related patents and report consulting or advisory roles with diagnostics and pharmaceutical firms.
The next step is larger prospective, multicenter trials. Science News adds that the group is testing whether the approach can detect kidney and prostate cancers, which also contact urine, and quotes Diehn as unsure that RNA-based testing will prove useful for cancer screening in the general population. Whether a test tuned on patients with known disease holds up in people with no symptoms is the question the trials will have to answer.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
AI, explained every day: Get The Deep View →
We may earn a commission if you sign up.
More from Morning Overview
