Could it guide treatment?

After surgery, many bladder cancer patients receive bacillus Calmette-Guérin, or BCG, a weakened bacterial strain originally developed as a tuberculosis vaccine. Doctors instill it into the bladder through a catheter once a week for six weeks, sometimes continuing as maintenance for up to three years. BCG rallies the patient's immune system against any lingering cancer cells.

BCG works only when a tumor already carries the immune activity needed to mount an attack. Until now, no reliable biomarker existed to predict who would benefit.

The Stanford team found that patients who later responded to BCG had pre-treatment urine RNA rich in genes tied to T-cell and immune-signaling activity. Non-responders instead showed RNA signatures of rapidly dividing tumor cells. From those patterns, the researchers built a biomarker that flags tumors more likely to respond to immunotherapy versus chemotherapy, the standard second-line therapy.

Reserving BCG for likely responders could ease a chronic global shortage, since a single manufacturer supplies most of the world.

What are the limitations?

The accuracy figures come from a single-center cohort. The authors plan prospective, multi-center trials to confirm both the diagnostic performance and the treatment predictions. Results in any sub-group, including treatment response, will require validation before clinical adoption.

Researchers from Resero Bio, a Stanford spinoff formed to commercialize the technology, contributed to the study. Lead author Kevin J. Liu and several co-authors hold ownership stakes in the company, and two co-authors are Resero employees. Senior authors Diehn, Joseph Liao, and Ash Alizadeh, along with others, have filed patents on urine cell-free RNA methods and report consulting or advisory roles with diagnostics and pharmaceutical firms.

Why switch from DNA to RNA?

Healthy lining cells in the bladder often accumulate their own mutations that overlap with those in a patient's tumor. A DNA test can mistake these bystander mutations, called the field effect, for cancer. Correcting for them typically requires first sequencing both the tumor and a normal tissue sample so analysts know which mutations to track and which to ignore.

"For DNA-based, tumor-informed approaches, you need to know exactly what the tumor's mutations look like to track disease," said Ash Alizadeh, M.D., Ph.D., professor of medicine at Stanford. "RNA doesn't have that problem. You're looking for a biological signature of disease, not an exact likeness, and urine is a noninvasive way to accurately detect and track bladder cancer."

What's next?

The team plans larger prospective trials at multiple centers, with the goal of pairing each patient with the therapy most likely to help.

"There are currently no biomarkers at all that are recommended or approved for surveillance. These treatments are incredibly taxing—physically and financially," said Joseph Liao, M.D., chair of urology at Stanford. "If we can guide those choices, we have the potential to help both patients and clinicians."