From Blood to Urine: A Borrowed Blueprint
The technology at the center of this story was not built for urine at all. It was built for blood. A team led by Maximilian Diehn, an oncologist at Stanford University, first developed a method to read tiny fragments of RNA floating in blood plasma — RNA being DNA’s molecular cousin, a short-lived messenger that carries working instructions rather than the permanent archive. The method was named RARE-seq: random priming and affinity capture of cell-free RNA fragments for enrichment analysis by sequencing. In plain terms, it grabs stray RNA out of a fluid, copies it, and reads it. The blood version was published in Nature in 2025, and it proved roughly 50 times more sensitive at detecting tumor-derived RNA than standard whole-transcriptome sequencing, with a detection floor of 0.05 percent. [1] That first application was lung cancer. In 437 plasma samples from 369 people — patients with cancer, people with non-malignant conditions, and controls — the test picked up non-small-cell lung cancer expression signatures at rates that climbed with disease stage: 30 percent in stage I, 63 percent in stage II, 67 percent in stage III, and 83 percent in stage IV, all at 95 percent specificity. The team also detected resistance mechanisms in patients with EGFR-mutant lung cancer who had stopped responding to tyrosine kinase inhibitors.
The shift to urine is not a cosmetic change of fluid. It is a change of physics and biology. Blood is a crowded highway — RNA from every tissue passes through, and the signal must be fished out of noise. Urine, by contrast, sits in direct contact with bladder tumors. Diehn calls it “a very attractive fluid” for exactly this reason. [2] The new method, called uRARE-seq, keeps the same random-priming and capture logic but tunes it for this closer, more intimate sample.
The Numbers That Reframe the Field
The urine test was applied to 683 urine samples from patients with cancer and controls. It reached 95 percent sensitivity at 90 percent specificity for detecting localized bladder cancer, and it outperformed urine tumor DNA analysis, the DNA-based approach that has dominated the liquid-biopsy conversation. [1].
For years the field’s default assumption has been that DNA is the reliable molecule — stable, durable, easy to amplify. RNA was treated as the fragile cousin, too quick to degrade to be trusted. This study inverts that hierarchy for a specific job. Diehn’s framing is that RNA “can tell you what the cancer is doing at any given moment.” [2] The test also sidestepped a known trap: it was unaffected by the presence of field-effect mutations. [1].
Beyond detection, the test returned biology. It offered hints about tumor type and risk of recurrence. It sensitively detected minimal residual disease — cancer left behind after treatment. And it distinguished complete molecular responses after surgery from responses after intravesical BCG, the bacteria-based therapy instilled directly into the bladder.
What the Urine Reveals Before Treatment Begins
The most striking result sits in samples taken before therapy started. In pretreatment urine from 114 patients, the RNA profiles split into two camps. Complete responders to BCG showed enrichment for T cell and other immune signatures — evidence of a preexisting antitumor immune response already at work. Nonresponders showed higher expression of proliferation-related genes.
The biomarker predicted the likelihood of response to BCG versus chemotherapy with an area under the curve of 0.93, and it was strongly associated with risk of recurrence. [1].
The researchers are now exploring whether the technique can pick up other cancers. Their evidence suggests it can spot cancer in the kidney and prostate — organs also washed by urine. Diehn imagines a multi-cancer urine test focused on tumors of the genitourinary tract. He is equally clear about the distance to the clinic: prospective studies are still needed to assess the test’s clinical utility, and because the technology is so new, the jury is still out on whether RNA-based testing will find a place in cancer screening. “But absolutely, I think the field needs to investigate it.” [2]