A new urine cell-free RNA (cfRNA) test for bladder cancer shows promise as a functional liquid biopsy to guide treatment decisions regarding the use of the first-line immunotherapy Bacillus Calmette-Guérin (BCG).

The new test distinguished which patients would benefit from intravesical BCG, and a trained model also predicted whether patients would be complete responders to the sometimes-rationed immunotherapy. The new test also outperformed cytology and DNA-based diagnostic tools, and it reliably detected minimal residual disease (MRD) in proof-of-concept testing.

Members of the Stanford University research team expressed optimism that a urine sample may ultimately reduce the need for cystoscopy throughout the disease course, which is sometimes done every 3 months and involves inserting a scope through the urethra under anesthesia.

They focused their preliminary work on noninvasive bladder cancer, using 683 urine samples from 515 individuals, including matched control participants. All participants enrolled via Stanford University or the Palo Alto VA in California. The findings were published online this month in Nature Medicine.

DNA-based liquid biopsies are already being used in bladder cancer detection and surveillance, but “when looking at RNA, we can also tell what the cancer is doing — what genes it’s currently expressing,” said senior author Max Diehn, MD, PhD, professor and vice chair of research in the Department of Radiation Oncology at Stanford University in Stanford, California. “That lets us pursue applications that weren’t really feasible with DNA methods, such as phenotyping the cancer” to differentiate low vs high grade and estimating risk for recurrence.

Bladder cancer diagnosis still largely depends on cystoscopy, although the field continues to search for more objective and less invasive tools.

“There’s no precision medicine without precision diagnostics, and there’s not a better, more easily accessible biofluid than urine,” said study co-author Joseph C. Liao, MD, professor and chair in urology at Stanford University School of Medicine, Palo Alto, California.

In a pretreatment sample of 93 patients known to have bladder cancer, the urine cfRNA test delivered a 96% sensitivity rate compared with 34% for urine cytology.

“Because we’re looking for molecules that come from cells, we don’t need to see a whole cell. So it makes sense in retrospect that we are able to develop a method that’s more sensitive than cytology,” Diehn said.

Using RNA overcomes a main challenge facing bladder cancer urine biopsies: the field effect. Because normal cells often carry mutations that are shed into the urine from the bladder lining, it can be difficult with DNA-based urine tests to differentiate cancer from normal cells that carry mutations never destined to become cancer. This was among the findings of an earlier study by the research team. The field effect can contribute to poor specificity, they wrote.

In another pretreatment cohort of 85 patients known to have bladder cancer, sensitivity for the urine cfRNA test was 99%, and DNA-based testing was 85%, the latter of which Diehn characterized as “also high.” When the researchers analyzed a small group of control participants (matched cases without cancer) known to have noncancer mutations, the specificity was also high — around 90% for the cfRNA test.

Precision Potential

The researchers also examined urine cfRNA capability of detecting MRD. The results were so compelling that they took a further step to see if they could predict BCG treatment response.

In a sample of 36 patients with intermediate- or high-risk non-muscle invasive bladder cancer, all of those without recurrence had detectable cfRNA presurgery, but 58% of them had no detection pre-BCG and none post-BCG. A subset of patients may have been cured by surgery, the authors suggested. High-grade recurrence cases always had detectable tumor cfRNA pre-BCG, and 92% had detections post-BCG.

After surgery, “patients who had undetectable bladder cancer RNA did extremely well — there were basically no recurrences in that subset. But [among] the patients who still had detectable bladder cancer RNA after the surgery, about half of those patients eventually developed recurrence,” Diehn said. “Then, when we looked after the BCG…the separation of those two groups got even bigger. Nearly everybody who did not have bladder cancer detected after BCG also never developed recurrence, but everybody who had a residual bladder cancer signal did develop recurrence. And so that’s sort of the proof of concept that this test could work for detecting minimal residual disease and how well treatments worked.”

The team then developed a biomarker to predict BCG response, which was particularly influenced by a T-cell immune signal among responders and by tumor cell proliferation in nonresponders.

“The investigators were able to extract remarkably rich biological information from urine cell-free RNA, a source that historically has been very difficult to analyze because the RNA is highly fragmented and present in small amounts,” said Dan Theodorescu, MD, PhD, director of the Cancer Center and professor of urology and cellular and molecular medicine at The University of Arizona in Tucson.

He said the BCG-responder observations impressed him the most: “That finding suggests the assay is measuring biologically relevant processes rather than simply detecting the presence of cancer cells.”

What’s Next

Large-scale prospective work is the next step, the Stanford team indicated.

Theodorescu said that work will need to offer further comparisons to the newest assays, “many of which claim similar performance characteristics as the assay described here for detection and monitoring.”

“The broader message is that urine is becoming a far more informative specimen than many of us appreciated even a few years ago,” Theodorescu said. “Traditionally, a urine test was largely viewed as a method for detecting blood, infection, or malignant cells. We are now entering a period where urine may provide a detailed molecular portrait of a patient’s tumor and its interaction with the immune system.”

This study was supported by the Stanford University and the National Institutes of Health.

Diehn and Liao reported having interest in the firm Resero Bio, which is commercializing the technology. Diehn and Liao reported filing patents related to urine cfRNA analysis and having consulting or advisory roles with multiple diagnostics and pharmaceutical companies. Theodorescu reported having no relevant financial disclosures.