Bladder cancer biomarker test accurately adjudicates atypical cytology

February 5, 2020

“The results of our retrospective study support taking advantage of the Cxbladder test to identify patients who should be further evaluated for cancer and to spare those who likely do not have cancer from an unnecessary workup," says Badrinath Konety, MD, MBA.

Findings from a retrospective analysis pooling data from four studies investigating the performance of a urinary bladder cancer biomarker test (Cxbladder) demonstrate that it accurately adjudicated atypical cytology, including in cases that also had equivocal cystoscopy.

The study also reaffirms the biomarker test’s performance for accurately ruling out bladder cancer in both patients undergoing evaluation for hematuria and those being monitored for recurrence of bladder cancer.

The research was summarized in a paper published in European Urology (2019; 76:238-43). “Previous studies have only evaluated the accuracy of Cxbladder for ruling out bladder cancer, but they did not analyze its value for atypia adjudication,” lead author Badrinath Konety, MD, MBA, told Urology Times. “Up to 20% to 25% of urine cytology samples may be classified as atypical by local cytology, and the AUA/SUO joint guideline for diagnosis and treatment of nonmuscle-invasive bladder cancer suggests it is reasonable to use a second test to adjudicate equivocal cytology.

Read - Bladder Ca: Genetic testing, targeted therapies signal a new era of precision medicine

“The results of our retrospective study support taking advantage of the Cxbladder test to identify patients who should be further evaluated for cancer and to spare those who likely do not have cancer from an unnecessary workup,” added Dr. Konety, professor of urology and director of the Institute for Prostate and Urologic Cancers, University of Minnesota Medical School, Minneapolis.

Cxbladder is a multiplex mRNA test that measures concentrations of five genes in unfractionated urine. There are three versions of the test that are used in different clinical settings and that incorporate different clinical information in the prediction algorithm that determines the test result.

The retrospective study pooled data from 1,784 patients who participated in three prospective clinical trials and one real-world clinical study investigating the diagnostic performance of the Cxbladder tests. Ultimately, data were analyzed for 852 samples from 775 patients for which there were results from both the urine biomarker test and local cytology.

Of the 852 samples, 436 were from patients being evaluated for hematuria and 416 were from patients previously diagnosed with urothelial cancer (UC). A subgroup of 153 samples from 146 samples had atypical cytology, of which 14 had both atypical cytology and equivocal cystoscopy results.

Next: Slightly higher negative predictive value seenSlightly higher negative predictive value seen

Analyses comparing the diagnostic performance of Cxbladder with cytology showed that the urinary biomarker test had a slightly higher negative predictive value, 97.4% versus 92.6%. More noteworthy, however, was the finding that Cxbladder had a much lower false-negative rate than cytology, missing only 8.5% of pathology-confirmed tumors compared with 63% that were missed by cytology.

Also see: Agent for high-grade NMIBC meets primary endpoint in phase III trial

In the subgroup of 153 samples with atypical cytology, 26 (17%) were confirmed by pathology as positive for UC. Cxbladder correctly identified all 26 tumors, including the two found in the 14 patients who had atypical cytology and equivocal cystoscopy.

“The latter two cases were high-risk tumors and included a high-grade ≥T1 UC and one carcinoma in situ,” said Dr. Konety. “Further study is warranted to investigate the value of Cxbladder in cases where cytology is atypical and cystoscopy is indeterminate.”

Dr. Konety conducts research for Pacific Edge, several of his co-authors are researchers/consultants for Pacific Edge, and one co-author is an employee of the company.

download issueDownload Issue : Vol 48 No 2