Skip to main content
Premium Trial:

Request an Annual Quote

Convergent Genomics Building Data on Urine-Based Assay to Diagnose, Monitor Urothelial Carcinoma


NEW YORK – Liquid biopsy company Convergent Genomics has developed a urine-based test to diagnose and monitor recurrence and help select therapies for urothelial cancer patients, and is currently building data to support eventual regulatory approval and insurance reimbursement.

The Oregon Health and Science University spinout also believes the test could help predict future urothelial carcinoma and presented data last month at the American Urological Association's annual meeting that showed its UroAmp test was able to predict urothelial carcinoma in 90 percent of future cancers that developed within five years. Urothelial carcinoma is the most common form of bladder cancer and accounts for about 90 percent of bladder cancer cases.

The firm's liquid biopsy test, currently available through its CLIA-certified lab in South San Francisco, California, is intended to support initial diagnosis, recurrence monitoring, risk stratification, and therapy selection by analyzing mutations in cell-associated and cell-free DNA across 60 genes associated with urothelial carcinoma. The test leverages publicly available data, such as that from the Cancer Genome Atlas, but the company hopes to help define and expand the available data, said CEO Trevor Levin. 

Thus far, it has "not been understood … what mutations are acquired in our genome as we age," he said. "We've been defining this atlas of what it means to move from a healthy young person to somebody who has cancer." 

Within the 60 genes it analyzes for a variety of indications, the firm has homed in on a subset of 10 genes that are predictive for the development of future cancer using samples from the 50,000-patient Golestan Cohort Study, led by Tehran University of Medical Sciences, the National Cancer Institute, and the International Agency for Research on Cancer. All patients entered the study healthy and were followed over the course of 10 years, and Convergent analyzed a subset of bladder cancer patients and control subjects to "identify these areas of the genome that become mutated up to more than a decade ahead of clinical diagnosis of cancer," Levin said.

According to the study's abstract, which was also presented at the 2023 European Association of Urology Annual Congress in March, a training cohort of 140 controls and 96 tumors was used to develop a screening model for preclinical urothelial carcinoma. Validation of the model was performed in two separate studies: one with 96 controls and 70 urothelial carcinoma cases from multiple institutions, and one from the Golestan cohort with 29 asymptomatic patients who later developed primary urothelial carcinoma and 98 matched controls. 

The screening model was trained to a sensitivity of 88 percent and specificity of 94 percent, and the first validation showed the model had a sensitivity of 71 percent and specificity of 94 percent. In the Golestan cohort, the model had a sensitivity of 66 percent and specificity of 94 percent, although when limited to patients who developed cancer within five years, sensitivity rose to 90 percent.

After its founding in 2015, Convergent was selected by Illumina Accelerator to be incubated within Illumina and spend six months refining its early testing process. The firm spent multiple years developing the optimal urine sample preservation buffer to keep DNA and RNA stable at ambient temperatures for up to two weeks, Levin noted, and also developed its own customized proprietary processes to extract nucleic acids. Part of the reason for the lack of urine liquid biopsy tests compared to blood-based tests it that there "are not well-validated and robust sample collection preservation solutions, and there are not very robust and automatable, robotics-friendly, large-volume urine extraction systems out there," he said.

The company uses Illumina sequencers and has developed custom machine learning algorithms to identify mutations and classify disease and risk into categories such as tumor-positive, tumor-negative, high risk of recurrence, or low risk of recurrence, Levin said. 

If a patient is high risk or tumor-positive, a third layer of algorithms look at molecular risk factors associated with prognosis, such as pathologic stage and cancer grade, he added. Right now, the company runs the UroAmp test in its CLIA lab, and has processed more than 4,000 clinical specimens, including both diagnostic and research samples, and usually returns results in a couple of days. Convergent is working with the Centers for Medicare and Medicaid Services to garner insurance coverage and is expecting the test to become more widely available in the fourth quarter of this year, he said.

Levin also said that the company will eventually develop kits and support technology transfers allowing the test to be performed by other laboratories.

To that end, the company is building data to support a submission to the US Food and Drug Administration for genomic profiling, surveillance, and minimal residual disease testing. Company researchers and collaborators at OHSU published a study in the Journal of Clinical Medicine last year validating the test methodology with 423 urine samples and the somatic variant quantification with 91 tumor and urine samples. They found that the test was able to detect mutations with 97 percent sensitivity and 80 percent positive predictive value, and that tumor single-nucleotide variants were found in the urine samples with median concordance of nearly 92 percent.

Convergent is also running its prospective INSITE study of 2,000 patients who provided urine samples at the time of their initial diagnosis of bladder cancer. Those patients have provided additional samples and clinical data over the course of about three years, and the company is helping risk-stratify the patients based on their genomic profiles, Levin said. The data hasn't been analyzed, but the firm is further building up its biobank of samples to develop its eventual FDA validation study. The test has also been validated in risk stratification of patients with hematuria suspected of urothelial carcinoma, he said.

The company has multiple ongoing studies for test applications such as recurrence monitoring and therapy selection. Convergent is collaborating with the Southwest Oncology Group to test patients undergoing PD-L1 immunotherapies, collecting urine samples from patients before treatment and at multiple time points during treatment. "With the depth of measurements that we're making, we have unique fingerprints of every single cancer that can be correlated with responders and non-responders to the drug," he said.

Convergent is also studying patients undergoing Bacillus Calmette-Guerin induction therapy with Loyola University and has a 600-patient validation of its disease classification algorithms with Massachusetts General Hospital that is under peer review.

Levin noted that the data presented at last month's AUA meeting showcases the future potential of UroAmp as a predictive tool, but that prediction is not the immediate focus. Getting to a point where cancer can be predicted years in advance will require significant additional work, as it would be a "paradigm shift," he said. 

"There's a lot of exciting potential there, [but] there's also lots of important studies and further validation work to be done in this space."