BERLIN (GenomeWeb) – Bladder cancer has a high survival rate when detected early and removed by surgery but there is no good treatment, and poor survival, once it has become invasive.
In addition, current molecular tests for bladder cancer do not perform well, so the diagnostic gold standard remains cystoscopy, where a thin, lighted tube with a lens called a cytoscope is inserted into the urethra and bladder.
Researchers at the Danish Cancer Society Research Center in Copenhagen have recently developed a urine-based DNA test to detect biomarkers of bladder cancer, and similar tests could potentially be used to assess prostate, kidney, and other types of cancer. At the Association for Molecular Pathology's AMP Global Congress in Berlin this week, Per Guldberg, head of the research center's cancer genetics group, presented data from a study that investigated whether the urine DNA test could detect signs of bladder cancer in individuals before they developed symptoms of the disease.
For their retrospective study, they took advantage of the Diet, Cancer and Health (DCH) cohort that the Danish Cancer Society assembled in the 1990s, signing up 57,000 individuals age 50 to 64 with no prior cancer diagnosis. Participants provided a variety of samples, including blood, urine, a fat tissue biopsy, and toe nail clippings, as well as long-term clinical follow-up data. Since enrollment, there have been 12,000 cancer incidents in the cohort, including 606 cases of bladder cancer.
Patients developed bladder cancer between three months and 15 years after enrolling, with a median time of 10 years, and Guldberg and his colleagues analyzed their baseline urine samples, which had been stored frozen.
The urine DNA test uses a filtration device, described in a publication in PLOS One in 2015, to separate tumor from non-tumor cells, which are smaller in size. And last year, Guldberg's team published a prospective blinded evaluation study of the test, which uses droplet digital PCR and MethyLight to look for eight mutations in the TERT and FGFR3 genes as well as DNA methylation markers in six other genes.
For the evaluation study, they analyzed 475 patients who had undergone flexible cystoscopy because of hematuria, or blood in the urine, and 99 were diagnosed with bladder cancer. The urine DNA test was able to detect the cancers with 97 percent sensitivity and 77 percent specificity – which Guldberg said is better than that of cystoscopy – and had a positive predictive value of 53 percent, suggesting that it could be used as a first-line test to determine which patients can forego cystoscopy.
For their study of the DCH cohort, the researchers looked for mutations in the TERT and FGFR3 genes in those who later developed bladder cancer and found that they were able to detect FGFR3 mutations in patients as far as 15 years before they were diagnosed.
Seventy percent of patients who were diagnosed within a year of enrolling had urine DNA mutations in either gene, as did 10 percent of those diagnosed after 15 years. "We knew that bladder cancer has a long latency time, but we did not expect mutations 15 years before diagnosis," Guldberg said.
As a control they also tested the urine of patients who were referred for other bladder conditions than cancer and did not detect any mutations in them. They also assayed several hundred participants without any bladder cancer so far and found mutations in two of them, which could either mean there is a background level of TERT and FGFR3 mutations in healthy individuals, or that these two individuals will develop bladder cancer in the future, Guldberg said, noting that the prevalence of background mutations in the general population still needs to be established.
Because the specificity of the test is not high, it might not be suitable for general population screening, but it could potentially be used for monitoring high-risk groups, he said, such as patients with hematuria.
The researchers have also used another version of the test, which looks for markers of renal cell carcinoma, in 240 DCH participants who subsequently developed kidney cancer. The results of that study are "not as spectacular" as those from the bladder cancer study, Guldberg said, but they were able to detect markers up to several years before patients were diagnosed, often at a late stage of the disease. For renal pelvis cancer in particular, the detection rate was high, so the test could potentially be used for early detection of this type of cancer, he said, though prospective studies are needed.