A blood-based DNA test to detect colorectal cancer has the potential to increase screening rates and the number of tumors found at an early stage, according to researchers at Epigenomics, Inc. the Seattle-based molecular diagnostics company that is developing the test (see graphic on page 1).
WASHINGTONA blood-based DNA test to detect colorectal cancer has the potential to increase screening rates and the number of tumors found at an early stage, according to researchers at Epigenomics, Inc. the Seattle-based molecular diagnostics company that is developing the test (see graphic on page 1).
In two large samples, the test for the methylated form of the Septin 9 gene correctly identified about half of the patients with colorectal cancer at any stage, said Catherine Lofton-Day, PhD, vice president for molecular biology at Epigenomics. Dr. Lofton-Day presented the results at the 97th Annual Meeting of the American Association for Cancer Research (abstract LB-224).
That 50% sensitivity is an improvement over the sensitivity generally seen with the fecal occult blood test (FOBT), the only currently available noninvasive screening test for colorectal cancer, she said. Studies of the sensitivity of FOBT have shown variable results, ranging from 15% to 70% but usually in the lower range. "The sensitivity of our test is, at this point, much higher than the typical sensitivity of FOBT," she said.
Dr. Lofton-Day also noted that because FOBT requires people to collect stool samples over a 3-day period, compliance with the procedure is low. In contrast, a blood test "would increase compliance and cause a lot more people to use the test and therefore detect more disease," she said.
To evaluate the Septin 9 test, the researchers used real-time polymerase chain reaction (PCR) to detect the methylated form of the gene in plasma from more than 1,000 subjects. Septin 9 is involved in cell division and the cell cycle, and is known to be methylated in some cancers, Dr. Lofton-Day said.
Two Sample Sets
The researchers first used the assay in plasma from a sample of 501 patients in order to determine its clinical performance and the threshold for positive/negative classification. The test was then validated on a blinded, independent set of 790 patients. Patient ages ranged from 40 to 80, but most were age 50 and older.
In both sets, the researchers looked for methylated Septin 9 in four different groups: patients with colorectal cancer, age-matched healthy subjects without colorectal cancer (verified by colonoscopy), patients with other cancers, and patients with noncancerous diseases.
In the first sample set, the PCR test found 72 of 127 colorectal cancers overall (57%); most patients in this group had stage III and IV tumors, in which the test was more sensitive. In the validation sample, the test found 104 of 209 colorectal tumors overall (50%), including 24 of 51 stage I tumors (47%) and 29 of 65 stage II tumors (45%). In both groups, the test detected tumors in different locations in the colon and rectum with similar sensitivity.
The Septin 9 test was highly specific, with only 4% of results being false positive in the first sample and 6% in the second (94% to 95% specificity overall). In patients with conditions such as gastritis, arthritis, respiratory infection, and early-stage cancers other than colorectal cancer, there were more false positives, but specificity was still high, at 92%.
Working to Increase Sensitivity
Dr. Lofton-Day said that the company was working on ways to increase sensitivity. One strategy is to find additional markers to complement methylated Septin 9, she said, and another might be to lower the threshold for a positive classification, which would decrease specificity and increase sensitivity. The researchers concluded that the Septin 9 test, followed by colonoscopy for people with positive results, has the potential to reduce mortality from colorectal cancer. Roche Diagnostics has licensed the marker, and development of the test will continue, Dr. Lofton-Day said.