Originally published March 17.
By Valerie Ross
In a recent review of the available research on ColoSure, Laboratory Corporation of America's fecal DNA test for colorectal cancer, researchers from the Centers for Disease Control and Prevention found insufficient evidence to evaluate the test's analytic validity, clinical validity, and clinical utility in a general risk population.
In a review published last week in PLoS Currents: Evidence for Genomic Applications Journal, the CDC's Office of Public Health Genomics said that it found a paucity of solid data to support the clinical use of the test.
As is common for most lab-developed genetic tests, “there aren’t a whole lot [of available studies] on the analytic performance of the test, in terms of how well it detects the DNA markers it’s supposed to assay,” CDC health scientist Renée Ned, an author of the review, told Pharmacogenomics Reporter. “The tests we found that looked at some of the other performance measures looked at small case control studies, if there was any evidence at all.”
ColoSure is a non-invasive single-marker test that detects methylation of the vimentin gene, which is associated with colorectal cancer and pre-cancerous adenomas. LabCorp is marketing it as a laboratory-developed test.
Exact Sciences, from whom LabCorp licensed intellectual property for the ColoSure test, is currently developing a next-generation multi-gene fecal DNA test for colorectal cancer, called Cologuard. Exact Sciences has done a preliminary clinical study of the test, and plans to start clinical trials later this year and to apply for premarket approval from the US Food and Drug Administration, according to Exact Sciences CEO Kevin Conroy.
If Exact Sciences' Cologuard is approved, it's likely it would displace ColoSure, CDC's Ned said. “As these tests get developed, they tend to take the old ones off the market, because the newer ones show better performance.”
Current colorectal cancer screening tests, such as colonoscopy and flexible sigmoidoscopy, are highly invasive, making non-invasive fecal occult blood tests a desirable option for doctors and patients.
Such tests are not intended to completely replace other types of colorectal cancer screening, however, or to be used in patients with elevated risk of the disease. Instead, a LabCorp brochure on ColoSure notes, the test is meant to “be a useful alternative for asymptomatic patients who are at average risk for developing colorectal cancer” and “are unwilling or unable to undergo a more invasive structural exam.”
Currently, ColoSure is the only fecal DNA test for colorectal cancer available in the US, and thus was the sole focus of the CDC's review. The study examined the evidence for the test in three areas: analytic validity, clinical validity, and clinical utility. LabCorp could not be reached for comment before press time.
Like CDC's Evaluation of Genomic Applications in Practice and Prevention Working Group, the Office of Public Health Genomics identifies and compiles evidence on genetic tests in clinical practice. While EGAPP makes a clinical practice recommendation after reviewing the evidence, however, the Office of Public Health Genomics focuses on evaluating and identifying gaps in the existing data. “We tried to put this test in context,” said Ned.
The CDC team could find no published data on the test's sensitivity or specificity for detecting methylated vimentin. In addition, a study published in Clinical Chemistry in 2007 showed that “the amount (total and relative) of methylated vimentin in stool samples can vary widely in patients with adenoma or colorectal cancer,” they wrote. As a result, the team determined that it could not ascertain the test's analytic validity.
In evaluating ColoSure's clinical validity, the CDC team found six clinical studies, all with case-control designs, comparing patients known to be positive or negative for colorectal cancer as confirmed by colonoscopy. Combining the two studies that they found to be most relevant in terms of procedure and design, along with LabCorp internal data mentioned in ColoSure materials, the CDC researchers found the data suggests that ColoSure has sensitivity of between 72 percent and 77 percent, and specificity of between 83 percent and 94 percent.
Because of the limited data, and the fact that fecal DNA screening using methylated vimentin markers has not been adequately compared to other methods of colorectal cancer screening, the CDC group wrote that it could not determine ColoSure's clinical validity for an average-risk population. The authors note that this is in line with LabCorp's materials about the test, which state that “[t]he detection rates for general population screening have not been determined.”
The CDC team further found no randomized, controlled trials using ColoSure that examined colorectal cancer incidence or clinical outcomes. ColoSure is non-invasive, can be done at home, and does not require preparation (such as a particular diet) beforehand, the review authors note, which an earlier study suggests may make it more attractive to patients than traditional screening methods.
However, the test has “uncertain disease detection benefit,” there is no specified screening interval for ColoSure, and it may not be cost-effective in comparison to other screening tests. Because of this, the CDC team said, it could not determine ColoSure's clinical utility.
Cologuard, the fecal DNA test being developed by Exact Sciences, is designed to detect multiple markers, one of which is methylated vimentin, the marker used in ColoSure. The test will also require a much smaller stool sample — 8 grams as opposed to the 36 grams required for ColoSure — and use different DNA extraction and detection chemistry methods, said Conroy.
In a preliminary study of the test, presented last fall at a colorectal cancer conference sponsored by the American Association of Cancer Research, researchers from Exact Sciences and the Mayo Clinic reported that they were able to detect colorectal cancer with a sensitivity of 85 percent and adenomas larger than one centimeter with a sensitivity of 64 percent, at a specificity cutoff of 90 percent.
Before applying for premarket approval from the FDA, Exact Sciences will start a clinical trial of Cologuard in the third quarter of this year, Conroy said. The trial will enroll between 8,600 and 12,000 patients at more than 30 doctor's offices and colonoscopy clinics, and screen samples at two or three testing sites.
Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at vross [at] genomeweb [.] com.