NEW YORK (GenomeWeb News) – An expert panel of the US Food and Drug Administration today reviewed clinical data on Exact Science's Cologuard and unanimously supported approval of the stool-based, non-invasive molecular screening test for colorectal cancer.
All 10 members of the Molecular and Clinical Genetics Panel of FDA's Medical Devices Advisory Committee found that Cologuard is safe and effective in the population for which it would be indicated. Additionally, all committee members were reasonably assured that the benefits of the test outweighed its risks in the intended use population.
Exact is seeking approval for Cologuard as an adjunctive tool that assesses non-invasively from a patient's stool sample whether he or she has colorectal cancer or pre-malignant colorectal neoplasia (or advanced adenoma). The test analyzes hemogloblin, multiple DNA methylation markers, mutations in the KRAS gene, and the total amount of DNA in the stool sample. The company does not seek to replace colonoscopy, which is the recommended screening method for the disease in several treatment guidelines, with Cologuard.
The positive review by the panel bodes well for marketing approval of the test. Although advisory committee recommendations are not binding on the FDA, the agency has historically taken the advice of its expert panels.
Exact has pursued a modular pre-market approval submission pathway for Cologuard and submitted different components of its application from December 2012 to July 2013. The company is taking part in a parallel review pilot program, under which data from its 10,000-patient DeeP-C trial on Cologuard is doing double duty to support FDA marketing approval for the test as well as Medicare reimbursement from the Centers for Medicare & Medicaid Services. Alberto Gutierrez, director of FDA's Office of In Vitro Diagnostics and Radiological Health, commended Exact Sciences for taking the "unusual" move by involving FDA and CMS in designing the pivotal trial for Cologuard.
Data from the DeeP-C trial was published recently in the New England Journal of Medicine, in which Exact was able to show that in those with an average risk of colorectal cancer, Cologuard had a 92 percent sensitivity (ability to detect true positives) to detect CRC and an 87 percent specificity (ability to detect true negatives) for gauging non-advanced adenomas or negative results, both meeting prespecified study goals.
In DeeP-C, Exact also wanted to compare Cologuard to fecal immunohistochemical testing, which is a standard CRC screening tool. In the study, FIT had 74 percent sensitivity, while specificity was 94 percent. But a comparison of specificity was not an endpoint in DeeP-C, and as such, Cologuard was considered non-inferior and superior to FIT in detecting CRC in the study. In detecting patients with advanced adenomas, sensitivity was 42.4 percent for Cologuard, which was found superior to FIT with a sensitivity of 23.8 percent. Additionally, 13 cancers missed by FIT were detected by Cologuard, and one cancer missed by Cologuard was detected by FIT; 170 advanced adenomas missed by FIT were picked up by Cologuard, and 29 positive by FIT were negative according to Cologuard.
Drilling further into DeeP-C data, there were finer points that the advisory panel discussed with the FDA and Exact in order to better gauge the trade-off between sensitivity and specificity in using Cologuard as a CRC screening tool and the comparative false-positive rate with FIT. For example, according to analysis by FDA reviewers, based on a screening population of 100,000 people, Cologuard will likely report nearly 7,600 more false positive tests than FIT. However, Cologuard would also gauge one more CRC patient and 11 more AA patients for 59 more false positives compared to FIT.
"Based on the DeeP-C study, the seriousness of a false-positive error relative to a false-negative error is found to be smaller for Cologuard at its operating point than FIT at its operating point," one FDA reviewer said during the meeting, noting "there are different trade-offs for false-positive and false-negative errors" between the two tests.
Committee members seemed to be assured that the sensitivity of Cologuard was robust, and several noted the importance of a screening test to be able to accurately assess true positives. Regarding "the hit to specificity, I'm willing to have a false-positive in 15 out of 100 people with this test, versus 5 out of 100 with the FIT test, in order to gain that sensitivity," said committee member Karen Weck of the University of North Carolina, Chapel Hill.
After reviewing and discussing the DeeP-C data, committee member Steven Skates from Massachusetts General Hospital said that Cologuard was "one of the biggest improvements to early [CRC] detection" he had seen.
During public presentations, speakers representing CRC advocacy organizations, physicians, and patients spoke in favor of Cologuard's approval and reimbursement, highlighting its ability to pick up pre-malignant polyps (advanced adenomas) that wouldn't otherwise be gauged by other tests and citing the need for a non-invasive screening test that can reach rural or at-risk populations. Exact Sciences sends to patients' homes a stool sample collection kit, which they send back to the lab for analysis.
The invasive nature of colonoscopy is a major deterrent for many patients to get CRC screening. "It requires someone who has no symptoms, who feels healthy, to undergo a fairly invasive procedure," Eric Hargis, CEO of the Colon Cancer Alliance, told committee members. Marcia Mullins, diagnosed with stage III rectal cancer at 58 years old, admitted that she should have gotten screened when she was 50, "but I didn't want to have a colonoscopy."
Exact has also proposed a prospective, post-marketing study of nearly 2,000 participants to assess the risk of CRC and advance adenoma associated with a positive Cologuard result at three-year follow up. In the study, those with a positive result at baseline will undergo colonoscopy and discontinue from the study. Participants with a negative Cologuard result at baseline will stay on the study and receive Cologuard testing and colonoscopy again at three years.
Some advisory committee members felt that the design of this post-marketing study should be changed to look at whether Cologuard should be done annually instead of every three years. Hopeful that Cologuard will become "part of the arsenal" for CRC detection, committee member Patricia Ann Furlong, representing the Parent Project Muscular Dystrophy and a consumer representative, said she was worried about the proposed three-year testing interval for the test, which may hinder timely cancer detection for some patients. Another committee member noted the need to consider whether more frequent screening with Cologuard would also increase its false-positive rate.
The day before reviewing Cologuard, the Molecular and Clinical Genetics Panel also discussed the clinical data supporting Epigenomics' PMA for Epi proColon. In that case, although the panel unanimously found the test was safe for the intended population, experts provided less straightforward guidance to the FDA regarding the diagnostic's efficacy and risk/benefit balance.