FDA Panel Unanimously Backs Cologuard Colorectal Cancer Test

Troy Brown, RN

March 27, 2014

An advisory panel of the US Food and Drug Administration (FDA) has unanimously recommended premarket approval for the Cologuard (Exact Sciences Corporation) colorectal cancer (CRC) diagnostic device today.

The Molecular and Clinical Genetics Panel of the FDA's Medical Devices Committee voted unanimously that the test is safe and effective and that its benefits outweigh its risks.

Cologuard is an in vitro diagnostic device that analyzes patients' stool to detect hemoglobin, multiple DNA methylation and mutational markers, and the total amount of human DNA contained in cells shed by CRC or advanced adenoma into the large bowel

The test is intended to be used as an adjunctive screening test to detect colorectal neoplasia-associated DNA markers and the presence of occult hemoglobin in human stool. A positive test result may indicate the presence of CRC or premalignant colorectal neoplasia. The device is not meant to be a replacement for colonoscopy and is intended to be used in conjunction with colonoscopy and other test methods according to recognized screening guidelines.

The vote follows a discussion of data from a pivotal prospective, multicenter trial that enrolled 12,776 patients aged 50 to 84 years with average risk for CRC. The study evaluated the safety and efficacy of Cologuard and fecal immunochemical test (FIT) compared with colonoscopy. The study was published this month by the New England Journal of Medicine.

The study's primary performance measures were Cologuard CRC sensitivity and Cologuard advanced-neoplasia (AN) specificity. Cologuard sensitivity for CRC was 92.3% (60/65), with a 1-sided 95% lower confidence bound of 84.5% (Clopper-Pearson method). It met the primary study objective of Cologuard CRC sensitivity greater than 65% and a 1-sided 95% lower confidence bound exceeding 65%.

Cologuard AN specificity (categories 3 - 6) was 86.6% (7967/9198), with 95% 1-sided lower confidence bound 86.0% (Clopper-Pearson method). It met the primary study objective of Cologuard AN specificity greater than 85% with a 1-sided 95% lower confidence bound exceeding 85%.

In a secondary analysis of data from 9989 patients with available Cologuard, PolyMedco FIT, and histology results, CRC sensitivity was 92.3% (60/65) for Cologuard and 73.8% (48/65) for PolyMedco FIT, showing a difference of 18.5%. The 1-sided 95% lower confidence bound on the difference was 8.0%. Thus, Cologuard met the secondary study objective and was noninferior to Poly FIT in CRC sensitivity with respect to protocol-specified noninferiority margin 5.0% because the lower bound of 8.0% is greater than −5.0%.

Although not a secondary objective, Cologuard specificity for AN was lower than FIT (86.6% vs 94.9%).

Sensitivity Superior to FIT

"The sensitivity data that are shown to be superior [to] FIT for colorectal cancer and for advanced neoplasia do indicate that this would be useful and appropriate as a screening test," said voting panel member Karen E. Weck, MD, professor and director of the Molecular Genetics Laboratory, University of North Carolina, Chapel Hill.

Among 100,000 patients in a screening population, 700 are expected to have CRC, 7580 are expected to have AA, and 91,720 are expected to be not AN (neither CRC nor AA). Cologuard is expected to detect 129 more patients with CRC and 1412 more patients with AA than PolyMedco FIT. However, the test is also expected to have 7594 more false-positives in non-AN patients than PolyMedco FIT. Compared with PolyMedco FIT, Cologuard is expected to detect 1 more patient with CRC and 11 more patients with AA for every 59 more false-positives on non-AN patients.

"I think this is a very impressive study, with very positive conclusions," said temporary voting panel member Steven Skates, PhD, associate in Biostatistics (Medicine), Cancer Center, Massachusetts General Hospital, Boston. Regarding whether or not the test's benefits outweigh its risks, he said, "I thought the trade-off was balanced and completely justified."

Alternative to FIT?

Voting panel member Mary B. Mahowald, PhD, professor emerita at the University of Chicago, Illinois, in the Department of Obstetrics and Gynecology, MacLean Center for Clinical Ethics, and the Committee on Genetics, said she would like to see the test recommended as an alternative to the FIT test rather than as an adjunctive to the FIT test. "I don't see it as adjunctive to the FIT test," she said.

"I think this is a phenomenal study.... You're looking at the large polyps you would not otherwise find," said panel chair Ronald Przygodzki, MD, Acting director, Biomedical Laboratory Research and Development in the Office of Research and Development, Department of Veterans Affairs, Washington, DC.

The voting members have disclosed no relative financial relationships.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: