FDA Panel Endorses Whole-Blood CRC Screening Tool by Narrow Margin

Colon

The blood-based colorectal cancer (CRC) screening test Epi proColon passed the scrutiny of the FDA’s Molecular and Clinical Genetics advisory panel in a close 5-4 vote with 1 abstention in support of the claim that the test’s benefits outweigh its risks.

In an uncommon scenario, the panel was convened to assess an application for premarket approval (PMA) that was submitted by the diagnostic’s developer, Epigenomics AG. During the meeting, the panel favorably voted 9-0, with 1 abstention, that the test could be used safely; however, when weighing the screening tools effectiveness, the panel was split. A tie-breaking decision from the chairperson resulted in a negative vote of 5-6. The panel’s recommendations will be considered in the FDA’s review process for the Epi proColon PMA, which is the most stringent type of device marketing application available.

The Epi proColon test is intended to screen patients for CRC who are defined as average risk. The test utilizes real-time polymerase chain reaction with a fluorescent hydrolysis probe for the detection of Septin9 methylated DNA in the plasma. Studies have linked the Septin9 biomarker to the occurrence of CRC, with heightened sensitivity in advanced stages. Following the detection of circulating methylated Septin9 DNA, patients still require further evaluation by colonoscopy.

In the documentation leading up to the meeting, the FDA cast doubt on the clinical data supporting the PMA for the “first of a kind” device, which was addressed during the panel meeting. As a result of these concerns, Epigenomics consented to conducting post-approval studies. Moreover, the company noted that it plans to discuss product-labeling changes in order to further address these concerns.

In the study that was submitted as support for the approval, the FDA reviewed data from 1544 samples out of 1623 that were evaluated in the study. These samples were processed in 134 batches. Valid results were reported for 115 batches (91.3%), which passed the predefined criterion of a 90% success rate.

The FDA documentation raised concerns that the test did not meet all of its primary endpoints. In the study, the sensitivity of the test was 68.2% for the detection of CRC, passing the predefined criterion of 65%. However, the lower bound of the 95% confidence interval was 53.4%, which was lower than the targeted point estimate of 65%. The specificity of the test was 78.8%, which did not meet the predefined criterion of 85%. The false positive rate of 21% was higher than expected, the FDA noted.

Further analyses of the test were conducted across tumor stage, location, ethnicity, and age. Regarding tumor stage, researchers found a positive detection fraction of 41% in stage I, 83% in stage II, 80% in stage III, and 100% in stage IV CRC. The analysis showed differences between age groups, suggesting age-related increases in methylation.

An additional analysis of the data was conducted by the FDA showing a non-statistically significant difference between patients enrolled in the US compared with Germany. These estimates showed 57.7% sensitivity for US patients compared with 83.3% in German patients (P= .10).

Following the panel recommendation, Epigenomics will meet with the FDA within four to six weeks to discuss the next steps in the approval process.