A data-driven algorithm based on fecal microRNA and hemoglobin concentrations had the potential to increase the effectiveness and efficiency of a colorectal cancer screening program, according to study results published in Clinical Gastroenterology and Hepatology.
The algorithm called “miRFec” integrated data on fecal miR-421 and miR-27a-3p levels, fecal hemoglobin concentration, and patient age and sex from 767 fecal samples collected between 2011 and 2017 in Barcelona, Spain. All participants underwent a colonoscopy to assess how well miRFec predicted colorectal cancer (CRC) risk.
The miRFec algorithm scored samples between 1 (low risk) and 4 (high risk). A total of 54% (n=414) of the participants had advanced colorectal neoplasia identified during colonoscopy (67 CRC and 347 advanced adenomas); 136 participants had nonadvanced adenomas; and 217 had normal scans. The investigators reported a high correlation between the miRFec scores and the participant’s colonoscopy results (P <.001) among the advanced colorectal neoplasia group. The algorithm differentiated between patients with cancer from patients without CRC with an area under the receiver operating characteristic curve of 90% (95% CI, 86%-94%).
The investigators reported that samples with an miRFec score higher than 3.09 were 8-fold more likely to have advanced colorectal neoplasia than samples with a score below 2.14. When a 50% specificity cutoff rate was applied to the miRFec scores, the algorithm correctly identified 97% of the participants with CRC and 79% of participants with advanced neoplasia. The investigators asserted that if this algorithm were used as a diagnostic tool, 264 participants (34.4%) could have avoided their colonoscopy examination.
A limitation of this study is the lack of external validation. The miRFec model was designed and tested on a single cohort, which may have caused some of these reported results to be overestimated.
The study authors concluded that their designed algorithm, miRFec, was able to differentiate between patients with CRC from persons with nonadvanced adenomas or normal colonoscopies. Although the algorithm required additional testing on fecal samples, the output score may help patients avoid unnecessary colonoscopies and ultimately improve the efficiency and accuracy of CRC screening programs.
Disclosures: Some authors declared receiving consulting or funding from the pharmaceutical industry. A complete list of disclosures can be found in the original study.
Duran-Sanchon S, Moreno L, Gómez-Matas J, et al. Fecal microRNA-based algorithm increases effectiveness of fecal immunochemical test-based screening for colorectal cancer [published online February 27, 2020]. Clin Gastroenterol Hepatol. doi: 10.1016/j.cgh.2020.02.043