CHICAGO – A novel plasma microRNA assay and prediction model appears to successfully differentiate colorectal neoplasia from other neoplasms and from controls.
The assay includes seven microRNAs that were selected, based on P value, area under the curve (AUC), fold change, and biological plausibility, from among 380 microRNAs screened using microfluidic array technology from a “training” cohort of 60 patients. The training cohort included groups of patients, 10 each, with colorectal cancer, advanced adenoma, breast cancer, pancreatic cancer, and lung cancer – cancers chosen because they frequently develop at similar ages as colon cancer – and 10 controls.
A panel of seven “uniquely dysregulated” microRNAs specific for colorectal neoplasia was evaluated using single assays in a “test” cohort of 120 patients. A mathematical model was developed to predict sample identity in a 150-patient blinded “validation” cohort using repeat-subsampling validation of the testing dataset with 1,000 iterations each to assess model detection accuracy, Dr. Jane V. Carter of the University of Louisville (Ky.) explained at the annual meeting of the American Surgical Association.
The area under the curve for test cohort comparisons with the assay was 0.91, 0.79, and 0.98 for comparison No. 1 (comparing any neoplasia vs. controls), comparison No. 2 (comparing colorectal neoplasia with other cancers) and comparison No. 3 (comparing colorectal cancer with colorectal adenomas) respectively, Dr. Carter reported.
“Our prediction model identified blinded sample identity with 69%-77% accuracy in comparison No. 1, 66%-76% accuracy in comparison No. 2, and 86%-90% accuracy in comparison No. 3,” she said, noting that the sensitivity and specificity of the assay compare very well with current clinical standards.
Colorectal neoplasms frequently develop in individuals at ages when other common cancers also occur. Current screening methods, including endoscopic and imaging studies and fecal testing have poor patient compliance. Fecal and blood tests lack sensitivity and specificity for the detection of adenomas, limiting their use as screening methods, she said.
But this novel assay, which builds on the earlier work identifying miR-21 as a potential marker for colorectal cancer, provides a useful tool for identifying colorectal neoplasms, she said.
Efforts are underway to confirm the findings in a larger study population. If the findings are confirmed, the assay may have other potential uses such as monitoring therapy by comparing microRNA expression before and after treatment, and also for predicting response to treatment such as following preoperative neoadjuvant chemoradiation, Dr. Carter suggested.
The current findings have significant implications for the development of a noninvasive, reliable, and reproducible screening test for detection of colorectal neoplasia.
“If we can improve early-stage detection, we can improve survival,” she said.
Dr. Carter reported having no relevant disclosures.
The complete manuscript of this presentation is anticipated to be published in the Annals of Surgery pending editorial review.