Evaluating the potential of micronuclei DNA from erythrocytes for early detection or recurrence risk prediction of colorectal cancer.

3577 Background: Current development in stool- and cell-free DNA (cfDNA)-based technologies have demonstrated promising potentials in colorectal cancer (CRC) diagnosis. However, early detection of CRC and advanced adenoma (AA) remains challenging. Also, better predicting the recurrence risk following initial treatment serves as an important factor influencing survival of CRC patients. Micronuclei (MN) are extranuclear bodies containing chromatin segments resulting from errors in DNA repair. Elevated levels of MN+ erythrocytes have been studied in genotoxicity testing and cancer diagnosis. We have developed a method (WO2021/228246 A1) for purifying and sequencing micronuclei DNA (MN-DNA) in erythrocytes from peripheral blood. By comparing MN-DNA from healthy donors (HDs) and CRC patients, we identified significant changes in read densities at specific genomic locations in patients, which were termed as tumor-associated MN-DNA (taMN-DNA) signatures. Here, we explored the potential of MN-DNA for early detection or recurrence risk prediction of CRC patients. Methods: Peripheral blood (1-2ml) was obtained from a training cohort of 1165 individuals, a validation cohort of 290 individuals, and a test cohort of 363 individuals. MN-DNA was isolated and sequenced from erythrocytes. Predictive models were developed using distinctive taMN-DNA features identified by genome-wide analysis to differentiate between HD and AA or CRC. A clinical trial comparing MN-DNA with an approved multitarget stool DNA test (sDNA-FIT) was also conducted on 60 individuals (42 HDs, 6 AAs, and 12 CRCs). In addition, we conducted a retrospective study on CRC patients after postoperative treatment for evaluating the potential of MN-DNA in predicting CRC recurrence risk, with 28 non-recurrent after initial surgery and treatment and 6 relapsed patients enrolled. Results: The predictive CRC model built on taMN-DNA features achieved an AUC of 97.93%, with 88.61% sensitivity at 95.05% specificity, especially 89.58% sensitivity for early-stage (stage I-II) CRC in the test cohort. The detection of AA achieved an AUC of 93.01%, with 68.29% sensitivity at 92.48% specificity in the test cohort. Comparing MN-DNA to sDNA-FIT showed superior overall sensitivity (83.33% vs. 61.11%), AA sensitivity (83.33% vs. 16.66%), with comparable specificity (85.71% vs. 88.10%). Finally, 92.86% patients who had no recurrence were detected with negative prediction results, whereas 83.33% patients who had relapsed were detected positive prediction result. Conclusions: Our results demonstrate that MN-DNA within erythrocytes from 1-2ml peripheral blood enable accurate detection of AA and early-stage CRC as well as recurrence risk prediction. A larger clinical study (NCT05875584) is ongoing and will further validate the application of MN-DNA in CRC detection.