A systematic evaluation of stool DNA preparation protocols for colorectal cancer screening via analysis of DNA methylation biomarkers.

Objectives Colorectal cancer (CRC) screening using stool samples is now in routine use where tumor DNA methylation analysis for leading markers such as NDRG4 and SDC2 is an integral part of the test. However, processing stool samples for reproducible and efficient extraction of human genomic DNA remains a bottleneck for further research into better biomarkers and assays. Methods We systematically evaluated several factors involved in the processing of stool samples and extraction of DNA. These factors include: stool processing (solid and homogenized samples), preparation of DNA from supernatant and pellets, and DNA extraction with column and magnetic beads-based methods. Furthermore, SDC2 and NDRG4 methylation levels were used to evaluate the clinical performance of the optimal protocol. Results The yield of total and human genomic DNA (hgDNA) was not reproducible when solid stool scraping is used, possibly due to sampling variations. More reproducible results were obtained from homogenized stool samples. Magnetic beads-based DNA extraction using the supernatant from the homogenized stool was chosen for further analysis due to better reproducibility, higher hgDNA yield, lower non-hgDNA background, and the potential for automation. With this protocol, a combination of SDC2 and NDRG4 methylation signals with a linear regression model achieved a sensitivity and specificity of 81.82 and 93.75%, respectively. Conclusions Through the systematic evaluation of different stool processing and DNA extraction methods, we established a reproducible protocol for analyzing tumor DNA methylation markers in stool samples for colorectal cancer screening.