To avoid misconception hereby, we define stool DNA test as a DNA-based diagnostic test for colorectal cancer whereas genetic testing as an analysis of germline DNA from blood lymphocytes to define a hereditary condition such as familial adenomatous polyposis or Lynch syndrome.

Point mutations of following three genes that are often implicated in CRC have been targeted for stool DNA markers: 1) Adenomatous polyopsis coli (APC) residing on chromosome 5q, tumor suppressor gene. Multiple point mutations occur early in the development of an adenoma. 2) K-ras, found on chromosome 12, oncogene. Multiple point mutations of this gene frequently appear during CRC development following APC mutations. 3) TP53, tumor suppressor gene, located on chromosome 17p. Mutations at this gene are often found later in cancer progression and associated with larger adenoma with more severe grade of dysplasia.

As consequence of mutations of mismatch repair gene products, microsatellites and repeating sequences of DNA become instable. One of microsatellites, BAT-26 which is single locus of 26 consecutive adenine nucleotides, is associated with loss of function of a mismatch gene. Thus, alteration in microsatellite on the BAT-26 locus has often been used as a maker in stool DNA test.1

Normally, genomic DNA is fragmented into short length (< 200 bp) by endonucleases during apoptosis. Longer fragment human DNAs (larger than 200 bp) exist more abundantly in stool-derived DNA samples because neoplastic cells resistant to apoptosis are exfoliated into colonic lumen. This observation suggests that DNA integrity may be used as a marker for stool DNA test.5-7

Small scale pilot study for testing multitarget genetic markers in stool for early detection of colorectal neoplasia includes 21 mutations in APC, K-Ras and p53 as well as BAT-26 and DNA Integrity Assay (DIA) which was developed by Exact Sciences and marketed as Pre-GenPlus (version I) by Lab Corp. These approaches detected adenocarcinoma and adenoma with sensitivity ranging from 62% to 91% and from 27% to 82%, respectively. Specificity was high, ranging from 93% to 98%, for subjects without colorectal neoplasia.7 However, in the most prominent large scale prospective screening trial, the same markers as used in version I test had a lower sensitivity of 52% for invasive and 18% for advanced cancer, although the specificity of this particular assay was 94%.8 This lower sensitivity was attributed to unexpected low rate of positivity for DIA. Later, however, it was known that the suboptimal performance of DIA was due to DNA degradation during specimen mailing to the lab. Indeed, better sensitivity of this marker was achieved with improved condition of sample preparation.9 According to Song et al. who estimated clinical and economic consequences of fecal DNA testing and compared them with conventional CRC screening using Markov model, the sensitivity of the panel of DNA markers in detecting adenocarcinoma and adenoma in stool needs to be increased to 65% and 40%, respectively.12 In order for DNA testing in stool to be cost effective, the cost per test needs to be adjusted to $195. Experimental strategies that have improved analytical sensitivity of PCR to detect very low number of copies of tumor DNA in stool enhanced the performance of fecal DNA testing further. BEAMing technology identified DNA mutations in stool (92%; 23 of 25 stool DNA from CRC patients) and a digital melt curve (DMC) assay showed a high level of sensitivity in detecting individuals with colorectal neoplasms (90%; 28 of 31 tumors) and better performance in detecting those with advanced adenoma (59%; 16 of 27 advanced adenoma).10,11 In these studies, however, they did not comment on the specificity. As a general rule, sensitivity is increased at the expense of clinical specificity. Thus, additional refinements that maximize sensitivity and specificity for both advanced adenomas and cancer are necessary before widespread clinical implementation.

Aberrant DNA methylation is a major epigenetic change that is early and common event in human tumorigenesis.13-16 Methylation at a specific cytosine residue of CpG site is a key component of the epigenetic mechanism that is associated with changes in gene expression without change of DNA sequence in tumorigenesis. In CRC, several genes have been identified that are aberrantly hypermethylated while not in normal mucosa. There have been several attempts to utilize such tumor-associated DNA methylation as a biomarker for potential early detection of colon cancer.

Epigenetic studies have focused on analyzing single gene methylation as a candidate for a potential biomarker for stool DNA-based screening of CRC. Muller et al.17 assessed stool DNA for SFRP2 methylation from three independent sets of patients with CRC with sensitivity of 77% to 90% and a specificity of 77%. Three different groups also reported comparable data for SFRP2 methylation in stool DNA, with a sensitivity of 87% to 94% and a specificity of 85% to 90%.18-20 Other single methylation marker assessment such as SFRP1 gene for methylation status in stool reported a sensitivity of 89% and specificity of 86% in stool samples from 36 patients with colorectal neoplasia (7 adenoma, 29 colorectal cancer) and 17 healthy control subjects.21 Chen et al. found that vimentin gene is transcriptionally inactive in normal colon epithelial cells with very low level of methylation on CpG region upstream of the first exon and continuing across the first two-thirds of the first exon. This region becomes highly methylated in primary tumor of colon cancer. They detected aberrant hypermethylation in 83% (38 of 46) and 53% (57 of 107) of primary tumor tissues from two independent groups of colon cancer. However, in its recapitulation test in stool-derived DNA, they detected only 46% from third set of 94 colon cancer patients versus 10% from normal colonoscopy.22 Itzkowitz and colleagues9 further validated vimentin methylation as a marker for cancer detection in stool DNA under better condition of DNA preparation and obtained a remarkably improved sensitivity of 72.5% (29 of 40 patients) and a specificity of 86.9% (16 of 122 normal individuals).

Imperiale et al.8 analysed the performance of combination markers for stool DNA test, namely as second generation assay, using the same markers of version I used in the previous multicenter study plus vimentin methylation as a marker. They improved the way of sample collection with new DNA stabilization buffer which was developed for preventing DNA degradation23 and gene-based DNA capture to enhance DNA extraction from stool.23,24 Stool samples from 40 patients with CRC and 122 subjects with normal colonoscopy were used for analysis. The sensitivity of the version I markers increased from 51.6% to 72.5% due to improved performance of DIA from 3.2% to 65%. Aberrant methylation of vimentin gene alone provided high values of sensitivity and specificity of 72.5% and 89.9%, respectively, that are very similar to those of panel of version I markers. The least complex assay, using combination of only two markers consisting of vimentin and the best marker of DIA components, yielded a maximum sensitivity of 87.5% and specificity of 82%. DNA integrity as a biomarker for CRC screening remains to be determined. It has been known that the stability of long-length genomic DNA is not constantly maintained during colon transit and fecal storage. Indeed, long-length DNAs are detected more frequently in left-side cancer than right-side one. Furthermore, long length DNA may not be specific for neoplasia of CRC, because its presence could be attributed to increased inflammation or field cancerlization characterized by enhanced proliferation and apoptosis resistance.7,25 Notably, single marker test for vimentin methylation has replaced 23 markers test which had previously been developed by Exact-Sciences and is currently the only commercially available stool DNA test through Lab Corp in the name of ColoSure™ ($240 per test), although the performance of vimentin methylation in detection of advanced adenomas has not yet been defined.9

Glockner et al.26 reported recently that hypermethylation of tissue factor pathway inhibitor 2 (TFPI2) gene in stool could be utilized as a marker for CRC screening. As a frequently methylated gene in human CRC, they used gene expression microarray-based strategy to identify the methylation of TFPI2 gene. This gene was reactivated in their microarray analysis by methylation inhibiting agent, DAC, and was frequently found to be methylated in colon cancer cell lines. Many previous studies showed aberrant methylation of TFPI2 gene in various cancers.27-32 They detected aberrant methylation of this gene by methylation specific PCR in 94% of invasive errated adenoma (16 of 17), 100% of tubular adenomas (17 of 17), and 100% of vilous adenomas (22 of 22), all of which represent preinvasive stages, and in 99% of invasive colorectal cancer states I to IV (114 out of 115), almost all colorectal adenoma. They observed TFPI2 methylation in stool DNA from stage I to III CRC patients with a sensitivity of 76% to 89% and a specificity of 79% to 93%. However, for adenoma, although more than 95% sensitivity was detected in primary tumor tissues, only 21% to 43% sensitivity was achieved for stool DNA samples, indicating that fewer cells are shed by adenomas.