Diagnostic Potential of Hypermethylated DNA in Colorectal Cancer

Colorectal cancer (CRC) is one of the most common forms of cancer and the second leading cause of cancer-related deaths in the western world, accounting for the annual death of 215.000 europeans in 2012 and an estimated 50.310 Americans in 2014.

Sporadic CRC develops through the accumulation of genetic and epigenetic alterations in the epithelial cell-lining. These alterations are thought to push the normal epithelial cells into the adenoma-carcinoma sequence. Detection of hypermethylated DNA can be made in blood or stool of cancer patients and it is believed, that hypermethylated cell-free DNA in blood and stool can be used as a cancer biomarker, exhibiting high sensitivity and specificity in CRC and other major cancer-types (e.g. lung, prostate, and breast cancer).

Hypermethylated cell-free DNA - measured in plasma - could prove to be a more specific and sensitive biomarker for CRC than both the detection of blood in stool and CEA in blood of CRC patients.

The aims of this project are to examine if hypermethylation of specific genes measured from cell-free DNA in plasma of CRC patients is a diagnostic biomarker for CRC. The results of this study will show, if a single blood sample can differentiate patients with CRC from patients without cancer.

Development of a reliable sensitive and specific biomarker for CRC will immensely improve the diagnostics and handling of CRC patients.

To propose the most sensitive and specific hypermethylated DNA promoter regions for CRC detection, a review of the literature currently available will be undertaken. The following search criteria (MeSH terms) will be used: "colorectal neoplasm", "DNA methylation or CpG islands", "tumour or stool sample or blood or serum or plasma", "real time PCR or microarray analysis or genome sequencing or biopsy or blood analysis". Based on this review a literature review will be written: "DNA Methylation in Blood and Stool as a Screening and Prognostic Marker for Colorectal Cancer".

Epigenetics is the study of heritable changes in gene expression, which does not arise due to changes in the DNA sequence. Epigenetic alterations include micro-satellite instability, histone modifications and DNA methylation. DNA methylation is a normal phenomenon in eukaryotic cells and occurs across the whole genome. Methylation target cytosine-residues followed by a guanine on the DNA strand (called CpG-islands) converting these to methyl-cytosine. Un-methylated CpG-islands exist in a high concentration in the promoter regions of human genes, and DNA promoter hypermethylation leads to decreased transcriptional activity. This event happens in tumour suppressors leading to gene silencing which is related to tumour genesis. CpG-Island hypermethylation is shown to be a non-random process and the pattern of gene hypermethylation appears to be specific to individual cancers. This lead to the definition of a CpG-Island Methylator Phenotype (CIMP) in CRC, a panel of genes being hypermethylated in CRC tumours, and there is a growing body of literature on the correlation between hypermethylated genes and CRC. This investigation has primarily focused on DNA obtained from CRC biopsies but in recent years focus has moved to hypermethylated DNA obtained from blood and stool samples.

Hypermethylated genes that have been analysed in blood of CRC patients are e.g. SDC2, THBD, HIC1, RASSF1A. Hypermethylated SEPT9 has been analysed in a large prospective study to evaluate its performance as a screening marker, but this only revealed a sensitivity of 48.2% with a specificity of 91.5%. However when hypermethylated promoter regions are analysed in concert with others as part of a biomarker panel, markedly better performance has been reported. A biomarker panel composed of RASSF1A, APC, MGMT, and Wif-1 reached a sensitivity of 86.5% and a specificity of 92.1% for CRC detection.

From these studies, it is apparent that the profile of hypermethylated DNA in blood samples from CRC patients is different from the profile measured in healthy controls.

Only two studies have been conducted, prospectively studying the relationship between DNA hypermethylation measured in blood of CRC patients and risk of cancer recurrence. These studies are limited by only analysing a single gene promoter region (p16) and a low number of study participants.

The studies above show that hypermethylated DNA can be detected in blood of CRC patients, promoting it as biomarker for CRC. No study has however prospectively analysed the performance according to treatment and relapse of disease in CRC patients.

Studies on hypermethylated DNA in stool of CRC patients have been conducted using single gene hypermethylation markers. The gene primarily analysed has been hypermethylated VIM which also has a commercially available kit for stool analysis. The focus in the stool studies has been DNA hypermethylation as a screening marker for CRC. One study combined the iFOBT test with hypermethylation of the PHACTR3 gene, extracted from the iFOBT fluid. A positive result in at least one of the two tests raised sensitivity to 95% (19/20) at high specificity of 94% (45/48) for CRC and advanced adenomas.

Most of the methods for analysis require bisulfite conversion of the DNA for detection of methylated cytosines, and Pedersen et al. (2012) recently published an article addressing this key time-consuming step lowering the analysis-time and increasing the recovery of sparse amounts of DNA (< 0.1 ng/ml), rendering the test more clinically applicable.

In summary, DNA collected from blood and stool samples of CRC patients has been analysed with the purpose of discovering hypermethylated genes that are CRC specific. Hypermethylation of these specific genes in blood samples have been associated with poor prognosis (p16, HLTF, TMEFF1, ALX4, VIM, and FBN2), early diagnosis (APC, TAC1, SEPT9, NEUROG1, RASSF1A, SDC2, and THBD) and potential recurrence markers for CRC patients (p16 and TFPI2). Especially THBD and SDC2 show great promise as markers of early stage CRC, and the fact that it seems possible to use DNA methylation in blood as a marker of disease recurrence warrants further research, using these genes individually or in concert with others. Hypermethylation of DNA in stool samples of CRC patients have not been analysed as thoroughly as in blood samples, though detection of hypermethylated DNA fragments in stool samples could be used when screening for CRC. SFRP2, TMEFF2, SPG20, TFPI2, and PHACTR3 have shown promise in the detection of CRC at early stages or in premalignant lesions. The fact that PHACTR3 could be detected in iFOBT-fluid (reaching increased sensitivity and specificity) indicates a possible new screening method, where DNA hypermethylation is analysed in one stool sample. There are however issues regarding the analytical sensitivity of hypermethylated DNA in stool-samples from CRC patients. The diagnostic sensitivity for right-sided tumours is lower than for left-sided tumours suggesting DNA degradation in the colonic lumen before stool excretion. This promotes blood as the media of choice for detection of hypermethylated genes as biomarkers in the diagnosis, prognosis and recurrence detection of CRC.

METHOD AND STUDY PARTICIPANTS

1. To propose the most sensitive and specific hypermethylated DNA promoter regions for CRC detection, a review of the literature currently available will be undertaken. The search criteria are listed in the previous section.
2. Colorectal cancer patients: Blood samples from approximately 210 patients diagnosed with colorectal carcinoma collected prospectively and consecutively at the Department of Surgical Gastroenterology, Aalborg University Hospital during the time period 2003-2006 will be used. The blood samples were collected at the time of diagnosis, before and after treatment (surgery and radio-chemo-therapy) and at regular intervals for a time period of two years after treatment. These blood samples were collected during a previous PhD-study: "Pre- and postoperative Venous Thromboembolism in Patients with Colorectal Cancer - implications of Radiotherapy" (ID-number VN 2003/102). In collaboration with the Section of Molecular Diagnostics, Clinical Biochemistry, Aalborg University Hospital, the methylation profile of the cell-free DNA promoter regions in plasma will be analysed using methylation specific PCR (Polymerase Chain Reaction) based on an accelerated bisulfite treatment method.
3. Inflammatory bowel disease patients In collaboration with The Department of Medical Gastroenterology, Aalborg University Hospital, there will be recruited approximately 100 patients with ulcerous colitis for blood sampling (28 ml). In collaboration with the Section of Molecular Diagnostics, Clinical Biochemistry, Aalborg University Hospital, the methylation profile of the cell-free DNA promoter regions in plasma will be analysed using methylation specific PCR (Polymerase Chain Reaction) based on an accelerated bisulfite treatment method.
4. Control group: Blood samples (8ml) from patients (n=142) who underwent work-up for lower gastrointestinal malignancy during the time period 2003-2006. Colonoscopy was performed, and no malignancy or premalignant lesions found. Blood samples were collected before/after colonoscopy. These blood samples were collected during a previous PhD-study: "Pre- and postoperative Venous Thromboembolism in Patients with Colorectal Cancer - implications of Radiotherapy" (ID-number VN 2003/102). The plasma samples will be analysed at the Section of Molecular Diagnostics, Clinical Biochemistry, Aalborg University Hospital, and the methylation profile of the cell-free DNA in plasma will be analysed, using methylation specific PCR based on the accelerated bisulfite treatment protocol.
5. External validation: Cases: Blood samples from 143 patients were collected prospectively and consecutively for a previous study on the effect of omega 3 fatty acids an postoperative complications after colorectal surgery. All study participants have provided written informed consent (N-VN-20050035). Controls: There will be recruited approximately 100 control subjects with a positive occult faecal blood but with a normal colonoscopy for blood sampling (28 ml). In collaboration with the Section of Molecular Diagnostics, Clinical Biochemistry, Aalborg University Hospital, the methylation profile of the cell-free DNA promoter regions in plasma will be analysed using methylation specific PCR (Polymerase Chain Reaction) based on an accelerated bisulfite treatment method. Inclusion criteria: Patients referred to The Department of Gastrointestinal Surgery, Aalborg University Hospital, with a positive stool test for CRC but with no sign of cancer on the subsequent colonoscopy.
6. Blood sampling: Blood samples will be drawn by venepuncture in accordance with the European concerted action on thrombosis (ECAT) procedures.
7. Methylation specific PCR: In collaboration with the Section of Molecular Diagnostics, Clinical Biochemistry, Aalborg University Hospital the cell-free DNA promoter regions methylation profile will be analysed in EDTA plasma samples by methylation specific PCR based on the accelerated bisulfite protocol.
8. Biobank Any residual material (blood samples) from the study will be stored for fifteen years and thereafter the remaining samples will be transferred to the Danish National Biobank.