Colorectal cancer is one of the most prevalent causes of cancer deaths worldwide, with an estimated 1500 deaths each year in Norway alone. Early detection of colorectal cancer may significantly reduce this number, and it is therefore of great interest to identify biomarkers that can be used in a reliable non-invasive test for early detection.
Aberrant promoter DNA methylation has great potential as diagnostic biomarkers. They are both prevalent in cancer and have been shown to be an early event in tumor development. These changes can further be detected in feces and blood, materials suitable for non-invasive testing.
The present thesis is part of an ongoing project where we have set up a step-wise experimental protocol to identify DNA methylation biomarkers for cholangiocarcinomas – a malignancy arising in the bile ducts. We and others have previously shown that gastrointestinal cancers frequently display similar epigenetic aberrations, and the main focus of this thesis was to evaluate whether the identified candidates could be used as biomarkers for early detection of colorectal cancer. From the cholangiocarcinoma approach, 43 genes were identified as potential epigenetically deregulated genes. These genes were investigated by methylation specific PCR (MSP) in cancer cell lines (n=24). Twelve- and thirteen genes were frequently hypermethylated in the cholangiocarcinoma- and colon cancer cell lines, respectively, and were selected for further analysis in a pilot of primary tumors and normal samples from the respective malignancies.
Four genes CDO1, DCLK1, ZNF331, and ZSCAN18 were found to be methylated in ≥75% of colorectal cancer samples and simultaneously weakly methylated/ unmethylated in ≥80% of the normal samples. These genes were subjected to quantitative real-time MSP (qMSP). Methylation of at least one of the four genes was observed in 62 of the 65 colorectal cancers analyzed (95% sensitivity), and in two out of the 50 normal mucosa samples (96% specificity), with a combined area under the Receiver Operating Characteristics (ROC) curve (AUC) of 0.976. The only significant association when comparing methylation status with clinicopathological features and tumor phenotype, was observed between ZSCAN18 and microsatellite instability (MSI), indicating that aberrant methylation of the four genes is present in all tumor subtypes independent of age, gender and stage. A patent application covering these markers has been filed, and we are currently collaborating with an industrial partner to develop a non- invasive test for early detection of colorectal cancer based on these and previously published results.
To conclude: CDO1, DCLK1, ZNF331, and ZSCAN18 have been identified as novel promising DNA methylation biomarkers for early detection of colorectal cancer.