The overall objective of this study is to identify patterns of gene-specific hypermethylation that are associated with the adenoma-carcinoma sequence in colorectal cancer development. The majority of colorectal cancer arises from adenomatous polyps following the adenoma-carcinoma sequence. However, factors that contribute to the multi-step pathogenesis of colorectal cancer have not been well characterized. In recent years, hypermethylation of tumor-suppressor genes has been found as a common event in cancer cells. A loss of gene function resulted from aberrant methylation can promote cell proliferation and lead to tumorigenesis. Several tumor-suppressor genes are found to be hypermethylated in tissues of patients with colorectal cancer. However, studies that simultaneously examined the methylation patterns in adenomas are limited. No prior studies used a prospective design to examine the association between gene-specific hypermethylation and colorectal cancer risk.
Taking advantage of the availability of prediagnostic blood samples collected from colorectal cancer and adenoma patients and their frequency-matched controls, the first two aims of this study is to determine whether hypermethylation of multiple tumor-suppressor genes at baseline contributes to the subsequent development of colorectal cancer and adenoma among study participants randomized to receive the flexible sigmoidscopy screening in the prostate, lung, colorectal and ovarian (PLCO) screening trial. We will use a nested case-control design to investigate the first aim among 500 colorectal cancer cases and 500 controls and the second aim among 400 adenoma cases and 400 controls. Although exploratory, we will further examine whether gene-specific hypermethylation is associated with a heterogeneous risk of colorectal adenoma/cancer according to the clinicopathological features of the lesion such as the size and villous histology of the adenoma and the location, stage and histology of the cancer.
The causes of methylation abnormalities are poorly understood. Although the impact of age, dietary folate intake and alcohol consumption on methylation has been investigated in colorectal cancer patients, little is known with regard to the determinants of methylation abnormalities in healthy individuals whose methylation status will not be affected by cancer occurrence or treatment allocation. Using a wide breadth of epidemiologic data collected in the PLCO trial, the third aim of this study is to evaluate the association between gene-specific hypermethylation and dietary and behavioral risk factors in 900 controls. Because dietary and behavioral risk factors are modifiable, these investigations may provide valuable information to develop prevention strategies against methylation abnormalities.
To summarize, we plan to investigate the following specific aims:
Specific Aim 1: To investigate whether peripheral blood cell promoter hypermethylation in key tumor-suppressor genes involved in colorectal carcinogenesis is associated with an increased risk of colorectal cancer using a prospective design. Specifically, we will:
Aim 1a: Examine whether baseline promoter hypermethylation of 12 tumor-suppressor genes in various cellular pathways of colorectal cancer development (including APC, CDH1, CDH13, p16INK4a, p14ARF, MLH1, MGMT, DPAK, RASSF1A, THBS1, HLTF and TPEF/HPP1) is associated with a subsequent increased risk of colorectal cancer, among 500 incident colorectal cancer cases and 500 frequency-matched controls.
Aim 1b: Examine whether gene-specific hypermethylation is associated with a heterogeneous risk of colorectal cancer according to the clinicopathologic features of the tumor such as location, stage and histological grade.
Specific Aim 2: To investigate whether peripheral blood cell promoter hypermethylation in key tumor-suppressor genes involved in colorectal carcinogenesis is associated with an increased risk of adenomatous polyps using a prospective design. Specifically, we will:
Aim 2a: Examine whether baseline promoter hypermethylation of 12 tumor-suppressor genes in various cellular pathways of colorectal cancer development (including APC, CDH1, CDH13, p16INK4a, p14ARF, MLH1, MGMT, DPAK, RASSF1A, THBS1, HLTF and TPEF/HPP1) is associated with a subsequent increased risk of colorectal adenomas, among 400 incident colorectal adenoma cases and 400 frequency matched controls;
Aim 2b: Examine whether gene-specific hypermethylation is associated with a heterogeneous risk of colorectal adenoma according to the clinicopathologic features of the adenomas such as size, villous histology and multiplicity.
Specific Aim 3: To assess in healthy controls whether modifiable risk factors such as diet, alcohol and tobacco use, physical activity, the use of non-steroidal anti-inflammatory drugs and the use of hormone replacement therapy among women are associated with hypermethylation in the genes under study.
Overall, the proposed study differs from previous studies in its prospective design and capability of assessing gene-specific hypermethylation of colorectal adenoma, cancer and controls who underwent the same screening procedure from the same source population. Investigations on the specific aims of this study are expected to provide important data regarding the role of gene-specific hypermethylation in multi-step colorectal cancer development, as well as to advance our knowledge on the determinants of methylation abnormalities.

Fang Fang (UNT Health Science Center)
Richard Hayes (NCI, DCEG)
Wen-Yi Huang (NCI, DCEG)
Regina Santella (Columbia University)
Ruth Pfeiffer (NCI, DCEG)
Ying Wei (Columbia University)
Yujin Zhang (Columbia University)