A genome wide association study on Newfoundland colorectal cancer patients’ survival outcomes

Abstract Background In this study we performed genome-wide association studies to identify candidate SNPs that may predict the risk of disease outcome in colorectal cancer. Methods Patient cohort consisted of 505 unrelated patients with Caucasian ancestry. Germline DNA samples were genotyped using t...

Full description

Bibliographic Details
Published in:Biomarker Research
Main Authors: Xu, Wei, Xu, Jingxiong, Shestopaloff, Konstantin, Dicks, Elizabeth, Green, Jane, Parfrey, Patrick, Green, Roger, Savas, Sevtap
Format: Article in Journal/Newspaper
Language:English
Published: 2015
Subjects:
Omni
Newfoundland
Online Access:http://hdl.handle.net/1807/86873
https://doi.org/10.1186/s40364-015-0031-6
Description
Summary:Abstract Background In this study we performed genome-wide association studies to identify candidate SNPs that may predict the risk of disease outcome in colorectal cancer. Methods Patient cohort consisted of 505 unrelated patients with Caucasian ancestry. Germline DNA samples were genotyped using the Illumina® human Omni-1quad SNP chip. Associations of SNPs with overall and disease free survivals were examined primarily for 431 patients with microsatellite instability-low (MSI-L) or stable (MSS) colorectal tumors using Cox proportional hazards method adjusting for clinical covariates. Bootstrap method was applied for internal validation of results. As exploratory analyses, association analyses for the colon (n = 334) and rectal (n = 171) cancer patients were also performed. Results As a result, there was no SNP that reached the genomewide significance levels (p < 5x10−8) in any of the analyses. A small number of genetic markers (n = 10) showed nominal associations (p <10−6) for MSS/MSI-L, colon, or rectal cancer patient groups. These markers were located in two non-coding RNA genes or intergenic regions and none were amino acid substituting polymorphisms. Bootstrap analysis for the MSS/MSI-L cohort data suggested the robustness of the observed nominal associations. Conclusions Likely due to small number of patients, our study did not identify an acceptable level of association of SNPs with outcome in MSS/MSI-L, colon, or rectal cancer patients. A number of SNPs with sub-optimal p-values were, however, identified; these loci may be promising and examined in other larger-sized patient cohorts.

Similar Items