Plant & Animal Genomes XIX Conference

January 15-19, 2011
Town & Country Convention Center
San Diego, CA

Predicting Prognosis Of Colorectal Cancer Patients: A Pathway-Based Approach To MicroRNA Genomic Variations

Hushan Yang¹ , Falin Qu² , Ronald E. Myers¹ , Feng Zhou² , Shao-gui Wan¹ , Xiaoying Fu¹ , Zhinan Chen² , Jinliang Xing²

¹  Division of Population Science, Department of Medical Oncology, Kimmel Cancer Center, Thomas Jefferson University, Philadelphia, PA 19107

²  State Key Laboratory of Cancer Biology, Cell Engineering Research Centre & Department of Cell Biology, Fourth Military Medical University, Xi'an, 710032, China

MicroRNAs (MiRNAs) have been significantly associated with the etiology and clinical outcome of many tumors including colorectal cancer (CRC). Genetic variations in miRNA-related genes have been reported to influence the production and functions of miRNAs, and thus affect the risk and outcome of a variety of tumors. However, the associations of these variations with CRC prognosis have not been evaluated. In this study, we analyzed the effects of eight single nucleotide polymorphisms (SNPs) in miRNA genes on the survival of a well-characterized population of 459 Chinese CRC patients. The most significant finding related to rs2043556, a variant in the pre-miRNA region of has-miR-605. Compared to the patients with the homozygous wild-type genotype, those who have the heterogeneous and homozygous variant genotype of rs2043556 exhibited an increased risk of cancer death with a hazard ratio [HR] of 1.47 (95% confidence interval [CI] 0.84-2.60) and 2.96 (95% CI 1.16-7.59), respectively (P for trend=0.02). This association remained significant in patients who received chemotherapy (P for trend=0.04) but not in those without chemotherapy (P for trend=0.34). In addition, rsrs2289030 in pre-miR492 was also significantly associated with CRC death in chemotherapy-treated patients (P for trend=0.03) but not in patients without chemotherapy (P for trend=0.79). Finally, we conducted higher-order gene-gene interaction analysis using survival tree which grouped patients into different risk categories based on the genotype combinations of the 8 miRNA SNPs. We found that compared to the low-risk individuals identified by survival tree, the HR (95% CI) was 4.42 (1.45-13.45) and 35.24 (11.01-172.99) for medium-risk and high-risk individuals, respectively (P for trend=2.86x10-18). Overall, we presented one of the first epidemiologic studies showing that miRNA genetic variations may individually and jointly impact CRC prognosis.