Evaluating the predictive value of genetic risk score in colorectal cancer among Chinese Han population

Ding Ye,Danjie Jiang,Simeng Gu,Yingying Mao,Sangni Qian,Shujuan Lin,Qilong Li,Jinhua Yang,Kunhong Zhong,Mingjuan Jin,Kun Chen
DOI: https://doi.org/10.1038/s10038-019-0703-4
2019-01-01
Journal of Human Genetics
Abstract:Increasing single nucleotide polymorphisms (SNPs) have been identified to be associated with colorectal cancer (CRC). We aimed to investigate whether genetic risk scores (GRS) that aggregate information from multiple genetic variants can predict the risk of CRC in a Chinese population. Fifty candidate SNPs were selected to explore the associations with CRC in a discovery sample with 1002 CRC cases and 999 healthy controls. We modeled the significant SNPs identified by the case–control study as a multilocus weighted GRS and estimated the association of GRS with CRC. Furthermore, 300 pairs of cases and controls were included as a validation sample to confirm the finding. Area under the receiver operating characteristic curve (AUROC) was used to evaluate the predictive power of GRS in CRC. A total of seven SNPs were found to increase the risk of CRC, and two SNPs were found to be negatively associated with CRC in the discovery sample. Relative to participants with the lowest quartile of GRS, those with the highest quartile had a 2.64-fold (95% CI: 1.99–3.51) higher risk for CRC. For every 0.1 point of GRS increase, the risk of CRC increase by 11% (95% CI: 8–14%). AUROC for GRS alone were 0.59 (95% CI: 0.57–0.62) and 0.52 (95% CI: 0.46–0.58) in the discovery and validation sample, respectively. AUROC increased to 0.62 (95% CI: 0.59–0.64) and 0.71 (95% CI: 0.65–0.76) by combining environmental risk factors. Our findings support an association between GRS and risk of CRC, which provides evidence of improved prediction model for CRC in China.
What problem does this paper attempt to address?