Predictive model for risk of gastric cancer using genetic variants from genome‐wide association studies and high‐evidence meta‐analysis

• Published in: Cancer medicine (Malden, MA) Vol. 9; no. 19; pp. 7310 - 7316
• Main Authors: Qiu, Lixin, Qu, Xiaofei, He, Jing, Cheng, Lei, Zhang, Ruoxin, Sun, Menghong, Yang, Yajun, Wang, Jiucun, Wang, Mengyun, Zhu, Xiaodong, Guo, Weijian
• Format: Journal Article
• Language: English
• Published: United States John Wiley & Sons, Inc 01.10.2020; John Wiley and Sons Inc; Wiley
• Subjects: Biomarkers, Tumor - genetics; Body weight; Cancer Prevention; China - epidemiology; Female; Gastric cancer; Gene-Environment Interaction; Genetic analysis; Genetic diversity; Genetic factors; Genetic Predisposition to Disease; Genetic Testing; Genome-wide association studies; Genome-Wide Association Study; Genomes; Humans; Ions; Male; Medical screening; Meta-analysis; Meta-Analysis as Topic; Middle Aged; Models, Genetic; Original Research; Overweight; Patients; Phenotype; Polymorphism, Single Nucleotide; Population; Prediction models; predictive model; Predictive Value of Tests; Prognosis; Quality control; Regression analysis; Reproducibility of Results; Retrospective Studies; Risk Assessment; Risk Factors; Single-nucleotide polymorphism; Stomach Neoplasms - epidemiology; Stomach Neoplasms - genetics; Stomach Neoplasms - pathology; Stomach Neoplasms - therapy; susceptibility; China; predictive model; susceptibility; gastric cancer; prognosis; genome-wide association study
• ISSN: 2045-7634; 2045-7634
• DOI: 10.1002/cam4.3354

Genome‐wide association studies (GWAS) have identified some single nucleotide polymorphisms (SNPs) associated with the risk of gastric cancer (GCa). However, currently, there is no published predictive model to assess the risk of GCa. In the present study, risk‐associated SNPs derived from GWAS and large meta‐analyses were selected to construct a predictive model to assess the risk of GCa. A total of 1115 GCa cases and 1172 controls from the eastern Chinese population were included. Logistic regression models were used to identify SNPs that correlated with the risk of GCa. A predictive model to assess the risk of GCa was established by receiver operating characteristic curve analysis. Multifactor dimensionality reduction (MDR) and classification and regression tree (CART) were applied to calculate the effect of high‐order gene‐environment interactions on risk of the cancer. A total of 42 SNPs were selected for further analysis. The results revealed that ASH1L rs80142782, PKLR rs3762272, PRKAA1 rs13361707, MUC1 rs4072037, PSCA rs2294008, and PLCE1 rs2274223 polymorphisms were associated with a risk of GCa. The area under curve considering both genetic factors and BMI was 3.10% higher than that of BMI alone. MDR analysis revealed that rs13361707 and rs4072307 variants and BMI had interaction effects on susceptibility to GCa, with the highest predictive accuracy (61.23%) and cross‐validation consistency (100/100). CART analysis also supported this interaction model that non‐overweight status and a six SNP panel could synergistically increase the susceptibility to GCa. The six SNP panel for predicting the risk of GCa may provide new tools for prevention of the cancer based on GWAS and large meta‐analyses derived genetic variants. We identifed a six‐SNPs panel from GWAS and large meta‐analysis for predicting risk of gastric cancer, which may provide new tools for gastric cancer prevention.