Genetic susceptibility to hepatocellular carcinoma in chromosome 22q13.31, findings of a genome‐wide association study
Zhanwei Wang, Anuradha Budhu, Yi Shen, Linda L. Wong, Brenda Y. Hernandez, Maarit Tiirikainen, Xiaomei Ma, Melinda L. Irwin, Lingeng Lu, Hongyu Zhao, Joseph K. Lim, Tamar H. Taddei, Lopa Mishra, Karen Pawlish, Antoinette M. Stroup, Robert S. Brown, Mindie H. Nguyen, Jill Koshiol, Maria O. Hernandez, Marshonna Forgues, Hwai‐I Yang, Mei‐Hsuan Lee, Yu‐Han Huang, Motoki Iwasaki, Atsushi Goto, Shiori Suzuki, Koichi Matsuda, Chizu Tanikawa, Yoichiro Kamatani, Dean L. Mann, Maria Guarnera, Kirti Shetty, Claire E. Thomas, Jian‐Min Yuan, Chiea Chuen Khor, Woon‐Puay Koh, Harvey Risch, Xin Wei Wang, Herbert Yu
Abstract
Abstract Background and Aim Chronic hepatitis C virus (HCV) infection, long‐term alcohol use, cigarette smoking, and obesity are the major risk factors for hepatocellular carcinoma (HCC) in the United States, but the disease risk varies substantially among individuals with these factors, suggesting host susceptibility to and gene–environment interactions in HCC. To address genetic susceptibility to HCC, we conducted a genome‐wide association study (GWAS). Methods Two case‐control studies on HCC were conducted in the United States. DNA samples were genotyped using the Illumian microarray chip with over 710 000 single nucleotide polymorphisms (SNPs). We compared these SNPs between 705 HCC cases and 1455 population controls for their associations with HCC and verified our findings in additional studies. Results In this GWAS, we found that two SNPs were associated with HCC at P < 5E‐8 and six SNPs at P < 5E‐6 after adjusting for age, sex, and the top three principal components (PCs). Five of the SNPs in chromosome 22q13.31, three in PNPLA3 (rs2281135, rs2896019, and rs4823173) and two in SAMM50 (rs3761472, rs3827385), were replicated in a small US case‐control study and a cohort study in Singapore. The associations remained significant after adjusting for body mass index and HCV infection. Meta‐analysis of multiple datasets indicated that these SNPs were significantly associated with HCC. Conclusions SNPs in PNPLA3 and SAMM50 are known risk loci for nonalcoholic fatty liver disease (NAFLD) and are suspected to be associated with HCC. Our GWAS demonstrated the associations of these SNPs with HCC in a US population. Biological mechanisms underlying the relationship remain to be elucidated.