A genome-wide association study of serum levels of prostate-specific antigen in the Japanese population

• Published in: Journal of medical genetics Vol. 51; no. 8; pp. 530 - 536
• Main Authors: Terao, Chikashi, Terada, Naoki, Matsuo, Keitaro, Kawaguchi, Takahisa, Yoshimura, Koji, Hayashi, Norio, Shimizu, Masakazu, Soga, Norihito, Takahashi, Meiko, Kotoura, Yoshihiko, Yamada, Ryo, Ogawa, Osamu, Matsuda, Fumihiko
• Format: Journal Article
• Language: English
• Published: England BMJ Publishing Group LTD 01.08.2014
• Subjects: Age; Antigens; Arrays; Asian Continental Ancestry Group - genetics; Drug dosages; Ethnicity; Genome-Wide Association Study; Genomes; Hospitals; Humans; Japan; Life expectancy; Male; Polymorphism, Single Nucleotide - genetics; Population; Principal components analysis; Prospective Studies; Prostate cancer; Prostate-Specific Antigen - blood; Prostate-Specific Antigen - genetics; Questionnaires; ROC Curve; Studies; Urology; Japan; Screening; Genome-wide; Cancer: prostate; Epidemiology
• ISSN: 0022-2593; 1468-6244; 1468-6244
• DOI: 10.1136/jmedgenet-2014-102423

Background Prostate-specific antigen (PSA) is a useful marker for prostate cancer (PCa) and is widely used for screening of PCa. Previous studies have shown that genetic components influence the levels of PSA, and some of these genetic components would lead to better diagnostic sensitivity and specificity to PCa. However, genetic studies for PSA from Asian countries are limited. Our aim was to identify genetic components influencing PSA levels in the Japanese population using genome-wide association study (GWAS) and to analyse whether genetic components would lead to better screening abilities of PCa. Methods We performed a GWAS comprising 1086 male subjects using 303 283 single nucleotide proteins, followed by a replication study of 1302 subjects. PSA levels were quantified by chemiluminescence immunoassay method. Quantitative linear regression analysis was performed to assess genetic components of PSA levels. A total of 413 subjects with prostate biopsies were analysed to examine whether genetic determinants would improve diagnostic ability. Results Rs16856139 in SLC45A3, the same region as the previous Chinese study, showed an overall significant association with PSA levels (p=2.4×10−11) along with rs1058205 in KLK3. In silico analysis revealed significant association between rs16856139 and expression of SLC45A3. Genetic scores of PSA showed a dose-dependent decrease of area under curve (AUC) of PCa and successfully subgrouped the individuals with significantly different AUC (p≤0.0097). Conclusions Rs16856139, associated with the expression of SLC45A3, is significantly associated with the levels of PSA in the Japanese population. Classification of subjects based on PSA genetic determinants would improve screening ability of PSA to detect PCa.