Mendelian randomization study of C-reactive protein and cancer risk : the Korean genome and epidemiology study

Abstract

Background and Purpose: C-reactive protein (CRP) is one of the acute phase reactive protein that is an indicator for the degree of overall inflammatory response of the body. Previous studies have shown that elevated CRP levels are associated with a variety of chronic diseases such as colorectal cancer, endometrial cancer and ovarian cancer as well as overall risks of general cancer. There are precedent studies that relate CRP levels to risks of cancer occurrence in proportion to and also independently; however, there is lack of research that has yet to reveal their causalities. Therefore, in this study, based on genetic instrumental variable, we tried to confirm the causal relationship between blood CRP levels and cancer through Mendelian Randomization (MR) analysis. Subject and Method: In the Ansan/Anseong cohort data of the Korean Genome and Epidemiology Study (KoGES), subjects with genetic information were analyzed using their epidemiological and gene information. In the first phase of the study, after excluding the subjects who had cancer in the baseline study, the relationship between blood CRP levels and cancer developed during the follow-up was confirmed by using the linear regression analysis. In the second phase, 347,503 single nucleotide polymorphisms (SNPs) and genome wide association studies (GWAS) were performed for CRP traits. After isolating the genomic DNA drawn from the blood of the Ansan/Anseong cohort participants, the genetic data were subject to purification and a total of 347,503 SNPs were obtained from the final 8,840 subjects; the data were then used for analysis. The traits that were used in the analysis were selected from subjects who had never developed cancer in the baseline study; GWAS were performed using their blood CRP levels as traits. Study Results: In the epidemiological analysis, CRP and cancer showed positive correlation when age, sex, and BMI were controlled but it was not statistically significant (OR, 1.004; 95% CI, 0.91-1.10). In addition, there is a total of four SNPs (#1 chromosome; rs7553007, rs3093077, #12 chromosome; rs2393791, rs2259816) that were identified in the blood CRP levels of Korean adults but the results of performing Mendelian Randomization (MR) showed no causal relationship between CRP and cancer (OR, 1.32; 95% CI, 0.65-2.68). Conclusion: In this study, there was no observation of C-reactive protein (CRP) in causal association with cancer. In the future, through the discovery of single nucleotide polymorphisms (SNPs) that are specific to CRP of Korean adults and cancer, and by increasing the number of study subjects through large-scale data, additional studies should be continued.