Objectives: To investigate the association of genetic polymorphisms of the interleukin-18 (IL-18) pathway to Barrett's esophagus (BE) and esophageal adenocarcinoma (EAC). Most cases of EAC arise in a background of reflux-induced BE. Genetic influences in this pathway are poorly understood. IL-18 is a multifunctional cytokine implicated in anti-tumor immunity. A number of polymorphisms of the IL-18 and IL-18 receptor-accessory protein (IL-18RAP) genes have been reported to alter gene expression and have recently been linked to inflammatory processes and various tumors, but have not heretofore been studied in BE and EAC.
Methods: Two IL-18 promoter polymorphisms -137 G/C and -607 C/A, (rs187238 and rs1946518) and one IL-18RAP polymorphism (rs917997, C/T) were analyzed. Each single-nucleotide polymorphism (SNP) was genotyped in the following groups: EAC, BE, reflux esophagitis (RE), and controls and analyzed for association with disease status.
Results: The IL-18RAP rs917997C allele is strongly associated with a protective effect in BE (P = 0.0002) and EAC (P = 6 × 10(-7)), which approaches genome-wide levels of significance for allele association without incurring significant multiple testing. The CC genotype at IL-18RAP locus rs917997 was associated with a protective effect against esophageal disease (P = 6 × 10(-4), odds ratio (OR) = 0.59, and 95% confidence interval (CI) 0.43-0.80 for BE; and P = 2 × 10(-6), OR = 0.46, and 95% CI 0.34-0.64 for EAC). The genotype frequencies of IL-18-607 C/A were weakly associated with BE (P = 0.02), and this trend was also seen between controls and EAC (P = 0.07). The CC genotype was associated with an increased risk of BE (OR = 1.45, 95% CI 1.07-1.98) and approached significance for EAC (OR = 1.34, 95% CI 0.98-1.82). Allele and genotype frequencies at these loci were not significantly different between the RE group and controls. Although no significant association was observed between the disease groups at the -137 G/C locus, the -137G/-607C haplotype was associated with increased risk of BE (P = 0.006) with haplotype frequencies of 55% in controls and 65% in BE.
Conclusions: These data show a strong association of the IL-18RAP SNP rs917997 locus with BE and EAC and suggestive association of the Barrett's population with the IL-18-607 C/A promoter polymorphism. As both of these SNPs have been demonstrated as expression quantitative trait loci affecting expression of the respective genes, this strongly implicates IL-18 signaling in susceptibility to BE and EAC.