Ovarian cancer risk associated with inherited inflammation-related variants

Abstract

The importance of inflammation pathways to the development of many human cancers prompted us to examine the associations between single-nucleotide polymorphisms (SNP) in inflammation-related genes and risk of ovarian cancer. In a multisite case-control study, we genotyped SNPs in a large panel of inflammatory genes in 930 epithelial ovarian cancer cases and 1,037 controls using a custom array and analyzed by logistic regression. SNPs with P < 0.10 were evaluated among 3,143 cases and 2,102 controls from the Follow-up of Ovarian Cancer Genetic Association and Interaction Studies (FOCI) post-GWAS collaboration. Combined analysis revealed association with SNPs rs17561 and rs4848300 in the interleukin gene IL1A which varied by histologic subtype (P(heterogeneity) = 0.03). For example, IL1A rs17561, which correlates with numerous inflammatory phenotypes, was associated with decreased risk of clear cell, mucinous, and endometrioid subtype, but not with the most common serous subtype. Genotype at rs1864414 in the arachidonate 5-lipoxygenase ALOX5 was also associated with decreased risk. Thus, inherited variation in IL1A and ALOX5 seems to affect ovarian cancer risk which, for IL1A, is limited to rarer subtypes. Given the importance of inflammation in tumorigenesis and growing evidence of subtype-specific features in ovarian cancer, functional investigations will be important to help clarify the importance of inherited variation related to inflammation in ovarian carcinogenesis.

Figure 1. Per-allele Risks of Ovarian Cancer by Study Site

Odds ratios and 95% confidence intervals adjusted for study site (where combined) based on analysis of self-reported whites only are plotted; N represents cases and controls combines; p-het based on heterogeneity test of discovery versus replication participants; Mayo Clinic (MAY) and North Carolina Ovarian Cancer Study (NCO); Brigham and Women’s Hospital (BWH); NCI Ovarian Case-Control Study in Poland (POL); Tampa Bay Ovarian Cancer Study (TBO); the Familial Ovarian Tumour Study (TOR); the UK Ovarian Cancer Population Study, the UK Studies of Epidemiology and Risk Factors in Cancer Heredity Ovarian Cancer Study, the UK Familial Ovarian Cancer Registry, and the Wellcome Trust Case Control Consortium (UK).

Figure 2. Per-allele Risks of Ovarian Cancer by Histological Subtype, IL1A rs17561

Odds ratios and 95% confidence intervals adjusted for study site based on analysis of self-reported whites only are plotted (N cases by subtype are displayed, N=3.043 controls); p-het based on heterogeneity test using polytomous regression.