Genomic Risk Factors for Cervical Cancer

Abstract

Cervical cancer is the fourth common cancer amongst women worldwide. Infection by high-risk human papilloma virus is necessary in most cases, but not sufficient to develop invasive cervical cancer. Despite a predicted genetic heritability in the range of other gynaecological cancers, only few genomic susceptibility loci have been identified thus far. Various case-control association studies have found corroborative evidence for several independent risk variants at the 6p21.3 locus (HLA), while many reports of associations with variants outside the HLA region remain to be validated in other cohorts. Here, we review cervical cancer susceptibility variants arising from recent genome-wide association studies and meta-analysis in large cohorts and propose 2q14 (PAX8), 17q12 (GSDMB), and 5p15.33 (CLPTM1L) as consistently replicated non-HLA cervical cancer susceptibility loci. We further discuss the available evidence for these loci, knowledge gaps, future perspectives, and the potential impact of these findings on precision medicine strategies to combat cervical cancer.

Keywords: GWAS; HLA; HPV; SNP; cervical cancer; cervical dysplasia; eQTL; genetic susceptibility; meta-analysis; papillomavirus.

Figure 1

Pathogenesis of cervical cancer. Epithelial cells in the transformation zone of the cervix acquire lesions upon persistent infection with high risk HPV (hrHPV). In some cases, the lesions resolve, whereas in others, upon viral integration, cells are transformed and progress from cervical intraepithelial neoplasia I to II and III (CIN1, CIN2, and CIN3). Viral proteins E6 and E7 are released and inhibit apoptosis mediated by TP53, cell cycle checkpoint by p21, T-cell response by toll-like receptors (TLR), and macrophage activation by cytokines. This leads to an insufficient immune response and viral replication, uncontrolled cell proliferation and genome instability, and further cancer in situ (CIS) or invasive cervical cancer (CC).

Figure 2

GWAS workflow from replication, validation, fine-mapping, and identifying biological mechanisms to clinically relevant outcomes. The various stages of a genome-wide association study, starting from genotyping on custom arrays, imputation on reference genomes, association analysis, and visualisation, followed by replication in an independent cohort, validation genotyping, and meta-analysis. The top loci are then fine-mapped and integrated with bioinformatic annotations before proceeding to functional experiments in relevant cell and tissue types such as promoter and enhancer luciferase assays, ChIP, 3C, 4C, 5C, Hi-C, ChIA-PET, eQTL analysis, and genome editing via the CRISPR/Cas system. Expected outcomes are the identification of relevant genes and pathways affected by the variant, and extraction of polygenic risk scores (PRS), Mendelian randomisation (MR), and genetic correlation with other traits.