A comprehensive analysis of germline predisposition to early-onset ovarian cancer

Abstract

The subset of ovarian cancer (OC) diagnosed ≤ 30yo represents a distinct subgroup exhibiting disparities from late-onset OC in many aspects, including indefinite germline cancer predisposition. We performed DNA/RNA-WES with HLA-typing, PRS assessment and survival analysis in 123 early-onset OC-patients compared to histology/stage-matched late-onset and unselected OC-patients, and population-matched controls. Only 6/123(4.9%) early-onset OC-patients carried a germline pathogenic variant (GPV) in high-penetrance OC-predisposition genes. Nevertheless, our comprehensive germline analysis of early-onset OC-patients revealed two divergent trajectories of potential germline susceptibility. Firstly, overrepresentation analysis highlighted a connection to breast cancer (BC) that was supported by the CHEK2 GPV enrichment in early-onset OC(p = 1.2 × 10^-4), and the presumably BC-specific PRS₃₁₃, which successfully stratified early-onset OC-patients from controls(p = 0.03). The second avenue pointed towards the impaired immune response, indicated by LY75-CD302 GPV(p = 8.3 × 10^-4) and diminished HLA diversity compared with controls(p = 3 × 10^-7). Furthermore, we found a significantly higher overall GPV burden in early-onset OC-patients compared to controls(p = 3.8 × 10^-4). The genetic predisposition to early-onset OC appears to be a heterogeneous and complex process that goes beyond the traditional Mendelian monogenic understanding of hereditary cancer predisposition, with a significant role of the immune system. We speculate that rather a cumulative overall GPV burden than specific GPV may potentially increase OC risk, concomitantly with reduced HLA diversity.

Keywords: Early-onset; Germline whole exome sequencing; HLA; Mutation burden; Ovarian cancer; Polygenic risk score.

Figure 1

Overall germline pathogenic variant (GPV) burden in early-onset OC patients and super-controls (non-cancer individuals without family cancer history). BTO, Borderline Tumors of Ovary.

Figure 2

Association of germline pathogenic variant (GPV) in ovarian cancer (OC; A) and hereditary breast/ovarian/pancreatic (HBOP; B) cancer predisposition genes with multiple primary tumors in early-onset OC patients, and with family cancer history (C), and association of positive personal cancer history beyond OC with hematological malignancies in family cancer history (D). CI, Confidence Interval; NS, not significant; OR, Odds Ratio.

Figure 3

Frequency of Human Leukocyte Antigen (HLA) loci homozygotes in early-onset ovarian cancer (OC) patients compared to female super-controls. *p < 0.05; **p < 0.001;***p < 10^–5.

Figure 4

Comparison of polygenic risk scores (PRS) distribution between subgroups of ovarian cancer (OC) patients and PRS controls. Only SNP sets significantly stratifying at least one subgroup of OC patients from PRS controls are shown; (A) early-onset OC patients, (B) pooled early-onset and histology/stage-matched OC patients, (C) late-onset high-grade serous OC (HGSC) patients. Details are provided in Supplementary Table S5. NS, not significant.

Figure 5

Survival analysis of ovarian cancer (OC) patients. Kaplan–Meier survival curves of (A) early-onset and late-onset OC patients; (B) Low-grade serous (LGSC) early- and late-onset OC. NS, not significant.