Germline rare deleterious variant load alters cancer risk, age of onset and tumor characteristics

Abstract

Recent studies show that rare, deleterious variants (RDVs) in certain genes are critical determinants of heritable cancer risk. To more comprehensively understand RDVs, we performed the largest-to-date germline variant calling analysis in a case-control setting for a multi-cancer association study from whole-exome sequencing data of 20,789 participants, split into discovery and validation cohorts. We confirm and extend known associations between cancer risk and germline RDVs in specific gene-sets, including DNA repair (OR = 1.50; p-value = 8.30e-07; 95% CI: 1.28-1.77), cancer predisposition (OR = 1.51; p-value = 4.58e-08; 95% CI: 1.30-1.75), and somatic cancer drivers (OR = 1.46; p-value = 4.04e-06; 95% CI: 1.24-1.72). Furthermore, personal RDV load in these gene-sets associated with increased risk, younger age of onset, increased M1 macrophages in tumor and, increased tumor mutational burden in specific cancers. Our findings can be used towards identifying high-risk individuals, who can then benefit from increased surveillance, earlier screening, and treatments that exploit their tumor characteristics, improving prognosis.

Fig. 1. Study design.

Flowchart of the study pipeline to identify rare deleterious variants (RDVs) and to perform RDV burden and RDV load analyses.

Fig. 2. Principal component analyses (PCA) of the study cohort and the gated study cohort.

PCA based on common SNPs (MAF ≥ 0.05) showing the top two principal components of (a) the study cohort together with 1000 Genomes and The Ashkenazi Genome Consortium (TAGC) samples and of (b) the gated samples from the study cohort with European ancestry (6371 cases and 6647 controls).

Fig. 3. Gene-set level rare, deleterious variant (RDV) burden in the discovery (blue) and validation (red) cohorts.

The whiskers span the 95% confidence interval for OR values (penalized logistic regression). The black circle outline indicates significant burden p ≤ 0.05.

Fig. 4. Cancer risk based on RDV load.

a Discovery cohort; b Validation cohort. The whiskers span the 95% confidence interval for OR values (penalized logistic regression). The black circle outline indicates significant burden p ≤ 0.05.

Fig. 5. Comparison of age of diagnosis based on germline RDV load in TCGA cases.

a Cancer predisposition genes b DNA damage repair genes c Somatic cancer driver genes d Fanconi Anemia genes. p-values are calculated based on Mann–Whitney U-test. Boxplot elements: center line indicates median; box limits represent lower (25th percentile) and upper (75th percentile) quartiles; whiskers extend to 1.5 times the interquartile range.

Fig. 6. Comparison of M1 macrophages cell fraction in tumor based on germline RDV load in TCGA cases.

a Cancer predisposition genes b DNA damage repair genes c Somatic cancer driver genes d Fanconi Anemia genes. The marker in the violin plot indicates mean.