Rare variants of large effect in BRCA2 and CHEK2 affect risk of lung cancer

Abstract

We conducted imputation to the 1000 Genomes Project of four genome-wide association studies of lung cancer in populations of European ancestry (11,348 cases and 15,861 controls) and genotyped an additional 10,246 cases and 38,295 controls for follow-up. We identified large-effect genome-wide associations for squamous lung cancer with the rare variants BRCA2 p.Lys3326X (rs11571833, odds ratio (OR) = 2.47, P = 4.74 × 10(-20)) and CHEK2 p.Ile157Thr (rs17879961, OR = 0.38, P = 1.27 × 10(-13)). We also showed an association between common variation at 3q28 (TP63, rs13314271, OR = 1.13, P = 7.22 × 10(-10)) and lung adenocarcinoma that had been previously reported only in Asians. These findings provide further evidence for inherited genetic susceptibility to lung cancer and its biological basis. Additionally, our analysis demonstrates that imputation can identify rare disease-causing variants with substantive effects on cancer risk from preexisting genome-wide association study data.

Figure 1. Genome-wide P-values (−log₁₀P, y axis) plotted against their respective chromosomal positions (x axis)

(a) All lung cancer, (b) AD and (c) SQ. Shown are the genomewide P-values (two-sided) obtained using the Cochran-Armitage trend test from analysis of 8.9 million successfully imputed autosomal SNPs in 11,348 cases and 15,861 controls from discovery phase. The red and blue horizontal lines represent the significance threshold of P=5.0×10⁻⁸ and P=5.0×10⁻⁶ respectively. Any region contains at least one association signal better than P=5.0×10⁻⁶ were selected for the in silico replication.

Figure 2. Plot of the odds ratios of lung cancer associated with 13q13.1 (rs11571833 and rs56084662), 22q12.1 (rs17879961) and 3q28 (rs13314271) risk loci (a-l)

All lung cancer based on 21,594 lung cancer cases and 54,156 controls (a-d), SQ based on 6,477 SQ and 53,333 controls (e-h) and AD based on 7,031 AD and 53,189 controls (i-l). Studies are weighted according to the inverse of the variance of the log of the OR calculated by unconditional logistic regression. Horizontal lines: 95% confidence intervals (95% CI). Box: OR point estimate; its area is proportional to the weight of the study. Diamond (and broken line): overall summary estimate, with confidence interval given by its width. Unbroken vertical line: at the null value (OR = 1.0).

Figure 3. Regional plots of association results and recombination rates for the 13q13.1 in SQ (a), 22q12.1 in SQ (b) and 3q28 susceptibility loci in AD (c)

SQ related panels (a, b) were based on 3,275 SQ and 15,038 controls from discovery phase; and AD related panel (c) was based on 3,442 AD and 14,894 controls from discovery phase. Association results of both genotyped (circles) and imputed (diamonds) SNPs in the GWAS samples and recombination rates for each locus: For each plot, −log₁₀P values (y axis) of the SNPs are shown according to their chromosomal positions (x axis). The top genotyped SNP in each combined analysis is a large diamond and is labeled by its rsID. The color intensity of each symbol reflects the extent of LD with the top genotyped SNP: white (r²=0) through to dark red (r²=1.0). Genetic recombination rates (cM/Mb), estimated using HapMap CEU samples, are shown with a light blue line. Physical positions are based on NCBI build 37 of the human genome. Also shown are the relative positions of genes and transcripts mapping to each region of association. Genes have been redrawn to show the relative positions; therefore, maps are not to physical scale.