Genome-wide association analysis of chronic lymphocytic leukaemia, Hodgkin lymphoma and multiple myeloma identifies pleiotropic risk loci

Abstract

B-cell malignancies (BCM) originate from the same cell of origin, but at different maturation stages and have distinct clinical phenotypes. Although genetic risk variants for individual BCMs have been identified, an agnostic, genome-wide search for shared genetic susceptibility has not been performed. We explored genome-wide association studies of chronic lymphocytic leukaemia (CLL, N = 1,842), Hodgkin lymphoma (HL, N = 1,465) and multiple myeloma (MM, N = 3,790). We identified a novel pleiotropic risk locus at 3q22.2 (NCK1, rs11715604, P = 1.60 × 10^-9) with opposing effects between CLL (P = 1.97 × 10^-8) and HL (P = 3.31 × 10^-3). Eight established non-HLA risk loci showed pleiotropic associations. Within the HLA region, Ser37 + Phe37 in HLA-DRB1 (P = 1.84 × 10^-12) was associated with increased CLL and HL risk (P = 4.68 × 10^-12), and reduced MM risk (P = 1.12 × 10^-2), and Gly70 in HLA-DQB1 (P = 3.15 × 10^-10) showed opposing effects between CLL (P = 3.52 × 10^-3) and HL (P = 3.41 × 10^-9). By integrating eQTL, Hi-C and ChIP-seq data, we show that the pleiotropic risk loci are enriched for B-cell regulatory elements, as well as an over-representation of binding of key B-cell transcription factors. These data identify shared biological pathways influencing the development of CLL, HL and MM. The identification of these risk loci furthers our understanding of the aetiological basis of BCMs.

Figure 1. Manhattan plots (−log₁₀(P)) by chromosome.

Innermost to outermost ring – chronic lymphocytic leukaemia (CLL)-UK1, CLL-UK2, Hodgkin lymphoma (HL)-UK, HL-GER, multiple myeloma (MM)-UK, MM-GER, and ASSET association test. For clarity, only data with P < 1 × 10⁻³ are shown.

Figure 2

(a) Forest plot of the ORs for the association between rs11715604 and BCM. Studies were weighted according to the inverse of the variance of the log of the OR calculated. Horizontal lines: 95% CI. Box: OR point estimate; box area is proportional to the weight of the study. Diamond: overall summary estimate, with CI given by its width. Unbroken vertical line: null value (OR = 1.0). (b) Regional plot of association and recombination rates. −log₁₀(P) (y axis) of the SNPs are shown according to their chromosomal positions (x axis). The sentinel SNP is shown as a large circle. The colour intensity of each symbol reflects the extent of LD with the sentinel SNP: white (r² = 0) through to dark red (r² = 1.0). Genetic recombination rates, estimated from the 1000 Genomes Project, 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 the region of association. The arcs represent Hi-C promoter contacts in GM12878 cells. The colour intensity of each contact reflects the interaction score. The bottom track represents the chromatin-state segmentation track (ChromHMM) for lymphoblastoid cells using data from the HapMap ENCODE Project.

Figure 3. Manhattan plot representation of the step-wise conditional logistic regression of risk of BCM in the HLA region.

(1) Unconditioned test of the HLA region. (2) Results of the HLA region after conditioning on rs9269081. (3) Results of the HLA region after conditioning on rs9269081 and HLA-DPB1:03. (4) Results of the HLA region after conditioning on rs9269081, HLA-DPB1:03 and Ser37 + Phe37 HLA-DRB1. (5) Results of the HLA region after conditioning on rs9269081, HLA-DPB1:03, Ser37 + Phe37 HLA-DRB1 and Gly70 HLADQB-1. The −log₁₀(P) of the combined logistic regression test P-values are plotted against their physical chromosomal position. The broken red line represents the genome-wide level of significance (P < 5 × 10⁻⁸).