Abstract
Both Hodgkin lymphoma (HL) and asthma are associated with relative isolation from early childhood infectious exposures (hygiene hypothesis) and a T-helper-2 (Th2)-skewed immune response. To test the hypothesis that there is some component of shared genetic etiology between HL and asthma, we examined the genetic overlap in data from genome-wide association study (GWAS) meta-analyses of the two diseases conducted in populations of European ancestry. The HL GWAS meta-analysis consisted of 1,816 HL cases and 7,877 controls, and the asthma GWAS meta-analysis consisted of 2,088 asthma cases and 2,743 controls. The combined HL-asthma GWAS data resulted in 904,634 common single nucleotide polymorphisms (SNPs) genotyped with both arrays. We observed a total of 9 common SNPs associated with both HL and asthma at p<0.0001, compared to 3 SNPs expected in both data sets at that significance level. Out of 66 genetic risk variants associated with HL at a genome-wide significance level of p<10-8, 5 variants replicated in the asthma GWAS dataset at p < 0.05. In a meta-analysis combining the HL and asthma GWAS data, we found genome-wide significant associations with two correlated SNPs (r2 =0.91) in the Th2 transcription factor gene GATA3 (rs422628, ORHL-asthma meta =1.25, p=3.36 x 10-9 and rs444929, ORHL-asthma meta = 1.26, p=2.09 x 10-8) that were not genome-wide significant in either disease alone (Table 1). The association with one SNP remained genome-wide significant in the meta-analysis combined with the nodular sclerosis subset (rs422628, ORHL-asthma meta =1.32, p=3.75 x 10-9), but not with other HL subtypes. An association with a SNP in IKZF3, a gene involved in B lymphocyte differentiation and proliferation, reached genome-wide significance only in the meta-analysis with the HL subset positive for Epstein-Barr virus (EBV) in the tumor (rs9909593, ORHL-asthma meta = 1.19, p=3.80 x 10-8). In a genetic diseasome analysis based on an ontological analysis of published GWAS data, HL, especially nodular sclerosis HL, was more closely related to asthma than to solid cancers. Recognition of overlap in genetic predisposition to HL and other immune diseases sheds light on the complex etiology of HL and may enable novel diagnostic and therapeutic approaches.
. | HL GWAS1 . | Asthma GWAS1 . | Combined GWAS1 . | ||||||
---|---|---|---|---|---|---|---|---|---|
SNP2 . | Chr3 . | BP4 . | Gene . | OR5 . | P-value6 . | OR5 . | P-value6 . | OR5 . | P-value6 . |
rs422628a | 10 | 8151415 | GATA3 | 1.24 | 1.13 x 10-5 | 1.31 | 6.39 x10-5 | 1.25 | 3.26 x 10-9 |
rs444929b | 10 | 8150030 | GATA3 | 1.26 | 3.24x 10-6 | 1.28 | 8.90 x 10-4 | 1.26 | 2.09 x 10-8 |
rs422628c | 10 | 8151415 | GATA3 | 1.33 | 7.77 x 10-6 | 1.31 | 6.39 x10-5 | 1.32 | 3.75 x 10-9 |
rs9909593d | 17 | 35223675 | IKZF3 | 1.11 | 2.60 x 10-1 | 1.25 | 5.06 x 10-8 | 1.19 | 3.80 x 10-8 |
. | HL GWAS1 . | Asthma GWAS1 . | Combined GWAS1 . | ||||||
---|---|---|---|---|---|---|---|---|---|
SNP2 . | Chr3 . | BP4 . | Gene . | OR5 . | P-value6 . | OR5 . | P-value6 . | OR5 . | P-value6 . |
rs422628a | 10 | 8151415 | GATA3 | 1.24 | 1.13 x 10-5 | 1.31 | 6.39 x10-5 | 1.25 | 3.26 x 10-9 |
rs444929b | 10 | 8150030 | GATA3 | 1.26 | 3.24x 10-6 | 1.28 | 8.90 x 10-4 | 1.26 | 2.09 x 10-8 |
rs422628c | 10 | 8151415 | GATA3 | 1.33 | 7.77 x 10-6 | 1.31 | 6.39 x10-5 | 1.32 | 3.75 x 10-9 |
rs9909593d | 17 | 35223675 | IKZF3 | 1.11 | 2.60 x 10-1 | 1.25 | 5.06 x 10-8 | 1.19 | 3.80 x 10-8 |
1Genome-wide association scan
2Single nucleotide polymorphism
3Chromosome
4Base pair position
5Odds ratio
6P-values are derived from a meta-analysis using a fixed effects model with weights proportional to the square root of the number of cases.
aAll Hodgkin lymphoma cases
bAll Hodgkin lymphoma cases, SNP association reported in Cozen et al., Nat Comm, 2014
cNodular sclerosis cases only
dEBV-positive cases only
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.