Kamatani Y, Wattanapokayakit S, Ochi H, Kawaguchi T, Takahashi A, Hosono N, et al. A genome-wide association study identifies variants in the HLA-DP locus associated with chronic hepatitis B in Asians. Nat Genet 2009;41:591–595. www.nature.com (Reprinted with permission.)
Chronic hepatitis B is a serious infectious liver disease that often progresses to liver cirrhosis and hepatocellular carcinoma; however, clinical outcomes after viral exposure vary enormously among individuals. Through a two-stage genome-wide association study using 786 Japanese chronic hepatitis B cases and 2201 controls, we identified a significant association of chronic hepatitis B with 11 SNPs in a region including HLA-DPA1 and HLA-DPB1. We validated these associations by genotyping two SNPs from the region in three additional Japanese and Thai cohorts consisting of 1300 cases and 2100 controls (combined P = 6.34 × 10−39 and 2.31 × 10−38, OR = 0.57 and 0.56, respectively). Subsequent analyses revealed risk haplotypes (HLA-DPA1*0103-DPB1*0501 and HLA-DPA1*0202- DPB1*0301, OR = 1.45 and 2.31, respectively) and protective haplotypes (HLA-DPA1*0103-DPB1*0402 and HLA-DPA1*0103-DPB1*0401, OR = 0.52 and 0.57, respectively). Our findings show that genetic variants in the HLA-DP locus are strongly associated with risk of persistent infection with hepatitis B virus.
Infection with hepatitis B virus (HBV) is a serious global public health issue with particular importance in the Asian regions. Chronic hepatitis B infection leads to liver cirrhosis and liver failure in more than a quarter of patients infected and accounts for more than two-thirds of hepatocellular carcinoma, the third most common cancer globally. With more than 400 million chronically infected individuals worldwide,1 the health burden from complications arising directly from chronic hepatitis B infection is enormous.
The natural history of HBV is complex, with a great need for clinical and genetic markers to aid in prediction of those who are most at risk of chronic hepatitis B infection, liver cirrhosis, and hepatocellular carcinoma. Inability to achieve viral immune clearance, with persistent high levels of HBV viremia, has been shown in multiple studies to be associated with poor clinical outcomes, including hepatocellular carcinoma.2, 3 Evidence demonstrates that individuals who clear HBV infection have a low risk of adverse outcomes.4
Epidemiological risk factors associated with failure to clear virus include age at infection, sex, chronic alcohol abuse,5 and coinfection with other hepatitis viruses.6 Other risks associated with development of cirrhosis or hepatocellular carcinoma include HBV genotype C and F infection, HBV DNA levels >2000 IU/mL, and basal core promoter mutations.7
To date, several genetic variants have been found to have an association with chronic HBV infection, including interferon-γ, tumor necrosis factor,8 vitamin D receptor,9 estrogen receptor 1,10 and several human leukocyte antigen (HLA) loci.11–14 However, some studies have yielded conflicting results. Understanding more about factors predisposing to chronic infection and inability to achieve immune clearance are vital to guide clinicians in the decisions of if and when to start antiviral therapy.
To discover genetic factors predisposing to the development of chronic hepatitis B virus infection, Kamatani et al.15 carried out an elegant two-stage association study for chronic hepatitis B using genomewide single-nucleotide polymorphisms (SNPs) as genetic markers.
In the first stage, 179 Japanese individuals with chronic hepatitis B and 934 controls were genotyped using Illumina HumanHap550 BeadChip. For the second stage, they selected 12,000 SNPs that had the lowest P values based on minimum P value for three genetic models (allelic, dominant, and recessive). Kamatani et al. then analyzed these subselected SNPs in an independent set of 607 cases and 1267 controls. This revealed 11 SNPs that were significantly associated with chronic hepatitis B (P = 3.62 × 10−8 ∼ 1.16 × 10−13) after Bonferroni correction. These SNPs were all located within or in close proximity to genes encoding HLA-DPA1 and HLA-DPB1 loci. The researchers proved statistically using the Cochrane-Armitage test for all 11 SNPs that there was a low false-positive association rate with this method (genetic inflation factor lambda = 1.06). An age-adjusted and sex-adjusted logistic regression was also performed, which confirmed the HLA-DP association was independent of other variables.
Kamatani et al. then validated these findings through replication analyses performed on three independent cohorts. Two SNPs with the strongest association from the HLA-DP locus were selected for the analysis: rs3077 on HLA-DPA1 and rs9277535 on HLA-DPB1. In both cohorts—an age-matched, sex-matched, and alcohol consumption–matched cohort from Biobank, Japan (274 cases, 274 controls) and another independent Japanese cohort (718 cases, 1280 controls)—each SNP was shown to be significantly associated with chronic hepatitis B infection (P = 1.06 × 10−16 ∼ 1.96 × 10−6). A further Thai cohort (308 cases, 546 controls) also confirmed significance of association for the two loci (rs3077, P = 6.53 × 10−6; rs9277535, P = 6.52 × 10−8). A final meta-analysis was performed to combine results of these studies using a fixed-effects model, Mantel-Haenszel method. Odds ratios (ORs) were highly similar across the four cohort studies and no heterogeneity was observed, illustrating strong concordance between studies for the SNP association results. Mantel-Haenszel P values for independence were 2.31 × 10−38 for rs3077 (OR: 0.56, 95% confidence interval [CI] = 0.51–0.61) and 6.34 × 10−39 for rs9277535 (OR = 0.57, 95% CI = 0.52–0.62).
The 11 SNPs were all located within a 50-kilobase region spanning HLA-DPA1 and HLA-DPB1. Knowing the high degree of polymorphism in HLA molecules, particularly at antigen-binding sites such as exon 2 in HLA-DP, Kamatani et al. went on to genotype HLA-DPA1 and HLA-DPB1 alleles by direct sequencing of exon 2. Haplotypes were inferred based on the 11 SNPs and nucleotide variants in exon 2. They found a significant association of chronic hepatitis B with HLA-DPA1*0103, HLA-DPA1*0202, HLA-DPB1*0402, and HLA-DPB1*0501 (P = 2.93 × 10−11, 4.45 × 10−8, 2.27 × 10−7, and 6.98 × 10−7, respectively). These areas were confirmed to have strong linkage disequilibrium with HLA-DPA1 and HLA-DPB1 and weak linkage disequilibrium with other HLA loci, suggesting that the associations were unlikely to be the result of chance chromosomal realignments.
Case-control studies using samples from first-stage and second-stage cohorts revealed two haplotypes conferring susceptibility to chronic hepatitis B infection: DPA1*0202-DPB1*0501 (P = 5.79 × 10−6, OR = 1.45, 95% CI = 1.16-1.81) and DPA1*0202-DPB1*0301 (P = 0.002, OR = 2.31, 95% CI = 1.39-3.84). Two protective haplotypes were also discovered: DPA1*0103-DPB1*0402 (P = 6.00 × 10−8, OR = 0.52, 95% CI = 0.35–0.75) and DPA1*0103-DPB1*0401 (P = 0.002, OR = 0.57, 95% CI = 0.33–0.96). Based on consistency of results within multiple cohorts, Kamatani et al. concluded that the HLA-DP locus is strongly associated with the risk of chronic hepatitis B infection.
The undeniable strength of this study is the identification of associated genetic polymorphisms within one large population and subsequent validation within three further large independent cohorts. The results of these four cohort studies were then combined and analyzed collectively, which allowed identification of two protective and two susceptibility haplotypes through nucleotide sequencing. The power and P value calculations confirm the high statistical significance of this study.
A common flaw of genetic-based variant association studies is the failure to confirm findings within further independent populations, limiting the broader applicability of results. Kamatani et al. have demonstrated replicability of their SNP detection method within several Asian populations, adding significant confidence to their conclusion.
HLA-DPA1 and HLA-DPB1 encode alpha and beta chains, respectively, of HLA-DP, a class II major histocompatibility complex molecule expressed on the cell surface that presents antigens to CD4+ T lymphocytes. HLA-DP has a structure similar to other HLA class II molecules, but their precise role in immune responses has not been fully elucidated. However, a study by Fontenot et al. did identify a role for HLA-DP in the immune response to berylliosis.16
HLA molecules show considerable polymorphic diversity, particularly in regions encoding antigen-presenting sites, such as exon 2 in HLA-DP. It is likely that such polymorphisms encode nucleotide variation in antigen-binding sites, affecting HBV antigen presentation by cells and the ensuing immune response. Genetic polymorphisms of other HLA molecules such as HLA-DR13 have previously been demonstrated to be significantly associated with chronicity of hepatitis B infection.10, 11
The clinical significance of the research of Kamatani et al. is two-fold. First, the novel association of HLA-DP loci with chronic hepatitis B suggests an important role for this molecule in HBV infection. HLA-DP–dependent antigen presentation and subsequent immune stimulation may in part determine the clinical course of HBV infection. This may provide new insight into the host immune response to hepatitis B infection and offer new potential targets for therapeutic agents.
Second, once greater understanding of the role of HLA-DP in chronic hepatitis B infection is developed, the potential for clinical risk stratification of patients through HLA testing may help guide therapeutic decision-making. Identifying those at greatest risk of chronic hepatitis B infection will assist decisions of when to start treatment rather than watching to see whether immune clearance will be achieved spontaneously. Targeting high-risk individuals for earlier treatment may reduce the duration of high-level viremia and attendant risks of covalently closed circular DNA incorporation and development of hepatocellular carcinoma.17–20
Further research is needed to fully elucidate the role of HLA-DP in hepatitis B infection. Prospective studies to determine clinical outcomes associated with these genetic loci are paramount. The mechanism of association between these polymorphisms and chronic hepatitis B infection remains unclear. Reduced vertical transmission of virus, improved clearance of virus after inoculation, or association with other factors influencing viral exposure are possibilities requiring further exploration.