THE IMPACT OF HLA CLASS I ON VIROLOGICAL OUTCOMES OF HBV Philippa Matthews Consultant in Infectious Diseases & Microbiology SUPPRESSION CO-EVOLUTION ESCAPE
Host factors associated with the clinical course of HBV infection Matsuura et al., J. Med Virol., 2015 HLA Class II associations with HBV outcome HBV persistence HLA-DPA1 (rs3077) HLA-DPB1 (rs9277535) HLA-DPA1*02:02-DPB1*05:01 HLA-DPA1*02:02-DPB1*03:01 HLA-DPB1*09:01 HLA-DQ (rs2856718 / rs7453920) HLA-DR*03 HLA-DR*07 HBV clearance HLA-DPA1*01:03-DPB1*04:02 HLA-DPA1*01:03-DPB1*04:01 HLA-DPB1*02:01 HLA-DR*04 HLA-DR*13 Clear effect of HLA Class II (especially DP) with persistent infection, with disease progression and carcinogenesis, and with response to vaccine. Less certain in non-asian cohorts. Chang et al., PLoS One, 2014 Kamatani et al., Nat. Genet., 2009 Matsuura et al., J Med Virol., 2015 Nishida et al., PLoS One, 2012 Yan et al., World J Gastroenterol., 2012
HLA Class I associations with HBV outcome Class I alleles associations with HBV persistence (e.g. HLA-A*33:03) or clearance (e.g. HLA-B*13:01, HLA-C*08:01) (Miao et al, Clin Dev Immunol., 2013; Tan et al., J. Virol., 2014) Polymorphism close to HLA-C locus associated with CHB by GWAS (Hu et al., Nat. Genet., 2013) Down-regulation of CD8 expression in CHB (Ouyang et al., Mol Immuno., 2013) HBV adaptation to T cell selection pressure (Kefalakes et al., Hepatology, 2015; Desmond et al., J. Virol, 2012). CD8+ T cell epitopes proposed as HBV vaccine antigens (Comber et al., Hep Res & Treat, 2015) HBV sequence variability: core protein Sequences downloaded from https://hbvdb.ibcp.fr/ Epitopes from Kefalakes et al., Hepatology, 2015; Comber et al., Hep Res & Treat, 2015
Hypothesis HLA Class I genotype has an impact on virological (and therefore clinical) outcomes of CHB in African populations. E.g. HBsAg clearance (0.1-2.3% / yr) HBeAg clearance HBV DNA suppression Rationale Immunotherapy for HBV that seeks to control and/or cure HBV replication is now a real possibility Prof Ellie Barnes Review: Therapeutic Vaccines in HBV Med Microbiol Immunol (2015) 204:79 86
>2 billion exposed / cleared; 350 million chronic infections 0.5 million new cases of HCC / year Paucity of data from Africa
Study cohort: HIV+ adults Thames Valley Cohort Gaborone, Botswana Kimberley, South Africa Durban, South Africa N=1,114
Predicting the impact of HLA on HBV 1,114 HIV+ adults from southern Africa (a) HLA Class I (b) HBV coinfection (c) HIV VL and HBV serum markers To what extent does (a) predict (b) & (c)? HIV RNA viral load HBsAg ±HBeAg ±HBV DNA Markers of HBV disease in an extended African HIV+ cohort Matthews et al., PLoS One, 2014
Markers of HBV disease in an extended African HIV+ cohort No cases of HDV co-infection Matthews et al., PLoS One, 2014 Predicting the impact of HLA Receiver operator characteristic (ROC) curve True positive rate 1.0 0.5 0 0 0.5 1.0 False positive rate Brilliant test Good test Reasonable test Useless test (chance alone) Numerically represented by Area Under the Curve (AUC)
ROC curves for HLA Class I as a predictor of disease outcome a"hiv$viraemic$control$ A control p=5.1x10-4 n=932 True positive rate HLA & cohort AUC=0.67 Cohort only AUC=0.59 (Proof of principle) Matthews et al., in review JID ROC curves for HLA Class I as a predictor of disease outcome a"hiv$viraemic$control$ A control p=5.1x10-4 n=932 b"hbsag$status$ B status p=0.21 n=1012 True positive rate HLA & cohort AUC=0.67 Cohort only AUC=0.59 HLA & cohort AUC=0.59 Cohort only AUC=0.57 (Proof of principle) Matthews et al., in review JID
ROC curves for HLA Class I as a predictor of disease outcome a"hiv$viraemic$control$ A control b"hbsag$status$ B status c"hbeag$status$ C p=5.1x10-4 n=932 p=0.21 n=1012 p=0.03 n=58 True positive rate HLA & cohort AUC=0.67 Cohort only AUC=0.59 HLA & cohort AUC=0.59 Cohort only AUC=0.57 HLA & cohort AUC=0.65 Cohort only AUC=0.44 (Proof of principle) (same effect including HBV DNA >2000 iu/ml) Matthews et al., in review JID ROC curves for HLA Class I locus (HLA-A vs HLA-B vs HLA-C) as predictor of HBeAg-status HLA-A AUC=0.71 p=0.004 HLA-B AUC=0.35 p=0.97 HLA-C AUC=0.41 p=0.91 Matthews et al., in review JID
Modelling the impact of HLA class I loci on disease outcome HLA-A alleles and several HLA-B alleles target conserved human proteins and DNA viruses most efficiently, while all HLA-B alleles studied efficiently target RNA viruses. HBV HIV HLA-A HLA-B Hertz et al., J. Virol. 2011 What is the mechanism of the effect?
Conclusions Diagnosis Antivirals Vaccination HBV eradication CURE? HLA Class I genes at the HLA-A locus are a significant predictor of HBeAg clearance Screening blood products Passive immunisation This is a new finding that potentially informs strategy for a therapeutic T cell vaccine. Acknowledgements Prof Paul Klenerman, University of Oxford Prof Ellie Barnes, University of Oxford Prof Philip Goulder, University of Oxford Prof Anna Maria Geretti, University of Liverpool Amna Malik, University of Oxford Dr Jonathan Carlson, Microsoft Research, USA Prof Roger Shapiro, Gaborone, Botswana Prof Thumbi Ndung u, Durban, South Africa Dr Pieter Jooste, Kimberley, South Africa Thames Valley Cohort Clinicians p.matthews@doctors.org.uk