uva-dare (digital academic repository) immune responses in ... · 10 chapter 1 hepatitis b virology...

UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Immune responses in hepatitis B and C virus infection

Stelma, F.

Link to publication

LicenseOther

Citation for published version (APA):Stelma, F. (2017). Immune responses in hepatitis B and C virus infection.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 09 Feb 2021

https://dare.uva.nl/personal/pure/en/publications/immune-responses-in-hepatitis-b-and-c-virus-infection(2fc21fe8-6d06-49a3-8671-0a5d492a4af3).html

Immune responses inhepatitis B and C

virus infection

Imm

une responses in hep

atitis B and

C virus infection

FEMKE STELMA

FE

MK

E S

TE

LMA

Immune responses In hepatItIs B and C vIrus InfeCtIon

Femke Stelma

Immune responses in chronic hepatitis B and C virus infection

Thesis, University of Amsterdam, the Netherlands.

ISBN: 978-94-6182-834-7

Cover design: Jorrit Stelma

Layout & printing: Off Page, Amsterdam

Copyright © 2017 Femke Stelma | All rights reserved. No part of this thesis may be reproduced, stored or transmitted in any form or by any means without the permission of the author or publishers of the included scientific papers.

The printing of this thesis was financially supported by: Academic Medical Center, Nederlandse Vereniging voor Hepatologie, Regulus Therapeutics, Gilead Sciences, Virology Education, Zambon Nederland B.V. and ChipSoft.

Immune responses In hepatItIs B and C vIrus InfeCtIon

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor aan de Universiteit van Amsterdam op gezag van de Rector Magnificus

prof. dr. ir. K.I.J. Maex ten overstaan van een door het College voor Promoties ingestelde

commissie, in het openbaar te verdedigen in de Agnietenkapel

op donderdag 30 november 2017, te 10:00 uur

door

Femke Stelmageboren te Opsterland

promotIeCommIssIe Promotores Prof. Dr. U.H.W. Beuers AMC - Universiteit van Amsterdam Prof. Dr. T.B.H. Geijtenbeek AMC - Universiteit van Amsterdam

Copromotores Dr. H.W. Reesink AMC - Universiteit van Amsterdam Dr. N.A. Kootstra AMC - Universiteit van Amsterdam

Overige leden Dr. L.C. Baak OLVG oost, Amsterdam Prof. Dr. E. Barnes University of Oxford Prof. Dr. P.L.M. Jansen AMC - Universiteit van Amsterdam Prof. Dr. R.A. van Lier AMC - Universiteit van Amsterdam Dr. D. Sprengers Erasmus Universiteit Rotterdam Prof. Dr. H.L. Zaaijer AMC - Universiteit van Amsterdam

Faculteit der geneeskunde

taBle of ContentsChapter 1 General introduction 9

part I Immune responses in patients with an acute hepatitis B infection

Chapter 2 Dynamics of the immune response in acute hepatitis B infection 25Open Forum Infectious Diseases 2017

part II Immune responses in patients with chronic hepatitis B infection

Chapter 3 HBsAg loss after peginterferon and nucleotide combination treatment in chronic hepatitis B patients: 5 years of follow-up 43

Journal of Viral Hepatitis 2017

Chapter 4 Restoration of T cell function in chronic hepatitis B patients upon treatment with interferon based combination therapy 59

Journal of Hepatology 2016

Chapter 5 Natural Killer cell characteristics in chronic hepatitis B patients are associated with HBsAg loss after combination treatment with peg-interferon alpha-2a and adefovir 79

Journal of Infectious Diseases 2015

Chapter 6 HLA-C and KIR combined genotype as new response marker for HBeAg-positive chronic hepatitis B patients treated with interferon-based combination therapy 101


Chapter 7 HBsAg loss in patients treated with peginterferon alfa-2a and adefovir is associated with SLC16A9 gene variation and lower plasma carnitine levels 117

Journal of Hepatology 2014

Chapter 8 Peg-interferon plus nucleotide analogue treatment versus no treatment in patients with chronic hepatitis b with a low viral load: a randomised controlled, open-label trial 135

The Lancet Gastroenterology and Hepatology 2017

Chapter 9 Editorial: Peginterferon add-on: an added benefit? 155The Lancet Gastroenterology and Hepatology 2017

Chapter 10 HBV RNA reductions and serum cytokine changes in HBeAg positive chronic hepatitis B patients treated with REP 2139 161

Submitted

part III Immune responses in patients with chronic hepatitis C infection

Chapter 11 Safety, tolerability, and antiviral effect of RG-101 in patients with chronic hepatitis C: a phase 1B, double-blind, randomised controlled trial 179

The Lancet 2017

Chapter 12 Immune phenotype and function of NK and T cells in chronic hepatitis C patients who received a single dose of anti-miR-122, RG-101 203

Hepatology 2017

Chapter 13 Immune responses in DAA treated chronic hepatitis C patients with and without prior RG-101 dosing 221

Antiviral Research 2017

part Iv Characterisation of intrahepatic t cells in patients with chronic hepatitis B or C infection

Chapter 14 Human intrahepatic CD69+CD8+ T cells have a tissue resident memory T cell phenotype with reduced cytolytic capacity 241

Scientific Reports 2017

Chapter 15 Summary and future perspectives 259

appendix Nederlandse samenvatting 269List of contributing authors 275List of publications 278PhD portfolio 281Dankwoord 283Curriculum vitae 286

General IntroduCtIon

10

Chapter 1

hepatItIs B virologyThe hepatitis B virus is a member of the Hepadnaviridae (hepatotropic DNA viruses), which preferentially infects hepatocytes.1 At least 10 viral genotypes are known and these are labelled A-J.2 The viral DNA is 3.2 kb in length, and codes for 4 proteins; viral polymerase, core protein (HBcAg), envelope protein (HBsAg) and X protein (HBx). Viral particles are also called ‘Dane-particles’, and measure 40-42 nm in diameter. Together with the viral polymerase, the viral DNA is enclosed in a capsule built op of HBcAg, which is enveloped by lipoproteins containing HBsAg (figure 1). In addition, non-infectious subviral particles containing lipoprotein and HBsAg are produced and abundantly released into the circulation. Similarly, HBe antigen (HBeAg), which is a derivative of HBcAg and produced when transcription is initiated in the region preceding the core region (pre-core region), is secreted into the blood. While the exact role of the abundantly secreted HBsAg and HBeAg is not well understood, both are believed to play are role in subverting the antiviral immune response.3,4

When an hepatocyte is infected, the viral relaxed circular (rc) DNA is transported to the nucleus and subsequently converted to covalently closed circular (ccc) DNA which serves as a template for the transcription of pregenomic and subgenomic RNA (figure 1).5 These viral mRNAs code for the viral proteins. Furthermore, replication of the viral genome occurs via the pregenomic HBV RNA intermediate which is encapsidated and reverse transcribed by the HBV polymerase.6 Newly formed core particles can either locate back to the nucleus to replenish the cccDNA pool or be enveloped and secreted to infect other hepatocytes.

figure 1. HBV replication cycle.

11

GENERAL INTRODUCTION

1epidemiology & natural history As much as one third of the global population has encountered HBV at some point in their life.7 Infection with HBV at early age is generally asymptomatic, but will lead to chronic infection in the majority of cases (>95%). As a result, there are regions such as China and most of Africa where HBV is endemic (prevalence >8%) and mainly transmitted vertically (mother-to-child) or horizontally early during childhood.7 The introduction of HBV vaccination in the standard vaccination program is leading to a dramatic decrease in the incidence of HBV in many of these countries.8 An acute HBV infection that is acquired during adulthood can be symptomatic or asymptomatic and is usually self-limiting. In less than 5% of cases however, infection during adulthood will lead to chronicity, defined as persistence of HBV beyond 6 months after initial infection. An estimated 240 million patients worldwide are chronically infected with HBV.9 These patients are at increased risk of developing cirrhosis and/or hepatocellular carcinoma (HCC).10,11 Chronic HBV is a dynamic disease that has an unpredictable course. Based on the presence of liver inflammation (ALT level), viral load level and the presence of HBeAg, several phases can be identified during chronic hepatitis B infection12:

Immune tolerance: Generally, chronic hepatitis B infection will remain in the immune tolerance phase during the first decades of life. Patients are HBeAg positive and the viral load is very high, but transaminase levels are normal. Due to the absence of clinical signs of liver inflammation it has been postulated that the immune system is tolerant in this phase, however, this view has been challenged recently as several signs of immune activity can be observed in these patients.13

Immune active: At some point patients usually progress from the immune tolerant phase to the immune active phase. This phase is characterised by hepatocyte damage and therefore increases in the transaminase levels, also called flares. The liver infiltrate consist of granulocytes, monocytes, NK cells and HBV-specific as well as non-HBV specific T cells.13,14 This phase can lead to HBeAg seroconversion within one year in 10-20% of cases.15

Inactive carriers: Inactive carriers are HBeAg negative and anti-HBeAg positive, have a low viral load and normal transaminases. Transient ALT increases can occur, which makes it difficult to classify truly inactive carriers. Generally, a viral load of <2,000 IU/mL and normal ALT is used to define these patients. The risk of developing HCC is still present in inactive carriers, albeit significantly lower than in patients with a high viral load.16 Spontaneous HBsAg loss will occur 0.5-1% yearly and a decline in HBsAg is predictive of HBsAg seroclearance in inactive carriers.17

HBeAg negative chronic hepatitis B: In roughly 15-20% of patients who are inactive carriers, HBV reactivation occurs, with increases in ALT levels and in viral load (>104 IU/mL).18 Progression to HBeAg negative chronic hepatitis B can occur decades after HBeAg seroconversion and is associated with an increased risk of HBV-related liver complications.19

12

Chapter 1

treatment Treatment for chronic hepatitis B infection is indicated in patients with a high viral load,

increased ALT levels and signs of fibrosis, as well as in patients with cirrhosis.12 The goal of

treatment is to prevent the development or progression of liver related complications such

as cirrhosis and HCC. Complete cure is challenging as the HBV mini-chromosome cccDNA

is difficult to eradicate from the nucleus of the hepatocyte.5 The most optimal outcome

is a ‘functional cure’ which is defined as HBsAg loss with or without the development of

anti-HBs antibodies, a state in which the cccDNA is not transcriptionally active. A functional

cure is associated with improved outcome of the disease.20–22 As current treatment options

only occasionally lead to functional cure, current therapeutic approaches are mostly

focused on reducing the viral load and the amount of liver inflammation.

Current treatment optionsTwo treatment modalities are currently used for the treatment of patients with chronic

hepatitis B infection. These include pegylated interferon alfa (pegIFN) and nucleos(t)ide

analogues (NUCs). PegIFN is an exogenous cytokine thought to boost endogenous antiviral

immune responses. PegIFN has been shown to directly interfere with the cccDNA,23 as well

as to activate natural killer (NK) cells.24 Treatment consists of a yearlong course of weekly

subcutaneous injections, and can be associated with severe flu-like symptoms. In HBeAg

positive patients, treatment with pegIFN will lead to HBeAg seroconversion and HBV DNA

reduction in approximately 30% of patients.25 In HBeAg negative patients, treatment with

pegIFN may lead to HBV DNA suppression and ALT normalization in about one third

of patients.26

NUCs directly interfere with the reverse transcription step of the replication cycle by

blocking viral polymerase. Treatment with NUCs is highly effective in viral load suppression

and ALT normalization and leads to improvement of outcome.27,28 However, upon

discontinuation, viral load will often rebound, with the possibility of developing a flare.29

In either of these treatment modalities, functional cure is rarely achieved.30

New treatment options In order to improve treatment outcome, several trials have investigated the effects of

combining pegIFN with a NUC.31 Furthermore, novel treatment modalities are currently

being developed. These include treatment with RNA-interference molecules, core-

assembly inhibitors, and HBsAg release inhibitors.30

hepatItIs CvirologyThe hepatitis C virus is a member of the Flaviviridae. It is a single-stranded RNA virus of 9.6

kb in length.32 The viral genome encodes for 3 structural (core and envelope glycoproteins

E1 and E2) and 7 non-structural proteins (p7, NS2, NS3 protease, NS4A protease, NS4B,

NS5A, and NS5B polymerase). Based on a sequence diversity of greater than 30%, seven

genotypes are described, which are labelled 1-7 and can be further subdivided. Following

13


1entry of the hepatocyte, the HCV RNA is released into the cytoplasm and translated into a polyprotein, which is cleaved by host and viral proteases. HCV RNA is replicated via an anti-sense intermediate by NS5B polymerase.33 As there is a high viral turnover and the viral polymerase is very error-prone, many HCV quasispecies exist within one patient. For stability, HCV is dependent on a host factor, micro-RNA-122 (miR-122), which can bind to the 5’ end of the HCV genome. miR-122 is a highly conserved, liver-specific micro-RNA which has important functions in the regulation of cholesterol and fatty acid synthesis.34 By binding to the HCV genome, it promotes virus stability and enables viral replication.35 Other than HBV, HCV is not present in a chromosomal form in the hepatocyte nucleus, nor is it integrated in the host DNA.

epidemiology & natural historyWorldwide, an estimated 170 million are chronically infected with the hepatitis C virus.36 Acute HCV infection acquired as an adult is generally asymptomatic but will lead to chronicity in 80-85% of cases. Chronic hepatitis C infection can lead to liver related complications such as cirrhosis and hepatocellular carcinoma. After 20 years of infection, 7-18% of patients will have developed cirrhosis.32 Furthermore, chronic HCV infection can lead to extrahepatic manifestations such as cryoglobulinaemia or HCV-associated lymphoma.

treatmentThe goal of treatment for chronic hepatitis C infection is prevention of disease progression. The end-point of treatment is achieving viral cure referred to as a sustained viral response or SVR, usually determined 12 weeks after end of treatment (SVR12) . In the last decade,

figure 2. HCV replication cycle.

14

Chapter 1

the treatment for chronic hepatitis C infection developed very rapidly. Previously, patients were treated with a combination of pegIFN and ribavirin for up to 48 weeks, which was associated with relatively low chances of viral clearance. Nowadays patients are treated with direct acting antivirals (DAAs) for 12-24 weeks, which directly target the viral machinery and are very effective in doing so. Per genotype, different classes of DAAs are indicated, including NS5A inhibitors (ending with -asvir), NS5B inhibitors (-buvir) and NS3/4A protease inhibitors (-previr). The development of successful interferon-free treatment, with a duration of only 12-24 weeks has been a major accomplishment in antiviral therapy of the last decade.

Other treatment options Even though many patients can be cured by treatment with DAAs, there is still need for alternative therapeutic options for the small group of patients infected with DAA resistant HCV strains.37 One of these alternatives lies in the inhibition of host factor miR-122. The inhibition of this miRNA is based on anti-sense oligonucleotides, which are thought to bind free miR-122 in the liver. Without miR-122 binding to its 5’ end, the HCV genome is instable and more prone to degradation by exonucleases.38

Immune responsesHBV and HCV are both non-cytopathic viruses. The liver damage observed during these infections is due to cytolytic antiviral activity of the immune system. These include effectors of the innate as well as the adaptive immune system.

acute infectionUpon viral infection, the first detection of a pathogen is by innate pattern recognition receptors (PRRs) within the infected cell that detect conserved pathogen-associated molecular patterns (PAMPs). Detection of these PAMPs will ultimately lead to interferon production, which drives the synthesis of hundreds of interferon stimulated genes (ISGs). These ISGs activate specific pathways of antiviral activity.39 In chimpanzees infected with HBV, no such induction of ISGs was observed, suggesting HBV could act as a ‘stealth’ virus, which is not recognised by innate immunesensors.40 In addition, it is thought that HBV actively evades initial immune activation, by blocking cell signalling and ISG production.41 In line with this, in patients with an acute HBV infection no induction of type I interferon (IFN-α), nor of type III interferon (IFN-λ) was observed.42 HCV has likewise developed escape strategies evading the interferon pathway, however, during acute HCV infection a strong interferon response can be observed.33 Furthermore, spontaneous clearance of HCV infection has been associated with a single nucleotide polymorphism (SNP) in the gene coding for IFN-λ (IL-28B), which shows the importance of the early interferon response.43

Natural killer (NK) cells are also part of the innate immune system, and are activated early in the course of infection. NK cells can sense abnormal cells, such as virally infected

15


1hepatocytes and kill these via secretion of cytotoxic molecules or binding of death receptors by the tumor necrosis factor(TNF)-related apoptosis-inducing ligand (TRAIL). Furthermore, NK cells can produce antiviral cytokines such as IFN-γ and TNF-α. Early during an acute infection with HBV42,44 and HCV,45,46 NK cells have been shown to upregulate the expression of activating receptors supporting an antiviral response.

Later in the course of infection, virus-specific CD4+ and CD8+ T cells as well as B cells are activated. Historically, virus-specific T cells have acquired a prominent role in antiviral immunity against these hepatotropic infections. Depletion studies in chimpanzees have clearly demonstrated that CD4+ and CD8+ T cells are crucial for viral clearance. Furthermore, broad CD8+ T cell responses are observed in spontaneous clearance of HBV47 and HCV48,49 virus infection. On the other hand, reduction of the viral load already commences when T cell responses are still weak suggesting a combined effort of various immune effectors is needed for successful viral clearance. Together, virus-specific T cells and antibodies produced by B cells should protect the host against recurring infections. While this is the case after clearance of HBV, recurring HCV infection is seen more frequently.

Chronic infectionDuring chronic hepatitis B and C infection, many aspects of the antiviral immune response are significantly altered. Chronic hepatitis B and C virus infection are associated with increased levels of interferon-inducible protein-10 (IP-10), which is an ISG that attracts immune cells expressing its receptor, CXCR3. After successful treatment, these levels decline to normal.50 NK cells are activated and the expression of TRAIL is increased during chronic viral hepatitis.51 Treatment with pegIFN can further activate NK cells in patients with chronic viral hepatitis, and change their subset distribution.24 After successful treatment of HCV with interferon-free regimens, NK cell activation decreases to normal levels.52

Virus-specific CD8+ T cells become exhausted during persistent hepatitis B and C virus infection, a state in which they progressively lose their functional activity and are ultimately deleted. This T cell exhaustion is driven by continuous triggering of T cells by high antigen levels.53 This is associated with the expression of multiple inhibitory receptors, including programmed-death-1 (PD-1), T cell immunoglobulin and mucin domain–containing molecule 3 (Tim-3), and cytotoxic T lymphocyte antigen 4 (CTLA-4).54 Furthermore, during persistent infection, T cell function can be inhibited by other mechanisms such as regulatory T cells and cytokines (IL-10 and TGF-β). In HCV more than in HBV, viral mutants can arise, so that the mounted T cell response will not recognise the virus anymore and becomes futile. Successful suppression of HBV has been associated with an improvement in T cell responses,55 as has successful treatment of HCV infection.56 However, it is unknown whether these T cells play a causative role in controlling the infection, or whether merely a byeffect of the decline in antigen load is observed. Understanding the mechanisms that lead to viral clearance will help to further determine how we can establish antiviral immunity.

16

Chapter 1

thesIs outlIneThe main focus of this thesis was to analyse immune responses in patients with hepatitis B and C virus infection and to better understand the role of the different compartments of the immune system in viral clearance.

part I. Immune responses in patients with an acute hepatitis B infectionThe first part of this thesis describes the immune response in patients who have an acute hepatitis B infection. In chapter 2, nine patients with an acute hepatitis B infection were included and followed for 24 weeks. Extensive analyses of the immune response in these patients aimed to identify factors that are associated with either spontaneous resolution of infection or development of chronicity.

part II. Immune responses in patients with chronic hepatitis B infection The second part of this thesis focuses on immune responses in patients who have a chronic hepatitis B infection. In chapter 3 we describe the long term follow-up results of a clinical study in which 92 CHB patients with a high viral load and treatment indication were treated with a combination of pegIFN and adefovir. In the initial study, high rates of HBsAg loss were observed in HBeAg positive as well as –negative patients. In the long term follow-up study, the sustainability of this response up to 5 years after end-of-treatment was analysed. Subsequently, this well described cohort of CHB patients was the subject of many fundamental exploratory studies. In chapter 4, HBV-specific T cells responses were analysed in the same cohort of CHB patients treated with pegIFN-based combination therapy, while chapter 5 describes the NK cell phenotype and function in these patients. The aim of chapter 6 was to analyse whether there was an association between treatment outcome and the combined KIR and HLA-C genotype, which influences NK cell activity. In chapter 7, the results of a genome wide association study (GWAS) are described. This GWAS identified a novel single nucleotide polymorphism (SNP) which was associated to HBsAg loss in patients treated with combination therapy. Supported by the results of our previous treatment study, a follow-up study was performed, which is described in chapter 8. In this study, chronic hepatitis B patients with a low viral load, who currently have no treatment indication, were randomized to receive either pegIFN plus nucleotide analogue versus no therapy. As therapeutic options for patients with chronic hepatitis B are currently limited, many groups are investigating options to combine pegIFN and nucleos(t)ide analogues to improve treatment outcome. In chapter 9, the results of a study by another group in which pegIFN was used as add-on therapy to nucleotide treatment were critically appraised.

New treatment options for chronic hepatitis B are currently evolving rapidly. One of these includes the HBsAg release inhibitor REP-2139, which was given to 12 patients with chronic hepatitis B infection in a proof of concept study. In chapter 10, the effects of this new therapy on circulating HBV RNA and cytokine levels was analysed.

17


1part III. Immune responses in patients with chronic hepatitis C infectionThe third part of this thesis revolved around chronic hepatitis C infection. In chapter 11, the results of a phase 1 study are described, in which chronic hepatitis C patients received a single subcutaneous injection of an oligonucleotide targeting miR-122; RG-101. Chapter 12 describes the changes in NK cell and T cell phenotype and function in response to dosing with RG-101. Currently, standard of care for patients with chronic hepatitis C is treatment with DAAs. In chapter 13 clinical and immunological responses are described in 29 patients successfully treated with DAA therapy.

part Iv. Characterisation of intrahepatic t cells in patients with chronic hepatitis B or CThe last part of this thesis concentrates on intrahepatic CD8+ T cells. A lot of the immunological research in the field of viral hepatitis is based on peripheral blood samples from patients with chronic viral hepatitis. However, less in known about immune cells that are resident to the liver. In chapter 14, the phenotype of CD8+ cells derived from the liver are compared to that of CD8+ T cells from the peripheral blood. Furthermore, the differences between healthy controls, patients with chronic hepatitis B infection and patients with chronic hepatitis C infection were analysed.

18

Chapter 1

referenCes 1. Ganem D, Prince AM. Hepatitis B virus infection-

natural history and clinical consequences. N Engl J Med 2004;350(11):1118–29.

2. Lin C-L, Kao J-H. The clinical implications of hepatitis B virus genotype: Recent advances. J Gastroenterol Hepatol 2011;26 Suppl 1:123–30.

3. Shi B, Ren G, Hu Y, Wang S, Zhang Z, Yuan Z. HBsAg Inhibits IFN-a Production in Plasmacytoid Dendritic Cells through TNF-a and IL-10 Induction in Monocytes. PLoS One 2012;7(9):e44900.

4. Cheng J, Imanishi H, Morisaki H, et al. Recombinant HBsAg inhibits LPS-induced COX-2 expression and IL-18 production by interfering with the NFkB pathway in a human monocytic cell line, THP-1. J Hepatol 2005;43:465–71.

5. Lucifora J, Protzer U. Attacking hepatitis B virus cccDNA - The holy grail to hepatitis B cure. J Hepatol 2016;64(1 Supp):S41-8.

6. Beck J, Nassal M. Hepatitis B virus replication. World J Gastroenterol 2007;13(1):48–64.

7. Trépo C, Chan HLY, Lok A. Hepatitis B virus infection. Lancet 2014;384(9959):2053–63.

8. Ni Y-H, Chang M-H, Wu J-F, Hsu H-Y, Chen H-L, Chen D-S. Minimization of hepatitis B infection by a 25-year universal vaccination program. J Hepatol 2012;57(4):730–5.

9. Schweitzer A, Horn J, Mikolajczyk RT, Krause G, Ott JJ. Estimations of worldwide prevalence of chronic hepatitis B virus infection: a systematic review of data published between 1965 and 2013. Lancet 2015;386(10003):1546–55.

10. Iloeje UH, Yang H-I, Su J, et al. Predicting cirrhosis risk based on the level of circulating hepatitis B viral load. Gastroenterology 2006;130(3):678–86.

11. Chen C-J, Yang H-I, Su J, et al. Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level. JAMA 2006;295(1):65–73.

12. EASL clinical practice guidelines: Management of chronic hepatitis B virus infection. J Hepatol 2012;57(2):167–85.

13. Bertoletti A, Kennedy PT. The immune tolerant phase of chronic HBV infection: new perspectives on an old concept. Cell Mol Immunol 2015;12(3):258–63.

14. Abrignani S. Bystander activation by cytokines of intrahepatic T cells in chronic viral hepatitis. Semin Liver Dis 1997;17(4):319–22.

15. Lok AS, Lai CL, Wu PC, Leung EK, Lam TS. Spontaneous hepatitis B e antigen to antibody seroconversion and reversion in Chinese patients with chronic hepatitis B virus infection. Gastroenterology 1987;92(6):1839–43.

16. Chen J De, Yang HI, Iloeje UH, et al. Carriers of Inactive Hepatitis B Virus Are Still at Risk for Hepatocellular Carcinoma and Liver-Related Death. Gastroenterology 2010;138(5):1747–1754.e1.

17. Chen YC, Jeng WJ, Chu CM, Liaw YF. Decreasing Levels of HBsAg Predict HBsAg Seroclearance in Patients With Inactive Chronic Hepatitis B Virus Infection. Clin Gastroenterol Hepatol 2012;10(3):297–302.

18. Funk ML, Rosenberg DM, Lok ASF. World-wide epidemiology of HBeAg-negative chronic hepatitis B and associated precore and core promoter variants. J Viral Hepat 2002;9(1):52–61.

19. Saikia N, Talukdar R, Mazumder S, Khanna S, Tandon R. Management of patients with HBeAg-negative chronic hepatitis B. Postgrad Med J 2007;83(975):32–9.

20. Moucari R, Mackiewicz V, Lada O, et al. Early serum HBsAg drop: A strong predictor of sustained virological response to pegylated

19


1interferon alfa-2a in HBeAg-negative patients. Hepatology 2009;49(4):1151–7.

21. Fattovich G, Olivari N, Pasino M, D’Onofrio M, Martone E, Donato F. Long-term outcome of chronic hepatitis B in Caucasian patients: mortality after 25 years. Gut 2008;57(1):84–90.

22. Feld JJ, Wong DKH, Heathcote EJ. Endpoints of therapy in chronic hepatitis B. Hepatology 2009;49(5 Suppl):S96–102.

23. Lucifora J, Xia Y, Reisinger F, et al. Specific and nonhepatotoxic degradation of nuclear hepatitis B virus cccDNA. Science 2014;343(6176):1221–8.

24. Micco L, Peppa D, Loggi E, et al. Differential boosting of innate and adaptive antiviral responses during pegylated-interferon-alpha therapy of chronic hepatitis B. J Hepatol 2013;58(2):225–33.

25. Lau GKK, Piratvisuth T, Luo KX, et al. Peginterferon Alfa-2a, lamivudine, and the combination for HBeAg-positive chronic hepatitis B. N Engl J Med 2005;352(26):2682–95.

26. Marcellin P, Bonino F, Lau GKK, et al. Sustained Response of Hepatitis B e Antigen-Negative Patients 3 Years After Treatment with Peginterferon Alfa-2a. Gastroenterology 2009;136(7):2169–79.

27. Papatheodoridis G V., Dimou E, Dimakopoulos K, et al. Outcome of hepatitis B e antigen-negative chronic hepatitis B on long-term nucleos(t)ide analog therapy starting with lamivudine. Hepatology 2005;42(1):121–9.

28. Marcellin P, Gane E, Buti M, et al. Regression of cirrhosis during treatment with tenofovir disoproxil fumarate for chronic hepatitis B: a 5-year open-label follow-up study. Lancet 2013;381(9865):468–75.

29. Papatheodoridis G, Vlachogiannakos I, Cholongitas E, et al. Discontinuation of oral antivirals in chronic hepatitis B: A systematic review. Hepatology 2016;63(5):1481–92.

30. Zoulim F, Durantel D. Antiviral Therapies and Prospects for a Cure of Chronic Hepatitis B. Cold Spring Harb Perspect Med 2015;5(4):a021501–a021501.

31. Viganò M, Invernizzi F, Grossi G, Lampertico P. Review article: the potential of interferon and nucleos(t)ide analogue combination therapy in chronic hepatitis B infection. Aliment Pharmacol Ther 2016;44:653–61.

32. Webster DP, Klenerman P, Dusheiko GM. Hepatitis C. Lancet 2015;385(9973):1124–35.

33. Horner SM. Activation and evasion of antiviral innate immunity by hepatitis C virus. J Mol Biol 2014;426(6):1198–209.

34. Esau C, Davis S, Murray SF, et al. miR-122 regulation of lipid metabolism revealed by in vivo antisense targeting. Cell Metab 2006;3(2):87–98.

35. Jopling CL, Yi M, Lancaster AM, Lemon SM, Sarnow P. Modulation of hepatitis C virus RNA abundance by a liver-specific MicroRNA. Science 2005;309(5740):1577–81.

36. WHO Fact sheet hepatitis C. http://www.who.int/mediacentre/factsheets/fs164/en/.

37. Pawlotsky J-M. Hepatitis C Virus Resistance to Direct-Acting Antiviral Drugs in Interferon-Free Regimens. Gastroenterology 2016;151(1):70–86.

38. Machlin ES, Sarnow P, Sagan SM. Masking the 5’ terminal nucleotides of the hepatitis C virus genome by an unconventional microRNA-target RNA complex. Proc Natl Acad Sci U S A 2011;108(8):3193–8.

39. Wong M-T, Chen SS-L. Emerging roles of interferon-stimulated genes in the innate immune response to hepatitis C virus infection. Cell Mol Immunol 2016;13(1):11–35.

40. Wieland S, Thimme R, Purcell RH, Chisari F V. Genomic analysis of the host response to hepatitis B virus infection. Proc Natl Acad Sci U S A 2004;101(17):6669–74.

20

Chapter 1

41. Hayes CN, Chayama K. Interferon stimulated genes and innate immune activation following infection with hepatitis B and C viruses. J Med Virol 2017;89(3):388–96.

42. Dunn C, Peppa D, Khanna P, et al. Temporal analysis of early immune responses in patients with acute hepatitis B virus infection. Gastroenterology 2009;137(4):1289–300.

43. Thomas DL, Thio CL, Martin MP, et al. Genetic variation in IL28B and spontaneous clearance of hepatitis C virus. Nature 2009;461(7265):798–801.

44. Webster GJM, Reignat S, Maini MK, et al. Incubation Phase of Acute Hepatitis B in Man: Dynamic of Cellular Immune Mechanisms. Hepatology 2000;32(5):1117–24.

45. Werner JM, Heller T, Gordon AM, et al. Innate immune responses in hepatitis C virus-exposed healthcare workers who do not develop acute infection. Hepatology 2013;58(5):1621–31.

46. Amadei B, Urbani S, Cazaly A, et al. Activation of Natural Killer Cells During Acute Infection With Hepatitis C Virus. Gastroenterology 2010;138(4):1536–45.

47. Fisicaro P, Valdatta C, Boni C, et al. Early kinetics of innate and adaptive immune responses during hepatitis B virus infection. Gut 2009;58(7):974–82.

48. Urbani S, Amadei B, Fisicaro P, et al. Outcome of acute hepatitis C is related to virus-specific CD4 function and maturation

of antiviral memory CD8 responses. Hepatology 2006;44(1):126–39.

49. Sung PS, Racanelli V, Shin E-C. CD8(+) T-Cell Responses in Acute Hepatitis C Virus Infection. Front Immunol 2014;5:266.

50. Meissner EG, Wu D, Osinusi A, et al. Endogenous intrahepatic IFNs and association with IFN-free HCV Treatment outcome. J Clin Invest 2014;124(8):3352–63.

51. Rehermann B. Pathogenesis of chronic viral hepatitis: differential roles of T cells and NK cells. Nat Med 2013;19(7):859–68.

52. Spaan M, van Oord G, Kreefft K, et al. Immunological Analysis During Interferon-Free Therapy for Chronic Hepatitis C Virus Infection Reveals Modulation of the Natural Killer Cell Compartment. J Infect Dis 2015;213:jiv391.

53. Wherry EJ. T cell exhaustion. Nat Immunol 2011;12(6):492–9.

54. Frebel H, Richter K, Oxenius A. How chronic viral infections impact on antigen-specific T-cell responses. Eur J Immunol 2010;40(3):654–63.

55. Boni C, Laccabue D, Lampertico P, et al. Restored function of HBV-specific T cells after long-term effective therapy with nucleos(t)ide analogues. Gastroenterology 2012;143(4):963–73.e9.

56. Martin B, Hennecke N, Lohmann V, et al. Restoration of HCV-specific CD8+ T-cell function by Interferon-free therapy. J Hepatol 2014;61:538–43.

Immune responses In patIents wIth an aCute hepatItIs B InfeCtIon

dynamICs of the Immune response In aCute hepatItIs B InfeCtIon

F. Stelma, S.B. Willemse, A. de Niet, M.J. Sinnige, H.L. Zaaijer, E.M.M. van Leeuwen, N.A. Kootstra, H.W. Reesink

Open Forum Infectious Diseases 2017

aBstraCtBackground Acute hepatitis B virus (AHB) infection in adults is generally self-limiting but may lead to chronicity in a minority of patients.

methods We included 9 patients with acute HBV infection and collected longitudinal follow-up samples. NK cell characteristics were analysed by flowcytometry. HBV-specific T cell function was analysed by in vitro stimulation with HBV peptide pools, and intracellular cytokine staining.

results Median baseline HBV DNA load was 5.12 log IU/mL and median ALT was 2,652 U/mL. Of 9 patients, 8 cleared HBsAg within 6 months whereas one patient became chronically infected. Early timepoints after infection showed increased CD56bright NK cells, and an increased proportion of cells expressing activation markers. Most of these had normalised at week 24, while the proportion of TRAIL-positive CD56bright NK cells remained high in the chronically infected patient. In patients that cleared HBV, functional HBV-specific CD8+ and CD4+ responses could be observed, whereas in the patient who developed chronic infection, only low HBV-specific T cell responses were observed.

Conclusions NK cells are activated early in the course of acute HBV infection. Broad and multi-specific T cell responses are observed in patients who clear acute HBV infection, but not in a patient who became chronically infected.

27

IMMUNE RESPONSES IN ACUTE HEPATITIS B

2

IntroduCtIonInfection with the hepatitis B virus (HBV) affects large numbers of individuals worldwide. As much as one third of the global population has encountered HBV at some point in their life. Infection with HBV at early age will lead to chronicity in the majority of cases (>95%), resulting in an estimated 240 million patients worldwide who are chronically infected with HBV.1,2 When an acute HBV infection is encountered later in life, the virus will be spontaneously cleared in most cases.3 Less than 5% of immunocompetent adult patients however, fail to clear the virus and become chronically infected with HBV. The mechanisms that lead to chronicity of hepatitis B infection are largely unknown. For clearance of the virus in the acute setting both the innate and the adaptive immune system are important.4–7 The innate immune system is responsible for early containment of the viruses and initial activation of adaptive immune responses. Although HBV has been shown to act as a ‘stealth’ virus in woodchucks and chimpanzees, evading early intrahepatic immune responses,8,9 it is uncertain whether these early innate responses are induced during acute HBV infection in man.10 Other players of the innate immune system, natural killer (NK) cells, are activated early during infection, before HBV-specific T cells arise.11,12 Later on during infection, functionally active HBV-specific T cells can be detected, which are thought to play an important role in viral clearance. In chimpanzees, depletion of CD8+ T cells at week 6 of infection lead to failure to clear the infection.5 During chronic infection, HBV-specific T cells are exhausted and their function is impaired.13 However, whether these HBV-specific T cells were functionally active during the initial phases of infection is unknown. In acute hepatitis C infection, patients with self-limited infection have significant T cell responses compared to little or no responses those who evolve to chronicity.14

Acute hepatitis B infection is asymptomatic in the majority of cases which makes it difficult to study. However, previous studies have focused on blood donors that became HBsAg positive (n=2) or a local outbreak (n=5) for the initial phases of infection.11,15 Here we examined the early dynamics of NK and HBV-specific T cell responses in symptomatic patients with acute HBV infection who presented at our clinic.

patIents and methods patientsPatients were included at the gastroenterology and hepatology department of the Academic Medical Center in Amsterdam. Acute infection was diagnosed based on HBsAg and HBV DNA positivity, further serology, biochemistry and anamnesis reporting risk of acquiring HBV. Patients were assessed at the outpatient clinic at first visit (clinical onset) which was defined as baseline (BL). Follow-up was at week 1, 4, 12 and 24. All patients were HIV seronegative and were not co-infected with hepatitis C or hepatitis delta virus. The study was approved by the Ethical Review Board of the Academic Medical Center Amsterdam and all patients gave written informed consent. The study was conducted in accordance with Declaration of Helsinki, Good Clinical Practice guidelines, and local regulatory requirements.

28

Chapter 2

laboratory testingBiochemical and virological analyses were carried out by local laboratories in accordance with good laboratory practice. Qualitative detection of serum hepatitis B surface antigen (HBsAg) and antibody to hepatitis B surface antigen (anti- HBs) was performed by enzyme immunoassay (AxSYM; Abbott Laboratories, Abbott Park, IL, USA). Quantitation of plasma HBV DNA levels was done by the Roche COBAS® TaqMan 48® assay (F. Hoffmann-La Roche Ltd, Diagnostics Division, Basel, Switzerland), with a dynamic range between 20 and 1.70x108 IU/mL. HBV genotype was determined by sequencing a part of the polymerase gene with dideoxynucleotide technology.

samplingPeripheral blood samples were obtained at baseline and during follow-up (week 1, 4, 12 and 24). Sampling included plasma as well as peripheral blood mononuclear cells (PBMCs) which were isolated using standard density gradient centrifugation and subsequently cryopreserved until the day of analysis. Eight healthy blood donors were included for comparison.

Cytokine measurements Ip-10 and Il-18Levels of IP-10 and IL-18 were measured in available plasma samples with a DuoSet ELISA (R&D Systems, Minneapolis, MN, USA). Values were extrapolated from standard curves (IP-10 range: 62.6- 4000 pg/ml, IL-18 range: 11.7-750 pg/ml). Plasma samples from 6 healthy controls were included in the analyses.

Immune phenotyping by flowcytometryPBMCs were washed in PBA (PBS containing 0.01% (w/v) NaN3, 0.5% (w/v) bovine serum albumin and 2 mM EDTA) and 1.0 x 106 cells were incubated for 30 min in the dark at 4°C with different combinations of fluorescent label-conjugated mouse monoclonal antibodies (mAbs). For phenotypic analysis, the following mAbs were used: CD3 V500, CD56 BUV395, CD16 BV786, CD16 BV421, CD27 BUV737, HLA-DR FITC, CD38 PE-Cy7, PD-1 BV421, CD14 PE-CF594, CD19 PE-CF594 (BD Biosciences, San Jose, USA), CD8 BV711 CD8 BV785, CD57 Alexa Fluor 647 (Biolegend), life/dead fixable red stain (Invitrogen, Camarillo, USA), NKp46 PerCP-efluor 710, CD45RA efluor 605 (eBioscience, San Diego, USA), NKG2A PE (Beckman Coulter, Fullerton, CA, USA) and TRAIL APC (Miltenyi Biotec, Bergisch Gladbach, Germany). For intracellular staining, cells were fixed after surface staining with FACS Lysing Solution (BD) and subsequently permeabilized (FACS Permeabilizing Solution 2 (BD)). Cells were incubated for 30 min in the dark at 4oC with one or more of the following antibodies: perforin FITC (BD Biosciences), granzyme B PE (Sanquin, Amsterdam, The Netherlands), Ki67 BV711 (Biolegend, San Diego, CA, USA), Eomes PerCP-efluor 710, and T-bet Pe-Cy7 (eBioscience, San Diego, USA). Measurements were done using LSR Fortessa flow cytometer (BD Biosciences, Europe) and FACS Diva Software. Analysis was done using FlowJo v10 (FlowJo LLC, Ashland, USA).

29


2

Intracellular staining hBv-specific t cellsPBMC were cultured for 10 days with 5 pools of in total 315 15-mer peptides (10 overlapping residues) covering all proteins of HBV genotype A (Chiron Mimotopes, Victoria, Australia) at 1 µg/ml. PBMC were restimulated for 6 hours at day 10 in the presence of CD107a PE (BD Biosciences, San Jose, CA, USA), Brefeldin A, and monensin. The production of cytokines was evaluated by intracellular staining with IFN-γ BV421, MIP-1β Pe-Cy7, TNF-α AF700, and IL-2 APC monoclonal antibodies (BD Biosciences, San Jose, CA, USA) after staining with surface markers as described above. Measurements and analyses were done as described above.

statistical analysesThe two-tailed Mann-Whitney U test was used for analysis of differences between groups. For longitudinal analysis in individual patients the Wilcoxon signed rank test was used. P values <0.05 were considered statistically significant. GraphPad Prism version 6.07 for Windows (GraphPad Software, La Jolla, CA, USA) was used for analyses.

results patients Nine patients with acute hepatitis B infection were included in the study (for baseline characteristics, see Table 1). Patients were infected with genotype A (n=5), or genotype B, D, E, or F (all n=1). At BL, median HBV DNA load was 5.12 log IU/mL (iqr: 3.80-6.64, Figure 1A) and median ALT was 2,652 U/mL (iqr: 1,554-3,390 U/mL Figure 1B). Six months after infection, 8 of 9 patients had spontaneously cleared the virus, of which 6 had formation of anti-HBs antibodies. One patient, infected with HBV genotype A, did not clear HBsAg within 6 months. At 6 months after initial presentation the viral load in this patient was > 1.7x108 IU/mL (upper limit of quantification) and ALT was 454 U/mL (Figure 1A,B). In a subset of patients, IP-10 and IL-18 levels were measured. At baseline, the plasma levels of IP-10 were increased in patients with acute HBV infection

table 1. Baseline characteristics

patient Genotype Gender ageBaseline hBv dna (10log Iu/ml) alt (u/l) Bilirubin (µmol/l)

Clearance < 6 months

1 D M 30 5.16 2652 n.a. yes2 B F 42 3.48 2126 78 yes3 A M 51 8.23 1631 30 no4 A M 29 4.06 690 393 yes5 A F 54 5.08 3514 194 yes6 A M 53 6.79 3016 246 yes7 F M 49 6.49 3970 257 yes8 A M 38 4.14 1528 338 yes9 E M 42 3.53 2686 219 yes

30

Chapter 2

as compared to healthy controls (median 1613.0 and 39.4 pg/mL respectively, p=0.0007, Figure 1C). In addition, baseline IL-18 levels were significantly increased in patients with AHB infection (median 1182 pg/mL) as compared to healthy controls (median 124.3 pg/ml, p=0.002, Figure 1D). At week 24, IP-10 levels had normalised, while IL-18 levels were still increased as compared to healthy controls (median 306.0 and 124.3 pg/mL respectively, p=0.007 Figure 1D).

Early time points after infection (BL and week 1) showed an increase in the proportion of CD56bright NK cells (BL median 7.4%) as compared to healthy controls (median 1.4%, p=0.0028, Figure 2A). Furthermore, the proportion of CD56dim NK cells was decreased (Figure 2B) early during infection, while the proportion of total NK and total CD8+ T cells was not significantly different from that of healthy controls (Figure 2C,D). There was no significant change in the proportion of effector and memory CD8+ T cells. However, the memory T cell population was activated, as demonstrated by the significant increase in PD-1, Ki67, HLA-DR/CD38, perforin and granzyme B positive memory T cells (Supplementary Fig 1).

figure 1. Course of HBV DNA (A), ALT (B), IP-10 (C) and IL-18 (D) in patients with acute HBV infection. Not all patient samples were available for IP-10 (n=5) and IL-18 (n=4) measurements. Statistical testing: Mann-Whitney U test. lloq; lower limit of quantification, uloq; upper limit of quantification; ns, non-significant; ***, p < 0.001; **, p < 0.01.

(a) (B)

(C) (d)

31


2

early nK cell activationTo investigate the role of NK cells at different time points during acute infection we measured the expression of several markers of NK cell activation. At baseline, the proportion of CD56bright NK cells expressing CD38 was significantly increased (median 44.0%) as compared to healthy controls (median 30.3%, p=0.01, Figure 3A). Similarly, CD38 was expressed on 56.2% of CD56dim NK cells in AHB patients at baseline as compared to 42.0% of HC CD56dim NK cells (p=0.03, Figure 3A). At baseline, HLA-DR was expressed on an increased proportion of CD56dim NK cells (median 18.3%) in AHB patients as compared to healthy controls (median 4.6%, p=0.0008). The proportion of HLA-DR+ CD56dim NK cells was still increased 1 week after presentation of AHB infection (Figure 3B). Ki67, a marker for proliferation, was not differentially expressed between baseline NK cells from AHB patients and healthy controls. At week 4 and week 12 however, the proportion of Ki67+ CD56dim NK cells was increased as compared to healthy controls (Figure 3C). The differentiation status of NK cells, as measured by CD57 and NKG2A expression, was

figure 2. Proportion of CD56bright (C) and CD56dim (D) NK cells, total NK cells (C), and total CD8+ T cells (D) in 9 patients with acute HBV infection as well as in healthy controls (HC). Statistical testing: Mann-Whitney U test. ns, non-significant; **, p < 0.01.

(a) (B)

(C) (d)

32

Chapter 2

not different in patients with acute hepatitis B infection as compared to healthy controls (Supplementary Fig 2).

figure 3. Markers of NK cell activation CD38 (A), HLA-DR (B), Ki67 (C) in CD56bright (left) and CD56dim (right) NK cells. Statistical testing: Mann-Whitney U test. ns, non-significant; ***, p < 0.001; *, p < 0.05.

(a)

(B)

(C)

33


2

long term increase in effector markersThe proportion of CD56bright NK cells expressing NKp46, an activating receptor, was significantly increased at all time points during acute HBV infection, as compared to healthy controls (AHB baseline median 96% and HC 81.6%, p<0.0001, Figure 4A). This was similar for CD56dim NK cells (Figure 4A). The TNF related apoptosis inducing ligand TRAIL is generally expressed by a minority of CD56bright NK cells. At baseline acute HBV infection, the proportion of TRAIL+ CD56bright NK cells was significantly increased as compared to healthy controls (median 11.7 and 2.2% respectively, p=0.0028, Figure 4B). Furthermore, TRAIL expression in the patient that was chronically infected was still elevated at week 24 after initial presentation (29.5% of CD56bright NK cells, Figure 4B). During acute HBV infection the proportion of granzyme B and perforin positive CD56dim NK cells was significantly increased compared to healthy controls (granzyme B baseline AHB 71.7 and HC 48.6%, p=0.04. Perforin baseline AHB 82.4% and HC 63.3%, p=0.01, Figure 4C,D). In addition, granzyme B was expressed by a considerate proportion of CD56bright NK cells, which are generally thought to play a lesser role in cytotoxicity (Figure 4C). In particular, a high proportion of granzyme B expressing CD56bright NK cells remained present in the patient who evolved to chronic infection (Figure 4C).

hBv-specific t cell responsesHBV-specific T cell responses, measured by cytokine production in response to HBV peptide pools, were analysed in 5 patients who were infected with HBV genotype A. Even though intra-individual differences were observed in the specificity and timing of the response, HBV-specific T cell responses could be observed in all patients that spontaneously cleared the HBV infection during the 6 months follow-up (Figure 5A-D). In the patient who became chronically infected however, very low HBV-specific T cell responses were observed at all time points (Figure 5E). Patients that cleared the HBV infection could have an early peak of HBV-specific CD8+ responses at week (patient 4 and 6) or a later peak (patient 5 and 8). The specificity of the T cell response also differed between patients. In patients 4 and 5 the highest observed sum of cytokine responses was observed in response to in the polymerase peptide pool, whereas in patient 6 the highest cytokine production was against the envelope pool, and in patient 8 against the core peptide pool. CD4+ T cell responses mostly followed the same patterns (Supplementary Figure 3).

dIsCussIonHere were show the temporal dynamics of NK cells and T cells in nine patients during acute hepatitis B infection, of which one developed chronicity. HBV is a non-cytopathic virus, however, immune responses mounted by the host can cause serious liver damage. Whereas viral clearance can occur without clinical symptoms or cell destruction,4,11 in our patient cohort hepatocyte damage was apparent by ALT elevations. As we observed a significant increase in the proportion of total peripheral memory T cells expressing makers of activity and cytolytic proteins, HBV-nonspecific bystander T cells could have

34

Chapter 2

figure 4. Effector markers NKp46 (A), TRAIL (B), Granzyme B (C) and Perforin (D) in CD56bright and CD56dim NK cells. Statistical testing: Mann-Whitney U test. ns, non-significant; ****, p < 0.0001; ***, p < 0.001; **, p < 0.01; *, p < 0.05.

(a)

(B)

(C)

(d)

35


2

played a role in hepatocyte injury,16,17 as seen in mouse models of acute fulminant HBV infection.18

At the time of presentation at the clinic, plasma levels of IP-10, an interferon stimulated gene (ISG), were significantly increased suggesting broad innate immune activation via the interferon pathway. Previous studies have shown no significant induction of type I interferons in the initial phases of acute HBV infection in chimpanzees8 or woodchucks.9

figure 5. HBV-specific CD8+ T cell functionality was assessed in patients who were infected with HBV genotype A. TNFα, IL-2, MIP-1β and IFNγ production by HBV-specific CD8+ T cells in 4 patients who cleared AHB infection (A-D) and 1 patient who evolved to chronicity (E).

(a) (B)

(C) (d)

(e)

36

Chapter 2

Also in man, type I interferon production was not observed early during infection,10 still, it has not been ruled out that other interferon types such as IFN-λ are induced.10 As we included patients who already were symptomatic, it is more likely that local danger signals or production of interferons by lymphocytes have activated the ISG pathway.8

Early during the infection, the proportion of CD56bright NK cells was significantly increased as compared to healthy controls. Furthermore, we observed a striking increase in the proportion of CD38, HLA-DR and Ki67 positive NK cells, indicating activation of these cells. Previously, it was shown that in patients with symptomatic infection, NK cell function in inhibited before peak viremia, and NK cells only become activated after peak viremia.10 As we only included patients after peak viremia, we may have missed the described inhibition of NK cells in the preclinical phase of infection. The proportion of NK cells expressing perforin and granzyme B was increased in patients with an acute HBV infection. While in mice, hepatocytes infected with a recombinant adenovirus have been suggested to be resistant to perforin mediated killing,19 the increase in the proportion of NK cells expressing cytotoxic molecules suggests they could have a role in hepatocyte damage. As such, an increase in perforin expression by HBV-specific T cells has been observed in acute HBV infection.20

As the proportion of TRAIL expressing CD56bright NK cells was significantly increased, TRAIL mediated killing of infected hepatocytes could also be responsible for ALT elevations.21 However, in two previously described patients who cleared the infection without any symptoms or ALT elevations, NK cells were activated early in the course of acute HBV infection, emphasizing that these cells can be important in non-cytopathic elimination of HBV.11 While Ki67, HLA-DR, and CD38 normalised after clearance of the virus, other NK cell markers were still increased at week 24 as compared to healthy controls, including NKp46, TRAIL, perforin, and granzyme B. In line with this, baseline elevated levels of IL-18, which is associated with NK cell activation,22 were still significantly elevated at week 24. Interestingly, in patients with chronic hepatitis B, TRAIL expression is significantly elevated.23 Similarly, in the patient who evolved to chronic infection, TRAIL remained expressed on a significant proportion of CD56bright NK cells. Whether this TRAIL positive population is a cause or a result of chronic infection remains unanswered. During chronic infection, TRAIL has been suggested to play a role in killing of infected hepatocytes,21 as well as in NK cell mediated clearance of HBV-specific T cells.24

Even though HBV-specific T cells seem to be inhibited by IL-10 and arginase early during infection,10,25 their presence is highly associated with viral clearance.5,10,26 From acute HCV infection, we know that evolution to chronicity is associated with weak and transient responses of HCV-specific T-cells.14,27 Clearance of the infection on the other hand, is associated with the appearance of multi-specific HCV-specific CD8+ T-cell responses against multiple epitopes.27 However, in two blood donors with acute HBV infection who were followed during asymptomatic infection, HBV-specific T cell responses reached their peak when HBV DNA was already declining, emphasizing the importance of other immune mechanisms for antiviral activity.11 Early T cell responses in these patients were directed against envelope and polymerase proteins.11 In 2 other patients, early responses

37


2

were directed against polymerase and the X protein, followed by responses against core and envelope.10 Here we observed broad reaction to all HBV proteins in 2 patients at the earliest time point (patient 4 and 6). A more delayed and narrow HBV- specific T cell response was observed in 2 other patients, while they already showed ALT elevation and viral load decline (patient 5 and 8). Previously, a lack of T cell responses was associated with persistently high HBV DNA levels and the need for treatment in an immunosuppressed patient.15 In the one patient who did not clear HBV infection, the observed narrow T cell response may have led to chronic infection. This is in line with evolution of chronicity in woodchucks infected with the woodchuck hepatitis virus,28 as well as observations in acute hepatitis C patients.14

Acute HBV infection can present in many different ways. With or without ALT elevations, and with or without symptoms. Therefore, the underlying mechanisms may also differ between cases. As one of the patients in this study developed chronic infection we had the unique opportunity to analyse early events that could be associated with failure to clear HBV. Our data suggests that the absence of a HBV-specific T cells response at all time points could play a role in progression to chronic infection. Furthermore a substantial population of CD56bright NK cells expressing TRAIL, as seen in chronic HBV infection, was present at all time points in this patient. A better understanding of the early course of acute infection and the evolution to chronicity may help the development of novel therapeutics targeting chronic hepatitis B virus infection.

38

Chapter 2

referenCes1. Schweitzer A, Horn J, Mikolajczyk RT, et

al. Estimations of worldwide prevalence of chronic hepatitis B virus infection: A systematic review of data published between 1965 and 2013. Lancet 2015;386:1546–55.

2. Ott JJ, Stevens GA, Groeger J, et al. Global epidemiology of hepatitis B virus infection: New estimates of age-specific HBsAg seroprevalence and endemicity. Vaccine 2012;30:2212–9.

3. Revill P, Yuan Z. New insights into how HBV manipulates the innate immune response to establish acute and persistent infection. Antivir Ther 2013;18:1–15.

4. Guidotti LG, Rochford R, Chung J, et al. Viral clearance without destruction of infected cells during acute HBV infection. Science 1999;284:825–9.

5. Thimme R, Wieland S, Steiger C, et al. CD8(+) T cells mediate viral clearance and disease pathogenesis during acute hepatitis B virus infection. J Virol 2003;77:68–76.

6. Ferrari C, Penna A, Bertoletti A, et al. Cellular immune response to hepatitis B virus-encoded antigens in acute and chronic hepatitis B virus infection. J Immunol 1990;145:3442–9.

7. Ferrari C. HBV and the immune response. Liver Int 2015;35:121–8.

8. Wieland S, Thimme R, Purcell RH, et al. Genomic analysis of the host response to hepatitis B virus infection. Proc Natl Acad Sci U S A 2004;101:6669–74.

9. Fletcher SP, Chin DJ, Ji Y, et al. Transcriptomic analysis of the woodchuck model of chronic hepatitis B. Hepatology 2012;56:820–30.

10. Dunn C, Peppa D, Khanna P, et al. Temporal analysis of early immune responses in patients with acute hepatitis B virus infection. Gastroenterology 2009;137:1289–300.

11. Fisicaro P, Valdatta C, Boni C, et al. Early kinetics of innate and adaptive immune

responses during hepatitis B virus infection. Gut 2009;58:974–82.

12. Rehermann B. Natural Killer Cells in Viral Hepatitis. C Cell Mol Gastroenterol Hepatol 2015;1:578–88.

13. Rehermann B, Lau D, Hoofnagle JH, et al. Cytotoxic T lymphocyte responsiveness after resolution of chronic hepatitis B virus infection. J Clin Invest 1996;97:1655–65.

14. Urbani S, Amadei B, Fisicaro P, et al. Outcome of acute hepatitis C is related to virus-specific CD4 function and maturation of antiviral memory CD8 responses. Hepatology 2006;44:126–39.

15. Webster GJM, Reignat S, Maini MK, et al. Incubation Phase of Acute Hepatitis B in Man: Dynamic of Cellular Immune Mechanisms. Hepatology 2000;32:1117–24.

16. Abrignani S. Bystander activation by cytokines of intrahepatic T cells in chronic viral hepatitis. Semin Liver Dis 1997;17:319–22.

17. Maini MK, Boni C, Lee CK, et al. The role of virus-specific CD8(+) cells in liver damage and viral control during persistent hepatitis B virus infection. J Exp Med 2000;191:1269–80.

18. Ando BK, Moriyama T, Guidotti LG, et al. Mechanisms of class I restricted immunopathology. A transgenic mouse model of fulminant hepatitis. J Exp Med 1993;178:1541–54.

19. Kafrouni MI, Brown GR, Thiele DL. Virally infected hepatocytes are resistant to perforin-dependent CTL effector mechanisms. J Immunol 2001;167:1566–74.

20. Urbani S, Boni C, Missale G, et al. Virus-specific CD8+ lymphocytes share the same effector-memory phenotype but exhibit functional differences in acute hepatitis B and C. J Virol 2002;76:12423–34.

21. Dunn C, Brunetto M. Cytokines induced during chronic hepatitis B virus infection

39


2

promote a pathway for NK cell–mediated liver damage. J Exp Med 2007;204:667–80.

22. Serti E, Werner JM, Chattergoon M, et al. Monocytes activate natural killer cells via inflammasome-induced interleukin 18 in response to hepatitis C virus replication. Gastroenterology 2014;147:209–220.e3.

23. Peppa D, Micco L, Javaid A, et al. Blockade of immunosuppressive cytokines restores NK cell antiviral function in chronic hepatitis B virus infection. PLoS Pathog 2010;6(12):e1001227.

24. Peppa D, Gill U, Reynolds G. Up-regulation of a death receptor renders antiviral T cells susceptible to NK cell–mediated deletion. J Exp Med 2013;210:99–114.

25. Sandalova E, Laccabue D, Boni C, et al. Increased levels of arginase in patients with

acute hepatitis B suppress antiviral T cells. Gastroenterology 2012;143:78–87.e3.

26. Boettler T, Panther E, Bengsch B, et al. Expression of the Interleukin-7 Receptor Alpha Chain (CD127) on Virus-Specific CD8+ T Cells Identifies Functionally and Phenotypically Defined Memory T Cells during Acute Resolving Hepatitis B Virus Infection. J Virol 2006;80:3532–40.

27. Sung PS, Racanelli V, Shin E-C. CD8(+) T-Cell Responses in Acute Hepatitis C Virus Infection. Front Immunol 2014;5:266.

28. Menne S, Roneker CA, Roggendorf M, et al. Deficiencies in the Acute-Phase Cell-Mediated Immune Response to Viral Antigens Are Associated with Development of Chronic Woodchuck Hepatitis Virus Infection following Neonatal Inoculation. J Virol 2002;76:1769–80.

supplementary dataSupplementary data is available on request.

Immune responses In patIents wIth ChronIC hepatItIs B InfeCtIon

hBsaG loss after peGInterferon and nuCleotIde ComBInatIon treatment In ChronIC

hepatItIs B patIents: 5 years of follow-up

F. Stelma, M.H. van der Ree, L. Jansen, M.W. Peters, H.L.A. Janssen, H.L. Zaaijer, R.B. Takkenberg, H.W. Reesink


aBstraCt Background and aims Combining peginterferon alfa-2a (pegIFN) with a nucleotide analogue can result in higher rates of HBsAg loss than either therapy given alone. Here we investigated the durability of the response to combination therapy in chronic hepatitis B (CHB) patients after 5 years of follow-up.

methods In the initial study, 92 CHB patients (44 HBeAg positive, 48 HBeAg negative) with HBV DNA > 100,000 c/mL (~20,000 IU/mL) and active hepatitis were treated 48 weeks with pegIFN 180 mcg/week and 10 mg adefovir dipivoxil daily. For the long term follow-up study, patients were followed for 5 years after the end of treatment. At year 5, 70 (32 HBeAg positive, 38 HBeAg negative) patients were still included.

results At year 5, 19% (6/32) of HBeAg positive patients and 16% (6/38) of HBeAg negative patients had HBsAg loss, and no HBsAg sero-reversion was observed. The 5-year cumulative Kaplan-Meier estimate for HBsAg loss was 17.2% for HBeAg positive patients and 19.3% for HBeAg negative patients. 14/16 patients who lost HBsAg at any time point during follow-up developed anti-HBs antibodies (>10 IU/L). At year 5, in total 63% (20/32) of HBeAg positive and 71% (27/38) of HBeAg negative patients were retreated with nucleos(t)ide analogues during follow-up. The cumulative Kaplan-Meier estimate for retreatment was 60% of patients at year 5.

Conclusion At year 5 of follow-up, 18% of CHB patients treated with pegIFN/ nucleotide analogue combination therapy had durable HBsAg loss and 88% of these had developed anti-HBs antibodies.

45

5 YEAR FOLLOW-UP AFTER COMBINATION THERAPY

3

IntroduCtIonHepatitis B virus (HBV) infection is a major public health issue. Worldwide over 240 million patients are chronically infected with HBV.1 Patients who have chronic hepatitis B (CHB) virus infection are at increased risk of developing hepatic complications such as cirrhosis and hepatocellular carcinoma.2,3 Current treatment options for patients with CHB consist of pegylated interferon alfa (pegIFN) or nucleos(t)ide analogue (NUC) therapy. PegIFN is thought to directly interfere with HBV and boost the immune system and is usually given as a yearlong course. NUCs have a direct antiviral mode of action and generally have to be given lifelong. As it is particularly difficult to eradicate the covalently closed circular HBV DNA (cccDNA) from the infected hepatocyte, the outcome closest to cure is currently clearance of HBV DNA with HBsAg loss, referred to as ‘functional cure’. This outcome is rare, therefore treatment is mainly focused on reducing the viral load and liver inflammation. Achieving these goals is referred to as a ‘combined response’, in which the HBV DNA viral load is under 2,000 IU/mL and ALT levels are normal.4

Experimental data suggests a combination of pegIFN and a NUC may have supplementary effects on outcome.5 To improve treatment outcome, several trials have been conducted with combination regimens of pegIFN combined with a NUC.6 Recently, a multicenter randomized controlled trial was performed which compared pegIFN or NUC monotherapy with combination therapy in HBeAg positive or –negative CHB patients. At six months follow-up after the end of treatment 9% of patients had HBsAg loss in the pegIFN + tenofovir combination arm, as compared to 2.8% in the pegIFN mono-treatment arm.7 In the past, we have completed a study in which HBeAg positive and negative CHB patients were treated with a combination of pegIFN and adefovir. In this study, similar rates of HBsAg loss (9%) were observed at 6 months after the end of treatment.8

In our study, we observed increasing rates of HBsAg loss (13/92, 14%) at two years of follow-up after end of treatment. However, whether this proportion can further increase after more years of follow-up is unknown. Furthermore, the durability of response to combination therapy is unknown. We therefore assessed the outcome of pegIFN and NUC combination treatment after five years of follow-up.

patIents and methodspatientsInitial study: Ninety-two CHB patients were included in the initial study in the Academic Medical Center (AMC) in Amsterdam and the Erasmus University Medical Center (EMC) in Rotterdam (controlled-trials.com; ISRCTN 77073364).8 Patients were treated for 48 weeks with pegylated interferon alfa-2a once a week (Pegasys® 180 mcg subcutaneously; Hoffman La Roche, Basel, Switzerland) and adefovir dipivoxil daily (Hepsera® 10 mg orally; Gilead Sciences, Foster City, CA, USA). Patients that could be included were aged 18 years or older, had chronic HBV infection (documented HBsAg positivity for longer than six months, HBeAg-positive or -negative), had a viral load of HBV DNA >100,000 copies/ml (~20,000 IU/ml) and had normal or increased alanine aminotransferase (ALT) levels (≤10x upper limit of normal (ULN)) or histological signs of chronic active hepatitis.

46

Chapter 3

Follow-up (FU) study: Patients attended the outpatient clinic once a year up to five years after end of treatment for laboratory testing and routine examination. Screening for hepatocellular carcinoma (HCC) was performed routinely by abdominal ultrasound. Both parts of the study were conducted according to the guidelines of the Declaration of Helsinki and with the principles of Good Clinical Practice and were approved by the local ethics committee. All patients gave written informed consent. All authors had access to the study data and reviewed and approved the final manuscript.

outcomeThe primary outcome of the long term follow-up (LTFU) study was sustainability of the functional cure i.e. HBsAg loss, with or without anti-HBs antibody formation (defined as anti-HBs >10 IU/L). The secondary outcomes were defined as the need for re-treatment and the sustainability of ‘combined response’ (CR) defined as sustained viral suppression (HBV DNA <2,000 IU/ml) with ALT normalization. In HBeAg positive patients this included HBeAg loss.

laboratory testingLocal laboratories carried out biochemical and hematological analyses in accordance with good laboratory practice. ALT levels were expressed relative to the ULN range (45 U/l for males and 34 U/l for females). Plasma HBV DNA was extracted by COBAS® Ampliprep (F. Hoffmann-LaRoche Ltd, Diagnostics Division, Basel, Switzerland) according to the manufacturer’s instructions. Quantitation of plasma HBV DNA levels was done by the Roche COBAS® TaqMan 48® assay (F. Hoffmann-LaRoche Ltd, Diagnostics Division, Basel, Switzerland), with a dynamic range between 20 and 1.70×108 IU/ml.

Qualitative detection of serum hepatitis B surface antigen (HBsAg), antibody to hepatitis B surface antigen (anti-HBs), hepatitis B e antigen (HBeAg) and antibody to hepatitis B e antigen (anti-HBe) was performed by enzyme immunoassay (AxSYM; Abbott Laboratories, Abbott Park, IL, USA), and expressed as sample to cutoff ratio (S/CO) with a lower limit of detection for HBsAg of 0.05 IU/ml. Quantitative detection of serum HBsAg was performed using the Architect quantitative HBsAg assay (Abbott Diagnostics, Abbott Park, IL, USA) with a lower limit of detection of 0.05 IU/ml.

statistical analysesStatistical analyses were performed in SPSS (IBM SPSS Statistics for Windows, Version 23.0 Armonk, NY: IBM Corp.). Analyses were either based on the per protocol study population, which included all patients who were still included in the study at a specific time-point, or the initial study population which was analysed with Kaplan Meier estimates. Patients who became eligible for retreatment were regarded as non-responders in the LTFU study. Patients with missing data (no show for the study visit) were classified as non-responders (qualitative data) or excluded from the analysis (quantitative data). For Kaplan-Meier analyses, patients who were lost to follow up as well as patients who had not reached the endpoint at year five were censored. For the identification of factors influencing

47


3

HBsAg loss univariable Cox regression analysis was used. The hazard ratio (HR) and 95% confidence interval (CI) are given for factors that were associated with HBsAg loss. Statistical significance was based on a p-value below 0.05.

resultspatientsIn the initial pegIFN and adefovir combination treatment study, 92 CHB patients were included (baseline characteristics are depicted in Table 1). At two years after the end of treatment, 85 patients (92%) were still participating. Five years after end of treatment, 70 patients (76%) were still in follow-up. A per-year overview of the patients included in the LTFU study is given in Supplementary Figure 1. At year five, 22 patients were excluded from the study of which 18 patients were lost to follow-up (including 6 drop-outs during treatment). Of these, three had HBsAg loss with the formation of anti-HBs antibodies during last visit, four had a combined response during last visit, three were retreated with NUC due to relapse of HBV DNA with ALT elevation, and one had a HBV DNA relapse without an indication for retreatment. In addition, due to participation in another experimental treatment study, two patients were excluded from the LTFU study who at the time of exclusion had a combined response. Furthermore, two patients had died of which one had HBsAg loss with anti-HBs formation, and one was previously retreated due to relapse.

table 1. Baseline characteristics

hBeag pos n=44

hBeag neg n=48

Age, median (range) 35 (19-55) 44 (24-69)Gender male, n (%) 35 (80) 33 (69)Ethnicity Caucasian, n (%) 16 (36) 12 (25) Asian, n (%) 20 (46) 14 (29) African, n (%) 8 (18) 22 (46)IFN naive 35 (80) 34 (71)ALT xULN mean (SD) 4.27 (5.47) 2.19 (1.68)HBVDNA, log10 IU/mL mean (SD) 8.05 (1.21) 5.53 (1.10)HBsAg, log10 IU/mL mean (SD) 4.31 (0.73) 3.35 (0.67)Genotype A, n (%) 18 (41) 11 (23) B, n (%) 8 (18) 7 (15) C, n (%) 7 (16) 5 (10) D, n (%) 9 (21) 18 (38) E, n (%) 2 (5) 7 (15)Liver cirrhosisa 5 (11.4) 9 (18.8)

a liver cirrhosis in biopsy.

48

Chapter 3

hBsag lossFrom year two to year five of follow-up, three additional patients lost HBsAg, of which two had development of anti-HBs antibodies. These patients all had a combined response before becoming HBsAg negative. One of these patients (HBeAg negative at baseline) lost HBsAg at year three after end of treatment, one at year four, and one at year five after end of treatment (the latter were both HBeAg positive at baseline). None of these patients received any additional antiviral treatment after the cessation of the study medication. No HBsAg sero-reversions were observed in this study.

In a per protocol analysis at year five, 18.8% (6/32) of HBeAg positive patients had HBsAg loss (of which five had developed anti-HBs), and 15.8% (6/38) of HBeAg negative patients had HBsAg loss (of which five had developed anti-HBs). A year to year overview of patients with HBsAg loss is depicted in Table 2.

At year five the cumulative Kaplan-Meier estimate for HBsAg loss was 17.2% for HBeAg positive patients and 19.3% for HBeAg negative patients (Figure 1). In total 16 patients lost HBsAg of which 12 patients reached end of follow-up (year five); the other four patients had died (n=2), or were lost to follow up (n=2). 14/16 patients (87.5%) who lost HBsAg developed anti-HBs antibodies at some point during follow-up. Patients with HBsAg loss were infected with HBV genotype A (n=9), B (n=1), C (n=2), D (n=2), and E (n=2).The baseline factors associated with HBsAg loss are depicted in Table 3. In patients who were HBeAg positive at baseline, factors associated with HBsAg loss were age and HBV genotype A. In patients who were HBeAg negative at baseline, HBsAg loss was significantly associated with HBV DNA, and HBsAg levels at baseline.

Quantitative hBsag During LTFU, plasma levels of HBsAg decreased in patients who were not retreated. Patients who were already HBsAg negative at year two of follow-up were excluded from

0

10

20

30

Years of follow-up

Kap

lan-

Mei

eres

timat

eof

HB

sAg

loss

(%) HBeAg negative

HBeAg positive

1 2 3 4 5EoTremaining

casesHBeAg +HBeAg -

4448

4045

4044

3739

3636

3333

2932

19.3%17.2%

figure 1. Kaplan-Meier estimate of HBsAg loss in HBeAg positive and HBeAg negative patients (intention-to-treat population).

49


3

this analysis. Likewise, in patients who were retreated with NUCs, HBsAg levels declined

over time (Supplementary figure 2), but in this group HBsAg loss was not observed.

ltfu outcome – combined response At year two of follow-up after cessation of therapy, in total 11 of 85 patients (12.9%)

had a combined response with HBV DNA < 2,000 IU/ml and ALT normalization, including

HBeAg loss for HBeAg positive patients. This excluded patients with HBsAg loss and

patients that met the criteria for combined response after having been retreated with

NUC. Of these 11 patients, two had a combined response at year five, while one patient

relapsed (HBV DNA > 2,000 IU/mL). The other eight patients with a combined response at

year two either lost HBsAg (n=3), were included in another treatment study (n=2), or were

lost to follow-up (n=3). Between years two and five of follow-up, three additional patients

(two HBeAg negative, one HBeAg positive at baseline) met the criteria for combined

response without being retreated.

table 2. Year to year overview of patients and outcome during follow-up in a per protocol analysis.

hBeag positive hBeag negative

year 2 n (%)

year 3 n (%)

year 4 n (%)

year 5 n (%)

year 2 n (%)

year 3 n (%)

year 4 n (%)

year 5 n (%)

Patients n= 41 38 34 32 44 42 39 38HBsAg loss 5 (12) 4 (11) 4 (12) 6 (19) 8 (18) 9 (21) 7 (18) 6 (16)Anti-HBs* 4 (80) 3 (75) 3 (75) 5 (83) 7 (88) 5 (56) 3 (43) 2 (33)HBeAg loss** 18 (44) 14 (37) 13 (38) 12 (38) n.a. n.a. n.a. n.a.HBeAg conversion 15 (37) 10 (26) 7 (21) 7 (22) n.a. n.a. n.a. n.a.Combined response 7 (17) 6 (16) 4 (12) 2 (6.3) 4 (9.1) 2 (4.8) 1 (2.6) 3 (7.9)Retreated 20 (49) 20 (53) 20 (59) 20 (63) 25 (57) 28 (67) 27 (69) 27 (71)HBsAg loss after NUC retreatment

0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

*anti-HBs as a percentage of patients with HBsAg loss. **n=1 HBeAg sero-reversion observed.

table 3. Univariable Cox regression analysis in HBeAg positive and –negative patients

factor

hBeag positive patients hBeag negative patients

hazard ratio CI p-value hazard ratio CI p-value

Age 1.09 1.00-1.19 0.048 1.06 0.98-1.14 0.15Baseline HBsAg log10 IU/mL 0.88 0.34-2.30 0.80 0.13 0.05-0.38 0.0002Baseline HBV DNA log10 IU/mL 1.10 0.59-2.05 0.77 0.47 0.22-0.98 0.04Baseline ALT (xULN) 1.00 0.88-1.14 0.99 1.03 0.69-1.52 0.90HBV genotype A 0.10 0.12-0.82 0.03 0.53 0.13-2.11 0.37Interferon naïve 0.58 0.07-4.78 0.61 3.45 0.92-12.87 0.07

50

Chapter 3

In a per protocol analysis, at the end of LTFU 5/70 (7%) of patients had a durable

combined response five years after having received combination therapy (two HBeAg

positive, three HBeAg negative) (Table 2).

ltfu outcome – retreatment Between year two and year five of follow-up, six additional patients became eligible for

retreatment due to an increase in HBV DNA and ALT level. These patients all had a relapse

of HBV DNA without the indication for retreatment at year two of follow-up. Only one of

these patients had a combined response at some time point after treatment (six months

after the end of treatment) but HBV DNA levels relapsed after that, meeting treatment

criteria at year three of follow-up. At two years follow-up after stopping treatment, 45/85

(52.9%) patients who were participating in the trial had been retreated with a NUC. At year

five, 62.5% (20/32) of HBeAg positive and 71.1% (27/38) of HBeAg negative patients were

retreated during follow-up (Table 2).

Of the 86 patients who completed 48 weeks of treatment, 51 patients were retreated.

The cumulative Kaplan-Meier estimate for retreatment was 60% (54.7% for HBeAg positive

and 64.9% for HBeAg negative patients, p=0.20) at year five (Figure 2). Half of the patients

had been retreated at year two of follow-up after the cessation of treatment (95% CI

0.6-3.4 years). No HBsAg loss was observed in any of the retreated patients (Table 2).

ltfu clinical outcome During follow-up two patients died, of which one patient had a myocardial infarction,

and one patient died due to a drug overdose. One patient, who had cirrhosis at baseline,

developed HCC during follow-up, shortly after the end of treatment. No other patients

developed HCC between year two and year five. Furthermore, no hepatic decompensation

was observed in 14 patients with cirrhosis (including two dropouts) during follow-up in

the study. At year five, nine patients with cirrhosis were still included in the study.

0

20

40

60

80

100

Years of follow-up

Kap

lan-

Mei

eres

timat

eof

retr

eatm

ent(

%) HBeAg negative

HBeAg positive

1 2 3 4 5EoTremaining

casesHBeAg +HBeAg -

4448

4145

3833

2124

2119

1914

1512

64.9%54.7%

figure 2. Kaplan-Meier estimate of retreatment in HBeAg positive and HBeAg negative patients (intention-to-treat population).

51


3

dIsCussIonIn this study we showed that the high rate of HBsAg loss that was observed 2 years after the cessation of treatment in CHB patients was sustained until year 5 of follow-up. From year two to five of follow-up, only three additional patients lost HBsAg resulting in 17 and 19% HBsAg loss after five years in HBeAg positive and -negative patients respectively. Durable HBsAg loss is the closest to a cure and associated with improved survival and a lower incidence of HCC.9–11

In the past, several controlled studies have analysed the effect of pegIFN and NUC combination treatment for patients with CHB.6,7,12,13 Long term follow-up data on the efficacy of such regimens however, is scarce. In a recent study, rates of HBsAg loss were 9.1% at six months after the end of treatment in patients treated with pegIFN and tenofovir, compared to 2.8% in the pegIFN monotherapy arm.7 This short term follow-up outcome reflects the rate of HBsAg loss (9%) that also was observed in our study 6 months after the cessation of treatment. This could suggest that, in contrast to when used as monotherapy, adefovir may not be inferior to tenofovir when used in combination with pegIFN.14 In another study in HBeAg negative patients, HBsAg loss was observed in 8.7% of patients three years after the end of treatment with pegIFN and lamivudine therapy, as compared to 0% in lamivudine-only treated patients.13 A study in HBeAg positive patients reported 15% HBsAg loss in patients treated with pegIFN (alfa-2b) and lamivudine combination treatment after a mean of three years follow-up, compared to 8% in patients treated with pegIFN monotherapy, however this difference was not significant.12 In the absence of a monotherapy arm in our study, comparison of data to these results is difficult, even so, the rates of HBsAg loss in the combination treatment arms appear to be comparable with the results of other combination therapy studies.

In our study, 11/85 (13%) of patients had a combined response at year 2 after the cessation of therapy (excluding patients with HBsAg loss). This outcome was sustainable as only one of these patients had HBV DNA relapse (>2,000 IU/mL). The remaining patients with a previous combined response had a durable combined response during LTFU, had HBsAg loss or were lost to follow-up.

One of the limitations of our study was that after five years of follow-up, only 76% of patients were still included, introducing the risk of a selection bias. However, the participation rate was still relatively high, as compared to other studies reporting 59-65% CHB patients still participating three years after end of treatment.12,13 To compensate for the cases lost to follow-up, we analyzed the data not only in a per protocol analysis but also calculated Kaplan Meier estimates in which all patients are taken into account, and patients lost to follow-up are censored.

In the past, many studies have assessed predictive markers of response to therapy.15–17 Due to a low number of patients, the study was underpowered for multivariable analyses of factors associated with HBsAg loss. Univariable analyses resulted in a high level of significance for low baseline HBsAg level as a predictor for HBsAg loss in HBeAg negative patients. Measuring quantitative HBsAg levels has previously been suggested to be of help in making clinical decisions. Others found that during therapy, an early drop in HBsAg

52

Chapter 3

levels during pegIFN treatment was associated with a sustained therapy response in HBeAg positive and negative patients.17–20 Here, baseline HBsAg levels were significantly correlated with HBsAg loss in HBeAg negative patients, confirming the data from our initial study. During LTFU, HBsAg levels showed a stable decline over time.

Other factors associated with HBsAg loss were HBV genotype A in HBeAg positive patients, as previously shown,12,21 while in HBeAg negative patients, this was not associated with response.8,13

None of the patients who became eligible for retreatment with NUC therapy during LTFU (n=6) had a combined response at two years after end of treatment. The cumulative probability for retreatment was 60% after five years. Retreatment occurred mostly in the first two years after the end of treatment, which is consistent with previous data on interferon based treatment in which HBV DNA relapse was usually seen within the first year after treatment.13

No HBsAg loss was observed in patients who were retreated with NUC therapy. Similarly, when comparing pegIFN and lamivudine combination therapy with lamivudine monotherapy in HBeAg negative patients, no HBsAg loss was observed in the lamivudine monotherapy arm.13

In conclusion, five years after the end of treatment with pegIFN and NUC combination treatment, a durable functional cure was achieved in approximately 18% of CHB patients. Even though high levels of HBsAg loss were observed in this study, the majority of patients had to be retreated with NUC therapy. There is an urgent need for improved treatment of CHB patients, in order to achieve a durable functional cure. Hopefully the recent advances in development of new drugs targeting the HBV replication cycle will lead to more effective treatment options for CHB patients to achieve a durable functional cure.

53


3

referenCes1. Ott JJ, Stevens GA, Groeger J, Wiersma

ST. Global epidemiology of hepatitis B virus infection: new estimates of age-specific HBsAg seroprevalence and endemicity. Vaccine. 2012 Mar 9;30(12):2212–9.

2. Iloeje UH, Yang H-I, Su J, Jen C-L, You S-L, Chen C-J. Predicting cirrhosis risk based on the level of circulating hepatitis B viral load. Gastroenterology. 2006 Mar;130(3):678–86.

3. Chen C-J, Yang H-I, Su J, Jen C-L, You S-L, Lu S-N, et al. Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level. JAMA. 2006 Jan;295(1):65–73.

4. European Association For The Study Of The Liver. EASL clinical practice guidelines: Management of chronic hepatitis B virus infection. J Hepatol. 2012 Jul;57(1):167–85.

5. Thimme R, Dandri M. Dissecting the divergent effects of interferon-alpha on immune cells: time to rethink combination therapy in chronic hepatitis B? J Hepatol. 2013 Feb;58(2):205–9.

6. Viganò M, Invernizzi F, Grossi G, Lampertico P. Review article: the potential of interferon and nucleos(t)ide analogue combination therapy in chronic hepatitis B infection. Aliment Pharmacol Ther. 2016 Oct;44(7):653–61.

7. Marcellin P, Ahn SH, Ma X, Caruntu FA, Tak WY, Elkashab M, et al. Combination of Tenofovir Disoproxil Fumarate and Peginterferon α-2a Increases Loss of Hepatitis B Surface Antigen in Patients With Chronic Hepatitis B. Gastroenterology. 2016 Jan;150(1):134–144.e10.

8. Takkenberg RB, Jansen L, de Niet A, Zaaijer HL, Weegink CJ, Terpstra V, et al. Baseline hepatitis B surface antigen (HBsAg) as predictor of sustained HBsAg loss in chronic hepatitis B patients treated with peginterferon alfa-2a and adefovir. Antivir Ther. 2013 May;18:895–904.

9. Moucari R, Korevaar A, Lada O, Martinot-Peignoux M, Boyer N, Mackiewicz V, et al. High rates of HBsAg seroconversion in HBeAg-positive chronic hepatitis B patients responding to interferon: A long-term follow-up study. J Hepatol. 2009;50(6):1084–92.

10. Fattovich G, Olivari N, Pasino M, D’Onofrio M, Martone E, Donato F. Long-term outcome of chronic hepatitis B in Caucasian patients: mortality after 25 years. Gut. 2008;57(1):84–90.

11. Feld JJ, Wong DKH, Heathcote EJ. Endpoints of therapy in chronic hepatitis B. Hepatology. 2009;49(5 Supp):S96–102.

12. Buster EHCJ, Flink HJ, Cakaloglu Y, Simon K, Trojan J, Tabak F, et al. Sustained HBeAg and HBsAg Loss After Long-term Follow-up of HBeAg-Positive Patients Treated With Peginterferon α-2b. Gastroenterology. 2008;135(2):459–67.

13. Marcellin P, Bonino F, Lau GKK, Farci P, Yurdaydin C, Piratvisuth T, et al. Sustained Response of Hepatitis B e Antigen-Negative Patients 3 Years After Treatment with Peginterferon Alfa-2a. Gastroenterology. 2009;136(7):2169–79.

14. Marcellin P, Heathcote EJ, Buti M, Gane E, de Man R a, Krastev Z, et al. Tenofovir disoproxil fumarate versus adefovir dipivoxil for chronic hepatitis B. N Engl J Med. 2008;359(23):2442–55.

15. Jansen L, de Niet A, Stelma F, van Iperen EPA, van Dort KA, Plat-Sinnige MJT, et al. HBsAg loss in patients treated with peginterferon alfa-2a and adefovir is associated with SLC16A9 gene variation and lower plasma carnitine levels. J Hepatol. 2014 Oct;61(4):730–7.

16. Jansen L, Kootstra NA, van Dort KA, Takkenberg RB, Reesink HW, Zaaijer HL. Hepatitis B Virus Pregenomic RNA Is Present in Virions in Plasma and Is Associated With a Response to Pegylated Interferon Alfa-2a and Nucleos(t)ide Analogues. J Infect Dis. 2016 Jan 15;213(2):224–32.

54

Chapter 3

17. Sonneveld MJ, Hansen BE, Piratvisuth T, Jia J, Zeuzem S, Gane E, et al. Response-guided peginterferon therapy in hepatitis B e antigen-positive chronic hepatitis B using serum hepatitis B surface antigen levels. Hepatology. 2013;58(3):872–80.

18. Moucari R, Mackiewicz V, Lada O, Ripault MP, Castelnau C, Martinot-Peignoux M, et al. Early serum HBsAg drop: A strong predictor of sustained virological response to pegylated interferon alfa-2a in HBeAg-negative patients. Hepatology. 2009;49(4):1151–7.

19. Rijckborst V, Hansen BE, Cakaloglu Y, Ferenci P, Tabak F, Akdogan M, et al. Early on-treatment prediction of response to

peginterferon alfa-2a for HBeAg-negative chronic hepatitis B using HBsAg and HBV DNA levels. Hepatology. 2010;52(2):454–61.

20. Marcellin P, Bonino F, Yurdaydin C, Hadziyannis S, Moucari R, Kapprell HP, et al. Hepatitis B surface antigen levels: Association with 5-year response to peginterferon alfa-2a in hepatitis B e-antigen-negative patients. Hepatol Int. 2013;7(1):88–97.

21. Buster EHCJ, Hansen BE, Lau GKK, Piratvisuth T, Zeuzem S, Steyerberg EW, et al. Factors that predict response of patients with hepatitis B e antigen-positive chronic hepatitis B to peginterferon-alfa. Gastroenterology. 2009 Dec;137(6):2002–9.

55


3

fundInGThis study was funded by Roche.

supplementary data

1

Supplementary figures to: HBsAg loss after peginterferon and nucleotide combination treatment in chronic hepatitis B patients: 5 years of follow-up.

Femke Stelma, Meike H. van der Ree, Louis Jansen, Martine W. Peters, Harry L.A. Janssen, Hans L. Zaaijer, R.

Bart Takkenberg, Hendrik W. Reesink

Supplementary figure 1 Year to year overview of included patients. Abbreviations: CR; combined response (HBV DNA < 2,000 IU/mL and normal ALT), PR; partial response (HBV DNA >2,000, no retreatment indication).

supplementary figure 1. Year to year overview of included patients Abbreviations: CR; combined response (HBV DNA < 2,000 IU/mL and normal ALT), PR; partial response (HBV DNA >2,000, no retreatment indication).

56

Chapter 3

supplementary figure 2. Quantitative HBsAg levels. (A) Slight drop in HBsAg levels between previous measurements and measurements for LTFU study. This is probably due to 4 year interval between measurements, i.e. new architect batch and longer storage time for samples. (B) Only samples measured in the same batch (year 3 year 4 and year 5) were analysed for the LTFU study. Left side: plasma HBsAg levels of patients who were not retreated during LTFU and not HBsAg negative at 2 years after cessation of therapy. Right side: plasma HBsAg levels of patients who were retreated at year 2 of follow-up.

2

HB

sAg

log 1

0 IU

/mL

year 3 year 4 year 5

1

2

3

4

5

6

HBsAg in retreated patients

Year of follow-up

**ns

neg

HB

sAg

log 1

0 IU

/mL

*ns

Supplementary figure 2 Quantitative HBsAg levels. (A) Slight drop in HBsAg levels between previous measurements and measurements for LTFU study. This is probably due to 4 year interval between measurements, i.e. new architect batch and longer storage time for samples. (B) Only samples measured in the same batch (year 3 year 4 and year 5) were analysed for the LTFU study. Left side: plasma HBsAg levels of patients who were not retreated during LTFU and not HBsAg negative at 2 years after cessation of therapy. Right side: plasma HBsAg levels of patients who were retreated at year 2 of follow-up.

restoratIon of t Cell funCtIon In ChronIC hepatItIs B patIents upon treatment wIth

Interferon Based ComBInatIon therapy

F. Stelma*, A. de Niet*, L. Jansen, M.J. Sinnige, E.B. Remmerswaal, R.B. Takkenberg, N.A. Kootstra, H.W. Reesink, R.A. van Lier, E.M. van Leeuwen

Journal of Hepatology 2016;64:539-546

* contributed equally

aBstraCtBackground & aims Chronic hepatitis B virus (HBV) infection is characterized by functional impairment of HBV-specific T cells. Understanding the mechanisms behind T cell dysfunction and restoration is important for the development of optimal treatment strategies.

methods In this study we have first analysed the phenotype and function of HBV-specific T cells in patients with low viral load (HBV DNA <20,000 IU/ml) and spontaneous control over the virus. Subsequently, we assessed HBV-specific T cells in patients with high viral load (HBV DNA >17,182 IU/ml) treated with peginterferon/adefovir combination therapy who had various treatment outcomes.

results HBV-specific T cells could be detected directly ex vivo in 7/22 patients with low viral load. These showed an early differentiated memory phenotype with reduced ability to produce IL-2 and cytotoxic molecules such as granzyme B and perforin, but with strong proliferative potential. In a cohort of 28 chronic hepatitis B patients with high viral load treated with peginterferon and adefovir, HBV-specific T cells could not be detected directly ex vivo. However, HBV-specific T cells could be selectively expanded in vitro in patients with therapy-induced HBsAg clearance (HBsAg loss n = 7), but not in patients without HBsAg clearance (n = 21). Further analysis of HBV-specific T cell function with peptide pools showed broad and efficient antiviral responses after therapy.

Conclusions Our results show that peginterferon based combination therapy can induce HBV-specific T cell restoration. These findings may help to develop novel therapeutic strategies to reconstitute antiviral functions and enhance viral clearance.

61

RESTORATION OF T CELL FUNCTION AFTER COMBINATION THERAPY

4

IntroduCtIonPatients who are chronically infected with hepatitis B (CHB) are at an increased risk of liver related morbidity and mortality. Treatment of these patients is aimed at prevention of progression to advanced liver disease.1–3 Ideally, hepatitis B surface antigen (HBsAg) loss and development of anti-HBs antibodies is achieved upon treatment. This outcome is considered the closest to cure and is associated with complete and durable control of the infection. Available data on acute, self-limiting HBV infection show strong and multispecific HBV-specific T cell responses.4,5 In patients with low viral load (LVL) CHB, hepatitis B virus (HBV) core-specific T cells can be detected directly ex vivo.5–7 How- ever, not much is known about the exact phenotype and function of these cells. In patients with high viral load (HVL) and active disease, low HBV-specific immune responses are found. HBV is thought to have evolved several mechanisms to avoid the development of an effective immune response.4,8–10 Repetitive T cell receptor stimulation by persistently high hepatitis B antigen (HBsAg and HBeAg) levels is believed to play a role in T cell exhaustion.11 Treatment options for patients with chronic active hepatitis B consist of peginterferon alpha 2a (pegIFN) and nucleot(s)ide analogues (NUC). The in vitro proliferative capacity of HBV-specific T cells can be restored upon successful treatment with NUCs, which inhibit viral replication.12–14 Recent data from CHB patients treated with pegIFN did not show restoration of HBV-specific T cells upon successful treatment, as these cells remained at low frequencies during and after treatment.15,16 In this study we first performed an extensive analysis of the phenotype and function of HBV-specific CD8+ T cells in patients with CHB with a LVL. These patients represent a unique CHB subpopulation because they spontaneously control the virus without the need for treatment and in general have minimal or no liver damage. Next, we longitudinally investigated HBV-specific CD8+ T cells in patients with active chronic HBV infection and a HVL, who were treated with combination therapy of pegIFN and adefovir. HBV-specific CD8+ T cell responses were compared in patients with HBsAg loss upon therapy vs. patients who did not respond to therapy. Our results reveal that combination therapy can induce reconstitution of HBV-specific T cells in patients with HBsAg loss. These data show that HBV-specific T cell activation and differentiation can be restored in CHB patients treated with pegIFN and adefovir and foster the development of combination therapy strategies to enhance viral clearance.

patIents and methods In order to achieve optimal reactivity with HBV-tetramers and genotype-specific peptides, a selection of CHB patients was made based on viral genotype (genotype A, D or E) and HLA-type (HLA-A2 positivity).

Twenty-two out of the 80 non-treated HBeAg negative patients with a LVL (alanine aminotransferase (ALT) <2x upper limit of normal (ULN) and HBV DNA <20,000 IU/ml) that visited our outpatient clinic, were selected for analysis in the present study (Table 1; Supplementary Figure 1).

62

Chapter 4

Ninety-two patients with HVL and active CHB (44 HBeAg positive and 48 HBeAg negative, HBsAg positivity >6 months, normal or increased (ALT) <10 x ULN and HBV DNA >17,182 IU/ml (>100,000 copies/ml)) were part of a clinical trial in which they received pegIFN alpha 2a (Pegasys?; Hoffman La Roche, Basel, Switzerland) 180 mg subcutaneously once a week and adefovir dipivoxil (Hepsera; Gilead Sciences, Foster City, CA, USA) 10 mg daily for 48 weeks and were subsequently followed-up for a period of 144–196 weeks (Table 1; Supplementary Figure 1).17

Response definitions of the clinical study, defined at week 72 (6 months after treatment) and follow-up (week 144–196) were as follows (with patient numbers included in the present study):

HBsAg loss (n = 7): persistently undetectable HBsAg combined with undetectable HBV DNA or HBV DNA <20 IU/ml. For HBeAg positive patients this also included HBeAg seroconversion (HBeAg loss with development of anti-HBe antibodies). Two patients had very low HBsAg levels at week 72 and lost HBsAg at week 84 and 144, and were therefore included in this analysis. All patients developed anti-HBs antibodies.

Combined response (CR) (n = 7): HBV DNA <17,182 IU/ml combined with persistently normal ALT values (<ULN) but with still detectable HBsAg. For HBeAg positive patients this also included HBeAg seroconversion.

Non-response (NR) (n = 14): HBV DNA >17,182 IU/ml in two consecutive measurements at least 3 months apart after cessation of treatment. These patients were all retreated with NUCs.

table 1. Baseline characteristics of subjects with chronic hepatitis B and healthy controls.

hvl hvl hvl lvl

healthy controls (hC)

Treatment peg-IFN + ADV

peg-IFN + ADV

peg-IFN + ADV

n.a. n.a.

Response HBsAg loss Combined response (CR)

Non-response (NR)

n.a. n.a.

Number of subjects 7 7 14 22 8Gender, male (%) 7 (100) 4 (57) 12 (86) 13 (59) 4 (50)Age, median (range) 47 (38-49) 42 (25-62) 42 (27-53) 46 (20-65) 32 (29-56)ALT U/L, median (range) 80 (26-275) 108 (26-199) 98 (22-1256) 26 (19-152) n.a.HBV DNA, log10IU/mL, mean ±sd

6.45 ±2.30 5.64 ±2.14 7.10 ±1.72 2.52 ±1.15 n.a.

HBeAg status, pos (%) 3 (42) 2 (28) 7 (50) 0 (0) n.a.Viral genotype n.a. A 4 2 8 7 B C D 1 3 4 7 E 2 2 3 3non measurable 5

63


4

All patients were HIV seronegative and were not co-infected with hepatitis C or

hepatitis delta virus.

For comparison, 8 HLA-A2 or HLA-B7-positive healthy controls were included.

These were all seropositive for cytomegalovirus (CMV), which enabled us to compare

characteristics of HBV-specific T cells (chronic infection with detectable viral load) with

CMV-specific T cells (latent infection with undetectable viral load).

viral assessmentsQuantification of plasma HBV DNA, HBsAg and HBV genotyping was assessed as described

previously.17 CMV serostatus was determined by anti-CMV IgG in serum using the AxSYM

microparticle enzyme immunoassay (Abbott Laboratories, Abbott Park, IL) according to

the manufacturer’s instructions.

peripheral blood mononuclear cells (pBmCs)Heparinized peripheral blood samples were obtained at baseline (all patients), and for CHB

patients with HVL subsequently during treatment (week 4, 12 and 48), after treatment (week

52 and week 72) and at long-term follow-up (LFU) (range 144–196 weeks). PBMCs were

isolated using standard density gradient centrifugation and subsequently cryopreserved

until the day of analysis.

Immunofluorescent staining and flow cytometryPBMCs were washed in PBA (PBS containing 0.01% (w/v) NaN3, 0.5% (w/v) bovine serum

albumin and 2mMEDTA). Thawed PBMCs (1.0 x 106 cells) were incubated for 30 min

in the dark at 4 °C with different combinations of fluorescent label conjugated mouse

antibodies. For phenotypic analysis, the following mAbs were used: CD45RA FITC, CD38

PE, CD161 PE, HLA-DR PerCP-Cy5.5, CD8 PerCP-Cy5.5, CD45RA PE-Cy7, CCR7 PE-Cy7

(BD Biosciences, San Jose, USA), CX3CR1 PE (MBL International, Naka-ku Nagoya, Japan),

CXCR3 PE (R&D Systems, Minneapolis, USA), CD3 PE-Alexa610, CD27 APC-Alexa750

(Invitrogen, Camarillo, USA), CD127 PerCP-Cy5.5, CD8 Alexa700, PD-1 PE (eBioscience,

San Diego, USA). For intracellular staining, cells were fixed after surface staining with

FACS Lysing Solution (BD) and subsequently permeabilized (FACS Permeabilizing

Solution 2 (BD)). Cells were incubated for 30 min in the dark at 4 °C with one or more

of the following antibodies: perforin FITC, Ki67 FITC (BD Biosciences), granzyme B PE

(Sanquin, Amsterdam, The Netherlands), granzyme K FITC (Immunotools, Germany) or

T-bet PerCP-Cy5.5 (eBioscience, San Diego, USA). Measurements were done using BD

FACS Canto or LSR Fortessa flow cytometer (BD Biosciences, Europe) and FACS Diva

Software. Analysis was done using FlowJo MacV8.6.3.

antigen-specific t cellsFor detection of antigen-specific CD8+ T cells, PBMCs were stained directly ex vivo or

after 10 days of culture with HBV peptides (HBV core18-27 (sequence, FLPSDFFPSV) or

HBV envelope335-343 (sequence; WLSLLVPFV)) and addition of 50 U/ml IL-2 at day 3 and

64

Chapter 4

day 6. PBMCs were incubated with tetrameric complexes for 30 min in the dark at 4 °C and subsequently phenotyped as described above. The following APC conjugated HLA-peptide tetrameric complexes were used: HLA-A2 tetramer loaded with HBVcore18-27, HLA-A2 tetramer loaded with HBV envelope335-343. HLA-A2 tetramer loaded with CMV pp65-derived NLVPMVATV peptide; HLA-B7 tetramer loaded with CMV pp65-derived TPRVTGGGAM peptide (Sanquin, Amsterdam, the Netherlands). HLA class I genotyping was carried out by PCR for HLA-A2 and HLA-B7.

functional assaysFor LVL patients’ functional assays, PBMCs were stimulated directly ex vivo for 4 h with phorbol 12-myristate 13-acetate (PMA)/ionomycine in the presence of Brefeldin A. For HVL patients, PBMC of 7 responders (HVL) and 7 matched non- responders (HVL) were stimulated with a panel of 315 15-mer peptides (HBV genotype A), overlapping by 10 residues, pooled in 5 mixtures, and restimulated for 6 h at day 10 in the presence of Brefeldin A, CD107a-PE and monensin. The production of cytokines was evaluated by intracellular staining with IFN-γ BV421, MIP-1β Pe-Cy7, TNF-α AF700, IL-2 APC IFN-γ FITC and IL-2 PE monoclonal antibodies (BD Biosciences, San Jose CA) after staining with surface markers – and in the case of LVL patients after staining with tetrameric complexes – as described above. Measurements and analyses were done as described above.

statistical analysisThe two-tailed Mann-Whitney U test was used for analysis of differences between groups. The Wilcoxon signed rank test was used for longitudinal analysis in individual patients. p values < 0.05 were considered statistically significant.

resultsphenotype of hBv-specific t cells in low viral load patientsHBV core-specific CD8+ T cells were detected directly ex vivo in 7 out of 22 LVL patients (median 0.052% of CD8+ T cells, range 0.02–0.12) and were selected for direct phenotyping and functional analyses. The phenotype of these HBV-specific CD8+ T cells was compared to the total CD8+ population and to CMV-specific CD8+ T cells of healthy controls (n = 8). In this way the differentiation status, level of exhaustion, homing capabilities, activation status and cytotoxic potential of HBV- and CMV-specific T cells were analysed (Figure 1; Supplementary Figure 2). In order to distinguish between different CD8+ T cells subsets along the human T cell differentiation pathway, a panel of cell surface markers was used.18 Analysing expression of CD27 and CD45RA allows discrimination of naive cells (CD27+CD45RA+) from memory cells (CD27+CD45RA-) and resting effector-type cells which have lost CD27 but re-expressed CD45RA. HBV-specific T cells predominantly showed a memory-like phenotype (CD27+CD45RA-) as described before19 and had high IL-7 receptor alpha (CD127) and low CCR7 expression (median 83.3 and 14.3%

65


4

respectively) consistent with an early differentiated phenotype(Figure 1B, C).19–21 Although no significant differences were found in the percentage of effector or memory T cells between HBV- and CMV-specific CD8+ T cells, which is in line with our earlier data,22 CMV-specific T cells were predominantly CD27 low and CD45RA high, and had significantly lower expression of CD127 and CCR7 (p = 0.0025 and p = 0.03 respectively, Figure 1B, C). The chemokine receptor CXCR3, associated with T cell recruitment to the liver, was significantly more expressed on HBV-specific cells of CHB patients compared to CMV-specific T cells of healthy controls. In contrast, CMV-specific cells showed high expression of CX3CR1 (fractalkine receptor), a chemokine receptor involved in T cell recruitment to activated endothelium(Figure 1D).23 HBV-specific T cells generally exhibited low levels of the activation markers HLA-DR and CD38, but were more activated than CMV-specific cells as expression of the cell-cycle marker Ki67 was significantly higher in HBV-specific cells

A BHBV LVL

HB

V c

ore

te

t

CD8

0.0168

CD

45

RA

CD27

92.3

2.560

CD8

CD

12

7

76.9

23.1

0

0

0.483

CM

V t

et

CD8

0 103

104

105

0

103

104

105

44.4

0.98447.9

CD

45

RA

CD270 10

210

310

410

5

0

103

104

105 62.6

37.4

0

0

CD8

CD

12

7

CC

R7

0

0

11.1

88.9

CD8

CD8

CC

R7

0

1.42

2.05

96.5

CMV

C

D

(a) (B)

(C)

(d)

figure 1. Phenotypic analysis of ex vivo HBV core-specific and total CD8+ T cells in LVL patients compared to ex vivo CMV-specific and total CD8+ T cells in healthy controls (HC). (A) Amount of detected virus-specific cells and (B) expression of differentiation markers with representative FACS plots of both HBV- and CMV-specific T cells. (C) Proportion of memory, effector, CD127 and CCR7 positive cells. (D) The amount of chemokine receptors and activation markers, amount of transcription factor T-bet and CD161, amount of exhaustion and cytotoxic markers.

66

Chapter 4

(p = 0.0047) (Figure 1D). The T-box transcription factor T-bet, which is seen as a regulator

of cell mediated immunity capable of controlling effector molecule gene expression of

CD8+ T cells,24 was low in HBV- specific cells but high in CMV-specific cells (p = 0.0003).

CD161, a marker associated with CXCR6 upregulation, which has been shown to be

highly expressed on hepatitis C virus-specific CD8+ T cells,25 was significantly higher on

HBV-specific cells compared to CMV-specific CD8+ T cells (Figure 1D). Except for CXCR3

which was expressed higher on total CD8+ T cells of HBV-infected individuals, no other

phenotypic differences were found between total CD8+ T cells of CHB patients or healthy

controls (Figure 1D). This is in line with our former study in which we demonstrated higher

expression of CXCR3 on total CD8+ T cells of HBV- and HCV-infected patients.22

During chronic viral infection, T cells can become exhausted due to continuous antigen

exposure and express the inhibitory receptor PD-1 and other inhibitory molecules. As we

analysed LVL patients with spontaneous control of the virus who have low or negative HBV

DNA level, it was not possible to correlate the expression of exhaustion markers with viral

load. Although on average 53.2% of HBV-specific T cells expressed PD-1, this was not

significantly higher than on CMV-specific T cells and could also indicate T cell activation

instead of exhaustion (Figure 1D).26

Cytotoxic potential of hBv-specific t cells in lvl patientsThe phenotypic data showed that HBV-specific T cells in LVL patients have an early

differentiated memory phenotype. We next investigated the intracellular presence of

granzyme B and perforin, reflecting cytotoxic potential. HBV-specific T cells contained

very little granzyme B or perforin, much less than in CMV-specific T cells, which are known

to have a strong cytotoxic potential (p = 0.0003)(Figure 1D).18,27 In contrast, about 50% of

HBV-specific T cells contained granzyme K, which is indeed mostly expressed by memory-

type T cells.28

function of hBv-specific t cells in lvl patientsIn order to assess cytokine production, HBV-specific T cells were stimulated with PMA/

ionomycine ex vivo. This induced IFN-γ production in 58.7% of HBV-specific T cells

whereas on average only 22.1% produced IL-2 (Figure 2A). A potent proliferative potential

is another hallmark of memory T cells. We therefore analysed the proliferation capacity of

HBV-specific T cells using an in vitro stimulation model with HBV-specific peptides. HBV-

specific T cells against core peptide could be detected in 19 out of 22 LVL patients after

in vitro expansion (median 1.56% of CD8+ T cells, range 0.06–14.9% of CD8+ T cells) and

showed a median fold increase of 21.9 (range 0.0–321.6) (Figure 2B–D). When LVL patients

were divided into a group consisting of patients with a HBV DNA load <2000 IU/ml and

HBsAg <1000 IU/ml and a group of patients with a HBV DNA load >2000 and/or HBsAg

>1000 IU/ml, no differences were observed between these two (Supplementary Figure 3).

These data show that indeed, HBV-specific memory T cells found in peripheral blood of

LVL patients are very well capable of clonal expansion.

67


4

proliferative potential of hBv-specific cells in hvl patients is related to treatment outcomeNext, we aimed to characterize HBV-specific T cells in CHB patients with a HVL and active disease. As these cells could not be detected ex vivo, we again used the in vitro expansion assay. To investigate whether viral control after combination therapy was associated with the restoration of HBV-specific T cells, we compared the proliferation of HBV core peptide- and HBV envelope peptide-specific T cells between three groups of patients with HVL at three different time points (week 0, 72 and LFU). These included HVL patients with HBsAg loss (n = 7), HVL patients with CR, n = 7) and HVL patients with non-response who were retreated with NUCs after week 72 (NR, n = 14). In the latter patient group, all NUC treatment resulted in long-term viral suppression with undetectable HBV DNA. HVL patients with CR and NR did not show HBV-specific proliferation during follow-up (Figure 3A–C). In HVL patients with HBsAg loss however, a strong HBV-specific response at week 72 and at LFU was observed (week 72: range 0.01–0.98%, LFU: range 0.01–5.8% of CD8+ T cells). During (weeks 4, 12 and 24) and at the end of therapy (week 48) however, very low amounts of HBV- specific T cells could be detected in these patients after in vitro expansion (Figure 3A; Supplementary Figure 4). Remarkably, when NR patients were

D

C

B

0.0158 1.96

0.4880.004H

BV

co

re te

t

CD8

A

ex vivo expanded

(a) (B)

(C) (d)

figure 2. Direct ex vivo detection and in vitro stimulation as well as in vitro proliferation of HBV-specific core CD8+ T cells LVL patients with spontaneous viral control. (A) Overview of cytokine production of HBV core-specific CD8+ T cells after direct ex vivo stimulation with PMA and ionomycin in LVL patients. (B) Overview of amount of HBV-core-specific CD8+ T cells of all LVL CHB patients, direct ex vivo and after 10 days of in vitro proliferation (Wilcoxon Signed rank test). (C) Fold change after 10 days in vitro proliferation of HBV core CD8+ specific T cells. (D) Two representative examples of detection of HBV core-specific CD8+ T cells direct ex vivo and after 10 days of proliferation.

68

Chapter 4

retreated with long-term NUC therapy, HBV- specific T cells did become detectable after expansion in 3 out of 12 patients during LFU (range 0.2–2.2% of CD8+ T cells) (Figure 3B, C). For all three groups of HVL patients, similar observations were made for HBV-specific T cells responses against the envelope epitope as for the core epitope (Figure 3B, C). Compared to the HVL patients, LVL patients had higher frequencies of HBV core peptide-specific T cells after in vitro expansion, however in none of the LVL patients a specific T cell response against the envelope peptide was observed (Figure 3B, C).

A

BC

NR

T0 T4 T48 T72

CD8

HB

V c

ore

tet

0.016 0.016 0.013 0.012

0.0230.019 0.053 0.032

HB

sA

g l

os

s

(a)

(B) (C)

figure 3. Longitudinal analysis of HBV-specific CD8+ T cells responses after 10 days of culture with HBV-specific peptide, measured with tetrameric complexes in patients with HBsAg loss, non-responders to trial therapy retreated with nucleot(s)ide retreatment from week 72 onwards (NR), combined responders with sustained viral response but positive HBsAg (CR), low viral load patients with spontaneous viral control (LVL). (A) Representative example of a patient with HBsAg loss (at week 36) vs. a non-responder patient without HBsAg loss, samples measured during treatment (week 0, 4, 48) and short-term follow-up (week 72) HBV core (B) and envelope-specific (C) CD8+ T cell responses at baseline (week 0), short-term follow-up (week 72) and long- term follow-up (week 144–196). Numbers of measured samples are indicated. Significant differences are indicated, Wilcoxon signed rank test (longitudinal analysis in individual patients) and Mann Whitney U test (comparing groups). LVL patients had significantly higher core responses (p value <0.05, not shown in graph), and no measurable envelope responses.

69


4

Broad cytokine responses of hBv-specific cells in responders to combination therapyThe breadth of the HBV-specific T cell response was measured upon expansion and re-stimulation using peptide pools in 7 patients with HBsAg loss and 7 matched non-responders during LFU. Analyses of cytokine production as well as degranulation by CD8+ T cells revealed reactivity against all parts of the HBV proteome and, for patients with HBsAg loss upon combination therapy, particular reactivity against envelope peptides (Figure 4A). Non-responders who were on long-term NUC treatment also showed cytokine responses in reaction to stimulation with peptide pools (Figure 4B), in accordance with the epitope specific data (Figure 3B, C) and earlier data.14 Similarly, broad CD4+ T cell responses were observed in patients with HBsAg loss upon combination therapy, and to a lesser extent in patients with non-response who were retreated with NUCs (Supplementary Figure 5).

proliferative potential of hBv-specific cells is related to antigen levelsThe proliferative potential of HBV core-specific CD8+ T cells was then related to

HBV DNA and HBsAg levels during therapy and LFU for the treated HVL patients and

A

B

C

CD

8

IFN MIP-1IL-2 TNFCD107a

1.26 0.62 0.07 1.04 0.43

(a)

(B)

(C)

figure 4 .Cytokine production by peptide-pool expanded hBv-specific t cells. The mean percentage of cytokine producing CD8+ T cells (>0.01% above background) after 10 days in vitro expansion and re-stimulation with overlapping peptide pools. Cytokine responses in 7 patients with HBsAg loss upon combination therapy (A) and 7 non-responders, retreated with NUCs (B). (C) Representative FACS plots showing CD107a expression and cytokine production of CD8+ T cells.

70

Chapter 4

the untreated LVL patients (Figure 5). The improved proliferative potential of HBV-specific T cells in patients with HBsAg loss coincided with the decline and clearance of HBsAg and HBV DNA (Figure 5A). A slightly improved proliferative potential of HBV-specific T cells was also observed during viral load suppression in NR retreated with NUCs, whereas in these patients HBsAg levels remained unchanged (Figure 5C). Furthermore, our results show that even though LVL patients with spontaneous partial control and HVL patients developing a CR after therapy have similar HBsAg levels, the proliferative potential of HBV-specific T cells was significantly higher in LVL patients compared to patients who needed therapy to achieve CR (Figure 5B, D).

dIsCussIonChronic HBV infection is characterized by functional impairment of HBV-specific T cells, which fail to successfully eradicate the virus from the body. It is not known what kind of antiviral T cell restoration is required (upon therapy) to achieve complete viral control and HBsAg loss. In patients with LVL, HBV-specific T cells can be detected directly ex vivo.5–7

figure 5. Proliferative potential of HBV-specific cells after 10-days of culture in relation to antigen levels. (A–D) HBV core-specific CD8+ T cell responses (right Y-axis) after 10 days of culture (as in Fig. 3B) in relation to antigen levels: viral HBV DNA load (IU/ml) and HBsAg antigen levels (IU/ml) (left Y-axis) for patients with HBsAg loss (A), combined response (B), non-response (C) and low viral load (D) respectively. Note the adjusted right Y-axes scales per patient group for core tetramer responses.

A B

C D

(a) (B)

(C) (d)

71


4

As these patients are able to control viral replication without the need for treatment, we were interested in the phenotype of these HBV-specific T cells. This might give insight into a distinct HBV-specific T cell profile that is associated with viral control. We therefore first analysed the phenotype and function of HBV-specific cells circulating in the blood of CHB patients with LVL relative to CMV-specific T cells of healthy controls. CHB patients with LVL predominantly displayed a HBV-specific T cell response directed against the core protein. These T cells had an early differentiated phenotype and a low ability to produce IL-2. Furthermore, low levels of cytotoxic proteins such as granzyme B and perforin were observed. During exhaustion, loss of effector function is believed to occur in a hierarchical manner. Functions such as IL-2 production, high proliferative capacity and ex vivo killing are lost first,9 whereas loss of homeostatic proliferation and physical deletion take place at the final stages. This could indicate that HBV-specific T cells in patients with LVL have preserved part of their function and are not terminally exhausted.

Next, we analysed longitudinal HBV-specific T cell dynamics in relation to HBsAg antigen load in patients with HVL and active disease who were treated with combination therapy of pegIFN and adefovir. Previous data has shown that a broad and strong HBV-specific T cell response can be detected after in vitro proliferation when HBsAg is cleared upon NUC therapy. Functional recovery however, was not observed for these HBV-specific T cells,14

In a small group of patients with HBsAg loss upon pegIFN monotherapy, proliferated HBV-specific cells were able to kill HBV-peptide specific target cells.29 However, pegIFN monotherapy does not seem to be able to revert proliferative capacity of HBV-specific T cells.15,16 Moreover, no downregulation of exhaustion markers such as PD-1 and CTLA-4 has been observed on HBV-specific T cells after pegIFN treatment.16 In the current study we have observed a significant increase in HBV-specific T cell proliferation in patients that cleared HBsAg upon combination treatment. While in LVL patients only HBV-specific CD8 T cell responses against the core peptide were observed, in HVL patients with HBsAg loss upon combination treatment, HBV-specific T cells that are specific for the core as well as envelope protein could be induced. Upon re-stimulation after in vitro expansion with HBV-peptide pools, broad production of cytokines was observed in response to different parts of the HBV proteome. This suggests a possible causal relationship between the action of HBV-specific T cells and HBsAg clearance in these patients. In patients with NR to therapy, no HBV-specific T cells could be detected at any time before retreatment with NUCs, and hardly any HBV-specific T cells were detectable in patients with CR, even though viral HBV DNA load and HBsAg levels in these patients were similar to LVL patients. The extent of antiviral T cell restoration that is required to achieve complete viral control and HBsAg clearance upon therapy in patients with HVL is currently unknown.

Our results show that HBsAg clearance after combination therapy is associated with a broad and functional HBV-specific T cell response in the circulatory compartment. This observation was made in HBeAg positive and negative patients with HBsAg loss. As we were only able to analyse a limited set of patients, we were not able to objectify any possible differences between these two entities. Furthermore, the exact phenotype of

72

Chapter 4

HBV- specific T cells in HVL patients with active disease could not be studied as these were only detected after in vitro expansion. It is possible that these cells express different phenotypic markers reflecting a more severe exhausted state and diminished proliferative potential than HBV-specific T cells in LVL patients. Further- more, our inability to detect specific T cells before therapy even after in vitro expansion, could be the result of allocation of HBV-specific T cells to the active site of inflammation, i.e. the liver. Detection of HBV-specific T cells could derive from relocation to the blood after cessation of inflammation and antigen decline. However, no increased frequency of HBV-specific T cells was observed previously in the liver during active infection.10 Still, whether the increase in HBV-specific T cell proliferation is a direct consequence of therapy on HBV-specific cells or an indirect result of therapy-induced antigen decline remains to be further investigated.

Our data may indicate that HBV-specific T cells in patients with HVL are present at frequencies below the threshold of detection but retain the flexibility to restore their antigen responsiveness. Indeed earlier advances in improving T cell function have been made by blockade of PD-1, indicating that these cells are not terminally differentiated and have the plasticity to become fully functional early differentiated memory cells.30,31 Furthermore, other lymphocyte subsets have proven to play a role in HBsAg clearance upon therapy.32 Regulatory T cells and natural killer cells that make up a considerable part of the lymphocytes in the liver, have not been analysed in this study. However, the increased proliferation observed in patients with HBsAg loss could be related to an alteration in NK cell function, as these cells have been identified as regulators of HBV-specific T cells in the liver.33

To summarize, our data indicates that in patients with LVL, relatively narrow HBV core-specific T cells with a resting effector memory phenotype and low ability to produce IL-2, but strong proliferative potential, seem to be sufficient to retain the virus at low levels. In patients with HBsAg loss upon treatment with pegIFN and adefovir, as well as in some patients on long-term NUC therapy, there is a partial recovery of HBV-specific T cells. In these patients a broader repertoire of HBV-specific T cell restoration was observed. The extent to which HBV-specific T cell restoration is needed, in order to achieve HBsAg loss, as well as the interaction with other players of the immune system remains to be elucidated. Future therapeutic strategies may be based on the modulation of protective T cell responses, and reconstitute their antiviral function enabling cure of CHB virus infection.

73


4

referenCes1. Lok AS. Chronic hepatitis B. N Engl J

Med 2002;346:1682–1683.

2. Chen CJ, Yang HI, Su J, et al. Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level. JAMA 2006;295:65–73.

3. Iloeje UH, Yang HI, Su J, et al. Predicting cirrhosis risk based on the level of circulating hepatitis B viral load. Gastroenterology 2006;130:678–686.

4. Thimme R, Wieland S, Steiger C, et al. CD8(+) T cells mediate viral clearance and disease pathogenesis during acute hepatitis B virus infection. J Virol 2003;77:68–76.

5. Ferrari C, Penna A, Bertoletti A, et al. Cellular immune response to hepatitis B virus-encoded antigens in acute and chronic hepatitis B virus infection. J Immunol 1990;145:3442–3449.


7. Webster GJ, Reignat S, Brown D, et al. Longitudinal analysis of CD8+ T cells specific for structural and nonstructural hepatitis B virus proteins in patients with chronic hepatitis B: implications for immunotherapy. J Virol 2004;78:5707–5719.

8. Wieland S, Thimme R, Purcell RH, et al. Genomic analysis of the host response to hepatitis B virus infection. Proc Natl Acad Sci U S A 2004;101:6669–6674.

9. Wherry EJ. T cell exhaustion. Nat Immunol 2011;12:492–499.

10. Boni C, Fisicaro P, Valdatta C, et al. Character- ization of hepatitis B virus (HBV)-specific T-cell dysfunction in chronic HBV infection. J Virol 2007;81:4215–4225.

11. Rehermann B, Nascimbeni M. Immunology of hepatitis B virus and hepatitis C virus infection. Nat Rev Immunol 2005;5:215–229.

12. Boni C, Penna A, Ogg GS, et al. Lamivudine treatment can overcome cytotoxic T-cell hyporesponsiveness in chronic hepatitis B: new perspectives for immune therapy. Hepatology 2001;33:963–971.

13. Boni C, Penna A, Bertoletti A, et al. Transient restoration of anti-viral T cell responses induced by lamivudine therapy in chronic hepatitis B. J Hepatol 2003;39:595–605.

14. Boni C, Laccabue D, Lampertico P, et al. Restored function of HBV-specific T cells after long-term effective therapy with nucleos(t)ide analogues. Gastroenterology 2012;143:963–973.

15. Penna A, Laccabue D, Libri I, et al. Peginterferon-alpha does not improve early peripheral blood HBV-specific T-cell responses in HBeAg-negative chronic hepatitis. J Hepatol 2012;56:1239–1246.

16. Micco L, Peppa D, Loggi E, et al. Differential boosting of innate and adaptive antiviral responses during pegylated- interferon-alpha therapy of chronic hepatitis B. J Hepatol 2013;58:225–233.

17. Takkenberg RB, Jansen L, de Niet A, et al. Baseline hepatitis B surface antigen (HBsAg) as predictor of sustained HBsAg loss in chronic hepatitis B patients treated with pegylated interferon- alpha2a and adefovir. Antivir Ther 2013;18:895–904.

18. van Lier RA, ten Berge IJ, Gamadia LE. Human CD8(+) T-cell differentiation in response to viruses. Nat Rev Immunol 2003;3:931–939.

19. Appay V, Dunbar PR, Callan M, et al. Memory CD8+ T cells vary in differentiation phenotype in different persistent virus infections. Nat Med 2002;8:379–385.

74

Chapter 4

20. Hamann D, Baars PA, Rep MH, et al. Phenotypic and functional separation of memory and effector human CD8+ T cells. J Exp Med 1997;186:1407–1418.

21. van Leeuwen EM, Remmerswaal EB, Heemskerk MH, et al. Strong selection of virus-specific cytotoxic CD4+ T-cell clones during primary human cytomegalovirus infection. Blood 2006;108:3121–3127.

22. de Niet A, de Bruijne J, Sinnige MJ, et al. Upregulation of CXCR3 expression on CD8+ T cells due to the pervasive influence of chronic hepatitis B and C virus infection. Hum Immunol 2013;74:899–906.

23. van de Berg PJ, Yong SL, Remmerswaal EB, et al. Cytomegalovirus-induced effector T cells cause endothelial cell damage. Clin Vaccine Immunol 2012;19:772–779.

24. Intlekofer AM, Takemoto N, Wherry EJet al. Effector and memory CD8+ T cell fate coupled by T-bet and eomesodermin. Nat Immunol 2005;6:1236–1244.

25. Northfield JW, Kasprowicz V, Lucas M, et al. CD161 expression on hepatitis C virus-specific CD8+ T cells suggests a distinct pathway of T cell differentiation. Hepatology 2008;47: 396–406.

26. Chen L. Co-inhibitory molecules of the B7-CD28 family in the control of T-cell immunity. Nat Rev Immunol 2004;4:336–347.

27. Gamadia LE, Rentenaar RJ, Baars PA, et al. Differentiation of cytomegalovirus-

specific CD8(+) T cells in healthy and immunosuppressed virus carriers. Blood 2001;98:754–761.

28. Remmerswaal EB, Havenith SH, Idu MM, et al. Human virus-specific effector-type T cells accumulate in blood but not in lymph nodes. Blood 2011;119:1702–1712.

29. Rehermann B, Lau D, Hoofnagle JH, et al. Cytotoxic T lymphocyte responsiveness after resolution of chronic hepatitis B virus infection. J Clin Invest 1996;97:1655–1665.

30. Fisicaro P, Valdatta C, Massari M, et al. Antiviral intrahepatic T-cell responses can be restored by blocking programmed death-1 pathway in chronic hepatitis B. Gastroenterology 2010;138: 682–693.

31. Bengsch B, Martin B, Thimme R. Restoration of HBV-specific CD8+ T-cell function by PD-1 blockade in inactive carrier patients is linked to T-cell differentiation. J Hepatol 2014;61:1212–1219.

32. Stelma F, de Niet A, Sinnige MJ, et al. Natural killer cell characteristics in patients with chronic hepatitis B virus (HBV) infection are associated with HBV surface antigen clearance after combination treatment with pegylated interferon alfa-2a and adefovir. J Inf Dis 2015;212:1042–1051.

33. Peppa D, Gill US, Reynolds G, et al. Up- regulation of a death receptor renders antiviral T cells susceptible to NK cell- mediated deletion. J Exp Med 2013;210:99–114.

75


4

aCKnowledGementsThe authors would like to thank J. Stelma for designing the graphical abstract and Prof. Dr. C.P. Engelfriet for critically revising the manuscript.

supplementary data

Chapter 4

Supplementary figures

Supplementary figure 1 | Patient breakdown. (A) Active CHB patients (HBV DNA>17,182 IU/mL, ALT <10X upper limit of normal (ULN)) were treated for 48 weeks with 180 ug peg‐interferon alfa‐2a weekly and 10 mg adefovir once daily and followed up in this analysis for a period of 96‐144 weeks (HVL patients). (B) Untreated LVL HBeAg negative CHB patients with low viral load and spontaneous viral control (HBV DNA <20.000 IU/mL, ALT <2x ULN) were used as control (LVL patients). (C) healthy controls were used as HBV‐negative controls. Selection of samples was based upon viral genotype (a,d,e) and HLA status (HLA‐A2 positivity) of the patients and on CMV seropositivity and HLA status (HLA‐A2 of HLA‐B7 positivity for healthy controls).

(a)

(B)

(C)

supplementary figure 1. Patient breakdown. (A) Active CHB patients (HBV DNA>17,182 IU/mL, ALT <10X upper limit of normal (ULN)) were treated for 48 weeks with 180 ug peg-interferon alfa-2a weekly and 10 mg adefovir once daily and followed up in this analysis for a period of 96-144 weeks (HVL patients). (B) Untreated LVL HBeAg negative CHB patients with low viral load and spontaneous viral control (HBV DNA <20.000 IU/mL, ALT <2x ULN) were used as control (LVL patients). (C) healthy controls were used as HBV-negative controls. Selection of samples was based upon viral genotype (a,d,e) and HLA status (HLA-A2 positivity) of the patients and on CMV seropositivity and HLA status (HLA-A2 of HLA-B7 positivity for healthy controls).

76

Chapter 4 Chapter 4

Supplementary figure 2 | Phenotypic analyses of HBV‐core‐specific CD8+ T cells in CHB patients with LVL and CMV‐specific CD8+ T cells in HC. Representative FACS staining of expression of phenotypic markers on HBV‐ and CMV specific T cells. HBV, hepatitis B virus; LVL, low viral load; CMV, cytomegalovirus; HC, healthy control.

supplementary figure 2. Phenotypic analyses of HBV-core-specific CD8+ T cells in CHB patients with LVL and CMV-specific CD8+ T cells in HC. Representative FACS staining of expression of phenotypic markers on HBV- and CMV specific T cells. HBV, hepatitis B virus; LVL, low viral load; CMV, cytomegalovirus; HC, healthy control.

supplementary figure 3. Graphical abstract.

natural KIller Cell CharaCterIstICs In ChronIC hepatItIs B patIents are assoCIated wIth hBsaG

loss after ComBInatIon treatment wIth peG-Interferon alpha-2a and adefovIr

F. Stelma, A. de Niet, M.J. Sinnige, L. Jansen, R.B. Takkenberg, H.W. Reesink, N.A. Kootstra, E.M. van Leeuwen

Journal of Infectious Diseases 2015;212:1042-1051

aBstraCtBackground The role of natural killer (NK) cells in the process of hepatitis B virus (HBV) surface antigen (HBsAg) clearance and whether their phenotype is related to treatment outcome in patients with chronic hepatitis B are currently unknown.

methods Patients with chronic hepatitis B (HBV DNA load, >17,000 IU/mL) were treated with pegylated interferon alfa-2a and adefovir for 48 weeks. NK cell phenotype and function were analysed in 7 responders (defined as individuals with HBsAg clearance by week 72; 3 HBVe antigen [HBeAg]-positive and 4 HBeAg-negative), 7 matched non-responders, and 7 healthy controls. Subsequently, 34 baseline samples from HBeAg-positive patients with chronic hepatitis B were analysed.

results During treatment, the percentage and absolute number of CD56bright NK cells increased significantly, whereas the percentage and absolute number of CD56dim NK cells decreased. At baseline, responders had a significantly lower expression of chemokine receptor CX3CR1 on CD56bright NK cells and inhibitory receptor NKG2A on CD56dim NK cells, compared with non-responders. In addition, responders had higher CD56bright TRAIL expression and interferon-γ production at end of treatment. These baseline differences were not found in HBeAg-positive patients who had HBeAg seroconversion without HBsAg clearance.

Conclusions Combination therapy significantly influences NK cell phenotype and function. Differences between patients with chronic hepatitis B with HBsAg clearance and non-responders suggest that NK cells play a role in the clearance of HBsAg during interferon-based combination therapy.

81

NK CELL FUNCTION AFTER COMBINATION THERAPY

5

IntroduCtIonHepatitis B virus (HBV) infection, with approximately 240 million chronic cases worldwide, is a global health problem.1 Patients with chronic hepatitis B (CHB) are at increased risk of developing serious hepatic complications, such as cirrhosis and hepatocellular carcinoma.2 Current treatment options for CHB include pegylated interferon alfa-2a (pegIFN) or nucleos(t)ide analogue (NUC) treatment. While NUCs can potently block viral replication, pegIFN exerts its action via direct antiviral activity and immunomodulatory effects.

The immunomodulatory effects of pegIFN have been investigated in innate and adaptive effector cells. PegIFN therapy has been shown to cause expansion and activation of CD56bright natural killer (NK) cells.3 T cells, on the other hand, remain present at low frequency in the peripheral blood, owing to the suppressive effects on bone marrow function and the anti-proliferative effects of pegIFN.4, 5 However, in patients with a long-term response to pegIFN therapy, a restoration of HBV-specific CD8+ T cell function has been shown,6 although no reversion of the exhausted phenotype was shown early during treatment.3, 7, 8 Conversely, the effect of NUC therapy on innate immune cell functionality is controversial,9, 10 whereas HBV-specific CD8+ T-cell responses are restored in patients in whom HBV surface antigen (HBsAg) is cleared during NUC treatment.11 Together, these data imply that combination treatment with pegIFN and a NUC is a promising option for patients with CHB, optimally restoring both the innate and adaptive functions of the immune system.12

Early studies of combination regimens in which a combination of pegIFN and lamivudine was used have not shown superior responses to those obtained by pegIFN monotherapy.13 Combined treatment with pegIFN and a newer-generation NUC, however, could give improved antiviral responses, as suggested in several studies.14–18 In our cohort of patients with CHB who received pegIFN combined with adefovir, we observed a relatively high rate of HBsAg clearance in both HBV e antigen (HBeAg)– positive patients and HBeAg-negative patients.14 We have previously shown that the function of HBV-specificCD8+ T cells is restored in responders to combination therapy.19 However, the effect of this treatment regimen on the innate immune response is unknown. Furthermore, the role of NK cells in the clearance of HBsAg has not yet been identified. In the present study, we characterized changes in the NK cell compartment during pegIFN/NUC combination therapy in patients with CHB and related these changes to treatment outcome.

patIents and methodspatient samples In a previously described investigator-initiated study (clinical trials registration ISRCTN 77073364),14 patients with CHB and a high HBV DNA load (>17 000 IU/mL) received combination therapy of pegylated interferon alfa-a2 and adefovir for 48 weeks. All patients gave informed consent, and the study was approved by the Ethical Review Board of the Academic Medical Center Amsterdam. Samples were obtained at baseline, during treatment (day 3, week 1, week 24, and week 48), and during follow-up (week

82

Chapter 5

72). Peripheral blood mononuclear cells (PBMCs) were isolated using standard density-

gradient centrifugation and cryopreserved for later analysis. From this cohort, 14 patients

with CHB were selected (Supplementary Figure 1A). Half of this group consisted of 7

responders with HBsAg clearance at week 72, defined according to European Association

for the Study of the Liver guidelines20 as having persistently undetectable HBsAg and

a negative result of HBV DNA testing, with or without development of antibody to

HBsAg (anti-HBs; for HBeAg-positive patients, this also included HBeAg seroconversion).

The other half comprised 7 non-responders (matched on the basis of genotype, ethnicity,

andHBeAg status at baseline) with no HBsAg clearance and, for HBeAg positive patients,

no HBeAg clearance (Table 1). All 7 non-responders had to be retreated with NUC within

2 years after the end of treatment (viral kinetics are depicted in Supplementary Figure 1B).

For comparison, 7 healthy controls (blood donors) were analysed. In addition, baseline

PBMCs from a separate cohort of 34 HBeAg-positive patients, of whom 9 had HBeAg

seroconversion without HBsAg clearance at week 72, were analysed. All 48 patients tested

negative for human immunodeficiency virus and were not co-infected with hepatitis C virus

(HCV) or hepatitis D virus. HBV genotype determination, quantification of plasma HBV

DNA and serum HBsAg, and detection of anti- HBs, HBeAg, and antibody toHBVe antigen

were performed as described before.14

flow Cytometry PBMCs were stained in phosphate-buffered saline containing 0.01% (w/v) NaN3 and, 0.5%

(w/v) bovine serum albumin with different combinations of fluorescent label-conjugated

figure 1. Combination treatment with peg-interferon alfa-2a and adefovir changes the natural killer (NK) cell compartment. NK cells were measured in 7 healthy controls (HC) and 14 patients with chronic hepatitis B (CHB) at baseline (BL) and the end of treatment (EoT). Frequency of total (A) and CD56bright and CD56dim (B) NK cells. (C) Representative fluorescence-activated cell-sorter staining of CD14−CD19−CD3− cells, with gating for total NK cells (CD56 and/or CD16 positive cells), CD56bright (CD16−/dim), and CD56dim (CD16+) NK cells. Frequencies represent percentages of CD56bright/total NK cells and CD56dim/ total NK cells. Statistical analyses of unpaired data were performed with the Mann–Whitney U test. Analyses of paired data were performed with the Wilcoxon test. *P < .05, **P < .01, and ***P < .001. Abbreviation: NS, nonsignificant.

0

10

20

30

40

50

HC CHB

BL

CHB

EoT

****

*

% N

K c

ells/

lym

ph

ocyte

s

CD16

CD

56

CHB BL

CHB EoT

A B C

0

20

40

60

80

100

HC CHB

BL

CHB

EoT

***

***ns

HC CHB

BL

CHB

EoT

***

***ns

CD56dim

CD56bright

% s

ub

set/

tota

l N

K c

ells

0 102

103

104

105

0

103

104

105

HC

0 102

103

104

105

0

103

104

105

0 102

103

104

105

0

103

104

105

54

80

81

31

9

7(a) (B) (C)

83


5

mouse monoclonal antibodies (mAbs) (Supplementary Materials). For intracellular staining, cells were fixed after surface staining, permeabilised, and stained with fluorescent label–conjugated mAbs (Supplementary Materials). All measurements were done on an LSR Fortessa or FACS Canto flow cytometer (BD Biosciences, San Jose, California) and analysed by FlowJoMacV9.7.5 software. When positive and negative expression could be identified for a specific marker, gates were used to indicate cell frequency. For markers that are gradually expressed, the geometricmean fluorescent intensity (gMFI) was used to define expression level.

nK Cell functional assays For functional measurements, PBMCs were cocultured with K562 cells for 3 hours (effector:target ratio, 5:1) in the presence of monensin (BD Biosciences), brefeldin (Invitrogen, Camarillo, California) and CD107a-FITC (Affymetrix eBioscience, San Diego, California) after overnight incubation with either medium or recombinant human interleukin 2. After surface staining, cells were fixed, permeabilized, and stained with interferon-γ (IFN- γ)–PE and tumor necrosis factor α (TNF-α)–PE-Cy7 (Affymetrix eBioscience). To assess target cell apoptosis, cells were stained with DiOC6. Measurements and analyses were performed as described above.

statistical analysis Statistical analyses were performed using GraphPad Prism 5 software. For paired data, the Wilcoxon signed rank test was used, and for non-paired data, the Mann–Whitney U test was used. For analysis of correlation, the Spearman correlation coefficient was calculated. 2-sided P values of <.05 were considered statistically significant.

table 1. Baseline characteristics of patients with chronic hepatitis B treated with combination therapy of pegIFN with adefovir

Case# sex age hBeaghBvdna log10Iu/ml

hBsag log10Iu/ml alt u/l Genotype Inf naïve t72 response

1 M 37 neg 4.05 2.41 36 D no HBsAg lossa

2 M 30 neg 5.92 3.63 79 D yes non-responder3 M 46 pos 8.72 4.91 81 A no seroconversionb

4 M 31 pos 8.65 4.75 1256 A yes non-responder5 M 39 pos 6.85 3.18 327 A yes seroconversion6 M 42 pos 7.28 4.63 106 A no non-responder7 M 46 neg 4.89 3.22 52 E yes seroconversion8 M 30 neg 7.72 4.04 214 E yes non-responder9 F 69 neg 3.63 2.54 50 C no HBsAg loss10 M 27 neg 5.37 3.44 49 C yes non-responder11 M 24 neg 5.48 1.62 38 A yes seroconversion12 M 27 neg 7.51 3.62 91 A yes non-responder13 M 48 pos 8.84 5.00 80 A yes seroconversion14 M 44 pos 9.81 5.27 105 A no non-responder

84

Chapter 5

resultsChanges in the nK cell compartment in patients with ChB receiving combination therapyNK cell characteristics were analysed in 14 patients with CHB (Table 1) who were treated with combination therapy and compared to those in 7 healthy controls. The gating strategy that was used excluded double, nonviable,CD3+,CD14+,and CD19+ cells. From these, the CD56+ and/or CD16+ cells were considered NK cells. NK cells were divided into the classical subsets CD56bright (CD16neg/dim) and CD56dim (CD16+) NK cells (Figure 1C and Supplementary Figure 2A). At baseline, patients with CHB had a significantly increased proportion of NK cells, compared with healthy controls, which is consistent with a previous report.21 The CD56bright and CD56dim subset distribution in patients at baseline was similar to that in healthy controls (Figure 1B). While the percentage of NK cells within the lymphocyte compartment in patients increased at the end of treatment (from 13.3% at baseline to 19.0% at the end of treatment; P < .05; Figure 1A), the absolute number of NK cells was unchanged (from 253 cells/µL at baseline to 231 cells/µL at the end of treatment; Supplementary Figure 2B). The percentages of NK cell subsets, however, changed considerably toward the end of treatment (Figure 1B and 1C), as well as their absolute numbers (Supplementary Figure 2B). This resulted in an increase in CD56bright NK cells (from 11.3% at baseline to 44.4% at the end of treatment; P < .0001) and a decrease in CD56dim NK cells (from 74.2% at baseline to 39.9%at the end of treatment; P < .0001). Longitudinal analysis of NK cell subsets showed no differences when measured at early time points (day 3 and week 1), suggesting a gradual change during combination therapy, with normalization during follow-up at week 72 (Supplementary Figure 2C).

nK cell proliferation and chemokine receptors To investigate the possible cause of the changes in NK cell subset distribution, the expression of various markers, including proliferation marker Ki67 and several chemokine receptors on NK cells, were analysed. The percentages of Ki67 positive CD56bright and CD56dim NK cells in patients was significantly higher at the end of treatment, compared with baseline (Figure 2A). Because the distribution of these subsets drastically changed, absolute numbers were also taken into account. In parallel, the absolute number of Ki67+CD56bright NK cells increased (from 3 cells/μL at baseline to 16 cells/µL at the end of treatment; P < .001; data not shown), and even though the CD56dim subset decreased in absolute number, the number of Ki67+CD56dim cells increased during therapy (from 14 cells/μL at baseline to 28 cells/µL at the end of treatment; P < .01; data not shown), illustrating an overall increased proliferation tendency in both subsets. CX3CR1, also known as the fractalkine receptor, which is associated with the migration of cells to sites of inflammation, is generally expressed at a higher level on CD56dim cells, compared with CD56bright NK cells.22 The expression of CX3CR1 on CD56dim NK cells in patients with CHB at baseline was higher than that in healthy controls. In the CD56bright subset, CX3CR1 expression was similar in healthy controls and patients with CHB at baseline, but

85


5

this increased after receipt of combination therapy (Figure 2B). The expression of CXCR3, which is associated with homing to sites of inflammation, including the liver, was higher in both subsets in patients with CHB at baseline as compared to healthy controls.23 During therapy, however, CXCR3 expression decreased in the CD56bright NK cell subset (Figure 2C). Expression of CXCR6, which is also associated with homing to the liver, was higher on CD56bright NK cells from patients with CHB at baseline, compared with those from healthy controls, but significantly decreased during combination therapy (from 27.5% at baseline to 4.6% at the end of treatment; P < .001; Figure 2D). A similar decrease in CXCR6 expression during therapy was seen in CD56dim NK cells (Figure 2D).

Increased natural cytotoxicity receptors and activation status The effector function of NK cells is regulated by the sum of activating and inhibitory signals from the environment, combined with the surrounding cytokine milieu. Inhibitory receptors include the heterodimer NKG2A/CD94, whereas CD16, NKG2C/ CD94, CD161, and the natural cytotoxicity receptors NKp30, NKp44, and NKp46 are activating receptors.

figure 2. Proliferation and expression of chemokine receptors is altered in natural killer (NK) cell subsets after treatment with combination therapy. Frequencies of Ki67+ cells (A), geometric mean fluorescence intensities (gMFIs) of CX3CR1 (B), gMFIs of CXCR3 (C), and frequencies of CXCR6 (D) in CD56bright and CD56dim NK cells from 7 healthy controls (HC) and 14 patients with chronic hepatitis B (CHB) at baseline (BL) and the end of treatment (EoT). (E) Representative dot plots show an overlay of CD56dim and CD56bright NK cells, with percentages of CXCR6+/CD56bright NK cells. Statistical analyses of unpaired data were performed with the Mann–Whitney U test. Analyses of paired data were performed with theWilcoxon test. *P < .05, **P < .01, and ***P < .001. Abbreviation: NS, nonsignificant.

A B

C D

0

20

40

60

HC CHB

BL

CHB

EoT

***

***ns **ns

HC CHB

BL

CHB

EoT

ns

CD56dim

CD56bright

% K

i67+

cells

0

1000

2000

3000

HC CHB

BL

CHB

EoT

HC CHB

BL

CHB

EoT

***

***ns ***

***ns

CD56dim

CD56bright

gM

FI C

X3C

R1

0

20

40

60

HC CHB

BL

CHB

EoT

HC CHB

BL

CHB

EoT

*

*ns

**

***

*

CD56dim

CD56bright

% C

XC

R6+

cells

0

500

1000

HC CHB

BL

CHB

EoT

HC CHB

BL

CHB

EoT

**

**ns**

**ns **

ns

*

CD56dim

CD56bright

gM

FI C

XC

R3

CXCR6

CD

56

BL

EoT

E

39

9

(a) (B)

(C) (d) (e)

86

Chapter 5

No differences were found in the expression of NKG2A, CD16, NKG2C, CD94, CD161, NKp44, and NKp46 on the different NK cell subsets in healthy controls, compared with patients with CHB at baseline. However, the expression of NKp30 on CD56dim NK cells was decreased in patients at baseline, compared with healthy controls (data not shown). During therapy, the expression of the inhibitory receptor NKG2A increased significantly on both CD56bright and CD56dim NK cells. The expression of CD94, which can form a heterodimer with NKG2A, followed a similar pattern. The expression of the activating receptors CD16 (Fc-γ receptor III), on CD56dim NK cells, and NKG2C, on the CD56bright subset, decreased, whereas the expression of the natural cytotoxicity receptors NKp30 and NKp46 increased significantly in both subsets during therapy (Supplementary Figure 3). CD161 expression, which has been implicated in triggering NK cell mediated cytotoxicity, decreased in both NK cell subsets during therapy. When the activation status (HLA-DR and CD38) of NK cells was assessed in patients with CHB at baseline and in healthy controls, no differences were found (data not shown). During therapy, the activation status of both CD56dim and CD56brgiht cells significantly increased, as illustrated by an increase in HLA-DR and CD38 expression (Figure 3A–C). The strongest increase was seen in the gMFI of CD38, which increased by a factor 1.5 for CD56bright, and a factor of 2.6 for CD56dim NK cells (P < .001; Figure 3A).

nK cell effector potential CD56dim NK cells from patients with CHB at baseline showed higher expression of both perforin and granzyme B, compared with healthy controls (Supplementary Figure 4D and 4E).During therapy, the expression of perforin further increased in this subset, while the level of granzyme B remained the same. In CD56bright NK cells, the expression of both of these cytotoxic proteins increased during therapy (Supplementary Figure 4A

0

50000

100000

150000

CHB

BL

CHB

EoT

CHB

BL

***

CHB

EoT

***

CD56dim

CD56bright

gM

FI C

D38

HLA-DR

CHB BL

CHB EoT

Internal negative

C

0

500

1000

1500

CHB

BL

CHB

EoT

CHB

BL

**

CHB

EoT

*

CD56dim

CD56bright

gM

FI H

LA

-DR

A B

CD56dim CD56bright

CD38

figure 3. Activation status in natural killer (NK) cells increases during combination therapy in patients with chronic hepatitis B (CHB). CD38 (A) and HLA- DR (B) expression in CD56bright and CD56dim NK cells. (C) Representative staining for CD38 and HLA-DR in CD56bright and CD56dim NK cells at baseline (BL) and the end of treatment (EoT). Statistical analyses were performed using the Wilcoxon test. *P < .05, **P < .01, and ***P < .001.

(a) (B) (C)

87


5

and 4B). TRAIL expression on CD56bright NK cells was significantly higher in patients, compared with healthy controls, and increased even further during therapy (gMFI, 284 at baseline and 385 at the end of treatment; P < .01; Supplementary Figure 4C). On CD56dim NK cells, TRAIL expression also increased significantly during therapy (Supplementary Figure 4F).

nK cell functionality upon target cell recognition Next, we evaluated the functionality of NK cells by measuring cytokine production (IFN-γ and TNF-α) and degranulation (CD107a) in response to target cells, as well as the killing capacity of NK cells, measured by apoptosis of these K562 cells in response to coculture with PBMCs. IFN-γ and TNF-α production, as well as CD107a expression, significantly decreased during therapy (Figure 4A; only IFN-γ data are shown). Apoptosis induced in K562 cells was significantly decreased at time points during treatment (day 3, week 24, and week 48), compared with baseline, but normalized during follow-up (Figure 4B). When activated overnight with interleukin 2, which, in vitro, has a similar effect as activation with interleukin 15,24 NK cell function increased, compared with unstimulated NK cells, however a decrease was seen when comparing baseline values to end-of-treatment values (Figure 4C and 4D). The percentage of NK cells capable of all 3 functions (ie, those positive for CD107a, IFN-γ, and TNF-α) also decreased (Figure 4C).

differential expression of traIl, CX3Cr1, nKG2a, and Ifn-γ in patients with hBsag clearance, patients with hBeag seroconversion, and non-respondersSubsequently, we analysed whether changes in NK cell phenotype and function differed in patients with CHB with regard to treatment outcome. Therefore, the group was divided into responders (7 subjects with HBsAg clearance, with or without anti-HBs formation at week 72) and 7 matched non-responders (Table 1). There were no differences in either total NK cell numbers or NK cell subsets between responders and non-responders (data not shown). At baseline, we observed a significantly lower expression of CX3CR1 on CD56bright NK cells in responders, compared with non-responders, but no significant differences were found at the end of treatment (Figure 5A). NKG2A expression on CD56dim cells was also significantly lower in responders (37.3%) than in non-responders (50.9%) at baseline (P < .05; Figure 5B). Furthermore, in responders, the expression of TRAIL on CD56bright NK cells was significantly higher at the end of treatment (Figure 5C). Even though there was no significant difference in IFN-γ production at baseline, IFN-γ production by activated NK cells at the end of treatment was twice as high in responders than in non-responders (14% vs 7.5%; P =.026; Figure 5D). A similar trend was seen for TNF-α production and CD107a expression (data not shown). We next investigated whether these markers were also differentially expressed in patients who developed a partial response to combination therapy, represented by HBeAg seroconversion without HBsAg clearance (Supplementary Table 1). In this second group of patients, however, no significant differences were observed in responders who had HBeAg seroconversion at week 72 and in patients who

88

Chapter 5

0

20

40

60

80E:T 2:1

E:T 1:1

0 d3 1 24 48 72

****

**

weeks% D

iOc

6-

K562 c

ell

s:

ap

op

tosis

C

A B

IFN

TNF

CD

56

D

0

10

20

30

40

**

CHB

BL

CHB

EoT

CHB

BL

CHB

EoT

CHB

BL

CHB

EoT

CD107a IFNγ TNFα

*** *** ***

polyfunct

CHB

BL

CHB

EoT

% p

osit

ive N

K c

ells

0 102

103

104

105

0

102

103

104

105

0 103

104

105

0

102

103

104

105

0 102

103

104

105

0

102

103

104

105

0 103

104

105

0

102

103

104

105

CHB BL CHB EoT

18

13

CD107a

0 102

103

104

105

0

102

103

104

105

0 102

103

104

105

0

102

103

104

105

31 28

8

13

figure 4. Combination therapy results in a decrease in functional activity of natural killer (NK) cells. (A) Longitudinal analysis of interferon γ (IFN-γ)– positive NK cells upon coculture with K562 cells. (B) K562 apoptosis as measured by DiOc6-negative cells (n = 11–14 per time point) in 2 effector:target (E:T) ratios. (C,D) Functionality of activated NK cells at baseline (BL) and the end of treatment (EoT; C), with representative fluorescence-activated cell-sorter plots at BL (gates identify positive cells among NK cells; D). Statistical analyses were performed using the Wilcoxon test. *P <.05, **P < .01, and ***P <.001. Abbreviations: CHB, chronic hepatitis B; d3, day 3 of treatment; NS, nonsignificant; polyfunct, polyfunctionality; TNF-α, tumor necrosis factor α; T24, week 24 of treatment.

did not have HBeAg seroconversion (Supplementary Figure 5A–F). Subsequent analyses

using combined response (ie, an HBV DNA load of < 2000 IU/mL with normalization in

the ALT level) at week 72 and HBeAg seroconversion at week 144 as an end point did not

change the outcome (data not shown).

association between Cd56bright nK cells and alt levels To investigate whether NK cells could have a role in liver injury, we studied the relation

between NK cell phenotype and ALT levels at baseline in HBeAg-positive patients.

The percentage of CD56bright NK cells at baseline was positively correlated with ALT

level (P = .0355; Figure 6A). Furthermore, TRAIL and CD38 expression on CD56bright NK

cells were also positively correlated with ALT level at baseline in HBeAg-positive patients

(P = .005 [Figure 6B] and P = .015 [Figure 6C], respectively). No correlation between NK

cell markers and viral parameters (ie, HBsAg levels and HBV DNA load) were found.

(a)

(d)

(B)

(C)

89


5

HC BL EoT BL EoT

0

200

400

600

8000.073

0.038

responders non-responders

gM

FI T

RA

IL C

D56

brig

ht N

Kcell

s

HC BL EoT BL EoT

0

100

200

300

400

500

0.026

0.224


gM

FI C

X3C

R1 C

D56

bri

gh

t NK

cell

sA B

C D

HC BL EoT BL EoT

0

20

40

60

80

100

0.038

0.073


%N

KG

2A

/ C

D56

dim

NK

cell

s

HC BL EoT BL EoT

0

10

20

30


0.209

0.026

% IF

Nγγ γγ+

cell

s/

NK

cells

figure 5. Differential expression of CX3CR1, NKG2A, TRAIL, and interferon γ (IFN-γ) between patients with chronic hepatitis B (CHB) with hepatitis B virus surface antigen clearance at week 72, compared with non-responders. Expression at baseline (BL) and the end of treatment (EoT) in responders and non-responders to therapy of CX3CR1 (geometric mean fluorescence intensity [gMFI]) on CD56bright cells (A), NKG2A (percentage) on CD56dim NK cells (B), TRAIL (gMFI) on CD56bright cells (C), and IFN-γ (percentage) among NK cells (D), compared with healthy controls (HC). Horizontal bars indicate means with standard errors of the mean. Statistical analyses were performed using the Mann–Whitney U test. *P < .05 and **P < .01. Abbreviation: NS, nonsignificant.

A B C

(a)

(a)

(d)

(B)

(B)

(C)

(C)

figure 6. Baseline CD56bright natural killer (NK) cells associated with the alanine aminotransferase (ALT) level in hepatitis B virus e antigen–positive patients. The CD56bright NK cell percentage correlated positively with the ALT level (A), the geometric mean fluorescence intensity (gMFI) of TRAIL on CD56bright NK cells correlated positively with the ALT level (B), and the gMFI of CD38 on CD56bright NK cells correlated positively with the ALT level (C). Spearman correlation coefficient (r) and P values are indicated.

90

Chapter 5

dIsCussIonImmunomodulatory effects have been reported in patients with CHB treated with NUCs,

as well as in those treated with pegIFN.3,7,8,10,11 We have previously reported restoration in

the function of HBV-specificCD8+ T cells upon successful treatment with a combination

of pegIFN and adefovir.19 However, the effect of combination treatment on the innate

immune compartment is currently unknown, and there is little understanding of its role

in HBsAg clearance. In this study, NK cell characteristics of patients with CHB treated

with pegIFN combined with adefovir were extensively analysed. High levels of HBsAg

clearance that were achieved in our previously described cohort,14 resulted in unique

patient material enabling us to study the mechanism behind HBsAg loss. As shown

previously for pegIFN mono-therapy,3 combination treatment with pegIFN and adefovir

induced a shift from primarily CD56dim NK cells to more-immature CD56bright NK cells

in our cohort. CD56bright NK cells are known as the precursors of CD56dim NK cells

and gradually differentiate via CD56dim NKG2a+ KIR−CD57− to terminally differentiated

CD56dim NKG2A− KIR+CD57+.25 We observed an increase inNKG2A expression on both

NK cell subsets in our cohort (Supplementary Figure 3), supporting a shift in the overall

NK cell compartment toward a more immature phenotype during combination therapy.

Further- more, decreased expression of CD16 on CD56dim NK cells and increased CD94,

NKp30, and NKp46 expression on this subset also suggests more-immature CD56dim

NK cells.25–27 These results are in line with a shift toward more-immature NK cells after

pegIFN treatment for other conditions, such as HCV and hepatitis delta infection and

myeloproliferative neoplasms.28–30

This skewing toward an immature NK phenotype in the periphery could have different

aetiologies, such as an increased egress of progenitors from the bone marrow, an increased

proliferation of immature cells, altered migration of NK cell sub- types, or a loss of mature

NK cells. When analysing Ki67 expression in NK cells, we found an increase in proliferation

of CD56bright NK cells, as shown previously.3 However, an increase of Ki67 expression was

also observed in CD56dim NK cells, whereas these cells decreased in absolute numbers

in the periphery. Immature CD56dim NKG2A+KIR−CD57− NK cells express more Ki67

than terminally differentiated CD56dim NKG2A−KIR+CD57+ NK cells. Thus, a shift within

the CD56dim NK cell compartment toward immaturity near the end of treatment with

interferon-based therapy could possibly explain the relatively increased expression of Ki67

in this subset.25 Furthermore, the observed changes in chemokine receptor expression

suggests that migration to other organs might contribute to the increased CD56bright NK

cells and loss of proliferating CD56dim NK cells. However, we were unable to study this

assumption because measurements of NK cells were performed only in the peripheral blood

and not in other compartments, such as the lymph nodes and liver. Last, the susceptibility

to apoptosis could be altered within specific NK cell subsets, such as CD56dim NK cells,

during interferon-based treatment, resulting in loss of these cells.

NK cells carry the interferon alfa receptor and can therefore be directly activated by

type I interferons. Indeed, an increase in the activation status was seen in both CD56bright

91


5

and CD56dimNKcell subsets during therapy, as has been shown previously only for CD56bright NK cells after pegIFN monotherapy.3

The functional activity of NK cells has been reported to be increased, as well as decreased, during pegIFN monotherapy.3, 29 In our cohort, combination therapy was associated with a functional impairment of NK cells. We did, however, find higher IFN-γ production in response to target cells in responders with HBsAg loss, compared with nonre- sponders, at the end of treatment (Figure 5D), suggesting that target cell–induced cytotoxicity and cytokine release by NK cells may play an important role in the viral clearance induced by pegIFN/adefovir combination therapy.

Another difference between responders and non-responders was found with regard to TRAIL expression (Figure 5C). TRAIL expression not only increased during treatment with combination therapy, as reported in pegIFN monotherapy,3, 31 it was also higher in responders, compared with non-responders, at the end of treatment. TRAIL expression on NK cells has contrasting functional abilities. Whereas expression of TRAIL on NK cells can induce apoptosis in virally infected hepatocytes and fibrosis-inducing stellate cells, it has also been suggested that NK cells can kill HBV- specificCD8+ T cells via the TRAIL pathway, potentially impairing viral clearance.32, 33 Higher expression of TRAIL on CD56bright cells in responders to therapy at end of treatment suggests a beneficial role of this ligand, in which HBV-infected hepatocytes and stellate cells are eliminated by NK cells. Furthermore, lower CD56bright CX3CR1 expression in responders may indicate that the migratory abilities of these cells to sites of inflammation are crucial for their action. Lower expression of the inhibitory receptor NKG2A on CD56dimNKcells at baseline, as found in responders in this study, has previously been related to long-term virologic response in HCV-infected patients treated with pegIFN and ribavirin34 and could render these cells more susceptible to activation. Furthermore, blocking of NKG2A has previously been shown to augment NK cell function in CHB.21 Future studies will have to be conducted to assess whether these NK cell characteristics could become predictive markers of response early during pegIFN based therapy.

In a second group, which consisted of HBeAg-positive patients who had CHB, with or without HBeAg seroconversion, no relation between NK cell phenotype and response was found. This underlines the fact that HBeAg seroconversion is driven by a different immunological mechanism than HBsAg clearance. A relation between ALT level at baseline and CD56bright NK cells, TRAIL expression, and CD38 expression of these cells was found in HBeAg-positive patients (Figure 6). This suggests that NK cells play a role in hepatocyte lysis, either directly or indirectly, similar to the relationship between CD56bright TRAIL expression and ALT level that has been shown previously.9, 33

Combination therapy with pegIFN and NUC for patients with CHB is a successful treatment option that can give rise to relatively high proportion of HBsAg clearance. Here, we show that combination therapy induces radical changes in the NK cell compartment, increasing their activation. Furthermore, significant differences were found in NK cell phenotype in patients with HBsAg clearance and patients who had no response to therapy. A correlation between NK cell phenotype and ALT level suggests a role for NK cells in

92

Chapter 5

liver damage. While the exact role of NK cells with respect to viral clearance remains to be determined, better understanding of the actions of this immunological subset may help generate better treatment options for patients with CHB in the future.

93


5

referenCes1. Ott JJ, Stevens GA, Groeger J, et al.

Global epidemiology of hepatitis B virus infection: New estimates of age-specific HBsAg seroprevalence and endemicity. Vaccine 2012; 30:2212–9.

2. Beasley RP. Hepatitis B virus. The major etiology of hepatocellular carcinoma. Cancer 1988; 61:1942–56.

3. Micco L, Peppa D, Loggi E, et al. Differential boosting of innate and adaptive antiviral responses during pegylated-interferon-alpha therapy of chronic hepatitis B. J Hepatol 2013; 58:225–33.

4. Binder D, Fehr J, Hengartner H, et al. Virus-induced transient bone marrow aplasia: major role of interferon-alpha/beta during acute infection with the noncytopathic lymphocytic choriomeningitis virus. J Exp Med 1997; 185:517–30.

5. Marshall HD, Urban SL, Welsh RM. Virus-induced transient immune suppression and the inhibition of T cell proliferation by type I interferon. J Virol 2011; 85:5929–39.

6. Rehermann B, Lau D, Hoofnagle JH, et al. Cytotoxic T lymphocyte responsiveness after resolution of chronic hepatitis B virus infection. J Clin Invest 1996; 97:1655–65.

7. Tan AT, Hoang LT, Chin D, et al. Reduction of HBV replication prolongs the early immunological response to IFNα therapy. J Hepatol 2014; 60:54–61.

8. Penna A, Laccabue D, Libri I, et al. Peginterferon-α does not improve early peripheral blood HBV-specific T-cell responses in HBeAg-negative chronic hepatitis. J Hepatol 2012; 56:1239–46.

9. Peppa D, Micco L, Javaid A, et al. Blockade of immunosuppressive cytokines restores NK cell antiviral function in chronic hepatitis B virus infection. PLoS Pathog 2010; 6:e1001227.

10. Tjwa ET, van Oord GW, Hegmans JP, et al. Viral load reduction improves activation and

function of natural killer cells in patients with chronic hepatitis B. J Hepatol 2011; 54:209–18.

11. Boni C, Laccabue D, Lampertico P, et al. Restored function of HBV- specific T cells after long-term effective therapy with nucleos(t)ide analogues. Gastroenterology 2012; 143:963–73.

12. Thimme R, Dandri M. Dissecting the divergent effects of interferon- alpha on immune cells: time to rethink combination therapy in chronic hepatitis B? J Hepatol 2013; 58:205–9.

13. Janssen HL, van Zonneveld M, Senturk H, et al. Pegylated interferon alfa-2b alone or in combination with lamivudine for HBeAg-positive chronic hepatitis B: a randomised trial. Lancet 2005; 365:123–9.

14. Takkenberg RB, Jansen L, de Niet A, et al. Baseline hepatitis B surface antigen (HBsAg) as predictor of sustained HBsAg loss in chronic hepatitis B patients treated with peginterferon alfa-2a and adefovir. Antivir Ther 2013; 18:895–904.

15. Lutgehetmann M, Volzt T, Quaas A, et al. Sequential combination therapy leads to biochemical and histological improvement despite low ongoing intrahepatic hepatitis B virus replication. Antivir Ther 2008; 13:57–66.

16. Wursthorn K, Lutgehetmann M, Dandri M, et al. Peginterferon alpha- 2b plus adefovir induce strong cccDNA decline and HBsAg reduction in patients with chronic hepatitis B. Hepatology 2006; 44:675–84.

17. Kittner JM, Sprinzl MF, Grambihler A, et al. Adding pegylated interferon to a current nucleos(t)ide therapy leads to HBsAg seroconversion in a subgroup of patients with chronic hepatitis B. J Clin Virol 2012;54:93–5.

18. Ouzan D, Pénaranda G, Joly H, et al. Add-on peg-interferon leads to loss of HBsAg in patients with HBeAg-negative chronic hepatitis and HBV DNA fully suppressed

94

Chapter 5

by long-term nucleotide analogs. J Clin Virol 2013; 58:713–7.

19. de Niet A, Sinnige MJ, Takkenberg RB, et al. 278 Augmented hepatitis B specific Tcell response after clearance of chronic hepatitis B. J Hepatol 2011; 54:S114.

20. EASL clinical practice guidelines: Management of chronic hepatitis B virus infection. J Hepatol 2012; 57:167–85.

21. Li F,Wei H,Wei H, et al. Blocking the natural killer cell inhibitory receptor NKG2A increases activity of human natural killer cells and clears hepatitis B virus infection in mice. Gastroenterology 2013; 144:392–401.

22. Bellora F, Castriconi R, Dondero A, et al. Human NK cells and NK receptors. Immunol Lett 2014; 161:168–73.

23. Oo YH, Shetty S, Adams DH. The role of chemokines in the recruitment of lymphocytes to the liver. Dig Dis 2010; 28:31–44.

24. Becknell B, Caligiuri M. Interleukin-2, interleukin-15, and their roles in human natural killer cells. Adv Immunol 2005; 86:209–39.

25. Bjorkstrom NK, Riese P, Heuts F, et al. Expression patterns of NKG2A, KIR, and CD57 define a process of CD56 dim NK-cell differentiation uncoupled from NK-cell education. Blood. 2010; 116:3853–64.

26. Lopez-verges S, Milush JM, Pandey S, et al. CD57 defines a functionally distinct population of mature NK cells in the human CD56dim CD16+ NK-cell subset. Blood 2010; 116:3865–74.

27. Yu J, Mao HC,Wei M, et al. CD94 surface density identifies a functional intermediary between the CD56bright and CD56dim human NK-cell subsets. Blood 2010; 115:274–81.

28. Markova A, Mihm U, Schlaphoff V, et al. PEG-IFN alpha but not ribavirin alters NK cell phenotype and function in patients with chronic Hepatitis C. PLoS One 2014; 9:e94512.

29. Lunemann S, Malone DFG, Grabowski J, et al. Effects of HDV infection and pegylated interferon α treatment on the natural killer cell compartment in chronically infected individuals. Gut 2015; 64:469–82.

30. Riley CH, Hansen M, Brimnes MK, et al. Expansion of circulating CD56 (bright) natural killer cells in patients with JAK2-positive chronic myeloproliferative neoplasms during treatment with interferon-α. Eur J Haematol 2015; 94:227–34.

31. Stegmann K, Björkström NK, Veber H, et al. Interferon-alpha-induced TRAIL on natural killer cells is associated with control of hepatitis C virus infection. Gastroenterology 2010; 138:1885–97.

32. Peppa D, Gill U, Reynolds G. Up-regulation of a death receptor renders antiviral T cells susceptible to NK cell–mediated deletion. J Exp Med 2013; 210:99–114.

33. Dunn C, Brunetto M. Cytokines induced during chronic hepatitis B virus infection promote a pathway for NK cell–mediated liver damage. J Exp Med 2007; 204:667–80.

34. Golden Mason L, Bambha K. Natural killer inhibitory receptor expression associated with treatment failure and interleukin-28B genotype in patients with chronic hepatitis C. Hepatology 2011; 54: 1559–69.

95


5

aCKnowledGementsWe thank dr. E.B.M. Remmerswaal and M.H. Jansen for technical assistance; and the Academic Medical Center Laboratory of Medical Immunology for providing K562 cells.

supplementary data Chapter 5

Supplementary figures

Supplementary figure 1 | (A) Schematic overview of selected patients. Definitions of response as mentioned in methods section. *Note that due to restriction by availability of samples only 7 out of initial 8 responders at week 72 were used in cohort I. **In cohort II the HBeAg positive patients from cohort I were excluded (n=6) (B) Viral and clinical parameters of responders and non‐responders of cohort I.

supplementary figure 1. (A) Schematic overview of selected patients. Definitions of response as mentioned in methods section. *Note that due to restriction by availability of samples only 7 out of initial 8 responders at week 72 were used in cohort I. **In cohort II the HBeAg positive patients from cohort I were excluded (n=6) (B) Viral and clinical parameters of responders and non-responders of cohort I.

(a)

(B)

96

Chapter 5 Chapter 5

Supplementary figure 2 | Absolute numbers of NK cell subsets at BL and EoT. Gating strategy (A) total NK cells, CD56bright and CD56dim NK cells (B) and longitudinal analysis of CD56bright and CD56dim NK cells subsets in patients with CHB (n=13‐14 per timepoint) (C). *p<0.05; **p<0,01 ***p<0,001; ns non‐significant. FSC, forward scatter; SSC, side scatter; H, hight; W, width; CHB, chronic hepatitis B patients; BL baseline; EoT, end of treatment.

Supplementary figure 3 | Combination therapy in CHB patients results in a change in expression of activating and inhibiting markers. Representative staining for CD16, NKG2A, CD94, NKG2C, NKp30, NKp46, CD161 (A), frequencies or gMFI levels of different NK cells markers on CD56bright (B) and CD56dim (C) NK cells. Graphs show 14 patients with CHB at BL and EoT. Statistical analyses Wilcoxon test: *p<0.05; **p<0.01; ***p<0.001. HC, healthy control; CHB, chronic hepatitis B; BL baseline; EoT, end of treatment; gMFI geometric mean fluorescence intensity. (Internal negative = CD14+/CD19+ cells for CD16, NKG2A, CD94, NKG2C, NKp30, CD161 and CD3+ cells for NKp46).

(a)

(B) (C)

Chapter 5

Supplementary figure 2 | Absolute numbers of NK cell subsets at BL and EoT. Gating strategy (A) total NK cells, CD56bright and CD56dim NK cells (B) and longitudinal analysis of CD56bright and CD56dim NK cells subsets in patients with CHB (n=13‐14 per timepoint) (C). *p<0.05; **p<0,01 ***p<0,001; ns non‐significant. FSC, forward scatter; SSC, side scatter; H, hight; W, width; CHB, chronic hepatitis B patients; BL baseline; EoT, end of treatment.

Supplementary figure 3 | Combination therapy in CHB patients results in a change in expression of activating and inhibiting markers. Representative staining for CD16, NKG2A, CD94, NKG2C, NKp30, NKp46, CD161 (A), frequencies or gMFI levels of different NK cells markers on CD56bright (B) and CD56dim (C) NK cells. Graphs show 14 patients with CHB at BL and EoT. Statistical analyses Wilcoxon test: *p<0.05; **p<0.01; ***p<0.001. HC, healthy control; CHB, chronic hepatitis B; BL baseline; EoT, end of treatment; gMFI geometric mean fluorescence intensity. (Internal negative = CD14+/CD19+ cells for CD16, NKG2A, CD94, NKG2C, NKp30, CD161 and CD3+ cells for NKp46).

(a)

(B)

(C)

supplementary figure 2. Absolute numbers of NK cell subsets at BL and EoT. Gating strategy (A) total NK cells, CD56bright and CD56dim NK cells (B) and longitudinal analysis of CD56bright and CD56dim NK cells subsets in patients with CHB (n=13-14 per timepoint) (C). *p<0.05; **p<0,01 ***p<0,001; ns non-significant. FSC, forward scatter; SSC, side scatter; H, hight; W, width; CHB, chronic hepatitis B patients; BL baseline; EoT, end of treatment.

supplementary figure 3. Combination therapy in CHB patients results in a change in expression of activating and inhibiting markers. Representative staining for CD16, NKG2A, CD94, NKG2C, NKp30, NKp46, CD161 (A), frequencies or gMFI levels of different NK cells markers on CD56bright (B) and CD56dim (C) NK cells. Graphs show 14 patients with CHB at BL and EoT. Statistical analyses Wilcoxon test: *p<0.05; **p<0.01; ***p<0.001. HC, healthy control; CHB, chronic hepatitis B; BL baseline; EoT, end of treatment; gMFI geometric mean fluorescence intensity. (Internal negative = CD14+/CD19+ cells for CD16, NKG2A, CD94, NKG2C, NKp30, CD161 and CD3+ cells for NKp46).

97


5

Chapter 5

Supplementary figure 4 | Perforin, granzyme B and TRAIL expression increase during therapy. Expression of perforin (A), granzyme B (B) and TRAIL (C) in CD56bright and CD56dim NK cells. (D) Representative staining for perforin, granzyme B and TRAIL expression in the total NK cell gate at BL and EoT. Statistical analyses unpaired data: Mann Whitney U test, paired data: Wilcoxon test: *p<0.05; **p<0,01; ***p<0,001. CHB chronic hepatitis B; BL baseline; EoT, end of treatment; gMFI geometric mean fluorescence intensity.

Supplementary figure 5 | HBeAg positive patients who achieve HBeAg seroconversion without HBsAg clearance at T=72 after therapy do not have different baseline NK cell markers or functional activity than non‐responders. Horizontal bars indicate mean with SEM. ns, non significant. BL baseline; EoT, end of treatment; gMFI geometric mean fluorescence intensity.

Chapter 5

Supplementary figure 4 | Perforin, granzyme B and TRAIL expression increase during therapy. Expression of perforin (A), granzyme B (B) and TRAIL (C) in CD56bright and CD56dim NK cells. (D) Representative staining for perforin, granzyme B and TRAIL expression in the total NK cell gate at BL and EoT. Statistical analyses unpaired data: Mann Whitney U test, paired data: Wilcoxon test: *p<0.05; **p<0,01; ***p<0,001. CHB chronic hepatitis B; BL baseline; EoT, end of treatment; gMFI geometric mean fluorescence intensity.

Supplementary figure 5 | HBeAg positive patients who achieve HBeAg seroconversion without HBsAg clearance at T=72 after therapy do not have different baseline NK cell markers or functional activity than non‐responders. Horizontal bars indicate mean with SEM. ns, non significant. BL baseline; EoT, end of treatment; gMFI geometric mean fluorescence intensity.

(a)

(a)

(B)

(B)

(C)

(C)

(d)

(d) (e) (f)

supplementary figure 4. Perforin, granzyme B and TRAIL expression increase during therapy. Expression of perforin (A), granzyme B (B) and TRAIL (C) in CD56bright and CD56dim NK cells. (D) Representative staining for perforin, granzyme B and TRAIL expression in the total NK cell gate at BL and EoT. Statistical analyses unpaired data: Mann Whitney U test, paired data: Wilcoxon test: *p<0.05; **p<0,01; ***p<0,001. CHB chronic hepatitis B; BL baseline; EoT, end of treatment; gMFI geometric mean fluorescence intensity.

supplementary figure 5. HBeAg positive patients who achieve HBeAg seroconversion without HBsAg clearance at T=72 after therapy do not have different baseline NK cell markers or functional activity than non-responders. Horizontal bars indicate mean with SEM. ns, non significant. BL baseline; EoT, end of treatment; gMFI geometric mean fluorescence intensity.

98

Chapter 5

supplementary table 1. HBeAg positive patients with and without HBeAg seroconversion (HBeAg loss with development of anti-HBe antibodies) at week 72 after start of treatment. Values indicate frequency with percentage or median with range. HBeAg, hepatitis B e antigen; HBV DNA, hepatitis B virus DNA; IU, international units; HBsAg, hepatitis B surface antigen; ALT, alanine aminotransferase; IFN, interferon based therapy.

Characteristic hBeag seroconversion no hBeag seroconversion

N 9 (36%) 25 (64%)Gender male 6 (67%) 19 (76%)Age (years) 41 (20-55) 32 (19-50)HBV DNA (log10IU/ml) 8.01 (5.50-9.00) 8.5 (4.78-10.45)HBsAg (log10IU/ml) 4.34 (3.00-4.89) 4.48 (1.89-5.19)ALT (U/L) 129 (50-499) 63 (25-1071)Genotype A 4 (44%) 6 (24%) B 2 (22%) 6 (24%) C 2 (22%) 5 (20%) D 1 (11%) 6 (24%) E - 2 (8%)IFN naive 7 (78%) 21 (84%)

hla-C and KIr ComBIned Genotype as new response marKer for hBeaG-posItIve ChronIC

hepatItIs B patIents treated wIth Interferon-Based ComBInatIon therapy

F. Stelma, L. Jansen, M.J. Sinnige, K.A. van Dort, R.B. Takkenberg, H.L. Janssen, H.W. Reesink, N.A. Kootstra

Journal of Viral Hepatitis 2016;23:652-659

aBstraCtBackground & aims Current treatment for chronic hepatitis B infection (CHB) consists of interferon-based therapy. However, for unknown reasons a large proportion of patients with CHB does not respond to this treatment. Hence there is a pressing need to establish response markers to select patients who will benefit from therapy and to spare potential non-responders from unnecessary side-effects of antiviral therapy. Here we assessed whether HLA-C and KIR genotypes were associated with treatment outcome for CHB.

methods Twelve SNPs in or near the HLA-C gene were genotyped in 86 CHB patients (41 HBeAg-positive; 45 HBeAg-negative) treated with peginterferon alfa-2a + adefovir. Genotyping of killer immunoglobin-like receptors (KIRs) was performed by SSP-PCR.

results One SNP in HLA-C (rs2308557) was significantly associated with combined response in HBeAg-positive CHB patients (p = 0.003). This SNP is linked to the HLA-C group C1 or C2 classification, which controls KIR binding. The combination of KIR2DL1 with its ligand HLA-C2 was observed significantly more often in HBeAg-positive patients with a combined response (13/14) than in non-responders (11/27, p = 0.001). Patients with the KIR2DL1/C2 genotype had significantly higher baseline ALT levels (136 vs 50 U/L, p = 0.002) than patients without this combination. Furthermore, KIR2DL1-C2 predicted response independent of HBV genotype and ALT at baseline.

Conclusions HLA-C and KIR genotype is strongly associated with response in HBeAg-positive CHB patients treated with interferon-based therapy. In combination with other known response markers, HLA-C/KIR genotype could enable the selection of patients more likely to respond to interferon-based therapy.

103

KIR/HLA-C GENOTYPE AND RESPONSE TO COMBINATION THERAPY

6

IntroduCtIonChronic hepatitis B (CHB), with over 240 million people infected worldwide, forms a global health burden.1 In patients with CHB, liver inflammation may progress to liver cirrhosis and/or hepatocellular carcinoma.2 To reduce the risk of these hepatic complications, patients can be treated with either pegylated interferon (pegIFN) or nucleos(t)ide analogues (NUCs). Treatment with pegIFN consists of a fixed treatment duration and acts via immunomodulation. NUCs on the other hand, potently inhibit viral replication by blocking reverse transcription and have to be taken life-long. A combination of these two treatment modalities potentially increases the percentage of patients with HBsAg loss, which is the closest to curing hepatitis B virus (HBV) infection.3 In our previously described cohort of CHB patients treated with a combination of pegIFN alfa-2a and adefovir, HBsAg loss was reported in 11% of HBeAg-positive and 17% of HBeAg-negative patients 2 years after the cessation of treatment.4 Recently, a large randomized controlled study in CHB patients showed HBsAg loss of 7.5% at 6 months of follow-up after 48 weeks of pegIFN alfa-2a and tenofovir combination therapy. This percentage was significantly higher than when either therapy was given alone.5 However, for unknown reasons, a large proportion of patients with CHB do not respond to pegIFN-based therapy. In addition, treatment with pegIFN is associated with significant side-effects. Hence there is a pressing need to establish response markers to aid clinicians in selecting those patients who will benefit from therapy and to spare potential non-responders from unnecessary side-effects and costs of antiviral therapy.

Possible markers include host factors related to the immune system. As the immune system plays a key role in the clearance of HBV, this is a promising area to investigate predictive markers of response. Not only HBV-specific T cells have been shown to be of importance in eradicating HBV,6 the activity of NK cells has also been shown to be important in determining treatment outcome.7,8

Human leukocyte antigen (HLA) class I molecules are important in the defence against viral infection. Peptides presented in HLA-C molecules can be recognized by cytotoxic T cells, which can then mount an effector response. Furthermore, HLA-C molecules can interact with killer immunoglobin-like receptors (KIRs) on the surface of natural killer (NK) cells, and either activate or inhibit these lymphocytes.

The course of various infectious diseases has been associated with polymorphisms in the HLA-C gene.9,10 In patients with acute hepatitis C infection, the HLA-C genotype in combination with KIRs has been associated with a spontaneous resolution of the infection.11 Furthermore, HLA-C genotype is associated with response to treatment with pegIFN and ribavirin in patients with chronic hepatitis C infection.12 In HBV, single nucleotide polymorphisms (SNPs) near HLA-C have been associated with chronicity.13,14 However, thus far, no relation with the response to treatment has been reported in CHB.

In this study we analysed HLA-C polymorphisms in relation to treatment outcome in a cohort of CHB patients treated with pegIFN and adefovir combination therapy in order to assess the potential importance of HLA-C genotype as a response biomarker.

104

Chapter 6

patIents and methods patientsIn a previous study, ninety-two CHB patients with a high viral load >17,182 IU/mL (>100,000 copies/mL) were treated with combination therapy consisting of peginterferon alfa-2a and adefovir for 48 weeks and followed by a treatment-free follow-up (week 72) (controlled-trials.com; ISRCTN 77073364). All patients gave written informed consent, and the study was approved by the Ethical Review Board of the Academic Medical Center Amsterdam. In total, 86 Patients (41 HBeAg-positive; 45 HBeAg-negative) completed the study. HBV genotype determination, quantification of plasma HBV DNA and serum HBsAg and detection of anti-HBs, HBeAg and anti-HBe were performed as described before.4 All patients were HIV negative and not co-infected with hepatitis C or hepatitis delta virus. Combined response was defined according to EASL guidelines as a persistently low HBV DNA (<2,000 IU/mL) with normalization of alanine aminotransferase (ALT). For HBeAg-positive patients this included loss of HBeAg.15

single nucleotide polymorphisms in the hla-C regionGenomic DNA, extracted from peripheral blood mononuclear cells, was available from 84 patients. Single nucleotide polymorphisms (SNPs) were genotyped with the HumanOmni1-Quad BeadChip (v1.0, Illumina Inc., San Diego, California, USA) at Hoffman-La Roche (Nutley, NJ, USA), as part of a previous genome-wide association study.16

Data on linkage disequilibrium of the included SNPs with proxy SNPs was calculated using the SNP Annotation and Proxy (SNAP) server.17 In this program, linkage disequilibrium data based on genotypes available from the International HapMap project are calculated.18

hla-C allele typingStatistical imputation of HLA-C alleles from SNP genotypes was performed using the HLA type imputation model, HLA*IMP:02.19 With this model we used 430 SNPs in the MHC region to impute HLA-C alleles at the 4-digit level. The mean posterior probability of the specified alleles was 0.96 ± 0.01, and alleles with a posterior probability of < 0.50 were excluded (n=6).

KIr genotypingThe presence or absence of four KIR genes known to have HLA-C as ligand (KIR2DL1, KIR2DL2, KIR2DL3, KIR2DS1) was determined by polymerase chain reaction using sequence specific primers (listed in supplementary table 1) as described before.20 Two sets of primers (Invitrogen) were used for each locus and when PCR results were indistinct, subsequent sequencing was performed.

statistical analyses Statistical analyses were performed using IBM SPSS Statistics v21.0.0.1 (IBM Corp. Armonk, NY). Frequency comparisons were made by χ2-test and two-sided Fisher’s exact test when

105


6

appropriate. p values <0.05 were considered statistically significant. Multivariable analyses were performed by logistic regression analyses.

results patient characteristicsBaseline characteristics of the 86 patients included in this study are summarized in table 1. At week 72 (24 weeks of treatment-free follow-up), 14/41 (35%) HBeAg-positive and 16/45 (36%) HBeAg-negative patients had a combined response.

association of snps in hla-C with therapy responseFor the present study, SNPs located within or 500 base pairs up- or downstream of the HLA-C gene and with a minor allele frequency >5% were included in the association analysis (n=12). Of the 12 SNPs included in this analysis, one - rs2308557 - was significantly

table 1. Baseline characteristics of CHB patients treated with peg-interferon/ adefovir according to HBeAg status.

demographicshBeag1-positive

(n=41)hBeag-negative

(n=45)

Female, n (%) 8 (19) 15 (32)Mean age, years (SD) 35.76 (9.45) 43.07 (9.84)Ethnicity Caucasian, n (%) 16 (38) 12 (26) Asian, n (%) 19 (45) 14 (30) African, n (%) 7 (17) 21 (45) IFN2 naïve, n (%) 33 (79) 33 (70)Laboratory characteristics Median ALT3, U/L (iqr) 97 (50-215) 63 (47-118)HBV Genotype A, n (%) 17 (40) 11 (23) B, n (%) 8 (19) 7 (15) C, n (%) 7 (17) 5 (11) D, n (%) 8 (19) 17 (36) E, n (%) 2 (5) 7 (15)Mean HBV-DNA4, log10 IU/mL (SD) 8.02 (1.23) 5.56 (1.08)Mean HBsAg5, log10 IU/mL (SD) 4.29 (0.74) 3.33 (0.67)Liverbiopsycharacteristics, n (%) Median inflammatory score HAI, (iqr) 5 (3-7) 5 (3-9) Median Ishak fibrosis score, (iqr) 1 (1-3) 1 (1-4)Therapy response HBeAg loss at week 72, n (%) 14 (35) - Combined response at week 72, n (%) 14 (35) 16 (36)

1 hepatitis B e antigen; 2 interferon-based treatment; 3 alanine aminotransferase; 4 hepatitis B virus DNA; 5 hepatitis B surface antigen.

106

Chapter 6

associated with a combined response in HBeAg-positive patients (p = 0.003, table 2), but not in HBeAg-negative patients. The other SNPs in the HLA-C region were not related to treatment outcome (all p>0.10).

snp rs2308557 genotype is linked to hla-C group 1 or 2As the interpretation of SNP data can be confounded by nearby, closely linked alleles, we performed a search for linkage disequilibrium of rs2308557 with other SNPs available in the 1000 genomes project using the SNAP server.17 This resulted in the identification of SNP rs17408553, that is in near complete linkage-disequilibrium with rs2308557 (Figure 1).

SNP rs2308557 (G/A) causes a serine to asparagine change at position of 77 of the α1domain of HLA-C, whereas SNP rs17408553 (C/A) determines the presence of either an asparagine or lysine at position 80 of this domain (Figure 1). Both amino acid positions are located in the peptide binding groove, however only the dimorphism at position 80 was shown to be crucial in determining the binding capacity of HLA-C to specific Killer Ig-like Receptor (KIR) groups.21 Hence, this amino-acid dimorphism divides HLA-C in two clinically important subgroups: HLA-C group 1 (C1) and group 2 (C2). HLA-C1 and HLA-C2 act as ligands, each recognized by a distinct set of KIRs; KIR2DL1 and KIR2DS1 bind to HLA-C2, whereas KIR2DL2 and KIR2DL3 bind to HLA-C1. Therefore, we next studied the HLA-C1/C2 and KIR2D genotype in our cohort.

accuracy of hla-C genotyping by snp analysis.SNP rs17408553, which determined HLA-C group 1 or 2 was not genotyped in our cohort. As SNP rs2308557 is in near complete L/D with SNP rs17408553, the former could be

table 2. Frequencies of SNPs in HBeAg-positive and HBeAg-negative responders and non-responders to combination therapy. Statistically significant differences were determined by chi-square test, corresponding p-values are given.

name position snp maf1

hBeag-positive hBeag-negative

Cr2 nr3 p Cr nr p

rs2524099 6:31268274 T/C 0.47 10 71% 18 69% 0.885 11 65% 21 78% 0.343rs2853951 6:31268338 T/C 0.31 6 43% 12 46% 0.842 8 47% 16 59% 0.429rs9394047 6:31268473 T/G 0.11 5 36% 7 27% 0.563 7 41% 7 26% 0.290rs34131062 6:31268633 A/G 0.14 3 21% 7 27% 0.702 2 12% 4 15% 0.774rs2524096 6:31268690 T/G 0.44 11 79% 19 73% 0.702 13 77% 14 52% 0.102rs2001181 6:31269221 T/C 0.12 1 7% 8 31% 0.124 2 12% 5 19% 0.551rs9264601 6:31269577 A/G 0.06 3 21% 1 4% 0.115 2 12% 2 7% 0.624rs9264602 6:31269629 T/C 0.26 6 43% 12 46% 0.842 8 47% 16 59% 0.429rs1131096 6:31270482 A/C 0.29 6 43% 13 50% 0.666 9 53% 19 70% 0.242rs2308557 6:31271640 A/G 0.39 12 92% 11 42% 0.003 10 59% 16 59% 0.977rs9366775 6:31272319 T/C 0.06 2 14% 3 12% 0.838 3 18% 5 19% 0.942rs2074489 6:31272351 A/G 0.27 7 50% 13 50% 1.000 4 24% 11 41% 0.241

1 minor allele frequency; 2 combined response; 3 non-response.

107


6

used to replace the latter to genotype HLA-C group 1 or 2. However, we wished to test

the accuracy of classifying patients into HLA-C group 1 or 2 in this way. It was done by

comparing imputed data for HLA-C alleles with the HLA-C classification based on SNP

rs2308557. Imputation of HLA-C alleles from a set of 430 SNP genotypes in the HLA

region was found to be in complete agreement with the HLA-C subgroup classification

based on SNP rs2308557 genotype (160/160 alleles) (Supplementary table 2).

distribution of hla-C and KIr genotypes in patients with ChBThe distribution of HLA-C1 and -C2 genotypes in HBeAg-positive patients was highly similar

to that in HBeAg-negative patients (table3). As described above, HLA-C is recognized by

KIR2D receptors, which were genotyped in our cohort of CHB patients. KIR genotype

distribution differed significantly between HBeAg-positive and –negative patients. Higher

frequencies of KIR2DL2 were observed in HBeAg-negative patients, whereas KIR2DL3 was

more prevalent in the HBeAg-positive group (table3).

When the HLA-C genotype was stratified by ethnic origin, Asian subjects were shown to

predominantly have the C1C1 genotype (21/32, 65.6%) and hardly any C2C2 (1/32, 3.1%),

whereas in Caucasian patients the genotype was mostly heterozygous (18/27, 66.7%). KIR

figure 1. Regional linkage disequilibrium (L/D) plot. The figure shows the L/D (y-axis) of SNP rs2308557 with other SNPs located within or near the HLA-C gene on chromosome 6 (x-axis). L/D data was calculated based on genotype information from the International HapMap CEU population (Utah residents with ancestry from northern and western Europe). Red diamonds represent SNPs included in the current association analysis, and grey dots represent additional SNPs from the HapMap database.

31,271,620 G T A G C C G C G C A G G T T C C G C A G G C T C A C T C G G T C C A T C G G C G C G T C C A A G G C G T C C G A G T G A G C C A G

Tyr Gly Arg Leu Asn Arg Leu Ser Val Arg Asp Lys Asn

A A rs17408553 rs2308557

31,271,630 31,271,640 31,271,650 - 3’ - 5’ 3’-

5’-

HLA-C α1 domain

31,268

HLA-C

31,269 31,270 31,271 31,272 (kb)

LD (R

-squ

ared

)

0.2

0.4

0.6

0.8

1.0

LD (R

-squared)

0.2

0.4

0.6

0.8

1.0

108

Chapter 6

genes were similarly distributed in Caucasian and Asian subjects, whereas KIR distribution in Asian patients differed (supplementary table 3).

association of hla-C and KIr genotype with therapy responseIn HBeAg-positive patients, the functionally relevant combination of KIR2DL1 with an HLA-C2 allele (KIR2DL1-C2) was observed significantly more often in responders (13/14) than in non-responders (11/27)(p = 0.002 OR=18.909, 95% CI=2.150-166.275) (figure 2). For the combination KIR2DL1 and homozygosity for HLA-C2 (KIR2DL1-C2C2), this was not the case.

The counterpart of the combination KIR2DL1-C2, is KIR2DL3-C1C1, which was present in most patients who were negative for KIR2DL1-C2. This combined genotype was more prevalent in HBeAg-positive non-responders than in HBeAg-positive patients with combined response (p = 0.003 OR=0.062, 95% CI=0.007-0.540) (figure 2).

Furthermore, in HBeAg-positive patients with KIR2DL1-C2, higher baseline ALT levels were observed than in patients without this combination, indicating a possible higher immunologic activity in these patients (median ALT 136 vs. 50 U/L, p = 0.002). Moreover, in baseline liverbiopsies, 7/19 of the patients with genotype KIR2DL1-C2 had an Histology Activity Index (HAI) score above 8, compared to 0 of the 12 patients who did not have the combination KIR2DL1-C2 (p = 0.026).

predictive value of the KIr-hla genotype in treatment responseTo determine whether the KIR-HLA genotype was an independent predictor of response to treatment in HBeAg-positive patients, the relation that was found between KIR2DL1 /HLA-C2 genotype and combined response to therapy was analysed in a logistic regression model. In a multivariable model, independent variables were added based on previous

table 3. HLA-C and KIR genotype distribution in patients with CHB, according to HBeAg status.

hla-C1 genotype

hBeag-positive n=41

hBeag-negative n=45

p valuen % n %

C1C1 17 41% 19 42% 0.982C1C2 17 41% 19 42% C2C2 7 17% 7 16% HLA-C1 allele 34 83% 38 84% 0.849HLA-C2 allele 24 59% 26 58% 0.943KIrs1 2DL1 41 100% 45 100% - 2DL2 14 34% 27 60% 0.017 2DL3 40 98% 38 84% 0.036 2DS1 14 34% 15 33% 0.937

1 Human leukocyte antigen C; 2 killer immunoglobin-like receptors.

109


6

associations with response (i.e. HBV genotype and ALT). In HBeAg-positive patients,

the presence of KIR2DL1-C2 predicted response independently of HBV genotype and ALT

levels at baseline (table 4).

dIsCussIonIn this study we showed that the genotypic combination of KIR and its cognate HLA-C

ligand is highly associated with the response to interferon-based therapy in HBeAg-

positive CHB patients. In an analyses of known SNPs within the HLA-C locus, one SNP

was associated with treatment response in HBeAg-positive, but not in HBeAg-negative

patients. Further investigation revealed a relationship of this SNP with the HLA-C group

1 and 2 classification. Combining the HLA-C groups with their cognate KIR genotypes

led to the identification of the favorable KIR2DL1-HLA-C2 combination in HBeAg-positive

figure 2. KIR-HLA-C compound genotype in HBeAg-positive patients with combined response (CR72) and non-response (NR72) at week 72. Level of significance as determined by chi-square test. **: p < 0.01.

table 4. Univariable and multivariable logistic regression analysis for the effect of independent variables on combined response at week 72 in HBeAg-positive patients in two models.

univariable multivariable model I multivariable model II

B1 s.e.2 p value B s.e. p value B s.e. p value

KIR2DL1-C2 2.940 1.11 0.008** 3.196 1.14 0.005** 2.880 1.15 0.012*ALT3 (xULN) 0.003 0.06 0.963 -0.070 0.07 0.327 - - -HBV gtA4 0.981 0.68 0.147 - - - 0.151 0.79 0.849

1 logistic regression coefficient; 2 standard error; 3 alanine aminotransferase; 4 Hepatitis B virus genotype A.

HBeAg positive

2DL1

-C2

2DL1

-C2C

2

2DL2

-C1

2DL2

-C1C

1

2DL3

-C1

2DL3

-C1C

1

2DS1

-C2

2DS1

-C2C

2

2DS2

-C1

2DS2

-C1C

1

0

20

40

60

80

100

CR72NR72

**

**Fr

eque

ncy

(%)

110

Chapter 6

patients who were treated with interferon-based combination therapy. Despite relatively small patient numbers, this association with response persisted after multivariable analysis in HBeAg-positive CHB patients.

The mechanism by which patients with this KIR-HLA-C combination are more likely to respond to therapy is currently unknown. Previous studies have shown a significant association of KIR-HLA-C interactions with the resolution of acute HCV infection. However this was only observed in adult patients who were exposed to a low dose of HCV,11 and it is still unknown how these interactions would change in a chronic setting, let alone in case of a different viral infection.

Here we have analyzed the effect of the KIR/HLA-C genotype on treatment outcome in CHB patients. However, the exact phenotype, i.e. the expression pattern of these KIRs in our patients is not known, neither is that of HLA-C. It is not exactly known how an NK cell forms its KIR expression repertoire and whether HLA molecules play a role in this. However, not every KIR by itself, but interactions between the complete KIR repertoire and the available HLA-C ligands have been shown to be of influence on the activity of the individual KIRs.22 Furthermore, these expression levels in the liver might constantly be influenced by the presence of chronic infection.

In addition, HLA class I molecules such as HLA-C can occur as a soluble form. It has been described that serum soluble HLA class I is markedly increased in viral hepatitis,23 possibly due to elevated levels of endogenous interferons. Furthermore, exogenous interferons can also lead to increased levels of soluble HLA class I molecules.24 Therefore, the function of NK cells expressing KIRs could be influenced by such a fluctuation in soluble HLA ligands upon the initiation of therapy.

We observed that the high affinity KIR2DL1-C2 combination is associated with the response to treatment in HBeAg-positive CHB patients and higher baseline ALT and HAI scores. The paradoxal inhibitory genotype combined with a more immuno-active phenotype might be explained by the NK cell education or licensing theory in which inhibitory KIR and HLA-C interactions early during NK cell development are thought to have an impact on NK cell activity. Licensing is a process in which the NK cell is educated to recognize self (HLA class I) in order to be able to respond to subsequent missing self on a target cell.25 This results in functional activity when an NK cell expresses an inhibitory KIR that recognizes self HLA class I, and hypo responsiveness in the case of NK cell inhibitory KIR expression that does not interact with self MHC class I molecules. 26 Our data show a higher frequency of KIR2DL1-C2 in responders to therapy, possibly representing a greater number of licensed NK cells that will have a lower threshold to become functionally reactive upon target cell recognition. The reason that the effect does not increase with KIR2DL2 in combination with homozygosity for HLA-C2 may be explained by the need for other KIR/HLA-C combinations, i.e. HLA-C heterozygous patients could license even more NK cells.

Thus far, several markers of response to therapy have been identified for HBeAg-positive as well as –negative CHB patients. For HBeAg-positive patients these include levels of HBV DNA, ALT27 and HBsAg/anti-HBs immune complexes at baseline,28 as well as the HBV genotype and HBsAg decline.29 In HBeAg negative patients, response to therapy

111


6

has been associated with baseline HBsAg4 and ALT levels30, specific SNPs16,31 and more recently, levels of HBV DNA and HBsAg early during therapy32–34 are taken into account. Thus far, no studies have been performed linking KIR-HLA-C genotype to treatment outcome for CHB. Our study shows a significant association between KIR-HLA-C genotype and combined response to immunomodulatory-based treatment in HBeAg-positive CHB patients. While the clinical significance of these findings has yet to be validated in further studies, KIR-HLA-C genotype in combination with other markers may enable the prediction of successful interferon-based treatment in HBeAg-positive CHB patients, thus enabling the identification of patients that might benefit from therapeutic interventions.

112

Chapter 6

referenCes1. Ott JJ, Stevens GA, Groeger J, Wiersma

ST. Global epidemiology of hepatitis B virus infection: new estimates of age-specific HBsAg seroprevalence and endemicity. Vaccine 2012; 30: 2212–2219.

2. Beasley RP. Hepatitis B virus. The major etiology of hepatocellular carcinoma. Cancer 1988; 61: 1942– 1956.

3. Thimme R, Dandri M. Dissecting the divergent effects of interferon-alpha on immune cells: time to rethink combination therapy in chronic hepatitis B? J Hepatol 2013; 58: 205– 209.

4. Takkenberg RB, Jansen L, de Niet A et al. Baseline hepatitis B surface antigen (HBsAg) as predictor of sustained HBsAg loss in chronic hepatitis B patients treated with peginterferon alfa-2a and adefovir. Antivir Ther 2013; 18: 895–904.

5. Marcellin P, Ahn SH, Ma X et al. 193 HBsAg loss with Tenofovir Disoproxil Fumarate (TDF) plus Peginterferon alfa-2a (PEG) in Chronic Hepatitis B (CHB): results of a global randomized controlled trial. Hepatology 2014; 60: 294A.

6. Maini MK,Boni C, Ogg GS et al. Direct ex vivo analysis of hepatitis B virus- specific CD8(+) T cells associated with the control of infection. Gastroenterology 1999; 117: 1386–1396.

7. Stelma F, de Niet A, Tempelmans Plat-Sinnige MJ et al. Natural killer cell characteristics in patients with chronic hepatitis B virus (HBV) infection are associated with HBV surface antigen clearance after combination treatment with pegylated interferon alfa-2a and adefovir. J Infect Dis 2015; 212: 1042– 1051.

8. Micco L, Peppa D, Loggi E et al. Differential boosting of innate and adaptive antiviral responses during pegylated-interferon-alpha therapy of chronic hepatitis B. J Hepatol 2013; 58: 225–233.

9. Blais ME, Dong T, Rowland-Jones S. HLA-C as a mediator of natural killer and

T-cell activation: Spectator or key player? Immunology 2011; 133: 1–7.

10. Zhang Y, Zhang H, Huang Y, et al. Human leukocyte antigen (HLA)-C polymorphisms are associated with a decreased risk of rheumatoid arthritis. Mol Biol Rep 2014; 41: 4103–4108.

11. Khakoo SI, Thio CL, Martin MP et al. HLA and NK cell inhibitory receptor genes in resolving hepatitis C virus infection. Science 2004; 305: 872–874.

12. Suppiah V, Gaudieri S, Armstrong NJ et al. IL28B, HLA-C, and KIR variants additively predict response to therapy in chronic hepatitis C virus infection in a European Cohort: a cross-sectional study. PLoS Med 2011; 8: e1001092.

13. Hu Z, Liu Y, Zhai X et al. New loci associated with chronic hepatitis B virus infection in Han Chinese. Nat Genet 2013; 45: 1499– 1503.

14. Jiang D, Ma X, Yu H et al. Genetic variants in five novel loci including CFB and CD40 predispose to chronic hepatitis B. Hepatology 2015; 62: 118–128.

15. EASL clinical practice guidelines. Management of chronic hepatitis B virus infection. J Hepatol 2012; 57: 167–185.

16. Jansen L, de Niet A, Stelma F et al. HBsAg loss in patients treated with peginterferon alfa-2a and adefovir is associated with SLC16A9 gene variation and lower plasma carnitine levels. J Hepatol 2014; 61: 730–737.

17. Johnson AD, Handsaker RE, Pulit SL, et al. SNAP: a web-based tool for identification and annotation of proxy SNPs using HapMap. Bioinformatics 2008; 24: 2938–2939.

18. The International HapMap Consortium. A second generation human haplotype map of over 3.1 million SNPs. Nature 2007; 449: 851–861.

113


6

19. Dilthey A, Leslie S, Moutsianas L et al. Multi-Population Classical HLA Type Imputation. PLoS Comput Biol 2013; 9: e1002877.

20. Kulkarni S, Martin MP, Carrington M. KIR genotyping by multiplex PCR-SSP. Methods Mol Biol 2010; 612: 365–375.

21. Mandelboim O, Reyburn HT, Pazmany L, et al. Protection from lysis by natural killer cells of group 1 and 2 specificity is mediated by residue 80 in human histocompatibility leukocyte antigen C alleles and also occurs with empty major histocompatibility complex molecules. J Exp Med 1996; 184: 913–922.

22. David G, Djaoud Z, Willem C et al. Large spectrum of HLA-C recognition by killer Ig-like receptor (KIR) 2DL2 and KIR2DL3 and restricted C1 specificity of KIR2DS2: dominant impact of KIR2DL2/KIR2DS2 on KIR2D NK cell repertoire formation. J Immunol 2013; 191: 4778–4788.

23. Puppo F, Scudeletti M, Indiveri F, et al. Serum HLA class I antigens: markers and modulators of an immune response? Immunol Today 1995; 16: 124–127.

24. Aulitzky WE, Grosse-Wilde H, Westhoff U et al. Enhanced serum levels of soluble HLA class I molecules are induced by treatment with recombinant interferon-gamma (IFN-gamma). Clin Exp Immunol 1991; 86: 236–239.

25. Kim S, Poursine-Laurent J, Truscott SM et al. Licensing of natural killer cells by host major histocompatibility complex class I molecules. Nature 2005; 436: 709–713.

26. Anfossi N, Andre P, Guia S et al. Human NK cell education by inhibitory receptors for MHC class I. Immunity 2006; 25: 331–342.

27. Buster EHCJ, Hansen BE, Lau GKK et al. Factors that predict response of patients

with hepatitis B e antigen-positive chronic hepatitis B to peginterferon-alfa. Gastroenterology 2009; 137: 2002–2009.

28. de Niet A, Jansen L, Zaaijer HL et al. Experimental HBsAg/anti-HBs complex assay for prediction of HBeAg loss in chronic hepatitis B patients treated with pegylated interferon and adefovir. Antivir Ther 2014; 19: 259–267.

29. Sonneveld MJ, Hansen BE, Piratvisuth T et al. Response-guided peginterferon therapy in hepatitis B e antigen-positive chronic hepatitis B using serum hepatitis B surface antigen levels. Hepatology 2013; 58: 872–880.

30. Marcellin P, Bonino F, Lau GKK et al. Sustained response of hepatitis B e antigen-negative patients 3 years after treatment with peginterferon alfa-2a. Gastroenterology 2009; 136: 2169–2179. e4.

31. Lampertico P, Vigano M, Cheroni C et al. IL28B polymorphisms predict interferon-related hepatitis B surface antigen seroclearance in genotype D hepatitis B e antigen-negative patients with chronic hepatitis B. Hepatology 2013; 57: 890–896.

32. Rijckborst V, Hansen BE, Cakaloglu Y et al. Early on-treatment prediction of response to peginterferon alfa-2a for HBeAg-negative chronic hepatitis B using HBsAg and HBV DNA levels. Hepatology 2010; 52: 454–461.

33. Brunetto MR, Marcellin P, Cherubini B et al. Response to peginterferon alfa-2a (40KD) in HBeAg-negative CHB: on-treatment kinetics of HBsAg serum levels vary by HBV genotype. J Hepatol 2013; 59: 1153–1159.

34. Moucari R, Mackiewicz V, Lada O et al. Early serum HBsAg drop: a strong predictor of sustained virological response to pegylated interferon alfa-2a in HBeAg-negative patients. Hepatology 2009; 49: 1151–1157.

114

Chapter 6

aCKnowledGementsThe authors would like to thank Prof.dr. C.P. Engelfriet for critically revising the manuscript. This work was supported by Roche.

supplementary dataSupplementary data associated with this article available can be found in the online version of this article.

hBsaG loss In patIents treated wIth peGInterferon alfa-2a and adefovIr Is

assoCIated wIth slC16a9 Gene varIatIon and lower plasma CarnItIne levels

L. Jansen, F. Stelma*, A. de Niet*, E.P. van Iperen, K.A. van Dort, M.J. Sinnige, R.B. Takkenberg, D.J. Chin, A.H. Zwinderman, U. Lopatin, N.A. Kootstra, H.W. Reesink

Journal of Hepatology 2014;61:730-737


aBstraCtBackground & aims Achievement of HBsAg loss remains the hallmark of chronic hepatitis B treatment. In order to identify host factors contributing to treatment-induced HBsAg loss, we performed a genome-wide screen of single nucleotide polymorphisms (SNPs) and studied its immunological consequence.

methods Chronic hepatitis B patients (40 HBeAg-positive and 44 HBeAg-negative) treated with peginterferon alfa-2a and adefovir were genotyped for 999,091 SNPs, which were associated with HBsAg loss at week 96 (n = 9). Plasma carnitine levels were measured by tandem-mass spectrometry, and the effect of carnitine on the proliferative capacity of hepatitis B virus (HBV)-specific and non-specific CD8 T cells was studied in vitro.

results One polymorphism, rs12356193 located in the SLC16A9 gene, was genome-wide significantly associated with HBsAg loss at week 96 (p = 1.84 x 10-8). The previously reported association of rs12356193 with lower carnitine levels was confirmed in our cohort, and baseline carnitine levels were lower in patients with HBsAg loss compared to patients with HBsAg persistence (p = 0.02). Furthermore, we demonstrated that carnitine suppressed HBV-specific CD8 T cell proliferation.

Conclusions In chronic hepatitis B patients treated with peginterferon and adefovir, we identified strong associations of SLC16A9 gene variation and carnitine levels with HBsAg loss. Our results further suggest that a lower baseline plasma carnitine level increases the proliferative capacity of CD8 T cells, making patients more susceptible to the immunological effect of this treatment. These novel findings may provide new insight into factors involved in treatment-induced HBsAg loss, and play a role in the prediction of treatment outcome.

119

SLC16A9 AND CARNITINE ASSOCIATED WITH HBSAG LOSS

7

IntroduCtIonChronic hepatitis B (CHB) is a major health problem, affecting more than 240 million people worldwide.1 Persons with prolonged hepatitis B virus (HBV) infection are at increased risk of developing liver fibrosis, cirrhosis and ultimately hepatocellular carcinoma (HCC) and death.2

The disease spectrum of CHB infection is diverse, and current guidelines recommend treatment of CHB patients with high viral load and active liver inflammation. There are two approved treatment options: a finite course of pegylated interferon-alfa (pegIFN) or long-term administration of nucleos(t)ide analogues (NUCs).3,4 Although NUCs potently suppress viral replication and prevent disease progression in most patients, they must be administered long-term (potentially for life) as they do not affect the viral reservoir – the covalently closed circular DNA (cccDNA) of HBV in hepatocytes. In contrast, patients on pegIFN appear to have a higher chance to achieve clearance of HBsAg compared with patients on NUCs. HBsAg loss is considered the closest outcome to cure, and is associated with improved survival.5 Nevertheless, while HBsAg loss occurs more frequently on pegIFN than on NUC based therapy, this clinical outcome is rarely achieved with rates of 3–7% in large clinical trials with or without lamivudine.6–8 In addition, clinical use of pegIFN is limited by significant side-effects.

It is therefore of utmost importance to be able to select those patients who are most likely to benefit from pegIFN monotherapy or combination with NUCs. Previously, higher rates of HBsAg loss were observed in HBeAg-positive patients with HBV genotype A and B,9 and in some studies, including the cohort studied here, lower baseline HBsAg levels were associated with HBsAg loss in HBeAg-negative patients.10,11

Unfortunately, these mainly virological markers account for only a small proportion of the variance in response to treatment. However, host genetic factors may also play a role in response of CHB to pegIFN. Genome-wide association studies (GWAS) have proven useful as a method for identifying genetic variations associated with diseases or treatment outcome. For example in hepatitis C, where genetic variation near the interleukin 28B (IL28B) gene is widely accepted as a strong predictor of viral clearance after treatment with pegIFN and ribavirin.12 In contrast, studies on the effect of the IL28B genotype in hepatitis B treatment have yielded conflicting findings.13–16 Although still under debate, the influence of the IL28B polymorphism is definitely not as pronounced in CHB as it is for chronic hepatitis C. Several other candidate gene studies have reported possible associations with IFN treatment response, but none had a level of significance that would lead to general acceptance of the finding. GWAS in East Asian populations did find that genetic variants in the HLA-DP locus were strongly associated with persistent HBV infection.17 However, the question of IFN response in patients with established chronic HBV infection was not addressed.

In this study we aimed to identify novel human genetic contributions to treatment-induced loss of serum HBsAg. We performed a genome-wide screen of single nucleotide polymorphisms (SNPs) in patients treated with pegIFN and adefovir. We identified a SNP (rs12356193), which is associated with HBsAg loss with genome-wide significance

120

Chapter 7

(p = 1.84 x 10-8) and confirmed its association with plasma carnitine levels.18 In concordance with this finding, we observed a strong association between lower carnitine levels in plasma and HBsAg loss. In addition, we demonstrated that the presence of carnitine limited the proliferation of HBV- specific T cells in vitro.

patIents and methods subjectsThis study was performed in 92 CHB patients who participated in an investigator-initiated study, carried out at the Academic Medical Center (AMC) in Amsterdam and the Erasmus University Medical Center (EMC) in Rotterdam, The Netherlands. The inclusion and exclusion criteria had been described elsewhere.11 In summary, patients had documented HBsAg positivity for longer than 6 months, were either HBeAg-positive or -negative, and had HBV-DNA levels above 17,182 IU/ml (100,000 copies/ml). All patients were treated with peginterferon alfa-2a 180 µg subcutaneously once a week, and adefovir dipivoxil 10 mg daily. After 48 weeks, treatment was discontinued and a treatment-free follow-up period started. The study was conducted according to the guidelines of the Declaration of Helsinki, with the principles of Good Clinical Practice and was approved by local ethics committees (controlled-trials.com; ISRCTN 77073364). All patients gave written informed consent.

Inclusion criteria for the present study were completion of the 48 weeks oftreatment and 48 weeks of follow-up (week 96), and availability of peripheral blood

samples for genotyping. Of the 92 patients treated in the initial study, 84 fulfilled these criteria (Supplementary data).

laboratory assaysRoutine biochemical and virological analyses were performed at the AMC in accordance with good laboratory practice. We determined quantitative HBsAg levels by the Architect (Abbott, IL, USA), and used the AxSYM (Abbott, IL, USA) as a qualitative test to define presence or absence of HBsAg. Both tests use a lower limit of detection of <0.05 IU/ml. Quantitation of plasma HBV-DNA levels was done by the Roche COBAS TaqMan 48 assay (F. Hoffmann-La Roche Ltd, Basel, Switzerland), with a dynamic range between 20 and 1.70 x108 IU/ml. Concentrations of plasma carnitine and carnitine derivatives were assessed retrospectively in plasma stored at -80 degrees Celsius, using a Quattro II triple-quadrupole mass spectrometer (Micromass, Manchester, UK) with 50 ll input, as described earlier.19

dna extraction and genotypingTotal DNA was extracted from peripheral blood mononuclear cells (PBMC) of patients by the Magna Pure LC (Roche Applied Science, Mannheim, Germany) and the DNA isolation kit I (Roche Diagnostics), according to the manufacturer’s recommendations. DNA samples were genotyped according to the Illumina-I- Infinium(r)-HD Assay Super manual protocol,

121


7

using the HumanOmni1-Quad BeadChip (v1.0, Illumina inc., San Diego, California, USA) at Hoffman-La Roche (Nutley, NJ, USA). In total 999,091 SNPs were genotyped in DNA samples from 84 patients and further analysed using the GenABEL package in R statistical software (v2.15.2). In total 743,673 (74.4%) SNPs and 84 (100%) patients passed all quality control criteria, and were used for further analysis (Supplementary data).

association analysisResponse was determined after 48 weeks of treatment free follow-up (week 96). Primary and secondary outcomes for both HBeAg-positive and HBeAg-negative patients were HBsAg loss (undetectable HBsAg levels by AxSYM; <0.05 IU/ml), and combined response (HBeAg negativity, HBV-DNA levels 62000 IU/ml, normalisation of alanine aminotransferase (ALT) levels), respectively. The genome-wide scan was performed with the ‘egscore’-function in GenABEL, which corrects for population stratification by principal components analysis (PCA) as implemented in the Eigenstrat algorithm.20 In our analysis, the first two principal components were used, which reflected well population ancestry (Supplementary data). Statistical significance was assessed with the 1-degree-of-freedom Cochran-Armitage trend test. Genomic control was applied to correct for any residual genomic inflation (pc values). Secondly, the same association analysis was performed using a model with sex, HBV genotype A, and HBV-DNA level as covariates. After Bonferroni correction for multiple testing, associations with a pc value of <6.7 x 10-8 were considered genome-wide significant. Odds ratios and confidence intervals were calculated using the effect allele as a reference. A linkage disequilibrium plot of rs12356193 was created using SNAP (BROAD Institute) based on phased genotype data from the International HapMap Project. Other statistical comparisons were performed using IBM SPSS Statistics (v19.0.0.1).

hBv-specific and non-specific t cell proliferation in presence of carnitineA detailed description of this assay is provided in the Supplementary data. In short, PBMC from seven CHB patients were isolated. To study the effect of carnitine on HBV-specific T cell proliferation, 0.5–1.0 x 106 cells were cultured for 10 days with the HBVcore18–27 peptide in the presence of 0, 10 or 20mM carnitine. After 10 days, cells were stained with APC-labeled-HBVcore18–27 tetrameric complexes. In addition, 0.5–1.0 x 106 CFSE labelled cells were cultured for five days with or without anti-CD3/anti-CD28 in the presence of 0, 10 or 20mM carnitine. All cells were analysed in a fluorescence-activated cell sorter (FACS Canto, BD Biosciences, San Jose, USA) with FlowJo-software MacV9.7.2 (TreeStar, Ashland, USA).

results patient characteristicsIn total, 84 patients were included in the study, and patient characteristics are summarised in Table 1. In total, nine (11%) patients achieved HBsAg loss at week 96, of which six (7%) had also developed anti-HBs. Of the other three patients, one had disappearance of

122

Chapter 7

anti-HBs after week 48, and two developed anti-HBs after week 96. Combined response (CR) at week 96 was attained by 25 patients (30%). The presence of HBV genotype A vs. non-A was the only significantly different baseline characteristic when comparing patients with HBsAg loss and patients without HBsAg loss (p = 0.02).

snp rs12356193 is genome-wide significantly associated with hBsag loss One SNP, rs12356193, was significantly associated with HBsAg loss (pc = 1.84 x 10-8) (Figure 1). The top 10 SNPs are summarised in Table 2, and all SNPs with pc <1.0 x 10-4 are listed in the Supplementary data. One other SNP, rs7094971 located upstream of rs12356193 in the SLC16A9 gene was in high linkage disequilibrium with rs12356193 and highly associated with HBsAg loss as well (pc = 1.27 x 10-6). No other SNPs genotyped are known to yield significant linkage to rs12356193 (Figure 1). After correcting for the calculated genomic inflation factor λ (1.09), SNPrs12356193 still reached genome-wide significance (pc value in Table 2). Incorporating sex with, either HBeAg status and HBV genotype A (pc = 5.93 x 10-8), baseline HBV-DNA (pc = 1.88 x 10-8), or baseline HBsAg (pc = 4.05 x 10-8) as covariates into the association analysis, resulted in similar pc values for rs12356193.

table 1. Baseline characteristics of all patients included in the association analysis.

CharacteristicshBsag loss (n = 9)

hBsag persistence(n = 75) p value

Demographics Mean age, years (SD) 44.9 (11.8) 38.9 (10.0) 0.10 1

Female, n (%) 1 20 0.44 3

Ethnicity 0.18 3

Caucasian, n (%) 5 (56) 22 (29) African, n (%) 3 (33) 23 (31) Asian, n (%) 1 (11) 30 (40) IFN naive, n (%) 5 (56) 57 (76) 0.19 3

Laboratory characteristics Median ALT, xULN (iqr) 1.5 (0.8-4.0) 2.0 (1.1-3.2) 0.31 2

HBeAg positive, n (%) 4 36 1.00 3

Mean HBV-DNA, log10 IU/ml (SD) 6.18 (2.22) 6.73 (1.67) 0.37 1

Mean HBsAg, log10 IU/ml (SD) 3.30 (1.30) 3.84 (0.79) 0.25 1

HBV genotype 0.13 3

A, n (%) 6 (67) 20 (27) 0.02 3

B, n (%) 0 (0) 14 (19) 0.35 3

C, n (%) 1 (11) 11 (15) 1.00 3

D, n (%) 1 (11) 22 (29) 0.43 3

E, n (%) 1 (11) 8 (11) 1.00 3

Liver biopsy characteristics, n (%) 8 (89) 60 (80)Inflammatory score (iqr) 6 (3.5-9.8) 5 (3-7.8) 0.35 2

Ishak fibrosis score (iqr) 2 (1.3-4.5) 1 (1-3) 0.19 2

1 p values are shown for Student’s t test.2 Mann-Whitney U test.3 Fisher’s exact or Chi-square test.

123


7

No SNP associated with CR at week 96 reached genome-wide significance

(Supplementary data).

ethnic distribution in rs12356193 genotypes and hBsag lossThe prevalence of the rs12356193 G allele was higher in Caucasians (0.22) than in Africans

(0.12) and Asians (0.02), which was comparable to those published by HapMap. The rate

of HBsAg loss in the three major ethnic groups within our study population appeared to be

associated with frequency of the G allele for rs12356193 (Supplementary data).

plasma carnitine levels are associated with snp rs12356193 and hBsag lossSNP rs12356193 is located on chromosome 10 in an intronic region of the solute carrier

family 16, member 9 (SLC16A9) gene, and was reported to be strongly associated with

plasma carnitine levels in other GWAS studies.18 This association was confirmed in

our study cohort: mean baseline plasma carnitine levels were significantly different in

figure 1. manhattan plot of association analysis for hBsag loss. p values of the Cochran-Armitage trend test after genomic control are shown as -log10 (p value). The threshold for genome-wide significance (pc <6.7 x 10-8) is depicted as a red striped line. The outlier circled in red corresponds to the significantly associated SNP rs12356193. A detailed plot of chromosome 10 is depicted in the lower figure, showing R-squared values of Linkage Disequilibrium (LD) for SNP rs12356193 with other SNPs genotyped (HapMap 1000 genomes database; based on reference populations of European ancestry).

124

Chapter 7

patients with the GG, AG, and AA genotype (p = 0.02, Figure 2A). In addition, carnitine levels were lower in patients with HBsAg loss than in patients with HBsAg persistence (p = 0.02, Figure 2A). This effect was independent of other predictors, such as viral genotype A and HBsAg at baseline and week 12 (Table 3). Receiver-operator characteristic analysis showed a comparable predictive value for baseline carnitine and HBsAg at week 12

table 2. Top 10 single nucleotide polymorphisms associated with HBsAg loss at week 96.

snp Chr.nearest gene

allele(1/2)

allele freq. (2)1

odds ratio2 pc-value3

hBsag loss

no hBsag loss

rs12356193 10 SLC16A9 A/G 0.56 0.06 19.6 (6.2-61.7) 1.84×10-8

rs11574 1 ID3 G/A 0.33 0.03 18.3 (4.5-73.7) 4.37×10-7

rs9926783 16 SHISA9 A/G 0.39 0.05 13.0 (3.9-43.8) 7.46×10-7

rs525413 6 SNORA18 G/A 0.50 0.07 12.6 (4.2-38.3) 9.51×10-7

rs169623804 16 SHISA9 G/A 0.50 0.06 15.7 (5.0-49.2) 1.09×10-6

rs7094971 10 SLC16A9 A/G 0.50 0.07 14.0 (4.5-43.1) 1.27×10-6

rs17241543 16 GNPATP G/A 0.67 0.17 9.5 (3.3-27.7) 1.44×10-6

rs122149484 6 EYS G/A 0.61 0.09 16.6 (5.5-50.0) 1.53×10-6

SNP6-667007494 6 EYS A/C 0.61 0.09 16.6 (5.5-50.0) 1.53×10-6

rs96544 19 SNAPC2 A/C 0.39 0.03 23.2 (5.9-91.7) 1.88×10-6

1 Allele frequency of effect allele (Allele 2).2 Unadjusted odds ratio and 95% confidence interval of effect allele (Allele 2).3 p value of Cochrane-Armitage’s trend test (1 d.f.) corrected for inflation factor λ (1.09).4 rs16962380, rs12214948, SNP6-66700749, and rs9654 violate Hardy-Weinberg equilibrium in the main study population (p < 0.05).

figure 2. association of carnitine with rs12356193 genotype and hBsag loss. (A) Baseline plasma carnitine levels according to rs12356193 genotype and HBsAg loss. (B) Baseline and week 42 (on-treatment) carnitine levels in 8 patients with HBsAg loss (genotype AG/GG) and 7 patients without HBsAg loss (genotype AA). Patients with HBsAg loss at week 96 are marked by open symbols. The lines represent mean and SEM. Reference values for carnitine are represented as dotted lines.

(a) (B)

125


7

(Supplementary data). Furthermore, there was an additive effect of carnitine to

rs12356193 genotype as patients with the favorable AG genotype and HBsAg loss had

significantly lower plasma carnitine levels than patients with the AG genotype and HBsAg

persistence (p = 0.05).

A similar association of lower baseline carnitine levels was found in patients with CR

compared to non-responders (33.9 vs. 37.8 µmol/L, respectively, p = 0.03). Values for all

carnitine derivatives are shown in the Supplementary data.

association of snp rs12356193 genotype and plasma carnitine levels with hBsag loss after longer follow-up durationDuring a longer (treatment-free) follow-up (week 144), 4 additional patients (1 HBeAg-

positive; 3 HBeAg-negative) became HBsAg undetectable. For this time point SNP

rs12356193 remained strongly associated with HBsAg loss, although not genome-wide

significant (pc = 1.14 x 10-5). In addition, the associations of baseline carnitine levels with

HBsAg loss at week 96 or week 144 were similar (Supplementary data).

factors influencing plasma carnitine levelsVarious factors are known to influence the level of plasma carnitine, e.g., dietary intake,

sex, age, and disease state.21,22 In our cohort, carnitine and acetylcarnitine level varied

mostly with sex and ethnicity, but not with age, creatinine, ALT, HBV-DNA or HBsAg level

(Supplementary data).

Most patients with HBsAg loss were non-Asian males (8/9). When analysed in

a non-Asian male subpopulation, mean plasma carnitine levels were significantly lower

in patients with HBsAg loss than those with HBsAg persistence (28.4 vs. 38.3 µmol/L,

respectively, p <0.001).

table 3. Multivariable analysis on the effect of plasma carnitine and other predictors in relation to HBsAg loss at week 96.

variablesodds ratio(95% CI) p value

odds ratio(95% CI) p value

odds ratio(95% CI) p value

Baseline carnitine 0.87

(0.77-0.98)

0.03 0.88

(0.78-0.98)

0.02 0.86

(0.75-0.99)

0.03

HBeAg positive 0.52

(0.10-2.65)

0.43 2.59

(0.44-15.2)

0.29 0.22

(0.03-1.75)

0.15

HBV genotype non-A 0.13

(0.02-0.71)

0.02 - - - -

Baseline HBsAg - - 0.36

(0.13-0.97)

0.04 - -

Week 12 HBsAg - - - - 0.19

(0.06-0.57)

<0.01

p values are shown for binary logistic regression including maximal 3 predictors.

126

Chapter 7

In addition, carnitine levels after 42 weeks of treatment were compared between eight non-Asian males with HBsAg loss (rs12356193 genotypes AG/GG) and seven sex and ethnicity matched patients without HBsAg loss (rs12356193 genotype AA). Carnitine levels increased in both groups compared to baseline (Figure 2B), but remained lower in patients with HBsAg loss compared to those without HBsAg loss (35.1 vs. 44.6 µmol/L, respectively, p = 0.01). A similar pattern was observed for acetyl- and propionylcarnitine levels (Supplementary data).

Carnitine limits proliferation of hBv-specific and non-specific Cd8 t cellsPreviously, in murine studies, carnitine was shown to inhibit T cell proliferation and functionality after non-specific stimulation in vitro.23,24 We explored the influence of carnitine on T cell responses in PBMC of seven CHB patients. The proportion of HBV tetramer-positive CD8 T cells in PBMCs after 10 days of HBV- peptide stimulation was significantly lower in the presence of 10 or 20 mM carnitine compared to untreated cells (Figure 3A). This effect was present in all CHB patients analysed (Figure 3B), which indicates that carnitine suppresses HBV-specific CD8 T cell responses. Furthermore, we observed less proliferation of total CD8 T cells during 5 days of stimulation with anti-CD3/anti-CD28 in the presence of 20 mM carnitine compared to untreated cells (Supplementary data). Figure 3C shows representative FACS plots of T cell responses in one CHB patient. There was no evidence for a toxic effect of carnitine on lymphocytes with the used concentrations (Figure 3D).

dIsCussIonIn this study we identified a SNP (rs12356193) in the SLC16A9 gene to be strongly associated with HBsAg loss in patients treated with pegIFN and adefovir. The functional relevance of this SNP was investigated, revealing lower carnitine levels in plasma of patients with the favourable rs12356193 genotype. Patients with HBsAg loss had lower carnitine levels, an effect which was independent of other confirmed predictors such as HBV genotype A. In addition, an inhibitory effect of carnitine was observed on the proliferative capacity of HBV-specific T cells.

Although several attempts have been made to find associations between candidate genes and IFN response in CHB therapy, no genome-wide screens have been published regarding this question to date. Although GWAS have the advantage of providing an unbiased screen of multiple genetic variations, it can be difficult to generate relevant biological hypotheses explaining the associations. Strikingly, a large GWAS on blood metabolite concentrations in healthy subjects identified a strong association of rs12356193 with plasma carnitine levels (p = 4.0 x 10-26).18 Another GWAS on metabolic traits confirmed this association, and provided evidence that SLC16A9 acts as a unidirectional carnitine efflux transporter.2525 Since SLC16A9 is predominantly expressed in kidney tissue it is possibly involved in the regulation of plasma carnitine levels by reabsorption of carnitine at the renal proximal epithelial cells. We are the first to confirm the association of rs12356193 with baseline plasma levels of carnitine in CHB patients, and observed

127


7

significantly lower plasma levels of carnitine in patients with treatment-induced

HBsAg loss.

Carnitine is an endogenous compound mainly known for its essential function in

fatty acid metabolism and cellular energy production.26 Carnitine is present in all body

tissues, and maintained by diet, synthesis in the liver, and tubular reabsorption.27 The liver,

together with the kidney, is a central organ for carnitine metabolism and distribution.

From the liver, carnitine can be distributed to other organs in form of free carnitine or as

a carnitine derivative. Therefore, it is not surprising that carnitine metabolism was found to

be altered in liver disease. However, findings regarding plasma carnitine levels in patients

figure 3. effect of carnitine on hBv-specific and non-specific Cd8 t cells. (A) Percentage of HBV tetramer-positive CD8 T cells among all CD8 T cells in PBMC from 7 CHB patients after 10 days of peptide stimulation in the presence of 0, 10, or 20 mM carnitine. Bars represent the mean values for 10 and 20mM carnitine relative to 0mM carnitine. (B) Individual plots of HBV tetramer-positive CD8 T cells for 0 and 20 mM carnitine. Each type of symbol represents one CHB patient. p values represent Wilcoxon signed-rank tests (*<0.05). (C) Representative FACS plot of one CHB patient. Numbers represent the percentage HBV tetramer- positive CD8 T cells (upper row), and the percentage CFSE-low CD8 T cells among all CD8 T cells (lower row). (D) Flow cytometric analysis of Live/Dead and Annexin-V staining in samples from (Figure 3A).

(a) (B)

(C) (d)

128

Chapter 7

with chronic hepatitis B or C have been somewhat contradictory, as different studies reported lower,28–30 unchanged,31 or even higher32 plasma carnitine levels compared to controls. Interestingly, two of these studies additionally reported lower carnitine levels in patients with a beneficial outcome to (peg)IFN based therapy.28,30

With the data presented here, a new model can be proposed in which a higher plasma carnitine level results in a less effective T cell response, which makes these patients less susceptible to the immune modulatory effect of pegIFN.

These findings may add to the knowledge on the complex virus-host interplay involved in CHB infection and pegIFN treatment. For example, interferon has been shown to have additional non-cytolytic antiviral effects, such as promotion of core particle decay and epigenetic cccDNA modifications.33 Furthermore, pegIFN and NUCs have been shown to have differential effects on the recovery of innate and adaptive immune responses.34

In this regard, part of the high rate of HBsAg loss found in our study cohort could be induced by the combination of pegIFN with adefovir, either by the direct anti-viral effect of adefovir or by an indirect effect caused by immune recovery with NUC therapy in general. While treatment with adefovir is described to lower plasma carnitine levels,35 IFN therapy was shown to increase plasma carnitine in HCV patients.30 We also observed an increase of carnitine levels during treatment. However, plasma carnitine levels remained lower in patients with HBsAg loss, suggesting that having lower carnitine levels throughout therapy is beneficial.

Support for this new model is provided in several lines of evidence suggesting that carnitine displays immune suppressive properties.23,24 Fortin et al. showed that carnitine treatment significantly inhibited both antigen presenting cells and CD4+ T cell function in carnitine-supplemented and deficient cell culture systems.23 Furthermore, they demonstrated that carnitine deficiency resulted in T cell hyperactivation in mice, which was reversed by carnitine supplementation. Next to this non-specific immune suppressive effect, we are the first to present data on the specific inhibitory effect of carnitine on human PBMC and HBV-specific CD8+ T cell proliferation.

Importantly, a direct relation between this inhibitory effect on HBV-specific CD8+ T cells and subsequent pegIFN induced HBsAg loss remains speculative. However, the role of carnitine in relation to other immunosuppressive mechanisms favouring viral persistence in CHB infection, such as the expression of inhibitory receptors on the surface of HBV-specific CD8+ T cells, the action of suppressive cytokines on NK cells and T cells, or the action of regulatory T cells,34 has to be elucidated in further studies.

Despite the relatively high rate of HBsAg loss found in our study, our cohort was small with respect to recently published genetic studies. An important tool to correct for possible confounding due to population stratification, which we applied, is the EIGENSTRAT algorithm incorporated in GenABEL.20 To further decrease the risk of false positive findings, Bonferroni adjustment and genomic control was applied to all p values. The Bonferroni adjustment is known to be a conservative approach because it does not account for dependency between single marker tests. However, with the inconsistent results of previous genetic association studies, we feel that this conservative approach

129


7

is appropriate. Furthermore, the association of SNP rs12356193 and carnitine levels with HBsAg loss appeared independent of other known predictors, such as HBV genotype A and baseline HBsAg, since introduction of these variables as covariates resulted in similar p values.

An interesting observation was the low allele frequency of SNP rs12356193 and high carnitine levels in people from Asian origin. This may make our findings of less practical use in the Asian population. On the other hand, our data suggest that rs12356193 genotype and higher carnitine levels could at least partly explain the low rate of HBsAg loss in the Asian patients in our cohort.

In addition to the effect of the rs12356193 genotype and carnitine levels, other yet undefined host genetic factors may play a role in achieving HBsAg loss in HBeAg-positive and HBeAg-negative patients. In our study, no other SNP reached the threshold of genome-wide significance, which might be attributed to the limited power of our study due to limited patient numbers.

Ideally, our results have to be confirmed in independent populations. Unfortunately, HBsAg loss in the treatment of CHB remains a rare event, and pegIFN in combination with adefovir or tenofovir has not been commonly used. Although a large randomised study comparing pegIFN and tenofovir in combination vs. monotherapies has been initiated (clinicaltrials.gov: NCT01277601), long-term off therapy results, as presented in our study, are not expected for several years. For this reason, further exploration of a possible biological consequence of genetic associations, as presented here, provides a more practical option.

In summary, the finding of both genetic and functional consequences of SLC16A9 gene variation on carnitine levels and HBsAg loss, in addition to the immune-suppressive potential of carnitine, suggests a possible role for these factors in pegIFN and adefovir induced viral clearance. First, these associations open novel avenues to broaden our understanding of the host immune response to IFN in CHB. Second, identifying those at greatest chance of HBsAg loss through SNP genotyping or carnitine measurement may assist in making the decision of whether to start with pegIFN based combination treatment, or to prefer NUC therapy only. Moreover, we could speculate that creating a therapeutic condition with low plasma carnitine could improve the immune response in pegIFN treated patients. Further research is needed to fully elucidate the role of the SLC16A9 gene and carnitine in hepatitis B infection.

130

Chapter 7

referenCes1. World Health Organization. Hepatitis B.

World Health Organization Fact Sheet 204 (Revised July 2012). World Health Organization 2012 ed 2012.

2. Fattovich G, Bortolotti F, Donato F. Natural history of chronic hepatitis B: special emphasis on disease progression and prognostic factors. J Hepatol 2008;48:335-352.

3. European Association For The Study Of The Liver. EASL Clinical Practice Guidelines: Management of chronic hepatitis B virus infection. J Hepatol 2012;57:167-185.

4. Lok AS, McMahon BJ. Chronic hepatitis B: update 2009. Hepatology 2009;50:661-662.

5. Lin SM, Yu ML, Lee CM, et al. Interferon therapy in HBeAg positive chronic hepatitis reduces progresssion to cirrhosis and hepato-cellular carcinoma. J Hepatol 2007;46:45-52.

6. Janssen HL, van Zonneveld M, Senturk H, et al. Pegylated interferon alfa-2b alone or in combination with lamivudine for HBeAg-positive chronic hepatitis B: a randomised trial. Lancet 2005;365:123-129.

7. Lau GK, Piratvisuth T, Luo KX, et al. Peginterferon Alfa-2a, lamivudine, and the combination for HBeAg-positive chronic hepatitis B. N Engl J Med 2005;352:2682-2695.

8. Marcellin P, Lau GK, Bonino F, et al. Peginterferon alfa-2a alone, lamivudine alone, and the two in combination in patients with HBeAg-negative chronic hepatitis B. N Engl J Med 2004;351:1206-1217.

9. Buster EH, Flink HJ, Cakaloglu Y, et al. Sustained HBeAg and HBsAg loss after long-term follow-up of HBeAg-positive patients treated with peginterferon alpha-2b. Gastroenterology 2008;135:459-467.

10. Manesis EK, Hadziyannis ES, Angelopoulou OP, et al. Prediction of treatment-related HBsAg loss in HBeAG-negative chronic

hepatitis B: a clue from serum HBsAg levels. Antivir Ther 2007;12:73-82.

11. Takkenberg RB, Jansen L, de Niet A, et al. Baseline hepatitis B surface antigen (HBsAg) as predictor of sustained HBsAg loss in chronic hepatitis B patients treated with peginterferon alfa-2a and adefovir. Antivir Ther 2013.

12. Prokunina-Olsson L, Muchmore B, Tang W, et al. A variant upstream of IFNL3 (IL28B) creating a new interferon gene IFNL4 is associated with impaired clearance of hepatitis C virus. Nat Genet 2013;45:164-171.

13. de Niet A, Takkenberg RB, Benayed R, et al. Genetic variation in IL28B and treatment outcome in HBeAg-positive and -negative chronic hepatitis B patients treated with Peg interferon alfa-2a and adefovir. Scand J Gastroenterol 2012;47:475-481.

14. Holmes JA, Nguyen T, Ratnam D, et al. IL28B genotype is not useful for predicting treatment outcome in asian chronic hepatitis B patients treated with peg-ylated-interferon-alpha. J Gastroenterol Hepatol 2013.

15. Lampertico P, Vigano M, Cheroni C, et al. IL28B polymorphisms predict interferon-related hepatitis B surface antigen seroclearance in genotype D hepatitis B e antigen-negative patients with chronic hepatitis B. Hepatology 2012.

16. Sonneveld MJ, Wong VW, Woltman AM, et al. Polymorphisms near IL28B and serologic response to peginterferon in HBeAg-positive patients with chronic hepatitis B. Gastroenterology 2012;142:513-520.

17. Kamatani Y, Wattanapokayakit S, Ochi H, 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.

18. Kolz M, Johnson T, Sanna S, et al. Meta-analysis of 28,141 individuals identifies common variants within five new loci that

131


7

influence uric acid concentrations. PLoS Genet 2009;5:e1000504.

19. Vreken P, van Lint AE, Bootsma AH, et al. Rapid diagnosis of organic acidemias and fatty-acid oxidation defects by quantitative electrospray tandem-MS acyl-carnitine analysis in plasma. Adv Exp Med Biol 1999;466:327-337.

20. Price AL, Patterson NJ, Plenge RM, et al. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet 2006;38:904-909.

21. Cederblad G. Plasma carnitine and body composition. Clin Chim Acta 1976;67:207-212.

22. Flanagan JL, Simmons PA, Vehige J, et al. Role of carnitine in disease. Nutr Metab (Lond) 2010;7:30.

23. Fortin G, Yurchenko K, Collette C, et al. L-carnitine, a diet component and organic cation transporter OCTN ligand, displays immunosuppressive properties and abrogates intestinal inflammation. Clin Exp Immunol 2009 ;156:161-171.

24. Athanassakis I, Mouratidou M, Sakka P, et al. L-carnitine modifies the humoral immune response in mice after in vitro or in vivo treatment. Int Immunopharmacol 2001; 1:1813-1822.

25. Suhre K, Shin SY, Petersen AK, et al. Human metabolic individuality in biomedical and pharmaceutical research. Nature 2011 ;477:54-60.

26. Bremer J. Carnitine--metabolism and functions. Physiol Rev 1983;63:1420-1480.

27. Engel AG, Rebouche CJ. Carnitine metabolism and inborn errors. J Inherit Metab Dis 1984;7 Suppl 1:38-43.

28. Selimoglu MA, Yagci RV. Plasma and liver carnitine levels of children with chronic hepatitis B. J Clin Gastroenterol 2004 ;38:130-133.

29. Zhou L, Wang Q, Yin P, et al. Serum metabolomics reveals the deregulation of fatty acids metabolism in hepatocellular carcinoma and chronic liver diseases. Anal Bioanal Chem 2012;403:203-213.

30. Malaguarnera M, Restuccia S, Di Fazio I, et al. Serum carnitine levels in chronic hepatitis C patients before and after lymphoblastoid interferon-alpha treatment. BioDrugs 1999;12:65-69.

31. Krahenbuhl S, Reichen J. Carnitine metabolism in patients with chronic liver disease. Hepatology 1997;25:148-153.

32. Eskandari GH, Kandemir O, Polat G, et al. Serum L-carnitine levels and lipoprotein compositions in chronic viral hepatitis patients. Clin Biochem 2001; 34: 431-433.

33. Wieland SF, Eustaquio A, Whitten-Bauer C, et al. Interferon prevents formation of replication-competent hepatitis B virus RNA-containing nucleocapsids. Proc Natl Acad Sci U S A 2005;102:9913-9917.

34. Thimme R, Dandri M. Dissecting the divergent effects of interferon-alpha on immune cells: time to rethink combination therapy in chronic hepatitis B? J Hepatol 2013;58:205-209.

35. Barditch-Crovo P, Toole J, Hendrix CW, et al. Anti-human immunodeficiency virus (HIV) activity, safety, and pharmacokinetics of adefovir dipivoxil (9-[2-(bis-pival-oyloxy-methyl)-phosphonylmethoxyethyl]adenine) in HIV-infected patients. J Infect Dis 1997; 176:406-413.

132

Chapter 7

fundInG This study was funded by Roche.

aCKnowledGementsWe thank Nico Abeling, Frédéric Vaz, and Ronald Wanders for the interpretation of carnitine measurements. We acknowledge Vincent Rijckborst and Harry Janssen for the inclusion of study participants.

supplementary dataSupplementary data associated with this article available can be found at http://dx.doi.org/10.1016/j.jhep.2014.05.004.

peG-Interferon plus nuCleotIde analoGue treatment versus no treatment In patIents

wIth ChronIC hepatItIs B wIth a low vIral load: a randomIsed Controlled, open-laBel trIal

A. de Niet*, L. Jansen*, F. Stelma*, S.B. Willemse, S.D. Kuiken, S. Weijer, C.M.J. van Nieuwkerk, H.L. Zaaijer, R. Molenkamp, R.B. Takkenberg, M. Koot,

J. Verheij, U. Beuers, H.W. Reesink

The Lancet Gastroenterology and Hepatology 2017;2:576-584


aBstraCtBackground Antiviral treatment is currently not recommended for patients with chronic hepatitis B with a low viral load. However, they might benefit from acquiring a functional cure (hepatitis B surface antigen [HBsAg] loss with or without formation of antibodies against hepatitis B surface antigen [anti-HBs]). We assessed HBsAg loss during peg-interferon- alfa-2a (peg-IFN) and nucleotide analogue combination therapy in patients with chronic hepatitis B with a low viral load.

methods In this randomised controlled, open-label trial, patients were enrolled from the Academic Medical Center (AMC), Amsterdam, Netherlands. Eligible patients were HBsAg positive and hepatitis B e antigen (HBeAg) negative for more than 6 months, could be treatment naive or treatment experienced, and had alanine aminotransferase (ALT) concentrations less than 5 × upper limit of normal (ULN). Participants were randomly assigned (1:1:1) by a computerised randomisation programme (ALEA Randomisation Service) to receive peg-IFN 180 µg/week plus adefovir 10 mg/day, peg-IFN 180 µg/week plus tenofovir disoproxil fumarate 245 mg/day, or no treatment for 48 weeks. The primary endpoint was the proportion of patients with serum HBsAg loss among those who received at least one dose of study drug or had at least one study visit (modified intention-to-treat population [mITT]). All patients have finished the initial study of 72 weeks and will be observed for up to 5 years of follow-up. This study is registered with ClinicalTrials.gov, number NCT00973219.

findings Between Aug 4, 2009, and Oct 17, 2013, 167 patients were screened for enrolment, of whom 151 were randomly assigned (52 to peg-IFN plus adefovir, 51 to peg-IFN plus tenofovir, and 48 to no treatment). 46 participants in the peg- IFN plus adefovir group, 45 in the peg-IFN plus tenofovir group, and 43 in the no treatment group began treatment or observation and were included in the mITT population. At week 72, two (4%) patients in the peg-IFN plus adefovir group and two (4%) patients in the peg-IFN plus tenofovir group had achieved HBsAg loss, compared with none of the patients in the no treatment group (p=0·377). The most frequent adverse events (>30%) were fatigue, headache, fever, and myalgia, which were attributed to peg-IFN dosing. Two (4%) serious adverse events were reported in the peg-IFN plus adefovir group (admission to hospital for alcohol-related pancreatitis [week 6; n=1] and pregnancy, which was electively aborted [week 9; n=1]), three (7%) in the peg-IFN plus tenofovir group (admission to hospital after a suicide attempt during a severe depression [week 23; n=1], admission to hospital for abdominal pain [week 2; n=1], and an elective laminectomy [week 40; n=1]), and three (7%) in the no treatment group (admission to hospital for septic arthritis [week 72; n=1], endocarditis [week 5; n=1], and hyperthyroidism [week 20; n=1]).

Interpretation In patients with chronic hepatitis B with a low viral load, combination treatment (peg-IFN plus adefovir and peg-IFN plus tenofovir) did not result in significant HBsAg loss compared with no treatment, which does not support the use of combination treatment in this population of patients.

138

Chapter 8

IntroduCtIonDespite the availability of a safe and effective vaccine for more than three decades, infection with hepatitis B virus is still a major public health problem, with an estimated 240 million chronically infected individuals worldwide.1

The goal of antiviral treatment in patients with chronic hepatitis B is to prevent progression to cirrhosis, hepatic decompensation, and hepatocellular carcinoma by reducing viral replication.2,3 Treatment is indicated in patients with a high viral load, progressive liver inflammation, and fibrosis. For patients with chronic hepatitis B with a low viral load and no signs of necroinflammatory activity or fibrosis, there is currently no indication for treatment, because viral replication levels are already low, and the progression of liver disease is slower.2,4,5 However, patients with chronic hepatitis B with a low viral load still have a substantial risk for the development of cirrhosis and hepatocellular carcinoma.4–11 As shown by the REVEAL cohorts,5,7 the risks for development of cirrhosis, hepatocellular carcinoma, or both, are highest in patients with the highest viral load, but are already increased at lower hepatitis B virus DNA levels. The cumulative incidence of hepatocellular carcinoma in hepatitis B surface antigen (HBsAg)-positive patients who are hepatitis B virus DNA negative and patients with chronic hepatitis B who have a viral load less than 20 000 IU/mL ranges from 1·3% to 3·57% at year 13, while the cumulative incidence of cirrhosis ranges from 4·5% to 9·8%.5,7 Furthermore, due to precore mutants, patients with chronic hepatitis B with a low viral load can progress to hepatitis B e antigen (HBeAg)-negative chronic hepatitis, which could lead to fulminant hepatitis or cirrhosis.12,13 The risk of hepatitis B virus reactivation in patients who receive immunosuppression is higher in patients who are HBsAg positive compared with patients who are HBsAg negative.14 The most favourable outcome in the treatment of patients with chronic hepatitis B is functional cure, defined as the clearance of HBsAg with or without formation of antibodies against hepatitis B surface antigen (anti-HBs), which is associated with improved outcome.15–18 Unfortunately, with a finite course of pegylated interferon alfa (peg-IFN) or long-term viral suppression with nucleos(t)ide analogues, this functional cure is rarely achieved.2

Previous attempts to improve response rates by combining peg-IFN and nucleos(t)ide analogue therapy have had varying results.19–21 Nevertheless, combining peg-IFN with more potent nucleos(t)ide analogues remains of interest because of the dual effect on both the innate and adaptive immune responses.22–24 In our previous study, in which patients with active chronic hepatitis B with a high viral load were treated with peg-IFN and adefovir, a relatively high proportion of HBsAg loss (17%) was noted in HBeAg-negative patients 2 years after therapy.25 A recent large randomised controlled trial24 in patients with active chronic hepatitis B showed a higher proportion of HBsAg loss in patients treated with peg-IFN and tenofovir than in those receiving either monotherapy.

Most (70–80%) patients with a chronic hepatitis B infection worldwide have a low viral load. By contrast with patients with a high viral load, patients with a low viral load have low HBsAg levels.26 Previous results in patients with a high viral load suggest that

139

COMBINATION TREATMENT FOR LOW VIRAL LOAD CHB

8

low baseline HBsAg levels are associated with a response to combination treatment.25,27 Furthermore, in patients who have a low viral load as a result of long-term nucleos(t)ide analogue suppression, low baseline HBsAg titers were associated with functional cure in response to peg-IFN add-on therapy.28 Hypothetically, patients with chronic hepatitis B with a low viral load might be more susceptible to hepatitis B virus treatment because their antiviral immune response is already capable of keeping hepatitis B virus DNA and HBsAg at low levels. This could be related to the finding that T-cell inhibition is less severe in the presence of low hepatitis B virus DNA levels.29,30

No randomised controlled studies have yet been done on the efficacy of peg-IFN or nucleos(t)ide analogues in patients with chronic hepatitis B with a low viral load. Therefore, we aimed to compare combination therapy of peg-IFN plus adefovir or tenofovir versus no treatment in HBeAg-negative patients with chronic hepatitis B with a low viral load, and to investigate the rate of HBsAg loss and markers of response to therapy.

Although adefovir is no longer indicated as monotherapy for viral suppression in patients with a treatment indication, based on our previous results with high rates of functional cure, a treatment group with peg-IFN and adefovir was included in the study. Because tenofovir (which is a stronger suppressor of hepatitis B virus DNA than adefovir, but with unknown immune modulatory effect31) had become available since our last study, we also included peg-IFN and tenofovir combination therapy. The effect of both combination therapies was compared to an untreated control group which is the current standard of practice for chronic hepatitis B patients with a low viral load.

methods study design and participants This investigator-initiated, prospective, open-label, randomised controlled trial was done at the Academic Medical Center (AMC), Amsterdam, Netherlands.

Patients with chronic hepatitis B aged 18–70 years with a low viral load were enrolled after assessment of eligibility. Hepatitis B virus DNA below a threshold of 20 000 IU/mL was chosen as low viral load according to local32 and international2,3 guidelines available at the time of study initiation. Inclusion criteria were documented HBsAg positivity and HBeAg negativity for more than 6 months. Exclusion criteria were concurrent infection with hepatitis C virus, hepatitis delta virus, or HIV; decompensated liver disease, hepatocellular carcinoma, or a history of bleeding from oesophageal varices; pregnancy or breastfeeding; and alanine aminotransferase (ALT) levels greater than five times the upper limit of normal (5 × ULN). Patients were either treatment naive, or had received peg-IFN or nucleos(t) ide analogues more than 6 months before inclusion. The full eligibility criteria and screening assessments are provided in the appendix (p 1–3).

The study complied with the Declaration of Helsinki and the principles of Good Clinical Practice and was approved by a legally instituted ethical committee. All patients gave written informed consent.

140

Chapter 8

randomisation and masking Patients were randomly assigned (1:1:1) to receive pegylated interferon alfa-2a (Pegasys; F Hoffman-La Roche, Basel, Switzerland) 180 µg/week, subcutaneously in combination with adefovir dipivoxil (Hepsera; Gilead Sciences, Foster City, CA, USA) 10 mg orally once daily for 48 weeks; peginterferon alfa plus tenofovir disoproxil fumarate (Viread; Gilead Sciences) 245 mg orally once daily for 48 weeks; or no treatment (current standard of care) for 48 weeks. Patients were enrolled by the subinvestigators and randomly assigned by a computerised randomisation programme (ALEA Randomisation Service). Randomisation was stratified according to hepatitis B virus genotype A, non-A (genotype B–G), or indeterminable genotype. Patients and investigators were not masked. After 48 weeks, treatment was discontinued and all patients were followed up until week 72.

procedures Routine examinations and laboratory tests were done with 2 weekly intervals for the first 2 months and 6 weekly intervals thereafter (appendix pp 4–7). The viral load was assessed before selecting or approaching eligible patients, if available from the medical chart. Furthermore, viral load was assessed at the time of screening and at day 1 of treatment, of which the latter was used as baseline hepatitis B virus DNA level. Due to subsequent changes in expert opinion on the classification of low viral load, a subgroup analysis on patients with hepatitis B virus DNA level less than 2000 IU/mL (often referred to as inactive carriers) was done.

Plasma hepatitis B virus DNA level was determined by the COBAS TaqMan assay (F Hoffmann-La Roche, Basel, Switzerland). Serum HBsAg level was quantified by using the Abbott HBsAg Architect assay (Abbott Diagnostics, Abbott Park, IL, USA). Qualitative detection of serum HBsAg, antibody to HBsAg (anti-HBs), HBeAg, and antibody to HBeAg (anti-HBe) was done by an enzyme immunoassay (AxSYM; Abbott Laboratories, Abbott Park, IL, USA). ALT levels were expressed as absolute values (U/L) or relative to the ULN range. ALT reference values were 45 U/L for men and 34 U/L for women. For histological assessment of liver biopsies the modified Ishak scoring system was applied, based on a 0–18 score for necroinflammation and a 0–6 score for fibrosis.33

outcomesThe primary objective was the proportion of patients with HBsAg loss. HBsAg loss was defined as undetectable serum HBsAg by AxSYM (<0·05 IU/mL). The secondary objectives were the proportion of patients with HBsAg loss who also had anti-HBs seroconversion (defined as anti-HBsAg >10 IU/L), and the identification of predictive markers of treatment response. Post-hoc analyses assessed HBsAg loss at week 48 and HBsAg decline at weeks 48 and 72. Patients were considered to be non-responders when not meeting the criteria for HBsAg loss or HBsAg decline. Adverse events were assessed at each study visit through anamnesis and laboratory testing.

141


8

statistical analysisThe sample size calculation was based on the primary endpoint (HBsAg loss at week 72).

The assumed response rates were 20% for patients treated in each combination group

versus 1% for patients in the control group (receiving no treatment).25 A group sample

of 44 patients in the control group was determined to detect a difference with either of

the treatment arms at the α level of 0·05 (two-sided Fisher’s exact test) with a power of

81%. Assuming a 10% drop-out rate, 50 patients or more were needed in each group.

In the primary analysis, differences in the proportion of patients with HBsAg loss in

each group were compared with a Pearson χ² test. In this analysis, a modified intention-

to-treat (mITT) model was applied, including all patients who received at least one dose

of study drugs (treatment groups) or had at least one study visit (no treatment group).

Patients who prematurely discontinued treatment were scored as non-responders. Patients

with missing data were considered non-responders. In the secondary analysis on HBsAg

decline, only patients who completed 48 weeks of treatment and 24 weeks of treatment-

free follow-up were included (per-protocol model). Use of the per-protocol population

for the secondary analysis allowed avoidance of a high proportion of missing data from

patients who discontinued.

Baseline and on-treatment variables were compared between study groups using

Welch’s t, Students t, Mann- Whitney U, Kruskall-Wallis, Pearson’s χ², or Fisher’s exact

tests. The associations between variables as potential predictors of HBsAg loss or HBsAg

decline were examined by multivariable logistic regression analysis. Data were analysed

with IBM SPSS Statistics, version 21. All p values are two sided and values below 0·05 were

considered statistically significant.

This study is registered with ClinicalTrials.gov, number NCT00973219.

role of the funding sourceThe funder of the study had no role in study design, data collection, data analysis, data

interpretation, or writing of the report. The corresponding author had full access to all

the data in the study and had final responsibility for the decision to submit for publication.

resultsBetween Aug 4, 2009, and Oct 17, 2013, 167 patients were screened for participation

in the study, of whom 151 patients were randomly assigned (nine did not meet inclusion

criteria and seven withdrew before randomisation).

52 participants were assigned to peg-IFN plus adefovir, 51 were assigned to peg-IFN

plus tenofovir, and 48 were assigned to no treatment (figure 1). 17 patients did not start

allocated intervention (reasons stated in figure 1). Between Sept 7, 2009, and Dec 12,

2013, 134 patients started intervention or no treatment and were included in the mITT

population: 46 of 52 in the peg-IFN plus adefovir group, 45 of 51 in the peg-IFN plus

tenofovir group, and 43 of 48 in the no treatment group. Of 134 patients, 12 prematurely

discontinued the study: five in the peg-IFN plus adefovir group, six in the peg-IFN plus

142

Chapter 8

tenofovir group, and one in the no treatment group (figure 1). Reasons for treatment

discontinuation in the peg-IFN plus adefovir group were alcohol-related pancreatitis

(n=1; at 6 weeks), dizziness (n=1; at 8 weeks), hair loss (n=1; at 24 weeks), back pain

(n=1; at 38 weeks), and laboratory abnormalities with concomitant alcohol abuse (n=1;

at 42 weeks). Treatment in the peg-IFN plus tenofovir group was discontinued because

of general interferon-related side-effects (n=2; at 1 and 8 weeks), nausea and vomiting

(n=1; at 12 weeks), depression (n=2; at 12 and 23 weeks), and hair loss (n=1; at 24 weeks).

One patient in the no treatment group was lost to follow-up after 6 weeks. Thus, the per-

protocol population consisted of 41 patients in the peg-IFN plus adefovir group, 39 in

the peg-IFN plus tenofovir group, and 42 in the no treatment group.

Patient characteristics are shown in table 1. Patients were predominantly men (75 [56%]

of 134), with a mean age of 43 years (SD 11). Ethnic background was mixed, as were

Figure 1. Consort flow diagram

* Patients who did not receive allocated intervention: Arm I: lost to follow-up (n=3), patient

withdrew consent (n=2), worsening of psychiatric symptoms (n=1). Arm II: lost to follow-up

(n=2), patient withdrew consent (n=1), worsening of psychiatric symptoms (n=1), initiation of

lamivudine (n=1), no health insurance (n=1). Arm III: lost to follow-up (n=2), moved abroad

(n=2), patient withdrew consent (n=1).

** Reasons for treatment discontinuation: are specified in the results section.

Discontinued intervention

(n=5)**

Discontinued intervention

(n=6)**

Discontinued (n=1)**

Received intervention:

Intention-to-treat (ITT)

population (n=46)



population (n=45)



population (n=43)

Completed intervention:

Per-protocol (PP)

population (n=41)

Completed intervention:

Per-protocol (PP)

population (n=39)

Completed:

Per-protocol (PP)

population (n=42)

Did not receive allocated

intervention (n=6)*


intervention (n=6)*


intervention (n=5)*

Screened for study (n=167)

Randomized (n=151)

Arm I: Peg-IFN +

adefovir (n=52)

Arm II: Peg-IFN +

tenofovir (n=51)

Arm III: No treatment

(n=48)

Excluded (n=16)

• Did not meet inclusion criteria (n=9)

• Withdrawn before randomization (n=7)

figure 1. Trial profile *Patients who did not receive allocated intervention; peg-IFN plus adefovir: lost to follow-up (n=3), patient withdrew consent (n=2), worsening of psychiatric symptoms (n=1); peg-IFN plus tenofovir: lost to follow-up (n=2), patient withdrew consent (n=1), worsening of psychiatric symptoms (n=1), initiation of lamivudine (n=1), no health insurance (n=1); no treatment: lost to follow-up (n=2), moved abroad (n=2) and patient withdrew consent (n=1). † Reasons for treatment discontinuation are specified in the Results section.

143


8

hepatitis B virus genotypes. Most patients were interferon naive, and none of the patients had received a nucleos(t)ide analogue within 6 months of entering the trial. All patients were HBeAg negative, with a low viral load (mean hepatitis B virus DNA 2·74 log10 IU/mL [SD 1·10]), normal ALT levels (median 27 U/L [IQR 21–37]), and minimal liver fibrosis (mean Fibroscan value 5·4 kPa [SD 1·9]). The modified hepatic activity index (HAI) was low (median 2 [IQR 2–3]) and none of the patients had evidence of liver cirrhosis; the maximum

table 1. Baseline characteristics of the modified intention-to-treat population (n=134)

Characteristics

treatment arms

arm I; peg-Ifn + adefovir n=46

arm II; peg-Ifn + tenofovir n=45

arm III; no treatment n=43

Women, n (%) 18 (39) 24 (53) 17 (40)Men, n (%) 28 (61) 21 (47) 26 (60)Age, years, mean (sd) 44 (12) 43 (12) 41 (10)ethnicity Caucasian, n (%) 11 (24) 14 (31) 17 (40) Asian, n (%) 7 (15) 11 (24) 7 (16) African, n (%) 20 (43) 11 (24) 14 (33) South American, n (%) 8 (17) 9 (20) 5 (12)IFN naïve, n (%) 43 (93) 42 (93) 43 (100)ALT, U/L, median (iqr) 27 (21-42) 25 (19-30) 30 (21-47)hBv Genotype Indeterminable, n (%) 9 (20) 11 (24) 8 (19) A, n (%) 11 (24) 10 (22) 8 (19) B, n (%) 4 (9) 3 (7) 2 (5) C, n (%) 2 (4) 1 (2) 3 (7) D, n (%) 10 (22) 13 (29) 12 (28) E, n (%) 10 (22) 7 (16) 9 (21) F, n (%) 0 (0) 0 (0) 0 (0) G, n (%) 0 (0) 0 (0) 1 (2) HBV-DNA, log10 IU/mL, mean (sd) 2·65 (1·23) 2.79 (1.03) 2.79 (1.04) HBV-DNA <2,000 IU/mL, n (%) 33 (72) 28 (62) 30 (70) HBsAg, log10 IU/mL, mean (sd) 3·21 (0.98) 3.31 (0.76) 3.06 (0.88) Fibroscans performed, n (%) 41 (89) 36 (80) 40 (93) Vvalue (kPa) mean (sd) 5.0 (1.8) 5.4 (1.8) 5.8 (2.0)liver biopsy Liver biopsies performed, n (%) 40 (87) 39 (87) 19 (44) Mean biopsy length, mm , median (iqr) 16 (12-22) 20 (15-26) 13 (11-15) Mean portal fields, n, median (iqr) 11 (7-14) 15 (8-18) 9 (7-14) Median inflammatory score, median (iqr) 2 (2-3) 2 (1-3) 2 (2-2) Median Ishak fibrosis score, median (iqr) 1 (1-1) 1 (1-1) 1 (1-1) Median steatosis, grade, median (iqr) 0 (0-1) 0 (0-1) 0 (0-1) Median % HBsAg staining, median (iqr) 10 (1-35) 25 (5-50) 10 (1-25)

Data are n (%), median (IQR), or mean (SD). Peg-IFN=peg-interferon-alfa-2a. ALT=alanine aminotransferase. HBV=hepatitis B virus. HBsAg=hepatitis B surface antigen.

144

Chapter 8

Ishak’s fibrosis score was 3 (or a maximum Fibroscan value of 8·1 kPa in patients with no liver biopsy available).

19 patients required peg-IFN dose reductions (11 in the peg-IFN plus adefovir group and eight in the peg-IFN plus tenofovir group).

At week 72, two (4%) patients in the peg-IFN plus adefovir group, two (4%) patients in the peg-IFN plus teno fovir group, and no patients in the no treatment group had HBsAg loss (p=0·377). Of the four patients who were HBsAg negative at week 72, one patient was not peg-IFN naive. Patients with HBsAg loss had hepatitis B virus genotype A (one patient in the peg-IFN plus adefovir group), genotype B (two patients in the peg- IFN plus tenofovir group), or indeterminable (one patient in the peg-IFN plus adefovir group). Three of four patients had anti-HBs higher than 10 IU/L (n=1 from peg-IFN plus adefovir group and n=2 from peg-IFN plus tenofovir group). HBsAg response rates at week 48 and week 72 are shown in table 2. At week 48, one patient (2%) in the peg-IFN plus adefovir group, three patients (7%) in the peg-IFN plus tenofovir group, and no patients in the no treatment group had HBsAg loss (p=0·171). Of four patients with HBsAg loss at week 48, three had anti-HBs higher than 10 IU/L. During follow-up, the patient without anti-HBs conversion (in the peg-IFN plus tenofovir group, indeterminable genotype) seroreverted to HBsAg positivity to levels around the detection limit at week 72. The course of virological and biochemical parameters per patient who achieved HBsAg loss during treatment, or follow-up, or both (n=5), is shown in the appendix (p 8).

table 2. HBsAg response rates

peg-Ifn+adv peg-Ifn+tdf untreated p value

Intention to treat (n=134) Number of patients 46 45 43 HBsAg loss Week48 1 (2%) 3 (7%) 0 0·171 a HBsAg seroconversion Week48 1 (2%) 2 (4%) 0 0·370 a

HBsAg loss Week72 2 (4%) 2 (4%) 0 0·377 a

HBsAg seroconversion Week72 1 (2%) 2 (4%) 0 0·370 a

per protocol (n=122) Number of patients 41 39 42 Week 48 HBsAg <10 IU/mL 4 (10%) 6 (15%) 1 (2%) 0·122 a

Week 48 >0·5 log10 IU/mL decline 15 (37%) 10 (26%) 0 0·0001 a

Week48 >1 log10 IU/mL decline 9 (22%) 8 (21%) 0 0·006 a

Mean HBsAg decline Week48, Log10 IU/mL (SD)

-0·61 (0·92) -0·61 (0·96) -0·06 (0·18) <0·0001 b

Week 72 HBsAg < 10 IU/mL 5 (12%) 5 (13%) 2 (5%) 0·393 a

Week 72 >0·5 log10 IU/mL decline 13 (32%) 14 (36%) 3 (7%) 0·005 a

Week72 >1 log10 IU/mL decline 5 (12%) 6 (15%) 0 0·037 a

Mean HBsAg decline Week72, Log10 IU/mL (SD)

-0·53 (0·77) -0·59 (0·85) -0·15 (0·22) 0·001 b

Data are n (%) or mean (SD). Differences in the proportion of patients between groups were compared by Pearson χ² test* or Welch’s t test†. Peg-IFN=peg-interferon-alfa-2a. HBsAg=hepatitis B surface antigen.

145


8

In a post-hoc per-protocol analysis, mean HBsAg level had declined significantly in all study groups at week 48 compared with at baseline: mean –0·61 log10 IU/mL (SD 0·92) reduction for the peg-IFN plus adefovir group (p=0·0001), –0·61 log10 IU/mL (0·96) reduction for the peg-IFN plus tenofovir group (p=0·0003), and –0·06 log10 IU/mL (0·18) reduction for the no treatment group (p=0·042; table 3). No difference in HBsAg decline was noted between the two treatment groups (p=0·990). However, HBsAg declined more strongly in the peg-IFN plus adefovir group (p=0·0005) and the peg-IFN plus tenofovir group (p=0·001) than in the no treatment group (figure 2A, B). An HBsAg decline of more than 1·0 log10 IU/mL was noted in 17 (21%) treated patients (nine [22%] in the peg-IFN plus adefovir group and eight [21%] in the peg-IFN plus tenofovir group), but in none of the no treatment patients (p=0·001).

During follow-up, HBsAg levels remained significantly lower than pretreatment levels in all groups: mean reduction –0·53 log10 IU/mL (SD 0·77) at week 72 for the peg-IFN

figure 2. Post-hoc analysis of HBsAg and HBV DNA decline according to treatment group. Mean log10 IU/mL decline in HBsAg (A) and HBV DNA (B) compared with baseline according to treatment group. Mean log10 IU/mL decline in HBsAg (C) and HBV DNA (D) compared with baseline in patients with an on-treatment HBsAg decline of >1 log10 IU/mL. Data are for the per-protocol population. HBsAg=hepatitis B surface antigen, HBV=hepatitis B virus.

Chapter 8

Figure 2 | Post‐hoc analysis of HBsAg and HBV DNA decline according to treatment group. Mean log10 IU/mL decline in HBsAg (A) and HBV DNA (B) compared with baseline according to treatment group. Mean log10 IU/mL decline in HBsAg (C) and HBV DNA (D) compared with baseline in patients with an on‐treatment HBsAg decline of >1 log10 IU/mL. Data are for the per‐protocol population. HBsAg=hepatitis B surface antigen, HBV=hepatitis B virus.

(a) (B)

(C) (d)

146

Chapter 8

tab

le 3

. Mul

tivar

iab

le a

naly

sis

of p

red

icto

rs fo

r o

n-tr

eatm

ent

HB

sAg

dec

line

uni

vari

able

ana

lysi

s

mul

tiva

riab

le a

naly

sis

mo

del

1m

od

el 2

Bse

pB

sep

Bse

p

Fem

ale

sex

-1·2

7(0

·63)

0·04

1-

--

--

-H

BV

gen

oty

pe

A-0

·29

(0·7

1)0·

683

--

--

--

Bas

elin

e A

LT (l

og

10 U

/L)

-0·0

2(0

·02)

0·39

7-

--

--

-B

asel

ine

HB

V-D

NA

(lo

g10

U/L

)-0

·23

(0·2

4)0·

341

--

--

--

Bas

elin

e H

BsA

g (l

og

10 U

/L)

-0·4

0(0

·28)

0·16

0-

--

--

-M

axim

um A

LT (l

og

10 U

/L)

3·30

(1·1

2)0·

003

--

-5·

02(1

·31)

0·00

01M

ax· f

old

-cha

nge

ALT

0·38

(0·1

4)0·

006

0·49

(0·1

6)0·

002

--

-W

eek

12 H

BV-

DN

A (l

og

10 U

/L)

-0·4

4(0

·52)

0·40

1-

--

--

-W

eek

12 H

BsA

g (l

og

10 U

/L)

-0·8

3(0

·27)

0·00

2-

--

-1·0

3(0

·31)

0·00

1W

eek

12 H

BsA

g d

eclin

e -3

·80

(1·1

2)0·

001

-3·9

1(1

·20)

0·00

1-

--

Bas

elin

e an

d e

arly

on-

trea

tmen

t va

riab

les

sig

nific

antly

ass

oci

ated

with

>1

log

HB

sAg

red

uctio

n at

wee

k 48

in u

niva

riab

le lo

gis

tic re

gre

ssio

n an

alys

is w

ere

asse

ssed

in

two

mul

tivar

iab

le m

od

els.

HB

V=

hep

atiti

s B

viru

s. A

LT=

alan

ine

amin

otr

ansf

eras

e. H

BsA

g=

hep

atiti

s B

sur

face

ant

igen

. B=

reg

ress

ion

coef

ficie

nt.

147


8

plus adefovir group (p<0·0001), mean reduction –0·59 (0·85) log10 IU/mL reduction at week 72 for the peg-IFN plus tenofovir group (p=0·0001), and mean reduction –0·15 (0·22) log10 IU/mL reduction at week 72 for the no treatment group (p<0·0001; table 3; figure 2A,B). Despite the slight increase in mean HBsAg levels during treatment-free follow-up of treated patients, the decline in HBsAg remained significantly larger in the peg-IFN plus adefovir group (p=0·004) and peg-IFN plus tenofovir group (p=0·004) than in the no treatment group.

The increase in HBsAg level during the treatment- free follow-up (HBsAg rebound) was particularly pronounced in patients with an HBsAg decline of more than 1·0 log10 IU/mL at week 48, and 13 patients in this group remained HBsAg-positive at this timepoint (figure 2C). Of these, 12 (92%) of 13 had an HBsAg rebound during the treatment-free follow-up (mean increase 0·84 log10 IU/mL at week 72 compared with week 48). The decline in hepatitis B virus DNA level and subsequent hepatitis B virus DNA rebound did not differ between patients with or without more than 1·0 log10 IU/mL HBsAg decline (figure 2D).

Baseline and early on-treatment variables were compared between treated patients with more than 1·0 log10 IU/mL reduction in HBsAg level at week 48 and those with less than 1·0 log10 IU/mL reduction (appendix p 9). Because the decline in HBsAg between the peg-IFN plus adefovir group and peg-IFN plus tenofovir group were similar, these were combined for this analysis. Baseline HBsAg levels were not significantly lower in treated patients with a strong HBsAg decline. However, significant predictors of HBsAg decline in univariable analysis were male sex (p=0·041), higher maximum on- treatment ALT level (p=0·003), and lower week 12 HBsAg level (p=0·002). Both on-treatment ALT increase and HBsAg level at week 12 were independent predictors of HBsAg decline at week 48, in different multivariable logistic regression models (table 3). Most patients had a baseline hepatitis B virus DNA level less than 2000 IU/mL (33 [72%] patients in the peg- IFN plus adefovir group, 28 [62%] in the peg-IFN plus tenofovir group, and 30 [70%] in the no treatment group; table 1), including all four patients with HBsAg loss at week 72. In this subgroup analysis, HBsAg declined more on average in the peg-IFN plus adefovir group (p=0·010) and peg-IFN plus tenofovir group (p=0·003) than in the no treatment group at week 72 (appendix p 10).

The most frequent adverse events (>30%) were fatigue, headache, fever, and myalgia, which were attributed to peg-IFN dosing (table 4). Two (4%) serious adverse events were reported in the peg-IFN plus adefovir group (admission to hospital for alcohol-related pancreatitis [week 6; n=1] and pregnancy, which was electively aborted [week 9; n=1]), three (7%) in the peg-IFN plus tenofovir group (admission to hospital after a suicide attempt during a severe depression [week 23; n=1], admission to hospital for abdominal pain [week 2; n=1], and an elective laminectomy [week 40; n=1]), and three (7%) in the no treatment group (admission to hospital for septic arthritis [week 72; n=1], endocarditis [week 5; n=1], and hyperthyroidism [week 20; n=1]). During treatment, 48 (53%) of 91 patients in the intervention groups (24 [52%] of 46 in the peg-IFN plus adefovir group and 24 [53%] of 45 in the peg-IFN plus tenofovir group) had ALT levels greater than 2 × ULN compared with three (7%) of 43 patients in the no treatment group (p<0·0001). During follow-up,

148

Chapter 8

table 4. Overview of adverse events in the modified intention-to-treat population

serious adverse events*

peg-Ifn+advn = 46

peg-Ifn+tdfn = 45

no treatment n = 43

2 3 3

General Fatigue 23 50% 28 62% 2 5% Headache 18 39% 19 42% 2 5% Fever 16 35% 18 40% 0 Myalgia 15 33% 16 36% 0 Other flu-like symptoms 9 20% 9 20% 2 5% Dizziness 8 17% 7 16% 5 12% Dyspnea 3 7% 1 2% 0 Back pain 2 4% 7 16% 3 7% Cough 2 4% 3 7% 0 Change in menstrual pattern 4 9% 4 9% 1 2%Digestive tract Abdominal pain 14 30% 10 22% 7 16% Change in stool consistency 10 22% 6 13% 1 2% Nausea 17 37% 3 7% 1 2% Loss of appetite 7 15% 13 29% 0 Dysgeusia 4 9% 3 7% 1 2% Gingivitis 1 2% 5 11% 0Dermatological Skin rash 7 15% 9 20% 1 2% Pruritus 9 20% 7 16% 0 Alopecia 8 17% 3 7% 0 Dry mucous membranes 3 7% 5 11% 0 Dry skin 2 4% 4 9% 0Psychiatric Depression, including mood changes 10 22% 12 27% 1 2% Concentration problems 2 4% 6 13% 0 Insomnia 5 11% 5 11% 0Laboratory Anemia (< 6.0 mmol/L) 4 9% 7 16% 0 Neutropenia (<0.75x10^9 cells/L) 15 33% 7 16% 1 2% Thrombocytopenia (<75x10^9 cells/L) 5 11% 0 0 Thyroid abnormalities 2 4% 5 11% 0 On-treatment ALT elevation** 24 52% 24 53% 3 7% Off-treatment ALT elevation** 4 9% 3 7% 4 9%

Peg-IFN=peg-interferon-alfa-2a. ALT=alanine aminotransferase. *Serious adverse events: peg-IFN plus adefovir: admission to hospital for alcohol-related pancreatitis (week 6; n=1), pregnancy, which was electively aborted (week 9; n=1); peg-IFN plus tenofovir: admission to hospital after a suicide attempt during a severe depression (week 23; n=1), admission to hospital for abdominal pain (week 2; n=1), and an elective laminectomy (week 40; n=1); no treatment: admission to hospital for septic arthritis (week 72), endocarditis (week 5), and hyperthyroidism (week 20). †Increase in ALT of more than 2 × the upper limit of normal (45 U/L for men, 34 U/L for women).

149


8

the rate of ALT levels greater than 2 × ULN was similar in the intervention groups (four [9%] of 46 in the peg-IFN plus adefovir group and three [7%] of 45 in the peg-IFN plus tenofovir group) versus the no treatment group (four [9%] of 43; p=0·751).

dIsCussIonOur results show no differences in functional cure, defined as HBsAg loss with or without formation of antibodies against anti-HBs between the treated and untreated patient groups. Our study is the first, to our knowledge, to analyse the effect of antiviral treatment in patients with chronic hepatitis B with a low viral load. The rationale for the study was based on the hypothesis that combining peg-IFN with a nucleos(t)ide analogue results in increased rates of HBsAg loss and anti-HBs seroconversion. Due to the relatively high rate of HBsAg loss in our cohort of patients with chronic hepatitis B with a high viral load treated with peg-IFN and adefovir, particularly in HBeAg-negative patients with low HBsAg levels,25 we explored the effect of this combination. In our previous study in patients with a high viral load, HBsAg loss was associated with low baseline HBsAg levels. Because HBeAg-negative patients with a low viral load generally have low HBsAg levels, we hypothesised that this patient group could benefit from combination treatment. Furthermore, we hypothesised that immunomodulation might be effective in patients with a low viral load, because they have residual hepatitis B virus-specific T-cell activity, possibly sensitive to a boost with immunomodulatory agents. However, despite the low baseline HBsAg levels26 and the significant residual T-cell function in this group,29,30 combination therapy did not lead to the anticipated increased levels of functional cure. Because the definition of patients with a low viral load (inactive carriers) changed over time, a subanalysis was done, taking into account only those patients with hepatitis B virus DNA less than 2000 IU/mL. Although all four patients with HBsAg loss were included in the group of patients with hepatitis B with a virus viral load less than 2000 IU/mL, no statistical difference was noted between treated and untreated patients with hepatitis B with virus DNA levels less than 2000 IU/mL.

When quantitative HBsAg level was taken into account, the decline in HBsAg was significantly greater in the treatment groups than in the no treatment group. This is in line with the findings of Bourlière and colleagues28 who reported a significant decline in HBsAg levels in nucleos(t)ide analogue suppressed patients treated with peg-IFN add-on compared with no peg-IFN add-on, while no difference in functional cure was observed between the groups. Although the exact mechanism might differ, boosting the immune system in patients with a low viral load and residual T-cell function does not seem to be effective in acquiring functional cure in either of these patient groups. Furthermore, Bourlière and colleagues28 showed no increase in the proportion of patients with HBsAg loss during long-term follow-up. Whether a strong HBsAg decline can lead to future HBsAg loss in patients with chronic hepatitis B with a low viral load will be determined in our subsequent 5-year follow-up study. Here, the outcome of HBsAg decline was associated with an on-treatment ALT increase and with low HBsAg level at week 12, indicating that the observed decline occurred early. Association of HBsAg loss and ALT flare has previously

150

Chapter 8

been observed in patients treated with combination therapy.24 However, because of low numbers and a rebound in HBsAg levels at week 72 in a proportion of patients, these results should be interpreted with caution. Furthermore, this study was done in a heterogeneous population. Although this accurately reflects the mixed European population, associations with ethnicity or hepatitis B virus genotype could have been underestimated.

Peg-IFN treatment is known to be associated with severe side-effects and safety issues. In our study, despite initial consent, 17 patients withdrew consent after randomisation. Even though these patients were equally distributed among the three randomisation groups, the acceptability rate of peg-IFN treatment was low, and similar to previous studies.28 Furthermore, judging by the high rate of adverse events and treatment discontinuation, peg-IFN was not well tolerated.

In this study, we have included two groups of combination therapy. In addition to its antiviral effect, adefovir has been shown to enhance innate immune functions, suggesting a possible synergistic effect when combined with peg-IFN.34,35 In our previous study, peg-IFN plus adefovir combination therapy resulted in high rates of HBsAg loss. Because adefovir has largely been replaced by tenofovir, which has a similar mechanism of action but is more potent in suppressing hepatitis B virus DNA, we also included a group with peg-IFN plus tenofovir combination.31 Because there are no previous data for treatment of patients with a low viral load, comparison of our results with peg-IFN or nucleos(t)ide analogue monotherapy data is not possible. However, we aimed to investigate the difference between combination therapy and no treatment. Based on natural history reports, HBsAg loss in untreated patients with chronic hepatitis B is 0·8–1·0% per year, which makes a control group indispensable.36,37

In conclusion, to our knowledge, this is the first treatment intervention study done in patients with chronic hepatitis B with a low viral load. Although HBsAg clearance rates did not differ significantly between the treatment and control groups, HBsAg decline was significantly greater in patients treated with peg-IFN and nucleos(t)ide analogue combination therapy at week 72 than in untreated patients. Although follow-up data will show whether a strong HBsAg decline could lead to higher rates of HBsAg loss in the treated patients in the long term, our findings do not support the use of combination treatment with peg-IFN and nucleos(t)ide analogue in patients with chronic hepatitis B with a low viral load. Pursuing functional cure for patients with chronic hepatitis B with a low viral load is supported by various benefits for the patient, including the discontinuation of close monitoring, a decreased risks of hepatitis B virus reactivation, and overcoming a stigma. The development of new antiviral hepatitis B virus compounds gives hope for future treatment for this large group of patients with chronic hepatitis B.

151


8

referenCes1. Ott JJ, Stevens GA, Groeger J, et al. Global

epidemiology of hepatitis B virus infection: new estimates of age-specific HBsAg seroprevalence and endemicity. Vaccine 2012; 30: 2212–29.

2. European Association for the Study of the Liver. EASL clinical practice guidelines: management of chronic hepatitis B virus infection. J Hepatol 2012; 57: 167–85.

3. Lok ASF, McMahon BJ. Chronic hepatitis B: update 2009. Hepatology 2009; 50: 661–62.

4. Chen C, Lee W, Yang H, et al. Changes in serum levels of HBV DNA and alanine aminotransferase determine risk for hepatocellular carcinoma. Gastroenterology 2011; 141: 1240–48.

5. Iloeje U, Yang H, Su J, et al. Predicting cirrhosis risk based on the level of circulating hepatitis B viral load. Gastroenterology 2006; 130: 678–86.

6. Tai D-I, Lin S-M, Sheen I-S, et al. Long-term outcome of hepatitis B e antigen-negative hepatitis B surface antigen carriers in relation to changes of alanine aminotransferase levels over time. Hepatology 2009; 49: 1859–67.

7. Chen C-J, Yang H-I, Su J, et al. Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level. JAMA 2006; 295: 65–73.

8. Kumar M, Sarin SK, Hissar S, et al. Virologic and histologic features of chronic hepatitis B virus-infected asymptomatic patients with persistently normal ALT. Gastroenterology 2008; 134: 1376–84.

9. Evans AA, Chen G, Ross EA, et al. Eight-year follow-up of the 90,000-person Haimen City cohort: I. Hepatocellular carcinoma mortality, risk factors, and gender differences. Cancer Epidemiol Biomarkers Prev 2002; 11: 369–76.

10. Beasley RP. Hepatitis B virus. The major etiology of hepatocellular carcinoma. Cancer 1988; 61: 1942–56.

11. Invernizzi F, Viganò M, Grossi G, et al. The prognosis and management of inactive HBV carriers. Liver Int 2016; 36: 100–04.

12. Fattovich G, Bortolotti F, Donato F. Natural history of chronic hepatitis B: special emphasis on disease progression and prognostic factors. J Hepatol 2008; 48: 335–52.

13. Papatheodoridis GV, Chrysanthos N, Hadziyannis E, et al. Longitudinal changes in serum HBV DNA levels and predictors of progression during the natural course of HBeAg-negative chronic hepatitis B virus infection. J Viral Hepat 2008; 15: 434–41.

14. Perrillo RP, Gish R, Falck-Ytter YT. Exam 3: American Gastroenterological Association Institute Technical Review on Prevention and Treatment of Hepatitis B Virus Reactivation During Immunosuppressive Drug Therapy. Gastroenterology 2015; 148: e16–17.

15. Chen Y, Sheen IS, Chu C, et al. Prognosis following spontaneous hbsag seroclearance in chronic hepatitis B patients with or without concurrent infection. Gastroenterology 2002; 123: 1084–89.

16. Yuen MF, Wong DKH, Sablon E, et al. HBsAg seroclearance in chronic hepatitis B in the Chinese: virological, histological, and clinical aspects. Hepatology 2004; 39: 1694–701.

17. Durantel D, Zoulim F. New antiviral targets for innovative treatment concepts for hepatitis B virus and hepatitis delta virus. J Hepatol 2016; 64: S117–31.

18. Moucari R, Mackiewicz V, Lada O, et al. Early serum HBsAg drop: a strong predictor of sustained virological response to pegylated interferon alfa-2a in HBeAg-negative patients. Hepatology 2009; 49: 1151–57.

19. Janssen HLA, van Zonneveld M, Senturk H, et al. Pegylated interferon alfa-2b alone or in combination with lamivudine for HBeAg-positive chronic hepatitis B: a randomised trial. Lancet 2005; 365: 123–29.

152

Chapter 8

20. Marcellin P, Bonino F, Lau GKK, et al. Sustained response of hepatitis B e antigen-negative patients 3 years after treatment with peginterferon Alfa-2a. Gastroenterology 2009; 136: 2169–79.

21. Viganò M, Invernizzi F, Grossi G, et al. Review article: the potential of interferon and nucleos(t)ide analogue combination therapy in chronic hepatitis B infection. Aliment Pharmacol Ther 2016; 44: 653–61.

22. Thimme R, Dandri M. Dissecting the divergent effects of interferon-alpha on immune cells: time to rethink combination therapy in chronic hepatitis B? J Hepatol 2013; 58: 205–09.

23. Tan AT, Hoang LT, Chin D, et al. Reduction of HBV replication prolongs the early immunological response to IFNα therapy. J Hepatol 2014; 60: 54–61.

24. Marcellin P, Ahn SH, Ma X, et al. Combination of tenofovir disoproxil fumarate and peginterferon α-2a increases loss of hepatitis B surface antigen in patients with chronic hepatitis B. Gastroenterology 2016; 150: 134–44.

25. Takkenberg RB, Jansen L, de Niet A, et al. Baseline hepatitis B surface antigen (HBsAg) as predictor of sustained HBsAg loss in chronic hepatitis B patients treated with peginterferon alfa-2a and adefovir. Antivir Ther 2013; 18: 895–904.

26. Jaroszewicz J, Calle Serrano B, Wursthorn K, et al. Hepatitis B surface antigen (HBsAg) levels in the natural history of hepatitis B virus (HBV)-infection: a European perspective. J Hepatol 2010; 52: 514–22.

27. Brunetto MR, Oliveri F, Colombatto P, et al. Hepatitis B surface antigen serum levels help to distinguish active from inactive hepatitis B virus genotype D carriers. Gastroenterology 2010; 139: 483–90.

28. Bourlière M, Rabiega P, Ganne-Carrie N, et al. Effect on HBs antigen clearance of addition of pegylated interferon alfa-2a to nucleos(t)ide analogue therapy versus nucleos(t)ide analogue therapy alone in patients with HBe antigen-negative chronic hepatitis B and sustained

undetectable plasma hepatitis B virus DNA: a randomised, controlled, open-label trial. Lancet Gastroenterol Hepatol 2017; 2: 177–88.

29. Boni C, Fisicaro P, Valdatta C, et al. Characterization of hepatitis B virus (HBV)-specific T-cell dysfunction in chronic HBV infection. J Virol 2007; 81: 4215–25.

30. de Niet A, Stelma F, Jansen L, et al. Restoration of T cell function in chronic hepatitis B patients upon treatment with interferon based combination therapy. J Hepatol 2015; 64: 539–46.

31. Marcellin P, Heathcote EJ, Buti M, et al. Tenofovir disoproxil fumarate versus adefovir dipivoxil for chronic hepatitis B. N Engl J Med 2008; 359: 2442–55.

32. Buster EHCJ, van Erpecum KJ, Schalm SW, et al. Treatment of chronic hepatitis C virus infection—Dutch national guidelines. Neth J Med 2008; 66: 292–306.

33. Ishak K, Baptista A, Bianchi L, et al. Histological grading and staging of chronic hepatitis. J Hepatol 1995; 22: 696–99.

34. Murata K, Asano M, Matsumoto A, et al. Induction of IFN-λ3 as an additional effect of nucleotide, not nucleoside, analogues: a new potential target for HBV infection. Gut 2016; published online Oct 26. DOI:10.1136/gutjnl-2016-312653.

35. Villani N, Caliò R, Balestra E, et al. 9-(2-Phosphonylmethoxyethyl) adenine increases the survival of influenza virus-infected mice by an enhancement of the immune system. Antiviral Res 1994; 25: 81–89.

36. Alward W, McMahon B, Hall D, et al. The long-term serological course of asymptomatic hepatitis B virus carriers and the development of primary hepatocellular carcinoma. J Infect Dis 1985; 151: 604–09.

37. Liaw Y, Sheen I, Chen T, et al. Incidence, determinants and significance of delayed clearance of serum HBsAg in chronic hepatitis B virus infection: a prospective study. Hepatology 1991; 13: 627–31.

153


8

fundInG This study was funded by Roche.

aCKnowledGementsWe would like to express great appreciation to Martine W Peters, Jeltje Helder, and Hadassa Heidsieck for patient follow-up and sampling. We thank Marjan J Sinnige for handling patient material and doing additional immunological analyses, as well as Meike H van der Ree for critical revision of the manuscript.


edItorIal:

peGInterferon add-on: an added BenefIt?

F. Stelma, H.W. Reesink

The Lancet Gastroenterology and Hepatology 2017;3:147-148

156

Chapter 9

With 240 million individuals infected worldwide,1 chronic hepatitis B virus (HBV) infection is a global cause of liver-related morbidity and mortality. Treatment options include pegylated interferon and nucleos(t)ide analogues. Pegylated interferon can stimulate antiviral immune mechanisms, whereas nucleos(t)ide analogues directly inhibit viral replication. These treatments are suffi cient for viral suppression, but generally fail to induce functional cure, which is defined as HBs antigen (HBsAg) loss with or without formation of anti-HBs antibodies. Achievement of functional cure is associated with an improved outcome.2–4 Combination of the two treatments could lead to increased efficacy of both drugs and has been the subject of many clinical trials.5 A large randomised controlled study by Marcellin and colleagues6 analysed the efficacy of pegylated interferon and nucleotide analogue combination therapy versus pegylated interferon or nucleotide analogue therapy given alone in previously untreated HBe antigen (HBeAg)-positive or HBeAg-negative patients with chronic hepatitis B. In this landmark study, 9.1% of patients with chronic hepatitis B had a functional cure 6 months after cessation of therapy in the combination treatment group, which was significantly higher than in the pegylated interferon monotherapy group (2.8%). In patients with chronic hepatitis B, virus-specific T cell responses are significantly impaired. Since virus-specific T cell responses are already partly restored after long-term viral load suppression by nucleos(t)ide analogues in patients with chronic hepatitis B,7 addition of an immune modulatory agent to nucleos(t)ide analogue therapy could hypothetically be beneficial for this group of patients, leading to increased functional cure.8

In The Lancet Gastroenterology & Hepatology, Marc Bourlière and colleagues9 present data for the PEGAN study. In this randomised, controlled, open-label trial, 185 patients with HBeAg-negative chronic hepatitis B who were on nucleos(t)ide analogue therapy for at least 1 year with undetectable HBV DNA were randomly allocated to receive pegylated interferon alfa-2a add-on therapy for 48 weeks or no additional therapy. The primary endpoint was HBsAg loss at week 96 by intention-to-treat analysis. Secondary endpoints included the proportion and grade of adverse events, quality of life as reported by patients, and willingness to receive pegylated interferon therapy.

In the pegylated interferon group, seven (7.8%) patients had lost HBsAg at week 96 as compared with three (3.2%) in the nucleos(t)ide analogue-only control group, which was not significantly different (difference 4.6% [95% CI –2.6 to 12.5]; p=0.15). Similar to the results from Marcellin and colleagues,6 the overall proportion of patients achieving a functional cure was low, and further analyses were done to identify factors associated with functional cure. These analyses revealed a significant association between low baseline HBsAg titre and HBsAg loss, as previously reported by others.10 The authors suggest that the add-on strategy in patients with HBeAg-negative chronic hepatitis B could potentially be meaningful among patients who have an HBsAg titre of less than 3 log10 IU/mL at baseline. In this select population (representing 42% of their patients), HBsAg loss was achieved in 23% of patients who received the full dose of pegylated interferon for 48 weeks. Assuming that such a proportion of functional cure could be maintained in a prospective study, select patients might benefit from add-on pegylated interferon treatment.

157

PEGINTERFERON ADD-ON THERAPY FOR CHB PATIENTS

9

The authors clearly state the limitations of their study. As these patients are nucleos(t)ide analogue suppressed, no information exists about HBV genotype. Furthermore, the study was underpowered because of a high degree of missing data and a higher than expected proportion of HBsAg loss in the control group. The absence of masking could have influenced patient behaviour, such as compliance with therapy. This patient behaviour, however, is the reality of clinical practice. The acceptability of pegylated interferon among nucleos(t)ide analogue-suppressed patients was 52%, indicating that a high barrier exists among patients with chronic hepatitis B to starting treatment with pegylated interferon. Furthermore, only 72% allocated to the pegylated interferon group received the full dose for 48 weeks.

Findings from this study clearly show the importance of reporting patient-related outcomes, such as barriers to initiation of treatment, the proportion and grade of adverse events, and drug discontinuation. This study is one of few randomised controlled trials on this topic that includes a large number of patients. Considering the observed high incidence of (serious) adverse events in the pegylated interferon group and the low occurrence of HBsAg loss in this study, Bourlière and colleagues9 conclude that add-on therapy with pegylated interferon is not a therapeutic strategy to treat chronic hepatitis B to establish functional cure.

Various novel HBV-specific compounds interacting with the replication cycle of HBV are under development. These direct-acting antivirals for HBV, eventually combined with less toxic immune modulators than pegylated interferon, should lead the way to treat chronic hepatitis B for achievement of functional cure in the future. In the development of new agents targeting HBV, the importance of patient-reported outcomes should be taken into consideration and critically assessed.11

158

Chapter 9

referenCes 1. Ott JJ, Stevens GA, Groeger J, Wiersma

ST. Global epidemiology of hepatitis B virus infection: new estimates of age-specifi c HBsAg seroprevalence and endemicity. Vaccine 2012; 30: 2212–19.

2. Moucari R, Korevaar A, Lada O, et al. High rates of HBsAg seroconversion in HBeAg-positive chronic hepatitis B patients responding to interferon: a long-term follow-up study. J Hepatol 2009; 50: 1084–92.

3. Fattovich G, Bortolotti F, Donato F. Natural history of chronic hepatitis B: special emphasis on disease progression and prognostic factors. J Hepatol 2008; 48: 335–52.

4. Feld JJ, Wong DKH, Heathcote EJ. Endpoints of therapy in chronic hepatitis B. Hepatology 2009; 49: S96–102.

5. Viganò M, Invernizzi F, Grossi G, Lampertico P. Review article: the potential of interferon and nucleos(t)ide analogue combination therapy in chronic hepatitis B infection. Aliment Pharmacol Ther 2016; 44: 653–61.

6. Marcellin P, Ahn SH, Ma X, et al. Combination of tenofovir disoproxil fumarate and peginterferon α-2a increases loss of hepatitis B surface antigen in patients with chronic hepatitis B. Gastroenterology 2016; 150: 134–44.

7. Boni C, Laccabue D, Lampertico P, et al. Restored function of HBV-specifi c T cells after long-term eff ective therapy with nucleos(t)ide analogues. Gastroenterology 2012; 143: 963–73.

8. Thimme R, Dandri M. Dissecting the divergent eff ects of interferon-alpha on immune cells: time to rethink combination therapy in chronic hepatitis B? J Hepatol 2013; 58: 205–09.

9. Boulière M, Rabiega P, Ganne-Carrie N, et al, for the ANRS HB06 PEGAN Study Group. Effect on HBs antigen clearance of addition of pegylated interferon alfa-2a to nucleos(t)ide analogue therapy versus nucleos(t)ide analogue therapy alone in patients with HBe antigen-negative chronic hepatitis B and sustained undetectable plasma hepatitis B virus DNA: a randomised, controlled, open-label trial. Lancet Gastroenterol Hepatol 2017;3:177-188.

10. Takkenberg RB, Jansen L, de Niet A, et al. Baseline hepatitis B surface antigen (HBsAg) as predictor of sustained HBsAg loss in chronic hepatitis B patients treated with peginterferon alfa-2a and adefovir. Antivir Ther 2013; 18: 895–904.

11. Sugarman J, Revill P, Zoulim F, et al. Ethics and hepatitis B cure research. Gut 2016;3:389-392.

hBv rna reduCtIons and serum CytoKIne ChanGes In hBeaG posItIve ChronIC hepatItIs B

patIents treated wIth rep 2139

F. Stelma, L. Jansen, A. Vaillant, M.J. Sinnige, K.A. van Dort, E.M.M. van Leeuwen, M. Bazinet, M. Al-Mahtab, N.A. Kootstra, H.W. Reesink

Submitted

aBstraCtBackground & aims Treatment with the HBsAg release inhibitor REP 2139 can lead to reductions in serum HBsAg levels in chronic hepatitis B (CHB) patients. Here, changes in HBV RNA and serum cytokine levels were analysed in patients treated with REP 2139.

methods Twelve HBeAg positive CHB patients received REP 2139 for 20-38 weeks in a phase 2 study (NCT02646189, REP 102). Nine responders (with >2 log decrease in serum HBsAg and HBV DNA) were subsequently treated with an add-on immunotherapy. HBV RNA and cytokine levels were assessed during the study.

results In 9 responders to REP 2139, HBV RNA levels declined in a similar fashion as HBV DNA and HBsAg levels. No decrease in serum HBV RNA was observed in non-responder patients. During REP 2139 treatment, IP-10, IFNγ and IL-21 significantly increased, while IL-7 serum concentration significantly decreased. Add-on immunomodulatory treatment did not further change the serum levels of these cytokines. No differences in cytokine concentrations were observed between responders and non-responders to REP 2139 treatment.

Conclusions HBsAg reduction induced by REP 2139 treatment is associated with HBV DNA and HBV RNA reductions. While serum levels of pro-inflammatory cytokines increased, the role of the immune system in clearance of HBV after treatment with REP 2139 needs to be further elucidated.

163

HBV RNA AND CYTOKINES IN REP2139 TREATED CHB PATIENTS

10

IntroduCtIonGlobally, 240 million people are chronically infected with the hepatitis B virus. Patients with chronic hepatitis B (CHB) infection are at increased risk of developing cirrhosis and hepatocellular carcinoma.1,2 Treatment of CHB patients aims at ‘functional cure’, i.e. HBsAg loss with or without anti-HBs antibodies, which is associated with an improved survival and lower incidence of hepatocellular carcinoma.3–5 Complete cure is challenging as the HBV mini-chromosome cccDNA is difficult to eradicate from the nucleus of the hepatocyte 6. Treatment options currently include nucleos(t)ide analogues (NUCs) which inhibit the viral reverse transcriptase, and peg-interferon, which modulates the immune response. These treatment options rarely lead to functional cure,7 hence there is a need for new compounds that can interfere with the hepatitis B virus replication cycle.8

Nucleic acid polymers (NAPs) are oligonucleotide-based antiviral compounds that exert antiviral action in HBV infection by blocking the release of HBsAg.9 In addition to virions, which can lead to newly infected hepatocytes, a 1000-10,000 fold excess of HBsAg-containing particles (or subviral particles) are secreted during hepatitis B infection.10 While the molecular target of NAPs has not yet been fully elucidated, these compounds act by selectively blocking the release of these HBV subviral particles from infected hepatocytes.11–13 In a phase 2 study, treatment with the NAP REP 2139 led to reductions in serum HBsAg levels > 1 log from baseline in 9 of 12 patients12 Furthermore, in 5 of these patients levels of anti-HBs became >10 IU/L during REP 2139 monotreatment. With the addition of a short course of immunomodulatory treatment, 8/9 of these responder patients achieved HBsAg loss (< 0.05 IU/mL) and anti-HBs > 150 IU/L. Four of these patients maintained control of the infection (HBsAg < 1 IU/mL, HBV DNA < 1000 c/ml) for 2 years after withdrawal of all therapy.12

The improved control of HBV infection with NAP therapy has been attributed to the clearance of HBsAg, which may have an inhibitory effect on the immune system. The elimination of circulating HBsAg may have an effect on the immune system. The large amount of HBsAg released into the circulation during CHB infection is thought to impair the activation and function of antigen presenting cells.14,15 In addition, HBsAg is shown to have an inhibitory effect on natural killer (NK) cells,16 and constant exposure of virus-specific T cells to HBsAg can leads to a state referred to as T cell ‘exhaustion’.17 These exhausted T cells are functionally impaired and ultimately deleted. The clearance of HBsAg could be the key to restoring the immune response against the hepatitis B virus. The effects of NAPs appear to be largely independent of any direct immune-stimulatory effect however, a residual pro-inflammatory effect of clinically active NAPs in human PMBCs in vitro is present which could contribute to the antiviral effects seen in patients.18

The effects of REP 2139 therapy on HBV RNA particles in currently unknown. Previous research has shown that treatment with NUCs affects HBV DNA levels more than HBV RNA levels, as the release of HBV RNA particles is not dependent on the reverse transcriptase which is blocked by NUCs.19–22 Treatment with a NUC in combination with peg-interferon on the other hand leads to a decline in HBV RNA similar to that observed in HBV DNA.19,22

164

Chapter 10

The aim of this study was to investigate changes in serum HBV RNA and cyto-/chemokines in patients with CHB who were treated with REP 2139.

patIents and methods patients in the previous phase 2 study (rep102 protocol)In a previously described phase 2 study (REP 102, NCT02646189), patients with CHB with a high viral load (>106 c/mL) and normal or elevated liver enzymes were dosed with the NAP: REP 2139.12 The study was conducted in accordance with the Declaration of Helsinki with local ethics approval and all patients gave written informed consent. Complete details of this study have been previously published.12 Twelve patients with HBeAg positive CHB infection (mean HBV DNA 7.51 log10 IU/ml) were included, and baseline characteristics are depicted in Table 1. In short, patients were median 25 years old (range 18-28), were predominantly male (9/12, 75%), with normal or elevated ALT levels (median 61 U/L, range 21-124 U/L) and infected with hepatitis B virus genotype A (n=1), C (n=7) or D (n=4). REP 2139 was administered as REP 2139-Ca (calcium chelate complex) and dosing was at 500 mg once weekly via 2 hour IV infusion. Responders to REP 2139 monotherapy were defined as patients experiencing HBsAg reductions > 1log from baseline. During monotherapy, all responders met the requirements for entry into combination therapy (defined as having a > 2 log reduction in serum HBV DNA and HBsAg) and were subsequently treated with a short course of add-on immunomodulatory agent (thymosin alpha 1 (thymosin) [n=5], or peginterferon alpha-2a (pegIFN) [n=4]) for 11-12 weeks (one patient was treated for 48 weeks). In the group of patients treated with thymosin, 2 patients subsequently received pegIFN treatment for 12 weeks (patients 6 and 7). All treatment was stopped 0-7 weeks after the cessation of immunomodulatory treatment (median 0 weeks).

table 1. Baseline characteristics.

Gender m/f age

hBv genotype

hBv dna load log10 Iu/ml

hBv rna load log10 c/ml

hBsag log10 Iu/ml

alt (u/l)

liver fibrosis score1

response to rep 2139 add-on

1 m 25 C 8,23 7,20 5,23 35 F0-F1 NR n.a.2 m 28 D 8,23 6,43 4,85 114 F3 R Thym3 f 26 C 7,46 6,89 4,13 52 F0-F1 R Thym4 m 20 A 7,65 6,14 3,54 89 F2 R Thym5 m 27 D 6,89 6,38 3,89 50 F2 NR n.a.6 m 19 C 7,47 6,24 4,71 124 F0-F1 R Thym/IFN7 m 22 C 8,23 n.a. 4,94 116 F0-F1 R Thym/IFN8 f 22 D 8,23 n.a. 4,86 41 F0-F1 R IFN9 m 26 D 5,09 n.a. 4,26 21 F1 R IFN10 m 18 C 8,23 7,79 5,09 36 F0-F1 NR n.a.11 m 18 C 8,23 8,00 5,10 70 F0-F1 R IFN12 m 26 C 6,23 6,05 3,18 87 F0-F1 R IFN

1liver fibrosis score (METAVIR) assessed by fibroscan.

165


10

sample selectionSerum samples consented for the purpose of additional analyses of the antiviral effects of REP 2139 were collected at the treatment site at multiple time points during the study and stored at -30 °C. Samples were sent to the Academic Medical Center (Amsterdam, The Netherlands) for analyses. Samples selected for luminex and HBV RNA analyses included (Supplementary figure 1): baseline (including 4 to 0 weeks before start of treatment, n=3 missing), during REP 2139 treatment: week 4 (including 4 to 7 weeks after start of therapy, n=1 missing), week 11 and week 20 (including week 19 to 23). Time points during add-on immunomodulatory treatment were selected 4 weeks after start of either thymosin (for patients 6 and 7, the first course of immunomodulatory treatment with thymosin was included) or pegIFN. Follow-up samples were median 24 (range 10-34) weeks after cessation of all (REP 2139 and immunomodulatory) therapy.

Clinical parametersMeasurements of HBV DNA, HBeAg, HBsAg, anti-HBs and ALT were performed previously12. Anti-HBs levels were considered positive when > 10 IU/L.

HBV RNA was detected in serum using a transcript-specific quantitative PCR (qPCR) assay as previously described 19. In short, total RNA was isolated from serum using the QIAamp Viral RNA Mini Kit (Qiagen), and treated with DNAse (Promega). RNA was concentrated using the NucleoSpin RNA Clean-up XS kit (Macherey-Nagel) and reverse transcribed with M-MLV reverse transcriptase (Promega) with an HBV-specific primer. The determination of HBV RNA levels was performed by qPCR in a LightCycler2000 system (Roche), using HBV RNA–specific primers (designed to detect both pgRNA and precore [PC] messenger RNA [mRNA]) and SYBR Green as a reporter dye. Complementary DNA was quantified by comparing the signals to a standard curve. The lower limit of quantification (LLOQ) was determined at 1000 copies/mL (3.00 log10 copies/mL).

luminex analysesSerum cytokine and chemokine levels were measured using a Luminex 27-plex immunoassay (Affymetrix eBioscience, San Diego, USA). These included IL-12, IL-23, IL-27, GM-CSF, IFN-α, IFN-γ, IL-1α, IL-1β, IL-1RA, IL-10, IL-13, IL-15, IL-17A, IL-18, IL-2, IL-21, IL-22, IL-31, IL-4, IL-5, IL-6, IL-7, IL-8, TNF-α, TNF-β, IL-10.

statistical analyses For differences between groups, the Mann–Whitney test (non-normal distribution) was used. For longitudinal analysis in individual patients the Wilcoxon signed rank test (non-normal distribution) was used. For analyses of correlation, Pearsons correlation coefficient was calculated. P-values below 0.05 were considered statistically significant. GraphPad Prism version 6.07 for Windows (GraphPad Software, La Jolla, CA, USA) was used for analyses.

166

Chapter 10

results Changes in hBv specific clinical parameters during rep 2139 dosing + immunotherapy The levels of pregenomic HBV RNA in the serum significantly decreased during treatment with REP-2139 in all 9 patients who responded to therapy with a decrease in HBsAg levels (Figure 1A,2A). As described previously, of the 9 responder patients, 5 patients became anti-HBs positive during REP 2139 monotherapy (Figure 2B). HBV DNA levels decreased, and transient elevations in ALT were observed (Figure 2C,D).

HBV RNA was below the limit of quantification at week 20 in all but one of the responding patients, where a marked decrease in HBV RNA (2.29 log10 c/mL) was observed between week 11 and 20, but no earlier HBV RNA levels were available. In the three patients who did not respond to REP 2139 treatment (HBsAg decline < 1 log from baseline), there was no difference in HBV RNA levels between baseline and week 20 (median 7.2 log10 c/mL

figure 1. Serum levels of HBV RNA (A) in patients dosed with REP 2139. Each line represents an individual patient, responders are depicted in black and non-responders in grey. During add-on, patients treated with thymosin are represented in blue triangles and patients treated with pegIFN are represented by green squares. Missing data HBV RNA: BL, n=3; week 4, n=1. (B) Levels of HBV RNA, HBV DNA, HBsAg and anti-HBs in non-responders retreated with entecavir during follow-up, n=3.

(a)

(B)

167


10

at BL to 7.9 log10 c/mL at week 20, p=0.75). In one of these patients HBV RNA was below the LLOQ at week 4, but increased to 4.28 log10 c/mL at week 11 and to baseline levels at FU (BL: 6.38 log10 c/mL, FU: 6.14 log10 c/mL). While HBV DNA levels decreased eventually upon initiation of entecavir treatment in these patients (Figure 2C), HBV RNA levels did not (Figure 1A,B).

The 9 patients who responded to REP 2139 ( >2 log decrease in serum HBsAg from baseline)(Figure 2A) were subsequently treated with an immunomodulatory agent (thymosin [n=5], or pegIFN [n=4]). During treatment with add-on immunotherapy, anti-HBs levels became > 10 IU/L in all 9 patients (Figure 2B) and HBV RNA levels remained < LLOQ in 8/9 patients (Figure 1A). At the treatment-free FU timepoint, HBV RNA was < LLOQ in 8/9 patients and 4 patients were HBsAg negative (<0.05 IU/mL) of which 3 patients had HBV DNA <LLOQ (<116 c/mL).

(a)

(C)

(B)

(d)

figure 2. Clinical parameters HBsAg (A), anti-HBs (B), HBV DNA (C) and ALT (D) in patients dosed with REP 2139. Each line represents an individual patient, responders are depicted in black and non-responders in grey. During add-on, patients treated with thymosin are represented in blue triangles and patients treated with pegIFN are represented by green squares. (Anti-HBs became positive later during add-on immunomodulatory treatment in 2/9 patients).

168

Chapter 10

serum Ip-10 and Ifn-γ levels increase upon dosing with rep 2139In all patients receiving REP 2139, serum interferon gamma-induced protein 10 (IP-10 or CXCL10) levels significantly increased from baseline (median 17.42 pg/ml) to week 4 (median 103.6 pg/ml) after start of therapy (p=0.0078, n=9 due to missing baseline samples, Figure 3A). IP-10 levels continued to increase during subsequent weeks of REP 2139 monotherapy in all patients (week 11; median 83.63 pg/ml, week 20: 107.28 pg/ml, p=0.0039). Addition of immunotherapy in the 9 responder patients with HBsAg decline did not further increase IP-10 levels; no significant differences were observed between the last time point before immunotherapy (week 20: median 111.3 pg/ml) and the first time point on immunotherapy treatment (median 194.5 pg/ml, p=0.13). During follow-up, IP-10 levels were still significantly increased compared to baseline (median 46.65 pg/ml and 17.42 pg/ml respectively, p=0.0078). There was no significant difference in IP-10 level increase in responders and non-responders to REP 2139 therapy (Figure 3B).

In addition, IFN-γ levels significantly increased from baseline (median 1.96 pg/ml) to week 4 (median 3.71 pg/ml) after start of therapy (p=0.027, n=9 due to missing baseline samples, Figure 3C). The level of IFN-γ was significantly increased at week 11 during REP 2139 monotherapy (median 4.27 pg/ml, p=0.0078) but not at week 20 (median 3.2 pg/ml, p=0.55). There was no difference between week 20 (median 6.05 pg/ml) and during immunomodulatory treatment (median 5.26 pg/ml, p=0.18). During follow-up IFN-γ levels were still significantly increased compared to baseline (median 5.68 pg/ml and 1.96 pg/ml respectively, p=0.0039). There was no significant difference in the change in IFN-γ from baseline to week 20 in patients who responded to REP 2139 treatment and patients who did not (Figure 3D).

no association is observed between Ip-10, Ifn-γ and hBsag levelsThe serum IP-10 levels at week 20 of REP 2139 treatment were not significantly correlated with the level of HBsAg at week 20 (p=0.64, Figure 4A). In addition, the serum levels of IFN-γ at week 20 of REP 2139 treatment were not significantly correlated to the level of HBsAg at week 20 (p=0.09, Figure 4B).

serum Il-7 levels decreased and Il-21 levels increased upon dosing with rep 2139Serum IL-7 levels, important for T cell survival and homeostatic proliferation, significantly decreased from baseline (median 4.11 pg/ml) to week 4 (median 3.28 pg/ml) after start of therapy (p=0.02, n=9 due to missing baseline samples, Figure 5A). Also at week 20, IL-7 levels were significantly decreased as compared to baseline (median 2.27 pg/ml and 4.11 pg/ml respectively, p=0.0039). No changes in IL-7 levels were observed before and after immunomodulatory treatment (add-on and FU). The levels of IL-21, which is a potent natural killer cell and T cell activator, were increased compared to baseline during REP 2139 monotherapy; median 12.65 pg/ml at baseline, 36.95 pg/ml at week 4 (p=0.16) and 17.62 pg/ml at week 11 (p=0.03)(Figure 5B), and were not influenced by immunomodulatory therapy.

169


10

serum cytokine changes in response to immunomodulatory treatment Serum levels of IL-10, TNF-α, IL-6, IL-1a, IL-1b and IL-1RA increased in patients during thymosin treatment (Supplementary Figure 2), but not during REP 2139 monotherapy or pegIFN add-on. Other measured cytokines did not show any changes during REP 2139 monotherapy, or with add-on immunotherapy (data not shown).

dIsCussIonIn this study, we have analysed changes in serum HBV RNA and serum cytokine and chemokine levels in patients with HBeAg positive CHB virus infection who were dosed with REP 2139. The goal was to characterize the HBV RNA and serum cytokine response to REP 2139 treatment.

(a)

(C)

(B)

(d)

figure 3. Serum levels of IP-10 (A) and IFN-γ (C) at various time points during REP 2139 monotherapy and during add-on immunomodulatory treatment. Differences in the change from baseline to week 20 in (B) IP-10 levels and (D) IFN-γ in responders (black) and non-reponders (grey) to REP-2139 treatment. During add-on, patients treated with thymosin are represented in blue triangles and patients treated with pegIFN are represented by green squares. Missing data: BL, n=3; week 4, n=1.

170

Chapter 10

(a)

(a)

(B)

(B)

figure 4. The level of IP-10 and HBsAg (A) and IFN-γ and HBsAg (B) at week 20 of REP 2139 treatment. R squared (R2) and p values are indicated.

figure 5. Serum levels of IL-7 (A) and IL-21 (C) at various time points during REP 2139 monotherapy and during add-on immunomodulatory treatment. During add-on, patients treated with thymosin are represented in blue triangles and patients treated with pegIFN are represented by green squares.

After dosing with REP 2139, HBsAg declined in 9/12 patients. In these patients, HBV DNA and HBV RNA levels also decreased, while ALT levels increased. While HBV DNA and HBV RNA particles contain HBsAg,21 previous studies have indicated that NAPs selectively block the release of HBV subviral particles, and not that of HBV DNA or HBV RNA particles.11,12 The observed decline in HBV DNA and HBV RNA could be caused by increased levels of free anti-HBs that appear in patients who clear HBsAg upon dosing with REP 2139, which would enhance the clearance of HBV DNA and HBV RNA containing particles from the circulation. In patients that did not respond to REP 2139 (i.e. patients who had HBsAg reductions < 1 log from baseline) persistently high levels of HBV DNA and HBV RNA were observed. The serum level of HBV RNA has previously been shown to correlate with intrahepatic cccDNA levels in treatment-naïve mice,22 suggesting a possible reduction of intrahepatic cccDNA in CHB patients with a decrease in HBV RNA.

HBsAg is known to have several direct and indirect immunomodulatory effects, such as alterations in cytokine signaling.15,23–26 Furthermore, an excess of HBsAg antigen can

171


10

induce tolerogenic properties in dentritic cells, which then become less able to activate HBV-specific T cells.14 Reducing HBsAg may also release the inhibition on virus-specific T cells, that become impaired upon prolonged stimulation by high antigen levels.27 As HBsAg is known to have these immune modulating properties, a decrease in HBsAg could abrogate these inhibitory effects. We therefore analysed serum cytokines in these patients as a reflection of immune activation. We have shown an increase in the levels of activating cytokines, however, the increase was observed in responders as well as non-responders to REP 2139.

An important consideration in assessing the antiviral effects of REP 2139 is the known potential for oligonucleotides to elicit immune responses via interaction with innate pattern recognition receptors. The innate immune system has various defense mechanisms that enable the body to recognize non-self molecular patterns. For instance, non-methylated CpG sequences, associated with bacterial or viral DNA, are recognized by TLR9.28 Similarly, double stranded RNA, only present in case of a virus infection, is recognized by TLR3.29 Activation of these receptors will ultimately lead to the production of cytokines such as type I interferons. The potential of exogenous oligonucleotides to elicit such responses is therefore always tested in preclinical studies. REP 2139 is optimized to prevent recognition by the innate immune response by the incorporation of naturally occurring nucleic acid modifications (5-methylcytosine and 2’O methyl ribose) known to block recognition of host nucleic acids by the innate immune response.9 Recently, a detailed analysis of pro-inflammatory and antiviral cytokine responses in primary human parenchymal and non-parenchymal liver cells as well as peripheral blood mononuclear cells was conducted with REP 213918 showing no significant immunoreactivity in the liver but a mild pro-inflammatory response in PBMCs.

Increases in serum IFN-γ, IP-10 and IL-21 levels were observed in all patients dosed with REP 2139, regardless of any change in HBsAg levels. These increases may potentially be connected with the mild pro-inflammatory effects of REP 2139 observed in PBMCs.18 Although the increases in these serum cytokines were not associated with reduction of HBsAg or the suppression of viremia off-treatment, the role these cytokine increases may have on establishing control of HBV infection remains to be understood and requires further clinical examination.

In conclusion, these results demonstrate that upon treatment with REP 2139 HBV RNA decreases in a similar fashion as HBV DNA. These antiviral effects occur only in patients who have a reduction in HBsAg > 1 log from baseline, and appear to be independent from the cytokine responses evaluated. After cessation of therapy, HBV DNA and HBV RNA were controlled in a selection of these patients. Furthermore, as some cytokine responses appear to be remodeled in all patients receiving REP 2139 therapy, further research is needed to analyse the contribution of the host immune response to the antiviral activity against hepatitis B occurring during REP 2139-mediated HBsAg reduction.

172

Chapter 10

referenCes1. Iloeje UH, Yang H-I, Su J, Jen C-L, You S-L,

Chen C-J. Predicting cirrhosis risk based on the level of circulating hepatitis B viral load. Gastroenterology 2006;130(3):678–86.

2. Chen C-J, Yang H-I, Su J, et al. Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level. JAMA 2006;295(1):65–73.

3. Moucari R, Korevaar A, Lada O, et al. High rates of HBsAg seroconversion in HBeAg-positive chronic hepatitis B patients responding to interferon: A long-term follow-up study. J Hepatol 2009;50(6):1084–92.

4. Fattovich G, Bortolotti F, Donato F. Natural history of chronic hepatitis B: special emphasis on disease progression and prognostic factors. J Hepatol 2008;48(2):335–52.

5. Feld JJ, Wong DKH, Heathcote EJ. Endpoints of therapy in chronic hepatitis B. Hepatology 2009;49(5 Supp):S96–102.

6. Lucifora J, Protzer U. Attacking hepatitis B virus cccDNA - The holy grail to hepatitis B cure. J Hepatol 2016;64(1 Supp):S41-8.

7. Marcellin P, Ahn SH, Ma X, et al. Combination of Tenofovir Disoproxil Fumarate and Peginterferon α-2a Increases Loss of Hepatitis B Surface Antigen in Patients with Chronic Hepatitis B. Gastroenterology 2016;150(1):134–44.

8. Liang TJ, Block TM, McMahon BJ, et al. Present and future therapies of hepatitis B: From discovery to cure. Hepatology 2015;62:1893–908.

9. Vaillant A. Nucleic acid polymers: Broad spectrum antiviral activity, antiviral mechanisms and optimization for the treatment of hepatitis B and hepatitis D infection. Antiviral Res 2016;133:32–40.

10. Ganem D, Prince AM. Hepatitis B virus infection: Natural history and

clinical consequences. N Engl J Med 2004;350(11):1118–29.

11. Noordeen F, Scougall CA, Grosse A, et al. Therapeutic antiviral effect of the nucleic acid polymer REP 2055 against persistent duck hepatitis B virus infection. PLoS One 2015;10(11):1–20.

12. Al-Mahtab M, Bazinet M, Vaillant A. Safety and Efficacy of Nucleic Acid Polymers in Monotherapy and Combined with Immunotherapy in Treatment-Naive Bangladeshi Patients with HBeAg+ Chronic Hepatitis B Infection. PLoS One 2016;11(6):e0156667.

13. Blanchet M, Vaillant A, Labonté P. Post-entry antiviral effects of nucleic acid polymers against hepatitis B virus infection in vitro. J Hepatol 2017;66:S257.

14. Op Den Brouw ML, Binda RS, Van Roosmalen MH, et al. Hepatitis B virus surface antigen impairs myeloid dendritic cell function: a possible immune escape mechanism of hepatitis B virus. Immunology 2009;126:280–9.

15. Vanlandschoot P, Van Houtte F, Roobrouck A, Farhoudi A, Leroux-Roels G. Hepatitis B virus surface antigen suppresses the activation of monocytes through interaction with a serum protein and a monocyte-specific receptor. J Gen Virol 2002;83(6):1281–9.

16. Chen Y, Wei H, Sun R, Tian Z. Impaired function of hepatic natural killer cells from murine chronic HBsAg carriers. Int Immunopharmacol 2005;5(13–14):1839–52.

17. Boni C, Fisicaro P, Valdatta C, et al. Characterization of hepatitis B virus (HBV)-specific T-cell dysfunction in chronic HBV infection. J Virol 2007;81(8):4215–25.

18. Real C, Werner M, Paul A, et al. Nucleic acid-based polymers effective against hepatitis B virus infection in patients do not harbour immune stimulatory properties in primary isolated blood or liver cells. J Hepatol 2016;64(2):S395.

173


10

19. Jansen L, Kootstra N a., van Dort K a., Takkenberg RB, Reesink HW, Zaaijer HL. Hepatitis B Virus Pregenomic RNA Is Present in Virions in Plasma and Is Associated With a Response to Pegylated Interferon Alfa-2a and Nucleos(t)ide Analogues. J Infect Dis 2015;213(2):224–32.

20. van Bömmel F, Bartens A, Mysickova A, et al. Serum hepatitis B virus RNA levels as an early predictor of hepatitis B envelope antigen seroconversion during treatment with polymerase inhibitors. Hepatology 2015;61(1):66–76.

21. Wang J, Shen T, Huang X, et al. Serum hepatitis B virus RNA is encapsidated pregenome RNA that may be associated with persistence of viral infection and rebound. J Hepatol 2016;65(4):700–10.

22. Giersch K, Allweiss L, Volz T, Dandri M, Lütgehetmann M. Serum HBV pgRNA as a clinical marker for cccDNA activity. J Hepatol 2016;66:460–2.

23. Cheng J, Imanishi H, Morisaki H, et al. Recombinant HBsAg inhibits LPS-induced COX-2 expression and IL-18 production by interfering with the NF??B pathway in a human monocytic cell line, THP-1. J Hepatol 2005;43:465–71.

24. Shi B, Ren G, Hu Y, Wang S, Zhang Z, Yuan Z. HBsAg Inhibits IFN-a Production in Plasmacytoid Dendritic Cells through TNF-a and IL-10 Induction in Monocytes. PLoS One 2012;7(9):e44900.

25. Woltman AM, den Brouw MLO, Biesta PJ, Shi CC, Janssen HL a. Hepatitis b virus lacks immune activating capacity, but actively inhibits plasmacytoid dendritic cell function. PLoS One 2011;6(1):e15324.

26. Wu J, Meng Z, Jiang M, et al. Hepatitis B Virus Suppresses Toll-like Receptor–Mediated Innate Immune Responses in Murine Parenchymal and Nonparenchymal Liver Cells. Hepatology 2009;49(4):1132–40.

27. Bertoletti A, Ferrari C. Innate and adaptive immune responses in chronic hepatitis B virus infections: towards restoration of immune control of viral infection. Gut 2012;(61):1754–64.

28. Krieg AM. CpG motifs in bacterial DNA and their immune effects. Annu Rev Immunol 2002;20(1):709–60.

29. Kawai T, Akira S. The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors. Nat Immunol 2010;11(5):373–84.

fundInG This study was funded by Replicor Inc.

174

Chapter 10

supplementary data

Supplementary data to:

HBV RNA reductions and serum cytokine changes in HBeAg positive chronic hepatitis B patients

treated with REP 2139.

F. Stelma, L. Jansen, A. Vaillant, M.J. Sinnige, E.M.M. van Leeuwen, M. Bazinet, Mamun Al‐Mahtab,

N.A. Kootstra, H.W. Reesink

Supplementary Figure 1. Overview of serum sampling.

*Add‐on: Pegasys™ 180 ug s.c. qW and/or Zadaxin™ (thymosin alpha‐1) 1.6mg s.c. 2qW.

supplementary figure 1. Overview of serum sampling. *Add-on: Pegasys™ 180 ug s.c. qW and/or Zadaxin™ (thymosin alpha-1) 1.6mg s.c. 2qW.

175


10

supplementary figure 2. Serum levels of IL-10 (A), TNF-α (B), IL-6 (C), IL-1a (D), IL-1b (E), and IL-RA (F) at various time points during REP 2139 monotherapy and during add-on immunomodulatory treatment. During add-on, patients were treated with thymosin alpha 1 represented in blue or peginterferon alpha-2a represented in green. Missing data: BL, n=3; week 4, n=1.

IL-1

0 (p

g/m

l)ns

nsns

ns

*

nsns

nsns

ns

nsns

nsns

ns

IL-1

a (p

g/m

l)

nsns

nsns

ns

nsns

nsns

ns

TNF-

(pg/

ml)

nsns

nsns

ns

Supplementary Figure 2. Serum levels of IL‐10 (A), TNF‐α (B), IL‐6 (C), IL‐1a (D), IL‐1b (E), and IL‐RA (F) at various time points during REP 2139 monotherapy and during add‐on immunomodulatory treatment. During add‐on, patients were treated with thymosin alpha 1 represented in blue or peginterferon alpha‐2a represented in green. Missing data: BL, n=3; week 4, n=1.

(a) (B)

(C) (d)

(e) (f)

Immune responses In patIents wIth ChronIC hepatItIs C InfeCtIon

safety, toleraBIlIty, and antIvIral effeCt of rG-101 In patIents wIth ChronIC hepatItIs C:

a phase 1B, douBle-BlInd, randomIsed Controlled trIal

M.H. van der Ree, J.M. de Vree, F. Stelma, S.B. Willemse, M. van der Valk, S. Rietdijk, R. Molenkamp, J. Schinkel, A.C. van Nuenen, U. Beuers, S. Hadi, M. Harbers,

E. van der Veer, K. Liu, J. Grundy, A.K. Patick, A. Pavlicek, J. Blem, M. Huang, P. Grint, S. Neben, N.W. Gibson, N.A. Kootstra, H.W. Reesink

The Lancet. 2017;389:709-717.

summary BackgroundmiR-122 is an important host factor for hepatitis C virus (HCV) replication. The aim of

this study was to assess the safety and tolerability, pharmacokinetics, and antiviral effect

of a single dose of RG-101, a hepatocyte targeted N-acetylgalactosamine conjugated

oligonucleotide that antagonises miR-122, in patients with chronic HCV infection with

various genotypes.

methodsIn this randomised, double-blind, placebo-controlled, multicentre, phase 1B study, patients

were randomly assigned to RG-101 or placebo (7:1). We enrolled men and postmenopausal

or hysterectomised women (aged 18–65 years) with chronic HCV genotype 1, 3, or 4

infection diagnosed at least 24 weeks before screening who were either treatment naive

to or relapsed after interferon-α based therapy. Patients with co-infection (hepatitis B virus

or HIV infection), evidence of decompensated liver disease, or a history of hepatocellular

carcinoma were excluded. Randomisation was done by an independent, unblinded,

statistician using the SAS procedure Proc Plan. The first cohort received one subcutaneous

injection of 2 mg/kg RG-101 or placebo; the second cohort received one subcutaneous

injection of 4 mg/kg or placebo. Patients were followed up for 8 weeks (all patients) and up

to 76 weeks (patients with no viral rebound and excluding those who were randomised to

the placebo group) after randomisation. The primary objective was safety and tolerability

of RG-101. This trial was registered with EudraCT, number 2013-002978-49.

findingsBetween June 4, 2014, and Oct 27, 2014, we enrolled 32 patients with chronic HCV

genotype 1 (n=16), 3 (n=10), or 4 (n=6) infections. In the first cohort, 14 patients were

randomly assigned to receive 2 mg/kg RG-101 and two patients were randomly assigned

to receive placebo, and in the second cohort, 14 patients were randomly assigned to

receive 4 mg/kg RG-101 and two patients were randomly assigned to receive placebo.

Overall, 26 of the 28 patients dosed with RG-101 reported at least one treatment-related

adverse event. At week 4, the median viral load reduction from baseline was 4·42 (IQR

3·23–5·00) and 5·07 (4·19–5·35) log10 IU/mL in patients dosed with 2 mg/kg RG-101 or 4

mg/kg RG-101. Three patients had undetectable HCV RNA levels 76 weeks after a single

dose of RG-101. Viral rebound at or before week 12 was associated with the appearance

of resistance associated substitutions in miR-122 binding regions in the 5’ UTR of

the HCV genome.

InterpretationThis study showed that one administration of 2 mg/kg or 4 mg/kg RG-101, a hepatocyte

targeted N-acetylgalactosamine conjugated anti-miR-122 oligonucleotide, was well

tolerated and resulted in substantial viral load reduction in all treated patients within 4

weeks, and sustained virological response in three patients for 76 weeks.

181

SAFETY, TOLERABILITY AND ANTIVIRAL EFFECT OF RG-101

11

IntroduCtIon Hepatitis C virus (HCV) is an enveloped, single-stranded RNA virus of the Flaviviridae family.1 Most people infected with HCV develop a chronic hepatitis C infection. Patients with this infection are at risk of developing complications such as liver cirrhosis and hepatocellular carcinoma.2 The goal of chronic hepatitis C treatment is to achieve a sustained virological response, defined as undetectable HCV RNA in blood 12–24 weeks after completing a course of antiviral treatment,3 which is associated with a reduced occurrence of liver-related complications and a prolonged survival.4,5 Most new anti-HCV drugs are direct-acting antivirals that target specific viral proteins, such as NS3A protease, NS5A protein, and NS5B polymerase.

Present treatment regimens with a combination of direct-acting antivirals for 12 to 24 weeks result in high, sustained, virological response rates in most patients with chronic hepatitis C. Although treatment with a combination of direct-acting antivirals for 8 weeks in a selected group of patients with chronic hepatitis C (ie, HCV genotype 1, treatment naive, and non-cirrhotic) results in high, sustained, virological response rates, attempts to shorten treatment duration to less than 8 weeks with a triple or quadruple direct-acting antiviral regimen has shown high variability of sustained virological response rates.6,7 An alternative treatment strategy is interference with host cell components that are essential for replication of HCV, such as cyclophilin A or microRNA (miRNA)-122.8,9 Host-targeting agents have broad pan-genotypic antiviral activity,10 and have a relatively high genetic barrier to resistance compared with direct-acting antivirals. Furthermore, host-targeting agents might act in a synergistic manner with direct-acting antivirals resulting in shortened direct-acting antiviral therapy, including in those patients who are difficult to treat with present direct-acting antivirals.11

miR-122 is an abundant, liver-specific miRNA that modulates the expression of genes involved in cholesterol and triglyceride metabolism 12–14 and in tumorigenesis.15–18 Additionally, miR-122 is a host factor that is essential for HCV replication. miR-122 binds to the 5’ untranslated region (5’ UTR) of the HCV genome and thereby promotes HCV RNA stability and accumulation,19,20 protects HCV RNA against degradation by cellular exoribonucleases 1 and 2 (Xrn1 and Xrn2),21–23 and induces an increase in rate of viral RNA synthesis, or a combination of these factors.24 Previously, repeated dosing with miravirsen, a locked nucleic acid DNA phosphorothioate oligonucleotide inhibitor of miR-122, resulted in a dose-dependent and prolonged decrease in HCV RNA levels in patients with chronic hepatitis C with HCV genotype 1 infection.9 RG- 101 is a mixed chemistry phosphorothioate oligonucleotide inhibitor of miR-122 that is linked to a multivalent N-acetylgalactosamine carbohydrate structure designed to enhance uptake of the oligonucleotide by hepatocytes through binding to the asialoglycoprotein receptor on hepatocytes.25 Conjugation to the N-acetylgalactosamine structure increases the potency of RG-101 by up to 10–20 times compared with the non-conjugated oligonucleotide.

The aim of this proof-of-concept study was to assess the safety and tolerability, pharmacokinetics, and antiviral effect of one dose of an N-acetylgalactosamine conjugated oligonucleotide against miR-122 (RG-101) in patients with chronic HCV genotype 1, 3, and 4 infection.

182

Chapter 11

methods study design In this randomised, double-blind, placebo-controlled, multicentre, phase 1B trial,

we enrolled 32 patients aged 18–65 years with chronic hepatitis C at two hospitals in

the Netherlands (Academic Medical Center, Amsterdam and University Medical Center,

Groningen, the Netherlands). Liver fibrosis was assessed by liver elastography (transient

elastography, known by the brand name FibroScan).

On study day 1, the first cohort received one subcutaneous injection of 2 mg/kg

(volume injection range 0·9–1·5 mL) RG-101 or placebo, and the second cohort received

one subcutaneous injection of 4 mg/kg (volume injection range 1·4–3·9 mL, or placebo).

In the main study, both dosing cohorts were followed up for 8 weeks after randomisation

(appendix). At the end of the main study, an independent physician determined which

patients qualified to be enrolled in an extended follow-up study in which patients were

followed up for a period of up to 76 weeks after dosing. At week 8, all patients treated with

placebo or with virological rebound confirmed at retest were not included in the extended

follow-up study. Virological rebound was defined as more than 1 log10 increase in HCV RNA

level from nadir (defined as the lowest HCV RNA level after baseline). During the extended

follow-up study, patients with virological rebound were excluded from further follow-up.

The study was approved by the regulatory authority and the independent ethics

committee at each participating site, and was done in compliance with the Declaration

of Helsinki, Good Clinical Practice guidelines, and local regulatory requirements. An

independent data and safety monitoring committee reviewed the safety of the patients in

the study.

patients Men and postmenopausal or hysterectomised women (aged 18–65 years) with chronic HCV

genotype 1, 3, or 4 infection diagnosed at least 24 weeks before screening were enrolled.

Eligible patients were treatment naive to, or relapsed after, interferon-α based therapy.

Other eligibility criteria were those with a platelet count of more than 100 × 109/L, a total

white blood cell count of more than 3·0 × 109/L, a haemoglobin concentration of more than

6·8 mmol/L for women and more than 7·4 mmol/L for men, an alanine aminotransferase

concentration of less than five times the upper limit of normal (ULN), a total and direct

bilirubin level within normal limits, a creatinine level within normal limits, and serum HCV

RNA ≥75 000 IU/mL at screening. Patients with co-infection (hepatitis B virus or human

immunodeficiency virus infection), evidence of decompensated liver disease, or a history

of hepatocellular carcinoma were excluded (appendix). Before enrolment and before any

study procedure, written informed consent was obtained from all patients.

randomisation and masking In each dose cohort (2 mg/kg or 4 mg/kg), patients were randomly assigned to receive

either RG-101 or placebo (7:1; appendix). Allocation to treatment groups (placebo or

183


11

active) was done randomly by an independent, unblinded, statistician, using the SAS procedure Proc Plan. Randomisation was done in blocks of eight patients, in a ratio of 1:7 (placebo:active). The randomisation list was kept by the unblinded pharmacist, and only provided to the study team after database hardlock and unblinding of the study at 8 weeks. Investigators and patients were blinded to treatment allocation and on-study HCV RNA results at least 8 weeks after RG-101 dosing.

outcomes The primary objective was to assess the safety and tolerability of one dose of subcutaneous RG-101 administered to patients with chronic HCV genotype 1, 3, and 4 infection. The secondary objective was to assess the pharmacokinetic profile and antiviral effect of RG-101 in patients with chronic hepatitis C.

procedures Safety was assessed in patients by physical examination, review of adverse events, laboratory testing of blood samples, and urinalysis (appendix). Calculated plasma and urine pharmacokinetic parameters included the maximum plasma concentration (Cmax), the time at which the maximum plasma concentration occurred (Tmax), the area under the plasma concentration curve up to the last measurable concentration (AUC0–t), and the percentage of the dose excreted in urine (as parent drug or major metabolite) during the first 24 h after receiving the treatment dose. Serum HCV RNA levels were measured with the Roche COBAS AmpliPrep/COBAS Taqman HCV (version 2.0) assay, with a reported lower limit of quantification of 15 IU/mL. Sustained virological response was defined as undetectable HCV RNA levels at 76 weeks after RG-101 treatment. HCV genotype was identified by sequence analysis of a fragment of the NS5B gene. Genotype of IFNL3 single nucleotide polymorphism rs12979860 was identified by in-house, developed, high-resolution melting curve analysis. Sequence analysis of the HCV 5’ UTR of HCV RNA was done by 5’ rapid amplification of complementary DNA ends (5’ RACE System, version 2.0), followed by sequencing and was done at baseline for all patients and at time of virological rebound (>1 log10 increase in HCV RNA from nadir) and with HCV RNA of more than 20 000 IU/mL (appendix). In patients with HCV RNA levels below the detection limit for sequence analysis at time of rebound, a retest sample was used or an additional follow-up sample was collected.

statistical analysis All patients were analysed in the treatment group for which they were originally allocated after randomisation. For the analysis of safety and tolerability, and antiviral effect of RG-101 all patients were included. Patients dosed with placebo were excluded for the pharmacokinetic analysis. All statistics were descriptive, and calculated for each treatment group and included measurement of HCV RNA levels, clinical safety laboratory tests, and pharmacokinetics. All HCV RNA levels were analysed after log10 transformation. Samples with HCV RNA levels below the limit of quantification were set to an arbitrary

184

Chapter 11

HCV RNA level of 7·5 IU/mL and changes in HCV RNA levels as compared with baseline (day 1) were calculated. Statistics were done with SPSS (version 22.0) and GraphPad Prism Software (version 6.0). The trial was registered with EudraCT, number 2013-002978-49.

role of the funding source The study was designed and done by the funder (Regulus Therapeutics) in collaboration with PRA Health Sciences and the principal investigators. PRA Health Sciences and the principal investigators collected the data, monitored study conduct, and analysed the data. The corresponding author had full access to all data in the study and had final responsibility for the decision to submit for publication.

results Screening began on June 4, 2014, with the last patient enrolled on Oct 27, 2014. Of 41 patients screened, 32 were randomly assigned and received the study drug or placebo (figure 1). No protocol violations were reported. Overall, most patients were male (24 [75%] of 32), white (29 [91%] of 32), and treatment naive to interferon-α based therapy (24 [75%] of 32; table 1). Most patients were infected with HCV genotype 1 (n=16), followed by genotype 3 (n=10), and genotype 4 (n=6). Almost half of the patients (n=15) had fibrosis stage F0–F1 (table 1).

No serious adverse events, including deaths, occurred during the main study or the extended follow-up study that were considered to be related to RG-101 administration. Overall, 26 (93%) of the 28 patients dosed with RG-101 reported at least one treatment-related adverse event (table 2). The most common adverse events were fatigue, emotional distress, insomnia, and local injection site reactions. The injection site reactions were mild and self-limiting. One patient, a 46-year-old man, who received a dose of 4 mg/kg RG-101 developed a severe intrahepatic cholestasis, with an alanine aminotransferase level of 219 U/L (ULN 68 U/L), aspartate aminotransferase level of 235 U/L (45 U/L), γ-glutamyl transferase level of 6557 U/L (59 U/L), alkaline phosphatase level of 481 U/L (129 U/L), total bilirubin level of 49 μmol/L (29 μmol/L), and direct bilirubin level of 38 μmol/L (7 μmol/L), most probably as a consequence of alcohol abuse; however, a role for RG-101 cannot be excluded here. This patient was followed up at the outpatient clinic and all laboratory sample abnormalities improved after alcohol intake was stopped.

At week 4, median serum alanine aminotransferase and median serum aspartate aminotransferase levels had normalised (table 3; appendix). An increase in median serum alkaline phosphatase levels was noted in patients in both dosing groups at week 8 (table 3, appendix). Of the 28 patients treated with RG-101, six patients in the 2 mg/kg group (up to 2·0 × ULN) and seven patients in the 4 mg/kg group (up to 3·7 × ULN) had one or more alkaline phosphatase values of more than the ULN (appendix). Median alkaline phosphatase values had not yet returned to baseline values at the time of follow-up (appendix). There was no significant change in median total serum bilirubin levels (appendix). There were no clinically significant findings with respect to clinical safety laboratory tests (including

185


11

clinical chemistry, haematological tests, coagulation, C-reactive protein, and urinalysis), complement activation, renal safety, vital signs, electrocardiograph, or physical examination after dosing with RG-101 (data not shown). A decrease in total plasma cholesterol levels was noted with a median reduction of 1·4 mmol/L (IQR 0·6–1·9) in patients dosed with 2 mg/kg RG-101 and 0·9 mmol/L (0·7–1·3) in patients dosed with 4 mg/kg RG-101 at week 4 compared with baseline, whereas median total cholesterol in patients treated with placebo increased compared with baseline (0·5 mmol/L [0·1–0·8]; appendix). Concentrations of HDL, LDL, and triglycerides all decreased proportionally in patients dosed with RG-101 (data not shown). At week 4, there was no correlation between change in HCV RNA level and total cholesterol level in patients dosed with 2 mg/kg RG-101 and 4 mg/kg RG-101 (R 0·34, p=0·076, appendix).

RG-101 was rapidly absorbed after subcutaneous administration, with a median plasma Tmax of 4 h for patients dosed with either 2 mg/kg or 4 mg/kg (appendix). After peaking, RG-101 clearance from plasma resulted in very low plasma concentrations (<0·17 μg/mL) observed by 24 h after administration (figure 2). Geometric mean Cmax and AUC0–t values for RG-101 increased more than dose proportional when increasing the dose from 2 mg/

figure 1. Flow chart of enrolment and follow-up.

186

Chapter 11

kg to 4 mg/kg (appendix). During the 0–24 h after drug administration time period, only the parent drug (RG-101) was observed in plasma, and less than 3% of the administered dose was eliminated as parent drug or its main active metabolite (RG1649) in the urine (appendix). No readily apparent differences in the plasma exposure of RG-101 were noted between treated patients with HCV who achieved HCV RNA levels below the limit of quantification and those who did not (data not shown).

table 1. Baseline patient characteristics

rG-101 (n=28)

placebo (n=4) total2 mg/kg 4 mg/kg

N = 14 14 4 32Male 13 7 4 24 (75)Age 51 (46-53) 54 (48-58) 55 (52-57) 52 (49-57)Weight 84 (71-98) 83 (69-89) 79 (74-83) 83 (72-92)Ethnicity Caucasian 13 12 4 29 (91) Asian 1 1 0 2 (6) Other 0 1 0 1 (3)IFNL3 CC genotype 3 3 2 8 (25)Prior IFN-based anti-HCV treatment a

None 9 12 3 24 (75) IFN+RBV 0 1 0 1 (3) pegIFN+RBV 2 0 1 3 (9) pegIFN+RBV+PI 3 1 0 4 (13)Baseline HCV RNA level (log 10, IU/mL)

6·12 (5·78-6·75) 6·19 (5·78-6·64) 6·38 (5·83-6·85) 6·16 (5·80-6·68)

HCV genotype 1a 8 4 2 14 (44) 1b 1 1 0 2 (6) 3 4 5 1 10 (31) 4 1 4 1 6 (19)ALT level 60 (41-139) 52 (44-61) 90 (35-232) 52 (39-124)Fibroscan (kPa) 6·8 (5·1-10·8) 7·7 (5·5-10·9) 5·1 (3·4-11·7) 7·5 (5·3-10·7)Fibrosis stageb

F0-F1 7 5 3 15 (47) F1-F2 3 5 0 8 (25) F3-F4 4 3 1 8 (25) F4 0 1 0 1 (3)

Data are given as median (IQR) or as frequency (percentage).a. none: naïve to IFN-alpha based therapy, IFN: interferon, RBV: ribavirin, peg-IFN: pegylated interferon, PI: protease inhibitor. b. Stage of fibrosis was determined by liver elastography (Fibroscan). Fibrosis score: F0-F1 (< 7 kPa), F1-F2 (7 - 8.8 kPa), F2-F3 (8.9 – 9.4 kPa), F3-F4 (9.5 - 14.5 kPa) and F4 (> 14.6 kPa).

187


11

Baseline HCV RNA levels were similar between the RG-101 study groups (table 1).

HCV RNA levels decreased rapidly in both dosing groups with a median decline of 2·63

log10 IU/mL (IQR 1·61–3·48) in patients dosed with 2 mg/kg RG-101 and 3·31 log10 IU/mL

(2·57–4·01) in patients with 4 mg/kg RG-101 at week 1 compared with baseline (appendix).

At week 4, the median reduction in HCV RNA level from baseline was 4·42 log10 IU/mL

(IQR 3·23–5·00) in patients dosed with 2 mg/kg RG-101 and 5·07 log10 IU/mL (4·19–5·35)

in patients dosed with 4 mg/kg RG-101 (figure 3A). Of the 14 patients treated with 2 mg/

kg RG-101, four had a virological rebound at week 8 and the other ten were included

in the extended follow-up study (figure 1, figure 3B). Of patients dosed with 4 mg/kg

RG-101, two patients had a virological rebound at week 8 and the other 12 patients were

included in the extended follow-up study (figure 1, figure 3C). Of those who were included

in the extended follow-up study, 15 of 22 patients had an HCV RNA level below the limit

table 2. Adverse events

main study (n=32)

extended follow-up (n=22)

2 mg/kg 4 mg/kg placebo 2 mg/kg 4 mg/kg

N = 14 14 4 10 12Any adverse event No. of events 49 39 4 10 11 No. of patients with event 13 13 3 4 7Adverse events occurring in >5% patients dosed with RG-101 Fatigue 8 6 0 0 0 Injection site reaction/irritation 4 2 0 0 0 Diarrhoea 2 2 1 0 0 Abdominal pain 1 3 0 0 0 Nausea 1 1 0 1 1 Dry mouth 1 1 0 0 0 Flatulence 0 2 0 0 1 Faeces discoloured 2 0 0 0 0 Insomnia 4 1 0 0 0 Emotional disorder/distress 4 2 0 0 0 Arthralgia 1 2 0 0 0 Myalgia/muscle spasm 2 1 1 3 0 Back pain 3 0 0 1 0 Headache 1 3 0 0 0 Dizziness 2 1 0 0 0 Skin irritation 1 2 0 0 0 Hyperhidrosis 1 1 0 0 0 Photopsia 0 0 0 2 0 Nasopharyngitis 2 1 0 2 2

188

Chapter 11

of quantification at week 8 (HCV genotype 1 [n=6], genotype 3 [n=5], and genotype 4 [n=4]). In the extended follow-up study, 16 of 22 patients had a virological rebound; these occurred at week 12 (n=8), week 16 (n=2), week 20 (n=1), week 28 (n=2), week 36 (n=2), and week 52 (n=1). Three patients had a sustained virological response at week 76 after one dose of RG-101; one patient with HCV genotype 3 and IFNL3 genotype CC who was treated with 2 mg/kg RG-101 and two patients with HCV genotype 4 and IFNL3 genotype CC or CT both treated with 4 mg/kg RG-101 (figures 3B, 3C). The remaining three patients withdrew consent (n=2) or were lost to follow-up by week 76 (n=1), of whom one had undetectable HCV RNA level at week 12 (figure 1).

At baseline, none of the patients had a mutation in any of the miR-122 binding sites located in the 5’ UTR of the HCV genome (S1, S2, or additional base-pair interactions) compared with the reference sequence (data not shown). In 14 of 22 RG-101 treated patients who had a viral rebound, matched samples (at baseline and time of viral rebound) with viral loads of more than 20 000 IU/mL were available. Among these 14 patients, six patients with HCV genotype 1 with viral rebound between weeks 5 and weeks 12 had a resistance associated substitution (RAS; figure 4, appendix). A single RAS, namely C3U, in the 5’ UTR miR-122 binding region was noted in five of these six patients. One of the six patients had viral rebound at week 8 (HCV RNA <20 000 IU/mL) and only a week 38 sample was available for analysis. This week 38 sample contained the C3U RAS and also a mixed population of C and U at position 27 (appendix). Additionally, two patients (HCV genotype 3 and 4) had a viral rebound between weeks 8 and weeks 12, and had HCV variants containing four nucleotide changes, including two RASs (C2G and C3U) and two polymorphisms (G1A and U4A) in the 5’ UTR miR-122 binding region (figure 4, appendix). The remaining six patients (HCV genotype 1 [n=1], 3 [n=4], and 4 [n=1]) had viral rebound between weeks 12 and 36 and had no RASs in the 5’ UTR miR-122 binding regions (figure 4, appendix). An additional polymorphism (G38A) was identified in one

figure 2. Pharmacokinetic profile. Error bars denote IQR. During the first 24 h after dosing, RG-101 was cleared from plasma.

189


11

patient (HCV genotype 4) with viral rebound at week 20. There was good agreement

between population based and deep sequence analyses (data not shown).

dIsCussIon In this study, one dose of 2 mg/kg or 4 mg/kg of RG-101 was well tolerated and no serious

adverse events were noted in patients with chronic hepatitis C. Treatment with RG-101

led to a substantial decrease in HCV RNA levels in all patients with chronic hepatitis C

infected with HCV genotype 1, 3, and 4. Median viral load reductions at week 4 were 4·42

log10 IU/mL for 2 mg/kg treatment and 5·07 log10 IU/mL for 4 mg/kg treatment and HCV

figure 3. Change in HCV RNA levels of patients dosed with RG-101 and placebo. (A) Error bars denote IQR. Median (IQR) change in HCV RNA level from baseline at week 4 compared with patients treated with placebo, 2 mg/kg RG-101, and 4 mg/kg RG-101. HCV RNA levels of individual patients dosed with 2 mg/kg RG-101 and placebo (B) and 4 mg/kg RG-101 and placebo (C). HCV=hepatitis C virus. LLOQ=lower limit of quantification.

(a)

(B)

(C)

190

Chapter 11

RNA levels were undetectable in three patients at week 76 of follow-up after one dose of RG-101.

Other host-targeting agents that have been studied previously in HCV infection include alisporivir, a cyclophilin A inhibitor, and miravirsen. Clinical development of alisporivir was stopped because of concerns about development of pancreatitis. Previous studies with miravirsen showed potential for therapeutic efficacy of a miR-122 inhibitor in patients who were treatment naive with HCV genotype 1. A transient viral load reduction of more than 2 log10 IU/mL occurred in 13 of 27 patients dosed with 3 mg/kg, 5 mg/kg, or 7 mg/kg miravirsen.9 RG-101 is the first in-human tested anti-miR oligonucleotide that is targeted for delivery to hepatocytes with N-acetylgalactosamine, a high-affinity ligand for the asialoglycoprotein receptor that is abundantly expressed on hepatocytes. N-acetylgalactosamine conjugation of an oligonucleotide results in enhanced targeting to hepatocytes and lower exposure in non-hepatic organs.25

In this study, there were no clinically significant findings with respect to the clinical safety laboratory tests described in the Methods. On the basis of clinical and non-clinical data with other antisense oligonucleotides, special attention was given to potential risks of complement activation, increased liver enzymes, activated partial thromboplastin

figure 4. Identification of 5’ UTR mutations associated with virological rebound in patients dosed with RG-101.Time of viral rebound (indicated by the bar) and presence of mutation(s) in 5’ UTR miR-122 binding sites (green bar=no mutation, red bar=mutation) in patients with virological rebound after RG-101 dosing. In patients with HCV RNA levels below the detection limit for sequence analysis at time of rebound, an additional follow-up sample was used for sequence analysis (time of sequencing indicated by diamond). Virological rebound prior to week 16 was associated with the presence of mutation(s) in miR-122 binding regions. The lower detection limit for HCV RNA level for 5’ RACE PCR and subsequent sequencing was 20 000 IU/mL. UTR=untranslated region. RAS=resistance associated substitution. VR=viral rebounder.

191


11

time, C-reactive protein, and renal safety. As expected from earlier studies with miR-122 inhibitors,9,12,14,20 RG-101 administration resulted in a modest and prolonged increase in serum alkaline phosphatase levels and a prolonged decrease in total plasma cholesterol levels in patients with chronic hepatitis C. The alkaline phosphatase gene (ALPL) contains multiple miR-122 seed matches in the 3’ UTR and is considered as a direct target of miR-122. Anti-miR-122 dosing resulted in hepatic ALPL de-repression in animal studies, and elevated serum alkaline phosphatase levels were noted in miR-122 knockout mice.26,27 One patient had transient increases in alkaline phosphatase and γ-glutamyl transferase levels, accompanied with increases in alanine aminotransferase, aspartate aminotransferase, and direct bilirubin levels. A possible relationship with alcohol consumption could not be excluded and future research might be needed to assess the effects of alcohol on treatment with RG-101. miR-122 has a tumour suppressive role and low miR-122 levels have been related to the development of hepatocellular carcinoma.15,26–28 Although the exact long-term risk of hepatocellular carcinoma development in patients with chronic hepatitis C after short-term miR-122 inhibition is unknown, a retrospective follow-up study of patients dosed with miravirsen showed no safety issues.29 An extension study of the current study is ongoing (EudraCT 2016-002069-77) with the intent to follow up patients treated with RG-101 for up to 3 years after dosing to assess long-term safety.

Animal studies have shown that RG-101 is rapidly absorbed into the systemic circulation and cleared from the plasma within 24 h to various tissues that include the liver.30 Once taken up by hepatocytes, RG-101 is efficiently metabolised to its main active metabolite (RG1649), which has a tissue half-life in animals of about 14 days. The findings in patients with chronic hepatitis C were consistent with findings in animal studies showing that RG-101 is stable in plasma and is cleared from plasma within 24 h, which most probably reflects extensive tissue uptake (including liver), rather than extensive metabolism or urinary elimination. The tissue half-life of RG-101 was not assessed in humans, but might be similar to that in animals given the prolonged pharmacodynamic effects (increased alkaline phosphatase levels and decreased cholesterol levels) and prolonged antiviral effect. In this proof-of-concept study, there were three patients with undetectable HCV RNA levels at 76 weeks after administration of one dose of RG-101. According to HCV guidelines, in which a sustained virological response is defined as undetectable HCV RNA levels at 12 weeks or 24 weeks after completion of therapy,3 these patients could be considered HCV cured. However, these results should be interpreted with caution considering the distinct mechanism of action and long tissue half-life of RG-101 compared with HCV direct-acting antivirals. Despite the finding that these three patients had sustained virological responses 76 weeks after a dose of RG-101, RG-101 should probably not be used as monotherapy but should be combined with direct-acting antivirals. An in-vitro study31 showed that combination treatment with anti-miR-122 and direct-acting antiviral had additive or synergistic antiviral effects.

Monotherapy with HCV direct-acting antivirals rapidly selects for resistant viruses, whereas host-targeting agents have a relatively high barrier to resistance compared with most direct-acting antivirals. In this study, we showed that early virological rebound

192

Chapter 11

(<12 weeks) after RG-101 dosing was associated with the emergence of a single RAS (C3U) in patients infected with HCV genotype 1, as has been noted previously in patients dosed with miravirsen.30 Additionally, we showed emergence of four nucleotide changes including two RASs (C2G and C3U) and two polymorphisms (G1A and U4A) in the 5’ UTR in patients with HCV genotype 3 or 4 with viral rebound after RG-101 dosing. Because the RASs appear in functional miR-122 binding regions,23 the emergence of these viral variants might alter the dependency of the virus on miR-122, however, this premise needs to be elucidated in future studies. A previous study32 showed that the C3U RAS in the 5’ UTR of HCV confers reduced susceptibility to miR-122 antagonists but has no effect on the susceptibility to various direct-acting antivirals, thus providing support for testing regimens combining RG-101 and one or more direct-acting antivirals. In patients with a virological rebound more than 12 weeks after RG-101 dosing, we observed no emergence of RASs in miR-122 binding regions, and viral rebound could probably result from low (subtherapeutic) RG-101 drug concentrations in the liver. A limitation of this study is that women of childbearing potential were excluded from participation, and therefore the study results might not be representative for this group of patients. Another limitation is that patients and investigators were unblinded for treatment allocation after week 8 of the study. This process might have influenced the adverse event reporting in the extended follow-up study.

The N-acetylgalactosamine linking to other oligonucleotides may have great potential for other liver diseases, such as chronic hepatitis B virus infection.33 The combination of a highly potent HCV direct-acting antiviral with RG-101 could potentially shorten HCV treatment duration to 4–6 weeks or less, which is of particular interest in view of the high costs of HCV direct-acting antivirals and potential patient non-adherence. An alternative application of miR-122 inhibitors, given the distinct mechanism of action, could be found in difficult to treat patients who previously did not respond to direct-acting antiviral therapy. In conclusion, RG-101 administered in 2 mg/kg or 4 mg/kg in a single subcutaneous injection to patients with chronic HCV infection genotype 1, 3, and 4 appeared safe, provided dose-dependent plasma exposure of the drug, and resulted in significant viral load reduction in all treated patients within 4 weeks, and sustained virological response in three patients at week 76 of follow-up. Phase 2 studies are underway (eg, EudraCT 2015-001535-21) to establish the efficacy of a combination of RG-101 with direct-acting antivirals to potentially shorten treatment duration.

193


11

referenCes 1. Lindenbach BD, Rice CM. Unravelling

hepatitis C virus replication from genome to function. Nature 2005; 436: 933–38.

2. Hajarizadeh B, Grebely J, Dore GJ. Epidemiology and natural history of HCV infection. Nat Rev Gastroenterol Hepatol 2013; 10: 553–62.

3. European Association for the Study of the Liver. EASL recommendations on treatment of hepatitis C 2015. J Hepatol 2015; 63: 199–236.

4. van der Meer A, Veldt BJ, Feld JJ, et al. Association between sustained virological response and all-cause mortality among patients with chronic hepatitis C and advanced hepatic fibrosis. JAMA 2012; 308: 2584–93.

5. Morgan RL, Baack B, Smith BD, Yartel A, Pitasi M, Falck-Ytter Y. Eradication of hepatitis C virus infection and the development of hepatocellular carcinoma: a meta-analysis of observational studies. Ann Intern Med 2013; 158: 329–37

6. Kohli A, Kattakuzhy S, Sidharthan S, et al. Four-week direct-acting antiviral regimens in noncirrhotic patients with hepatitis C virus genotype 1 infection: an open-label, nonrandomized trial. Ann Intern Med 2015; 163: 899–907.

7. Kohli A, Osinusi A, Sims Z, et al. Virological response after 6 week triple-drug regimens for hepatitis C: a proof-of-concept phase 2A cohort study. Lancet 2015; 385: 1107–13.

8. Flisiak R, Jaroszewicz J, Flisiak I, Łapiński T. Update on alisporivir in treatment of viral hepatitis C. Expert Opin Investig Drugs 2012; 21: 375–82.

9. Janssen HLA, Reesink HW, Lawitz EJ, et al. Treatment of HCV infection by targeting microRNA. N Engl J Med 2013; 368: 1685–94.

10. Li Y-P, Gottwein JM, Scheel TK, Jensen TB, Bukh J. MicroRNA-122 antagonism against hepatitis C virus genotypes 1-6 and reduced

efficacy by host RNA insertion or mutations in the HCV 5’ UTR. Proc Natl Acad Sci USA 2011; 108: 4991–96.

11. Pawlotsky J-M. What are the pros and cons of the use of host-targeted agents against hepatitis C? Antiviral Res 2014; 105: 22–25.

12. Esau C, Davis S, Murray SF, et al. miR-122 regulation of lipid metabolism revealed by in vivo antisense targeting. Cell Metab 2006; 3: 87–98.

13. Krützfeldt J, Rajewsky N, Braich R, et al. Silencing of microRNAs in vivo with “antagomirs”. Nature 2005; 438: 685–89.

14. Elmén J, Lindow M, Schütz S, et al. LNA-mediated microRNA silencing in non-human primates. Nature 2008; 452: 896–99.

15. Bai S, Nasser MW, Wang B, et al. MicroRNA-122 inhibits tumorigenic properties of hepatocellular carcinoma cells and sensitizes these cells to sorafenib. J Biol Chem 2009; 284: 32015–27.

16. Zeng C, Wang R, Li D, et al. A novel GSK-3 beta-C/EBP alpha-miR-122-insulin-like growth factor 1 receptor regulatory circuitry in human hepatocellular carcinoma. Hepatology 2010; 52: 1702–12.

17. Nassirpour R, Mehta PP, Yin M-J. miR-122 regulates tumorigenesis in hepatocellular carcinoma by targeting AKT3. PLoS One 2013; 8: e79655.

18. Xu J, Zhu X, Wu L, et al. MicroRNA-122 suppresses cell proliferation and induces cell apoptosis in hepatocellular carcinoma by directly targeting Wnt/β-catenin pathway. Liver Int 2012; 32: 752–60.

19. Jopling CL, Yi M, Lancaster AM, Lemon SM, Sarnow P. Modulation of hepatitis C virus RNA abundance by a liver-specific MicroRNA. Science 2005; 309: 1577–81.

20. Lanford RE, Hildebrandt-Eriksen ES, Petri A, et al. Therapeutic silencing of microRNA-

194

Chapter 11

122 in primates with chronic hepatitis C virus infection. Science 2010; 327: 198–201.

21. Sedano CD, Sarnow P. Hepatitis C virus subverts liver-specific miR-122 to protect the viral genome from exoribonuclease Xrn2. Cell Host Microbe 2014; 16: 257–64.

22. Li Y, Masaki T, Yamane D, McGivern DR, Lemon SM. Competing and noncompeting activities of miR-122 and the 5’ exonuclease Xrn1 in regulation of hepatitis C virus replication. Proc Natl Acad Sci USA 2013; 110: 1881–86.

23. Machlin ES, Sarnow P, Sagan SM. Masking the 5’ terminal nucleotides of the hepatitis C virus genome by an unconventional microRNA-target RNA complex. Proc Natl Acad Sci USA 2011; 108: 3193–98.

24. Masaki T, Arend KC, Li Y, et al. miR-122 stimulates hepatitis C virus RNA synthesis by altering the balance of viral rnas engaged in replication versus translation. Cell Host Microbe 2015; 17: 217–28.

25. Prakash TP, Graham MJ, Yu J, et al. Targeted delivery of antisense oligonucleotides to hepatocytes using triantennary N-acetyl galactosamine improves potency 10-fold in mice. Nucleic Acids Res 2014; 42: 8796–807.

26. Tsai W-C, Hsu S-D, Hsu C-S, et al. MicroRNA-122 plays a critical role in liver homeostasis and hepatocarcinogenesis. J Clin Invest 2012; 122: 2884–97.

27. Hsu S, Wang B, Kota J, et al. Essential metabolic, anti-inflammatory, and anti-tumorigenic functions of miR-122 in liver. J Clin Invest 2012; 122: 2871–83.

28. Coulouarn C, Factor VM, Andersen JB, Durkin ME, Thorgeirsson SS. Loss of miR-122 expression in liver cancer correlates with suppression of the hepatic phenotype and gain of metastatic properties. Oncogene 2009; 28: 3526–36.

29. van der Ree MH, van der Meer AJ, de Bruijne J, et al. Long-term safety and efficacy of microRNA-targeted therapy in chronic hepatitis C patients. Antiviral Res 2014; 111: 53–59.

30. Neben S, Liu K, Berman C, et al. P0907: pharmacokinetics, pharmacodynamics, and toxicity profile of RG-101, a novel GalNAc-conjugated hepatocyte-targeting inhibitor of microrna-122, in rodents and cynomolgus monkeys. J Hepatol 2015; 62: S684–85.

31. Liu F, Shimakami T, Murai K, et al. Efficient suppression of hepatitis C virus replication by combination treatment with miR-122 antagonism and direct-acting antivirals in cell culture systems. Sci Rep 2016; 6: 30939.

32. Ottosen S, Parsley TB, Yang L, et al. In vitro antiviral activity and preclinical and clinical resistance profile of miravirsen, a novel anti-hepatitis C virus therapeutic targeting the human factor miR-122. Antimicrob Agents Chemother 2015; 59: 599–608.

33. Sepp-Lorenzino L, Sprague AG, Mayo T, et al. Parallel 4: hepatitis B: novel treatments and treatment targets: 36—GalNAc-siRNA conjugate ALN-HBV targets a highly conserved, pan-genotypic X-orf viral site and mediates profound and durable HBsAg silencing in vitro and in vivo. Hepatology 2015; 62: 222A–25A.

aCKnowledGementsWe thank Hadassa Heidsieck, Jeltje Helder, and Martine Peters for their invaluable

support and care in this study. Furthermore, we thank Bert Baak, Louis Jansen, Andre van Vliet, Joanna Udo de Haes, and Peter Jansen for their support and advice.

fundInGRegulus Therapeutics, Inc.

195


11

supplementary materIals

supplementary table 1. Pharmacokinetics RG-101 measured in plasma and urine

2 mg/kg rG-101 4 mg/kg rG-101

N = 14 14Plasma tmax (h)* 4.0, (2.0 – 4.0), (2.0 - 16.0) 4.0, (2.0 – 4.0), (0.50 - 10.0) Cmax (µg/mL) 0.73 [57.9%], (0.25 - 2.02) 3.03 [85.5%], (0.88 - 16.3) AUC0-t (µg∙h/mL) 6.5 [28.4%], (3.5 - 11.7) 28.5 [35.0%], (14.8 - 43.6) t½ (h) 17.0 [214.9%], (6.6 – 76.1)** 7.2 [64.2%], (4.4 – 18.8)***Urine fe0-24 (%)**** (0 – 0) (0 - 0) (0 – 0), (0 - 2.07)

For Cmax, AUC0-t and t½, the geometric mean (geometric coefficient of variation [CV%]) and (minimum-maximum) are presented. * For tmax and fe0-24, the median (IQR) and (minimum-maximum) is presented. **n=3; estimated t½ values at the higher end of the reported range were not precisely estimated and thus should be interpreted with caution. ***n=6. ****: %fe0–24: percentage of the dose excreted in urine (as parent drug or major metabolite) during the first 24 h after receiving the treatment dose

supplementary figure 1. (1A) In each dose group (2 or 4 mg/kg), patients were randomised to receive either RG-101 or placebo (ratio 7:1). (1B) Patients received a single subcutaneous injection of RG-101 or placebo at day 1 (=baseline). All patient cohorts were followed 8 weeks after randomisation. Patients with no virological rebound at week 8 were included in the extended follow-up study in which patients were followed up to 76 weeks after dosing.

Chapter 4

Supplementary Figures Supplementary Figure 1 (1A) In each dose group (2 or 4 mg/kg), patients were randomised to receive either RG‐101 or placebo (ratio 7:1). (1B) Patients received a single subcutaneous injection of RG‐101 or placebo at day 1 (=baseline). All patient cohorts were followed 8 weeks after randomisation. Patients with no virological rebound at week 8 were included in the extended follow‐up study in which patients were followed up to 76 weeks after dosing.

196

Chapter 11

Chapter 4

Supplementary Figure 2 Change in laboratory values in patients dosed with 2 and 4 mg/kg RG‐101 and placebo. (2A): ALT level, (2B) AST level, (2C) alkaline phosphatase level, (2D): total bilirubin level.

2A 2B

2C 2D

supplementary figure 2. Change in laboratory values in patients dosed with 2 and 4 mg/kg RG-101 and placebo. (2A): ALT level, (2B) AST level, (2C) alkaline phosphatase level, (2D): total bilirubin level.

(2a) (2B)

(2C) (2d)

197


11

Chapter 4

Supplementary Figure 3 Alkaline phosphatase levels of individual patients dosed with 2 mg/kg RG‐101 (3A) and 4 mg/kg RG‐101 (3B) throughout the entire study period.

3A.

3B.

supplementary figure 3. Alkaline phosphatase levels of individual patients dosed with 2 mg/kg RG-101 (3A) and 4 mg/kg RG-101 (3B) throughout the entire study period.

(3a)

(3B)

198

Chapter 11

Chapter 4

Supplementary Figure 4 (4A) Change in total cholesterol level in patients dosed with RG‐101 and placebo. (4B) No significant correlation between change in total cholesterol level and change in HCV RNA level in patients dosed with RG‐101. 4A.

4B.

supplementary figure 4. (4A) Change in total cholesterol level in patients dosed with RG-101 and placebo. (4B) No significant correlation between change in total cholesterol level and change in HCV RNA level in patients dosed with RG-101.

(4a)

(4B)

199


11

supplementary figure 5. Change in HCV RNA levels within the first 7 days in patients dosed with 2 mg/kg RG-101 and placebo (5A) and 4 mg/kg RG-101 and placebo (5B).

(5a)

(5B)

200

Chapter 11

Chapter 4

Supplementary Figure 6 HCV RNA levels of patients with matched sample (at baseline and viral rebound) available for sequence analysis. In patients with HCV RNA levels below the detection limit for sequence analysis (20,000 IU/mL) at time of viral rebound (indicated by a solid square), an additional follow‐up sample was used for sequence analysis. (6A) HCV RNA levels of individual patients with viral rebound and HCV genotype 1. Identification of C3U RAV in miR‐122 binding sites (open dots with X), C3U RAV in combination with C27U/C RAV (open squares with X), or wild‐type sequences (solid dots) are indicated. (6B) HCV RNA levels of individual patients with viral rebound and HCV genotype 3 and 4. Identification of C2G and C3U RAVs in miR‐122 binding sites (open dots with X) or wild‐type sequences (solid dots) are indicated.

6A

6B.

supplementary figure 6. HCV RNA levels of patients with matched sample (at baseline and viral rebound) available for sequence analysis. In patients with HCV RNA levels below the detection limit for sequence analysis (20,000 IU/mL) at time of viral rebound (indicated by a solid square), an additional follow-up sample was used for sequence analysis. (6A) HCV RNA levels of individual patients with viral rebound and HCV genotype 1. Identification of C3U RAV in miR-122 binding sites (open dots with X), C3U RAV in combination with C27U/C RAV (open squares with X), or wild-type sequences (solid dots) are indicated. (6B) HCV RNA levels of individual patients with viral rebound and HCV genotype 3 and 4. Identification of C2G and C3U RAVs in miR-122 binding sites (open dots with X) or wild-type sequences (solid dots) are indicated.

(6a)

(6B)

201


11

supplementary figure 7. Sequence analysis of the 5’UTR (nucleotides 1 to 270) were performed by 5’RACE from 14 patients who experienced a virological rebound (VR) and whom matched samples were available for analysis. Sequence data of nucleotide 1-50 are shown for each patient (VR 1-14) aligned to the HCV genotype specific consensus sequence (GT 1, GT 3 and GT 4). miR-122 binding regions are indicated by the boxes (S1: seed site 1, S2: seed site 2). 5’UTR binding site mutations are shown in red, and viral polymorphisms are shown in orange. Y stands for mixed population of C and U.

Immune phenotype and funCtIon of nK and t Cells In ChronIC hepatItIs C patIents who

reCeIved a sInGle dose of antI-mIr-122, rG-101

F. Stelma, M.H. van der Ree, M.J. Sinnige, A. Brown, L. Swadling, J.M.L. de Vree, S.B. Willemse, M. van der Valk, P. Grint, S. Neben,

P. Klenerman, E. Barnes, N.A. Kootstra, H.W. Reesink

Hepatology. 2017;66:57-68.

aBstraCtMicroRNA-122 (miR-122) is an important host factor for the hepatitis C (HCV) virus. Treatment with RG-101, a GalNAc conjugated anti-miR-122 oligonucleotide, resulted in a significant viral load reduction in patients with chronic hepatitis C (CHC) infection. Here, we analyzed the effects of RG-101 therapy on antiviral immunity. 32 CHC patients HCV genotype 1, 3 and 4 received a single subcutaneous administration with RG-101 at 2 mg/kg (n=14), 4 mg/kg (n=14) or placebo (n=2 per dosing group). Plasma and PBMCs were collected at multiple time points and comprehensive immunological analyses were performed. Following RG-101 administration, HCV RNA declined in all patients (mean decline at week 2: 3.27 log10 IU/mL). At week 8 HCV RNA was undetectable in 15/28 patients. Plasma IP-10 levels declined significantly upon dosing with RG-101. Furthermore, the frequency of NK cells increased, the proportion of NK cells expressing activating receptors normalized and NK cell IFN-γ production decreased, after RG-101 dosing. Functional HCV-specific IFN-γ-T cell responses did not significantly change in patients who had undetectable HCV RNA levels by week 8 post RG-101 injection. No increase in the magnitude of HCV-specific T cell responses was observed at later time points, including 3 patients who were HCV RNA negative 76 weeks post dosing.

ConclusionsDosing with RG-101 is associated with a restoration of NK cell proportions and a decrease of NK cells expressing activation receptors. However, the magnitude and functionality of ex vivo HCV-specific T cell responses did not increase following RG-101 injection. Our data suggests that NK cells, but not HCV adaptive immunity may contribute to HCV viral control following RG-101 therapy.

205

NK AND T CELL FUNCTION AFTER RG-101 TREATMENT

12

BaCKGroundChronic hepatitis C (CHC) virus infection is a global health problem. Patients with CHC infection are at increased risk of developing liver-related complications such as hepatocellular carcinoma and cirrhosis.1 Successful HCV treatment reduces the risk of complications and improves survival.2,3 Recently, direct acting antivirals (DAAs) have become available, which can directly interfere with the HCV viral machinery. These DAAs have drastically changed the field of CHC treatment, by achieving sustained virological response (SVR) in a high proportion of patients.4 With such high cure rates for HCV, the next challenge of research needs to focus on vaccine development and optimizing current treatment regimens and delivering them cost effectively.5 The cost of DAAs is considerable and combining DAAs with compounds that have a different mechanism of action may reduce both duration and cost of treatment. An alternative therapeutic option lies in inhibition of an important host factor utilised by HCV; microRNA-122 (miR-122).

miR-122 is a highly conserved, liver-specific micro-RNA which has important functions in the regulation of cholesterol and fatty acid synthesis.6 In addition, miR-122 can bind to the HCV genome and thereby promote virus replication.7,8 Targeting this host factor with an antisense oligonucleotide has been proven effective in inhibiting HCV across several genotypes.9–11 A single subcutaneous dose of the GalNAc-conjugated anti-miR-122 oligonucleotide RG-101 resulted in substantial decreases in HCV RNA in all treated patients and HCV RNA negativity for at least 76 weeks in 3 patients.9

The exact mechanism of HCV inhibition by RG-101 is not yet known. As binding of miR-122 protects HCV RNA from degradation by exonucleases, blocking miR-122 could uncover the genome to these innate defence pathways.(8,12–14) Furthermore, HCV RNA replication could be disturbed by knocking down miR-122.14,15 Inhibition of HCV replication may lead to natural killer (NK) cell normalisation.16,17 Furthermore, a reduction in viral antigen expression could restore HCV-specific T cell responses, and thereby contribute to viral clearance.

Since former treatment options all included interferon-alfa, an immune modulator, the effects of new therapies on the immune system are substantially different. Exogenous interferon-alfa treatment of patients with CHC enhanced interferon stimulated gene (ISG) expression and activated NK cells, but did not lead to recovery of the T cell compartment.18–20 The inhibition of host factor miR-122 by RG-101 acts through a different mechanism and could potentially restore antiviral immunity.

The aim of this study was to investigate whether treatment with RG-101 would change important immune effectors in CHC infection and whether restored antiviral immunity could play a role in long term virologic impact observed in patients treated with a single dose of RG-101.

patIents and methodspatientsWe included 32 CHC patients at two sites in The Netherlands (Academic Medical Center Amsterdam and University Medical Center Groningen)(table 1). Males and post-

206

Chapter 12

menopausal females with chronic HCV genotype 1, 3 or 4 infection were enrolled. Patients were treatment naïve or had previously experienced a virological relapse after interferon-based therapy. Patients with co-infection (hepatitis B virus or human immunodeficiency virus infection), evidence of decompensated liver disease, or a history of HCC were excluded. The study was approved by the regulatory authority and the independent ethics committee at each participating site. All patients gave written informed consent, and the study was conducted in accordance with the Declaration of Helsinki, Good Clinical Practice guidelines, and local regulatory requirements. All authors had access to the study data and reviewed and approved the final manuscript.

studyIn this randomised, double-blind, placebo-controlled, phase 1b trial,21 patients received a single subcutaneous injection of RG-101, a GalNAc-conjugated oligonucleotide antagonizing miR-122 (EudraCT number 2013-002978-49). Dosage was 2 mg/kg (n=14) or 4 mg/kg (n=14), and 2 patients in each group received placebo. The initial follow-up was 8 weeks after dosing. At week 8, only patients with > 2 log 10 decrease from baseline, and < 1 log 10 increase in HCV RNA level from nadir were included in an extended follow-up

table 1. Baseline characteristics9

rG-101 (n=28)

placebo (n=4) total2 mg/kg 4 mg/kg

N = 14 14 4 32Male (n) 13 7 4 24 (75)Age (years) 51 (46-53) 54 (48-58) 55 (52-57) 52 (49-57)Weight (kg) 84 (71-98) 83 (69-89) 79 (74-83) 83 (72-92)Ethnicity (n) Caucasian 13 12 4 29 (91) Asian 1 1 0 2 (6) Other 0 1 0 1 (3)IFNL3 CC genotype (n) 3 3 2 8 (25)IFN naïvea (n) 9 12 3 24 (75)Baseline HCV RNA level (log10, IU/mL)

6·12 (5·78-6·75) 6·19 (5·78-6·64) 6·38 (5·83-6·85) 6·16 (5·80-6·68)

HCV genotype (n) 1a 8 4 2 14 (44) 1b 1 1 0 2 (6) 3 4 5 1 10 (31) 4 1 4 1 6 (19)ALT level (U/L) 60 (41-139) 52 (44-61) 90 (35-232) 52 (39-124)Fibroscan (kPa) 6·8 (5·1-10·8) 7·7 (5·5-10·9) 5·1 (3·4-11·7) 7·5 (5·3-10·7)HLA-A2+ (n) 4 5 2 11 (34)

Data are given as median (IQR) or as frequency (percentage). a. naïve to IFN-alpha based therapy.

207


12

study (n=22, including n=10 2 mg/kg and n= 12 4 mg/kg ), of these, 15 patients (n=6 2 mg/kg and n= 9 4 mg/kg ) had HCV RNA levels below the lower limit of quantification (< LLOQ) at week 8. Patients who did not meet above criteria were excluded from the study at week 8 (n=10, including placebo treated patients). The extended follow-up lasted until week 76 (supplementary figure 1). If patients had a virological rebound (defined as > 1 log 10 increase) during follow-up, a retest was performed, after which patients were excluded from the study. In the extended follow-up, rebounds occurred at week 12 (n=8), week 16 (n=2), week 20 (n=1), week 28 (n=2), week 36 (n=2) and week 52 (n=1) and 3 patients were lost to follow-up (supplementary figure 2). At week 76, 3 patients had undetectable HCV RNA levels after a single dose of RG-101. HCV RNA levels were measured using Roche COBAS AmpliPrep/COBAS Taqman HCV v2.0 assay, with a reported LLOQ of 15 IU/mL.

samplingPlasma samples from CHC patients for cytokine analyses were collected at baseline, day 3, weeks 1, 4 and 8. For comparison, healthy control plasma (n=6) was added (not matched for age, gender or ethnicity). Peripheral blood mononuclear cells (PBMCs) were collected at baseline, weeks 2, 8, 20, 28 and 52, at time of viral rebound or retest (between week 12 to 52) and at end of follow-up (week 76) (supplementary figure 1). PBMCs from healthy blood donors were used as heathy controls (n=13). PBMCs were separated by density gradient and cryopreserved for later analyses.

luminex analysesPlasma cytokine levels were measured using a Luminex 20-plex immunoassay (Affymetrix eBioscience, San Diego, CA, USA). This included IL-12, CCL2, CCL3, CCL4, CD54, Interferon gamma-induced protein 10 (IP-10 or CXCL10), GM-CSF, IFN-alpha, IFN-gamma, IL-1alpha, IL-1 beta, IL-10, IL-13, IL-17A, IL-4, IL-6, IL-8, sCD62E, sCD62P and TNF-alpha. Plasma levels of IL-18 were measured with a DuoSet ELISA (R&D Systems, Minneapolis, MN, USA). Plasma samples from 6 healthy controls were included in the analyses.

flow cytometryPBMCs were thawed and stained for 30 minutes at 4°C with different combinations of fluorescent label-conjugated mouse anti-human monoclonal antibodies (mAbs); CD56 BUV-395, CD27 BUV-373, HLA-DR FITC, CD3 V500, CD16 BV786, CD16 BV421, CD14 PE-CF594, CD19 PE-CF594, CD38 PE-Cy7 (BD, San Jose, CA, USA) NKp46 PerCP-efluor710, CD45RA eFluor 605NC (eBioscience, San Diego, CA, USA) CXCR6 BV-421, CD8 BV-711, CD8 BV-785, NKp30 APC (Biolegend, San Diego, CA, USA) Live/Dead fixable dead cell stain RED (Invitrogen, Life Technologies, Carlsbad, CA, USA) NKG2A PE, (Beckman Coulter, Fullerton, CA, USA) TRAIL/CD253 APC (Miltenyi Biotec, Bergisch Gladbach, Germany).

For intracellular staining, cells were fixed after surface staining, permeabilized (FoxP3/Transcription Factor Staining buffer set, eBioscience, San Diego, CA, USA), and stained

208

Chapter 12

with mAbs; Perforin FITC (BD, San Jose, CA, USA) Eomes PerCP-efluor710, T-bet PE-Cy7 (eBioscience, San Diego, CA, USA) Ki67 BV711, (Biolegend, San Diego, CA, USA) Granzyme B PE (Sanquin, Amsterdam, The Netherlands).

Ex vivo CD8+ T cell frequencies were determined in HLA-A2 positive patients (n=11 patients, genotype 1a (n=6), 3a (n=3), 4a (n=1) and 4d (n=1)) by HLA-class I multimers. Multimers were loaded with dominant epitopes NS31073-1081 (CINGVCWTV) (APC) (Sanquin, Amsterdam, The Netherlands), NS41406-1415 (KLVALGINAV)(PE) and NS5B2594-2602 (ALYDVVTKL)(FITC) (Immudex, Copenhagen, Denmark). Gating was based on the background level observed in healthy non-HCV infected controls (Supplementary Figure 3). As control, we stained for CMV-pp65495–504 (NLVPMVATV) and EBV BMLF-1259-267 (GLCTLVAML) tetramers (Sanquin, Amsterdam, The Netherlands). All measurements were performed on an LSR Fortessa cytometer (BD Biosciences, San Jose, CA, USA) and analyzed by FlowJoMacV9.7.5 software.

nK cell functionPatients were selected when samples were available for day 1 as well as week 8 (n=13), and healthy controls (n=13) were added as controls. PBMC were thawed and incubated overnight with IL-12 and IL-15 or no stimulus in the presence of CD107a FITC (eBioscience, San Diego, CA, USA). For the last 3 hours, monensin and brefeldin A were added, after which cells were stained with live/dead stain, CD3 V500 and CD56 BUV-395 as above. After fixing the cells, intracellular staining was performed (IFN-γ BV421, TNF-α AF700, MIP-1β Pe-Cy7 BD, San Jose, CA, USA) and measurements were done on an LSR Fortessa. The proportion of IL-12/IL-15 activated cells was calculated by subtracting the number of positive cells in the unstimulated condition.

Ifn-γ-elIspot assays IFN-γ-ELISpot assays were performed ex vivo in duplicate at 2x105 PBMCs/well. Thawed PBMC were rested overnight (37°C + CO2) and were stimulated with panels of 15 mer peptides that overlapped by 11 amino acids corresponding to HCV genotypes 1a, 1b, 3a or 4a (described in Barnes et al. 2012 and Kelly et al. 22,23). Patient PBMCs were stimulated with the panel of peptides that matched their own viral genotype and subtype (where possible). The peptides were arranged into 10 pools corresponding to core, E1, E2, P7&NS2, NS3p, NS3h, NS4, NS5A, NS5B I, and NS5B II. Each peptide was used at a final concentration of 3µg/ml. Internal controls were: dimethyl sulfoxide (DMSO) (Sigma-Aldrich, UK) as a negative control and concanavalin A (Sigma-Aldrich, UK) as a positive control. Other antigens used were a pool of MHC class 1 restricted epitopes of influenza A, EBV and CMV (BEI Resources, Manassas, VA, USA), and a lysate of CMV infected cells (Virusys Corp, Taneytown, MD, USA). Spot forming units (SFU) were calculated per 106 PBMC and background levels were subtracted. Positive responses were defined as (i) the mean of responses to a pool minus background being greater than 48 SFU /106 PBMC and (ii) the mean of responses to a well exceeding 3× background. The background level

209


12

of 48 SFU /106 PBMC per pool was determined previously in 74 healthy controls, which was the mean + 3 SD 22.

statistical analyses For differences between groups, the T-test (normal distribution) or Mann–Whitney test (non-normal distribution) was used. For longitudinal analysis in individual patients the Wilcoxon signed rank test was used. P values <0.05 were considered statistically significant. GraphPad Prism version 6.07 for Windows (GraphPad Software, La Jolla, CA, USA) was used for analyses.

resultsIp-10 levels decrease in patients dosed with rG-101HCV RNA levels declined in all patients dosed with RG-101 (n=28). At week 4, the mean reduction was 4.1 log10 IU/ml in patients dosed with 2 mg/kg and 4.8 log10 IU/ml in patients dosed with 4 mg/kg, as compared to 0.0 log10 IU/ml in placebo treated patients (figure 1a). Plasma IP-10 and IL-18 levels were significantly higher in CHC patients at baseline as compared to healthy controls (for IP-10: median 58.1 pg/ml and 19.3 pg/ml, p<0.0001; for IL-18: 322.1 pg/ml and 140.9 pg/ml , p<0.0001, figure 1B) and no significant differences were observed within the 3 groups of CHC patients at baseline (placebo, 2 mg/kg and 4 mg/kg; data not shown).

IP-10 levels decreased significantly in patients dosed with RG-101 (figure 1C). At week 1, the median decline in IP-10 levels was 23.9 pg/mL in 2mg/kg dosed patients, 42.3 pg/mL in 4 mg/kg treated patients (p=0.02 and p=0.007 compared to placebo, respectively), and 1.4 pg/mL in placebo treated patients. IL-18 levels however, increased after dosing with RG-101 (Supplementary Figure 4). 15 of 28 patients had undetectable HCV RNA 8 weeks post-dosing, and IP-10 levels significantly decreased in both of these groups (Figure 1D). However, IL-18 levels only significantly increased in the group of patients that had HCV RNA levels < LLOQ at week 8 (Figure 1D). None of the other measured cytokines or chemokines showed a significant change upon dosing with RG-101, including interferon-γ, interferon-α, tumor necrosis factor-α, and IL12p70(supplementary figure 4).

traIl expression decreases on Cd56bright nK cellsAt baseline, CHC patients and healthy controls had comparable frequencies of NK cells in the blood (median 7.4% and 7.0% respectively, p=0.71, figure 2a, full gating in supplementary figure 5a). In patients dosed with RG-101, the proportion of NK cells increased significantly from baseline 7.4% to median 10.1% (p=0.02) at week 8 and 10.3% at follow-up week 12-28 (which include patients with viral rebound defined as >1 log increase in viral load from nadir between week 12 and 28, and patients with HCV RNA < LLOQ at week 28 (n=5)), p=0.02. In placebo-treated patients, no changes were observed in NK cell proportions during follow-up (figure 2a). Furthermore, no changes in lymphocyte count were observed in the blood (supplementary figure 5B). Patients who had HCV RNA

210

Chapter 12

levels < LLOQ at week 8 had significantly lower proportions of NK cells as compared to

patients who had HCV RNA > LLOQ at week 8 (8.4% and 13.1% respectively, p=0.01

figure 2a). NK cells can be divided into two subsets, depending on their expression of

CD56; these include CD56bright and CD56dim NK cells (figure 2B). Upon dosing, CD56dim

NK cells significantly increased from median 91.7% to 94.8 % at week 12-28 (p=0.0007),

whereas CD56bright NK cells decreased (median 8.3% at baseline and 4.8% at follow-up

week 12-28, p=0.001, figure 2B). The expression of TNF-related apoptosis inducing

ligand (TRAIL) on CD56bright NK cells was significantly upregulated in CHC patients at

baseline as compared to healthy controls (median 13.3% and 7.3% of CD56bright NK cells

respectively, p=0.001, figure 2C). TRAIL expression significantly decreased as soon as

(a) (B)

(C) (d)

figure 1. (a) Mean change in HCV RNA levels between baseline and week 4 in patients dosed with placebo (black dots), 2 mg/kg (blue dots) and 4 mg/kg (red dots) RG-101.9 (B) Plasma IP-10 and IL-18 levels in healthy controls (HC) and CHC patients (HCV) at baseline (BL), bars indicate median. (C) Change in plasma IP-10 levels (median) in patients dosed with placebo (black dots), 2 mg/kg (blue dots) and 4 mg/kg (red dots) RG-101, level of significant difference indicated between RG-101 dosed (2 mg/kg and 4 mg/kg) and placebo dosed patients. (d) Plasma IP-10 and IL-18 levels at baseline and week 8 in patients with HCV RNA levels < LLOQ and > LLOQ at week 8.. Statistical testing; Mann–Whitney test (B,C) and Wilcoxon matched pairs test (D).

211


12

week 2 in patients upon dosing with RG-101 (from median 13.3% at baseline to 6.4% of CD56bright NK cells at week 2, p<0.0001, figure 2C,d). No changes in TRAIL expression were observed in placebo treated patients (not shown). No difference in TRAIL expression was observed in patients who had HCV RNA levels < LLOQ versus patients who had HCV RNA levels > LLOQ at week 8 (figure 2C green and red dots).

Decreased expression of NK cell activation receptorsIn CHC patients, the expression of Fcγ-receptor CD16 on NK cells was significantly higher as compared to healthy controls (median 80.8% and 74.9%, respectively, p=0.04), and this decreased after dosing with RG-101 (80.8% at baseline to 77.0% at week 8, p= 0.0019, figure 3a). Furthermore, the expression of activating receptors NKp30 and NKp46 on NK cells decreased in patients

(a) (B)

(C) (d)

figure 2. (a) Proportion of NK cells as a percentage of total lymphocytes in healthy controls (HC) and CHC patients dosed with RG-101 at baseline (bl), week (wk) 2, 8 and 12-28 (including patients with viral rebound between week 12 and 28 and patients with HCV RNA < LLOQ at week 28). Data is shown grouped in patients who are HCV RNA positive (in red) or have HCV RNA levels < LLOQ (in green), as well as in patients receiving placebo (in grey), bars indicate median. (B) Proportion of CD56bright and CD56dim cells as a percentage of total NK cells in RG-101 dosed patients at baseline(bl), week (wk) 2, 8, and 12-28, bars indicate median. Representative dot plot showing NK cell gating in total lymphocyte population. CD56bright and CD56dim NK cell gates are shown (C) Percentage of TRAIL+ cells within CD56bright NK cells in healthy controls and patients dosed with RG-101, bars indicate median. The same groups as in (A) are used. (d) Representative dot plot showing TRAIL gating, percentages of TRAIL positive cells within the CD56bright NK cell population (black) are shown. CD56dim NK cells are shown in grey for comparison. Statistical testing; Mann–Whitney test (A,C) and Wilcoxon matched pairs test (B,C).

212

Chapter 12

(a) (B) (C)

(d) (e) (f)

figure 3. Proportion of (a) CD16, (B) NKp30, (C) NKp46, (d) T-bet, (e) Eomes and (f) Ki67 in healthy controls (HC) as well as in CHC patients dosed with RG-101 (n=28) at baseline (bl), week (wk) 2 and 8, bars indicate median. Statistical testing; Mann–Whitney test and Wilcoxon matched pairs test (A-F). Patients who have HCV RNA levels < LLOQ are shown in green dots, patients with HCV RNA levels > LLOQ are in red dots. Statistical testing; Mann–Whitney test (A-F, HC vs. CHC) and Wilcoxon matched pairs test (A-F).

who were dosed with RG-101 (NKp30; 75.0% at baseline to 62.8% at week 8, p=0.01 and NKp46; 74% at baseline and 67.2% at week 8, p= 0.01, figure 3B,C). The expression of T-box transcription factors T-bet and Eomes decreased upon dosing with RG-101 (from 94.7% at baseline to 89.2% at week 8 for T-bet p=0.0003 and from 74.6% to 60.8% for Eomes, p=0.001, figure 3d,e), whereas NK cell IFN-γ production significantly decreased (supplementary figure 6). The expression of the cell cycle marker Ki67 in CHC patients was 5.9% at baseline and significantly decreased at week 2 (3.8%, p<0.0001) and week 8 (4.9%, p=0.02)(figure 3f). Other markers expressed on NK cells showed minor or no changes (supplementary figure 7). Changes in T-bet, Eomes and Ki67 expression were also observed in the CD8+ T cell and CD56+ NKT cell compartments (supplementary figure 8). No differences were observed between patients with HCV RNA levels < LLOQ and patients with HCV RNA levels > LLOQ at week 8 (figure 3a-f).

hCv-specific t cells responses did not change in patients dosed with rG-101Next, we investigated whether HCV-specific T cells play a role in the (long term) viral load reduction after RG-101 dosing. The frequency of HCV-specific T cells was measured in all HLA-A2+ patients (n=11, including 2 placebo treated patients). Baseline ex vivo

213


12

(a) (B)

(C) (d)

figure 4. (a) Ex vivo multimer positive cells as a proportion of total CD8+ T cells in RG-101 dosed patients at baseline (bl), week (wk) 2, wk 8 (including n=2 HCV RNA <LLOQ, excluding n=2 due to missed visit). Each dot represents one epitope, a positive response was defined as > 5 events with a lower threshold of 10,000 measured CD8+ T cells, bars indicate median. (B) ELISpot responses at baseline and week 8 in patients dosed with RG-101 (left) and placebo (right). Patients’ PBMCs were stimulated ex vivo with genotype specific HCV-peptide pools. (C) Baseline ELISpot responses in patients dosed with RG-101 with HCV RNA levels > LLOQ (red, left) and < LLOQ (green, right) at week 8, and in patients receiving placebo. (d) IFN-γ-ELISpot responses in patients dosed with RG-101 who had HCV RNA levels > LLOQ (in red) and < LLOQ (in green). Statistical testing; Mann–Whitney test (C) and Wilcoxon matched pairs test (A,B,D).

HCV-specific CD8+ T cell frequencies were low in CHC patients (median 0.009%, range 0% - 0.21%, figure 4a) and did not change upon dosing with RG-101 (median 0.008%, range 0% - 0.19% at week 2 and median 0.007%, range 0% - 0.05% at week 8, p=ns), or in patients who received placebo (supplementary figure 9a). Similarly, CMV and EBV specific T cell proportions did not change upon dosing with RG-101 (supplementary figure 9B). The functional capacity of HCV-specific T cells, as measured by interferon-γ-ELISpot, did not change significantly in patients who were treated with RG-101 as compared to placebo (figure 4B). Baseline IFN-γ T cell responses were low in patients with CHC (with 8/32 patients having positive responses), and there was no

214

Chapter 12

difference in the baseline T cell responses of patients who had HCV RNA levels < LLOQ and patients who were HCV RNA positive at week 8 (p=0.56) (figure 4C). Furthermore, no changes were observed in IFN-γ T cell responses between baseline and week 8 in patients who had HCV RNA levels < LLOQ at week 8 (p=0.13) or HCV RNA levels > LLOQ at week 8 (p=0.88) (figure 4d).

Three patients were tested HCV RNA negative up to 76 weeks after dosing with RG-101. In addition, 5 patients had a late viral rebound at week 28 (n=2), week 36 (n=2) and week 52 (n=1). In none of these patients, an increase in HCV-specific T cells responses was observed (figure 5).

dIsCussIonIn this unique phase 1b study we assessed the effects of dosing with the GalNAc conjugated anti-miR-122 oligonucleotide RG-101 on antiviral immunity in CHC patients for the first time in man. We show that, after a single monotherapy dose of RG-101, as a consequence of HCV RNA decline, IP-10 levels decrease, NK cell proportions increase, and the expression of NK cell activation receptors decreases. Furthermore, NK cell IFN-γ production significantly decreased, and no restoration of ex vivo HCV-specific T cell functionality was observed after viral load decline, nor after long term HCV RNA negativity.

In CHC patients, continuous immune activation is demonstrated by elevated levels of ISGs, such as IP-10 (or CXCL10), and viral load suppression during DAA treatment is associated with a decrease in IP-10 levels.17,24,25 In patients who have viral relapse, IP-10 levels have been shown to increase again.24 In our study, IP-10 levels decreased in all patients who were dosed with RG-101, irrespective of their treatment outcome. We did not observe a subsequent increase in IP-10 levels in patients who had experienced a viral rebound; however, since patients were not followed after rebound, IP-10 levels may have increased at a later time. Alternatively, RG-101 could induce other mechanisms leading to IP-10 decrease, irrespective of viral load decline.26 In in vitro studies, miR-122 has been shown to have pro-inflammatory effects which can be attenuated by inhibition with anti-miRNA-122.27 No other measured cytokines or chemokines were altered in the plasma of patients dosed with RG-101, which is consistent with preclinical data showing RG-101 does not elicit an undesirable systemic immune responses.28

NK cells play an important role in HCV viral infection. NK cells significantly increased in frequency in the blood of patients after dosing with RG-101. Interestingly, at week 8, the frequency of NK cells was higher in patients who were HCV RNA positive as compared to the patients who had HCV RNA levels < LLOQ. Possibly, a rebound in viral load with a subsequent rise in IL-18 levels in these patients could have led to NK cell expansion. After RG-101 dosing, CD56bright NK cells decreased while CD56dim NK cells increased, similar to data from DAA treated patients.17 TRAIL is an important ligand expressed by NK cells which can induce apoptosis in target cells via the TRAIL receptor (TRAIL-R1 or TRAIL-R2). Target cells include virally infected hepatocytes, which have increased TRAIL-receptor expression in CHC infection.29 Furthermore, TRAIL expression on NK cells is upregulated in HCV infected patients30,31 and has been shown to decrease upon viral load reduction

215


12

figure 5. IFN-γ-ELISpot responses in patients who were HCV RNA negative for 76 weeks (upper graphs, patients #1-3) as well as in patients who had HVC RNA levels < LLOQ for 20 or more weeks (lower graphs, patients #4-8). The responses considered above background (>48) are in colour, the other responses are in grey. (I): patient #4 had < 1 log10 increase in HCV RNA at week 20 (PBMC missing), and was negative again at week 24. Rebound was at week 28. (II): patient #5 had > 1 log10 increase in HCV RNA at week 24 and < 1 log10 increase in HCV RNA at retest (week 26, last available PBMCs), rebound was at week 36.

216

Chapter 12

by DAA treatment.17,25 We observed a similar decrease in TRAIL expression on CD56bright NK cells after a single dose of RG-101. In addition, the expression of the Fc-gamma-receptor, CD16, as well as cell cycle marker Ki67 decreased in patients dosed with RG-101, suggesting a reduction of the activated NK cell phenotype. Other markers also indicated decreased NK cell activation upon RG-101 dosing, including NKp30 and NKp46 which have previously been shown to decrease upon DAA treatment17, and T-bet.32 As some of these changes in NK cell markers showed similar changes on bulk CD8+ T cells and CD56+ NK T cells, the observed changes in the NK cell compartment might be the result of overall decrease in immune activation upon a decline in HCV viral load. However, as miR-122 has been implicated in immune regulation, direct effects of miR-122 inhibition could also contribute to the changes in NK cells observed.26,27,33,34

Chronic CHC infection leads to dysfunctional (‘exhausted’) HCV-specific CD8+ T cells. In patients successfully treated with interferon-free therapy, a restoration of the proliferative capacity of these HCV-specific CD8+ T cells has been observed.35 However, whether the function of these HCV-specific CD8+ T cells also improves is unknown, nor whether a restoration of HCV immunity could contribute to HCV viral control. We therefore analyzed HCV-specific CD8+ T cells in RG-101 dosed patients, and we did not observe an increase in the ex vivo magnitude of HCV-specific CD8+ T cells as measured by HLA-A2 restricted multimers. To overcome the limitation of only measuring three specificities with multimers, which could be not optimally genotype-matched23 or could have undergone viral escape, we subsequently analysed T cell functionality by stimulating with overlapping peptide pools and assessing IFN-γ production in ELISpot assays. This also allowed us to measure the functionality of both CD8+ and CD4+ HCV-specific T cells. The magnitude of HCV-specific T cell responses did not change in patients who were HCV RNA negative at week 8, suggesting T cells do not contribute to HCV RNA decline after RG-101 dosing. In line with this, we observed no specific change in the HCV-specific T cell responses in patients who were long term HCV RNA negative, suggesting other mechanisms play a role in the long term viral load reduction after RG-101 dosing.

In conclusion, one dose of anti-miR-122 RG-101 leads to a decrease in HCV RNA levels in all patients and sustained virological response >76 weeks in 3 patients. Dosing with RG-101 does not elicit systemic immune activation. A decrease in IP-10 levels and normalization of NK cell phenotype is observed which are likely the result of HCV RNA decline. Lastly, our data suggests that HCV-specific T cell recovery does not play a role in the decline or long term negativity of HCV RNA in patients dosed with RG-101.

217


12

referenCes1. Westbrook RH, Dusheiko G. Natural

history of hepatitis C. J Hepatol. 2014 Nov;61(1 Suppl):S58-68.

2. van der Meer AJ, Veldt BJ, Feld JJ, Wedemeyer H. Association Between Sustained Virological and Advanced Hepatic Fibrosis. J Am Med Assoc. 2012;308(24):2584–93.

3. Morgan RL, Baack B, Smith BD, Yartel A, Pitasi M, Falck-Ytter Y. Eradication of hepatitis C virus infection and the development of hepatocellular carcinoma: a meta-analysis of observational studies. Ann Intern Med. 2013 Mar 5;158(5 Pt 1):329–37.

4. Pawlotsky JM. New hepatitis C therapies: The toolbox, strategies, and challenges. Gastroenterology. 2014;146(5):1176–92.

5. Lindenbach BD. What’s next for hepatitis C virus research? Hepatology. 2016 May;63(5):1408–10.

6. Esau C, Davis S, Murray SF, Yu XX, Pandey SK, Pear M, et al. miR-122 regulation of lipid metabolism revealed by in vivo antisense targeting. Cell Metab. 2006 Feb;3(2):87–98.

7. Jopling CL, Yi M, Lancaster AM, Lemon SM, Sarnow P. Modulation of hepatitis C virus RNA abundance by a liver-specific MicroRNA. Science. 2005 Sep 2;309(5740):1577–81.

8. Machlin ES, Sarnow P, Sagan SM. Masking the 5’ terminal nucleotides of the hepatitis C virus genome by an unconventional microRNA-target RNA complex. Proc Natl Acad Sci U S A. 2011 Feb 22;108(8):3193–8.

9. van der Ree MH, de Vree JM, Stelma F, Willemse S, van der Valk M, Rietdijk S, et al. Safety, tolerability, and antiviral effect of RG-101 in patients with chronic hepatitis C: a phase 1B, double-blind, randomised controlled trial. Lancet (London, England). 2017 Jan 10;

10. Janssen HL, Reesink HW, Lawitz EJ, Zeuzem S, Rodriguez-Torres M, Patel K, et al. Treatment of HCV infection by targeting microRNA. N Engl J Med. 2013;368(18):1685–94.

11. Li Y-P, Gottwein JM, Scheel TK, Jensen TB, Bukh J. MicroRNA-122 antagonism against hepatitis C virus genotypes 1-6 and reduced efficacy by host RNA insertion or mutations in the HCV 5’ UTR. Proc Natl Acad Sci U S A. 2011 Mar 22;108(12):4991–6.

12. Li Y, Yamane D, Lemon SM. Dissecting the roles of the 5’ exoribonucleases Xrn1 and Xrn2 in restricting hepatitis C virus replication. J Virol. 2015 May 11;89(9):4857–65.

13. Li Y, Masaki T, Yamane D, McGivern DR, Lemon SM. Competing and noncompeting activities of miR-122 and the 5’ exonuclease Xrn1 in regulation of hepatitis C virus replication. Proc Natl Acad Sci U S A. 2013 Jan 29;110(5):1881–6.

14. Thibault PA, Huys A, Amador-Cañizares Y, Gailius JE, Pinel DE, Wilson JA. Regulation of Hepatitis C Virus Genome Replication by Xrn1 and MicroRNA-122 Binding to Individual Sites in the 5’ Untranslated Region. J Virol. 2015 Jun 15;89(12):6294–311.

15. Masaki T, Arend KC, Li Y, Yamane D, McGivern DR, Kato T, et al. miR-122 stimulates hepatitis C virus RNA synthesis by altering the balance of viral RNAs engaged in replication versus translation. Cell Host Microbe. 2015 Feb 11;17(2):217–28.

16. Serti E, Chepa-Lotrea X, Kim YJ, Keane M, Fryzek N, Liang TJ, et al. Successful Interferon-Free Therapy of Chronic Hepatitis C Virus Infection Normalizes Natural Killer Cell Function. Gastroenterology. 2015 Mar;(May):1–11.

17. Spaan M, van Oord G, Kreefft K, Hou J, Hansen BE, Janssen HLA, et al. Immunological Analysis During Interferon-Free Therapy for Chronic Hepatitis C Virus Infection Reveals Modulation of the Natural Killer Cell Compartment. J Infect Dis. 2016 Jan 15;213(2):216–23.

18. Feld JJ, Nanda S, Huang Y, Chen W, Cam M, Pusek SN, et al. Hepatic gene expression during treatment with peginterferon and

218

Chapter 12

ribavirin: Identifying molecular pathways for treatment response. Hepatology. 2007 Nov;46(5):1548–63.

19. Markova AA, Mihm U, Schlaphoff V, Lunemann S, Filmann N, Bremer B, et al. PEG-IFN alpha but not ribavirin alters NK cell phenotype and function in patients with chronic hepatitis C. PLoS One. 2014 Jan;9(4):e94512.

20. Barnes E, Harcourt G, Brown D, Lucas M, Phillips R, Dusheiko G, et al. The dynamics of T-lymphocyte responses during combination therapy for chronic hepatitis C virus infection. Hepatology. 2002 Sep;36(3):743–54.

21. Stelma F, de Niet A, Tempelmans Plat-Sinnige MJ, Jansen L, Takkenberg RB, Reesink HW, et al. NK Cell Characteristics in Chronic Hepatitis B Patients are Associated with HBsAg Loss after Combination Treatment with Peg-interferon Alpha-2a and Adefovir. J Infect Dis. 2015;1–23.

22. Barnes E, Folgori A, Capone S, Swadling L, Aston S, Kurioka A, et al. Novel adenovirus-based vaccines induce broad and sustained T cell responses to HCV in man. Sci Transl Med. 2012 Jan 4;4(115):115ra1.

23. Kelly C, Swadling L, Brown A, Capone S, Folgori A, Salio M, et al. Cross-reactivity of hepatitis C virus specific vaccine-induced T cells at immunodominant epitopes. Eur J Immunol. 2015 Jan;45(1):309–16.

24. Meissner EG, Wu D, Osinusi A, Bon D, Virtaneva K, Sturdevant D, et al. Endogenous intrahepatic IFNs and association with IFN-free HCV treatment outcome. J Clin Invest. 2014 Aug;124(8):3352–63.

25. Serti E, Chepa-Lotrea X, Kim YJ, Keane M, Fryzek N, Liang TJ, et al. Successful Interferon-Free Therapy of Chronic Hepatitis C Virus Infection Normalizes Natural Killer Cell Function. Gastroenterology. 2015 Jul;149(1):190–200.e2.

26. Li A, Qian J, He J, Zhang Q, Zhai A, Song W, et al. Modulation of miR-122 expression affects the interferon response in human hepatoma cells. Mol Med Rep. 2013 Feb;7(2):585–90.

27. Momen-Heravi F, Bala S, Kodys K, Szabo G. Exosomes derived from alcohol-treated hepatocytes horizontally transfer liver specific miRNA-122 and sensitize monocytes to LPS. Sci Rep. 2015 May 14;5(November 2014):9991.

28. Neben S, Liu K, Berman C, Tay J, Esau C, Kaiser R, et al. Pharmacokinetics, pharmacodynamics, and toxicity profile of RG-101, a novel GalNAc-conjugated hepatocyte targeting inhibitor of micro-RNA-122, in rhodentsand cynomolgus monkeys. J Hepatol. 2015;62(2):S684–5.

29. Mundt B, Kühnel F, Zender L, Paul Y, Tillmann H, Trautwein C, et al. Involvement of TRAIL and its receptors in viral hepatitis. FASEB J. 2003 Jan;17(1):94–6.

30. Wandrer F, Falk CS, John K, Skawran B, Manns MP, Schulze-Osthoff K, et al. Interferon-Mediated Cytokine Induction Determines Sustained Virus Control in Chronic Hepatitis C Virus Infection. J Infect Dis. 2016 Mar 1;213(5):746–54.

31. Serti E, Park H, Keane M, O’Keefe AC, Rivera E, Liang TJ, et al. Rapid decrease in hepatitis C viremia by direct acting antivirals improves the natural killer cell response to IFNα. Gut. 2016 Jan 4;0:1–12.

32. Burchill MA, Golden-Mason L, Wind-Rotolo M, Rosen HR. Memory re-differentiation and reduced lymphocyte activation in chronic HCV-infected patients receiving direct-acting antivirals. J Viral Hepat. 2015 Dec;22(12):983–91.

33. Nakamura M, Kanda T, Sasaki R, Haga Y, Jiang X, Wu S, et al. MicroRNA-122 Inhibits the Production of Inflammatory Cytokines by Targeting the PKR Activator PACT in Human Hepatic Stellate Cells. PLoS One. 2015;10(12):e0144295.

219


12

34. Hsu SH, Wang B, Kota J, Yu J, Costinean S, Kutay H, et al. Essential metabolic, anti-inflammatory, and anti-tumorigenic functions of miR-122 in liver. JClinInvest. 2012;122(1558–8238 (Electronic)):2871–83.

35. Martin B, Hennecke N, Lohmann V, Kayser A, Neumann-Haefelin C, Kukolj G, et al. Restoration of HCV-specific CD8+ T-cell function by Interferon-free therapy. J Hepatol. 2014 Jun;61:538–43.

fundInGThis study was funded by Regulus Therapeutics, Inc. Paul Klenerman received financial support from Wellcome Trust (WT091663MA), the Medical Research Council (STOP HCV, MR/K01532X/1), the NIHR Biomedical Research Centre, Oxford, the Oxford Martin School, NIH (U19AI082630) and an NIHR Senior Fellowship. Eleanor Barnes is funded through an MRC Senior Fellowship award and the Oxford NIHR Biomedical Research Centre.

aCKnowledGementsThe authors extend their grateful thanks to Dr. E.B.M. Remmerswaal and Dr. B. Hooibrink for technical assistance, and to all the participating patients.

supplementary dataSupplementary data associated with this article available can be found in the online version of this article.,

Immune responses In daa treated ChronIC hepatItIs C patIents wIth and

wIthout prIor rG-101 dosInG

F. Stelma*, M.H. van der Ree*, S.B. Willemse, A. Brown, L. Swadling, M. van der Valk, M.J. Sinnige, A.C. van Nuenen, J.M.L. de Vree,

P. Klenerman, E. Barnes, N.A. Kootstra, H.W. Reesink

Antiviral Research 2017;146:139-145.


summaryBackgroundWith the introduction of DAA’s, the majority of treated chronic hepatitis C patients (CHC) achieve a viral cure. It is unknown however, whether patients previously treated with an anti-miR-122 oligonucleotide can be successfully treated with DAA’s. Furthermore, the exact mechanisms by which the virus is cleared after successful therapy, is still unknown.

aimTo assess the role of the immune system and miRNA levels in acquiring a sustained virological response after DAA treatment in CHC patients with and without prior RG-101 dosing.

methodsIn this multicenter, investigator-initiated study, 29 patients with hepatitis C virus (HCV) genotype 1 (n=11), 3 (n=17), or 4 (n=1) infection were treated with sofosbuvir and daclatasvir ± ribavirin. 18 patients were previously treated with RG-101. IP-10 levels were measured by ELISA. Ex vivo HCV-specific T cell responses were quantified in IFN-γ-ELISpot assays. Plasma levels of miR-122 were measured by qPCR.

resultsAll patients had an SVR12. IP-10 levels rapidly declined during treatment, but were still elevated 24 weeks after treatment as compared to healthy controls (median 53.82 and 39.4 pg/mL, p=0.02). Functional IFN-γ HCV-specific T cell responses did not change by week 12 of follow-up (77.5 versus 125 SFU /106 PBMC, p=0.46). At follow-up week 12, there was no difference in plasma miR-122 levels between healthy controls and patients with and without prior RG-101 dosing.

ConclusionsOur data shows that successful treatment of CHC patients with and without prior RG-101 dosing results in reduction of broad immune activation, and normalization of miR-122 levels (EudraCT: 2014-002808-25).

223

IMMUNE RESPONSES IN DAA TREATED CRONIC HEPATITIS C PATIENTS

13

IntroduCtIonChronic hepatitis C virus (HCV) infection is a major public health problem affecting an estimated 170 million people worldwide.1 During chronic infection, plasma CXCL10 (or interferon gamma-induced protein 10, IP-10) levels are increased, reflecting a state of ongoing interferon-α signaling and prolonged immune activation.2,3 In addition, constant exposure to viral antigens results in so-called ‘exhausted’ virus-specific CD8+ T cells, T cells that have lost some or all of their effector functions.4,5 Functional virus-specific CD8+ T cells are crucial in viral clearance of acute HCV infection.6 However, interferon-α induced clearance of chronic HCV infection is not associated with improvement of HCV-specific T cell function.7 Current treatment regimens for chronic hepatitis C (CHC) patients include interferon-free combinations of direct acting antivirals (DAAs) which result in high sustained virological response (SVR) rates in the majority of patients. The exact mechanisms that lead to viral clearance in these patients treated with DAAs are not well known. Recent data suggests that the proliferative potential of exhausted T cells is improved after successful treatment with DAAs.3,8 However, the extent of T cell restoration is unknown, as is the role of HCV-specific T cells in viral clearance after DAA treatment.

Inhibition of microRNA-122 (miR-122), an important host factor for replication of HCV, is an alternative therapeutic approach to clear HCV infection.9–11 Recently, it was shown that a single dose of RG-101, an N-acetylgalactosamine conjugated anti-miR-122 oligonucleotide, resulted in substantial viral load reduction in CHC patients, and in SVR for 76 weeks in 3 of 28 patients participating in a phase 1b study.11 Viral load reduction and eradication was not associated with restoration of HCV-specific T cell function.12 Virological rebound following RG-101 dosing was associated with the emergence of resistance associated substitutions (RAS) in 5’ UTR miR-122 binding sites of the HCV genome.11 It is unknown if 5’ UTR C3U and C2G/C3U RAS persist and if these viruses are susceptible to DAAs in vivo. Furthermore, the effect of RG-101 and DAA treatment on plasma miR-122 levels has not been established.

The primary objective of this study was to analyse the impact of DAA treatment on immune activation and functionality of HCV-specific T cells in CHC patients with and without prior RG-101 dosing. The secondary objective was to assess treatment success, and circulating miR-122 levels prior and after DAA treatment.

materIal and methodsstudy designIn this investigator-initiated, open-label, multicenter study, we included CHC patients with genotype 1, 3, or 4 infection at two hospitals in the Netherlands; Academic Medical Center Amsterdam (AMC), and University Medical Center Groningen (UMCG). Patients with HCV genotype 1 and 4 were treated with sofosbuvir (SOF) and daclatasvir (DCV) for 12 weeks, and patients with HCV genotype 3 were treated with SOF, DCV and ribavirin (RBV) for 12 or 24 weeks (patients with cirrhosis were treated for 24 weeks) (figure s1). All patients were followed for 24 weeks after treatment cessation.

224

Chapter 13

patientsTwenty-nine CHC patients (n=25 in AMC and n=4 in UMCG) were included in this study. Of these, 18 patients had a virological rebound following RG-101 dosing,11 and were offered retreatment as part of this study. Only males and post-menopausal females were enrolled. Eligible patients were treatment-naïve (other than previous RG-101 dosing) or had previously had a relapse after antiviral therapy other than combination of SOF+NS5A inhibitor with or without ribavirin. Patients with co-infection (hepatitis B virus or human immunodeficiency virus infection), evidence of decompensated liver disease, or a history of HCC were excluded. Before enrolment and before any study procedure, written informed consent was obtained from all patients.

study oversightThe study was approved by the independent ethics committee at each participating site, and was conducted in compliance with the Declaration of Helsinki, Good Clinical Practice guidelines, and local regulatory requirements. The trial was registered with EudraCT, number 2014-002808-25.

study samplingPlasma and heparinized peripheral blood samples were collected at various time points before (< 90 days before start), during, and after treatment. PBMCs were isolated from heparinized blood using standard density gradient centrifugation and subsequently cryopreserved until the day of analysis.

study assessmentsChemistry and viral assessmentsSerum HCV RNA levels were measured using the Roche COBAS AmpliPrep/COBAS Taqman HCV v2.0 assay, with a reported lower limit of quantification (LLOQ) and lower limit of detection (LLOD) of 15 IU/mL. Effective antiviral treatment was defined by undetectable HCV RNA levels 12 weeks after the cessation of antiviral therapy (SVR12). Safety assessment was based on adverse events reporting and laboratory testing of blood samples (e.g. clinical chemistry, hematology, coagulation).

Sequence analysisSequence analysis of the HCV 5’ UTR of HCV RNA was performed by 5’ rapid amplification of cDNA ends (5’ RACE System, version 2.0, Life Technologies), followed by population-based sequencing and was done at baseline for patients with 5’UTR RAS in miR-122 binding sites at virological rebound in a previous phase 1b study.11

Immunological assaysPlasma levels of IP-10 were measured with a DuoSet ELISA (R&D Systems, Minneapolis, MN, USA) with a lower limit of quantification of 62.5 pg/mL. IFN-γ-ELISpot assays were

225


13

performed ex vivo in duplicate at 2x105 PBMCs/well at the Peter Medawar Building for Pathogen Research, Oxford, UK. Thawed PBMC were rested overnight (37˚C + CO2) and were stimulated with panels of 15 mer peptides that overlapped by 11 amino acids corresponding to HCV genotypes 1a, 1b, 3a, or 4a.13,14 The peptides were arranged into 10 pools, and each peptide was used at a final concentration of 3 µg/ml. Internal controls were dimethyl sulfoxide (DMSO) (Sigma-Aldrich, UK) as a negative control and concanavalin A (Sigma-Aldrich, UK) as a positive control. Other antigens used were a pool of MHC class 1 restricted epitopes of influenza A, EBV and CMV (BEI Resources, Manassas, VA, USA), and a lysate of CMV infected cells (Virusys Corp, Taneytown, MD, USA). Spot forming units (SFU) were calculated per 106 PBMC and background levels (responses in matched negative control wells) were subtracted. Positive responses were defined as (i) the mean of responses to a pool minus background being greater than 48 SFU /106 PBMC and (ii) the mean of responses to a well exceeding 3× background. Background subtracted data is shown. The background level of 48 SFU /106 PBMC per pool was determined previously in 74 healthy controls, which was the mean + 3 standard deviation (SD).13

Plasma miR-122Plasma levels of miR-122 were measured by RT-qPCR at baseline and throughout the study period, and were compared to plasma miR-122 levels of healthy controls (n=10) (supporting methods). Plasma miR-122 levels were normalized to mean levels of miR-425-5p and miR-103a-3p using the ∆Cq method (=Cq miRNA of interest – Cq reference miRNA) and are presented as the log 10 2 - ∆Cq.15 The difference in miR-122 level between baseline and follow-up week 12 was calculated with the comparative Cq-method (= 2 - (∆Cq baseline - ∆Cq FU week 12)) and expressed as the fold change level.15

study objectivesThe primary objective of this study was to analyse the impact of DAA treatment on immune activation and functionality of HCV-specific T cells in CHC patients with and without prior RG-101 dosing. The secondary objective was to assess treatment success, and circulating miR-122 levels prior and after DAA treatment.

data and statistical analysisThe results are presented as median with interquartile range (IQR) or frequency with percentage. All HCV RNA levels were analysed after log 10 transformation. We tested for differences between study groups (CHC patients with and without prior RG-101 dosing, and healthy controls) using the Mann-Whitney U-test, Wilcoxon test, and one-way ANOVA using SPSS (IBM SPSS Statistics for Windows, Version 23.0, Armonk, NY, USA) and GraphPad Prism Software (Version 7.0, La Jolla, CA, USA). Missing data and time points after patients had drop out were excluded from the analyses.

226

Chapter 13

resultsdaa treatment is highly effective in patients with virological rebound following rG-101 dosingScreening began on November 25, 2014, with the last patient enrolled on December 9, 2015. Of 30 patients screened, 29 were enrolled in the study (figure s1). In total, 22 patients (76%) were male, and most patients were infected with HCV genotype 3 (n=17), followed by genotype 1 (n=11), and genotype 4 (n=1) (table 1). Almost all patients (27/29, 93%) had fibrosis stage F3 or higher, which was assessed by liver elastography (Fibroscan) (table 1).

Eighteen patients (62%) included in this study had viral rebound following RG-101 treatment and started SOF+DCV±RBV treatment after a median time of 26 weeks (IQR: 19 - 36 weeks). Six of these patients had C3U or C2G/C3U RAS in miR-122 binding regions at time of viral rebound and/or at start of SOF+DCV±RBV treatment.11 In two of these patients, the persistence of 5’UTR RAS could not be assessed because no extra follow-up sample was available. Of the other four patients with 5’ UTR RAS at time of viral rebound following RG-101 dosing, mutation(s) had disappeared in 3 patients at start of treatment (range: 7 - 22 weeks), whereas the C3U mutation was still present in one patient at the start of treatment (12 weeks after identification) (figure s2).

table 1. Baseline patient characteristics

rG-101 (n=18)

no rG-101 (n=11)

total (n=29)

Male 13 (72) 9 (82) 22 (76)Weight 83 (70-98) 81 (73-91) 81 (71-93)HCV RNA level (log 10 IU/mL) 5.79 (5.16-6.50) 6.56 (5.74-6.65) 6.18 (5.42-6.59)HCV genotype 1a 9 (50) 1 (9) 10 (35) 1b 1 (6) 0 1 (3) 3 7 (39) 10 (91) 17 (59) 4 1 (6) 0 1 (3)ALT level 62 (32-85) 120 (79-151) 79 (46-123)Fibroscan result (kPa) 11.0 (10.3-13.9) 14 (11.9-22.6) 12.0 (10.7-14.8)Fibrosis stage F0-F1 2 (11) 0 2 (7) F1-F2 0 0 0 F2-F3 0 0 0 F3-F4 12 (67) 7 (64) 19 (66) F4 4 (22) 4 (26) 8 (28)

Data are given as median (IQR) or as frequency (percentage). Stage of fibrosis was determined by liver elastography (Fibroscan). Fibrosis score: F0-F1 (< 7 kPa), F1-F2 (7 - 8.8 kPa), F2-F3 (8.9 – 9.4 kPa), F3-F4 (9.5 - 14.5 kPa) and F4 (≥ 14.6 kPa).

227


13

(a) (B)

figure 1. HCV RNA levels of patients with and without previous RG-101 dosing. (1A) HCV RNA level at baseline compared between patients with (black dots) and without (grey dots) previous RG-101 dosing, median and interquartile range are shown. Mann-Whitney U test was used to compare study groups. (1B) Median HCV RNA levels throughout the study period up to follow-up week 12, median and interquartile range are shown. LLOQ, lower limit of quantification. * p < 0.05.

At baseline, median HCV RNA levels were lower in CHC patients who received RG-101 compared to those who did not (median: 5.79 (IQR: 5.16 - 6.50) log 10 IU/mL versus 6.56 (IQR: 5.74 - 6.65) log 10 IU/mL, p=0.043) (figure 1a). Median HCV RNA levels at week 1 and onwards did not differ between patients with and without prior RG-101 dosing (figure 1B). 26/29 (90%) patients achieved SVR12 according to treatment in the study protocol (intention to treat analysis). The remaining three patients achieved SVR12 outside the scope of this study. One HCV genotype 3 patient, fibrosis stage F3, was excluded at week 4 because of non-adherence to the study protocol and treatment was continued at the outpatient clinic resulting in SVR12. One cirrhotic HCV genotype 3 patient discontinued treatment at week 18 due to side effects (anaemia and peripheral neuropathy) and also achieved SVR12. In one patient with HCV genotype 1a, fibrosis stage F3, prior RG-101 treatment, SOF+DCV treatment was prolonged outside the scope of this study because of detectable HCV RNA levels at study week 8 and 12 (2.22 and 1.30 log 10 IU/mL, respectively). This patient achieved SVR12 after 12 additional weeks of SOF+ DCV treatment. These three patients were excluded from further analyses.

daa treatment is well tolerated and results in an improved fibroscan resultNone of the patients experienced a serious adverse event. 10/12 (83%) patients treated with SOF+DCV and 16/17 (94%) patients treated with SOF+DCV+RBV reported at least one adverse event (table 2). The most frequent reported adverse events were fatigue (41%), headache (40%), insomnia (31%), nausea (21%), and dizziness (21%). The total number of adverse events was higher in patients who were treated with RBV (mean number of 6 AEs/patient) compared to those who were not (mean number of 3 AEs/patient). The dose of

228

Chapter 13

RBV was decreased from 1200 mg to 800 mg at week 8 in six patients due to anaemia or other side effects.

In 26/29 patients who completed the study, median ALT level normalised from 79 U/L (IQR: 38 – 121) before treatment to 27 U/L (IQR: 19 – 38) 24 weeks after treatment, p<0.0001 (figure s3). In 23 of these patients, a Fibroscan was performed before treatment and at follow-up week 24. A significant decline in Fibroscan result was observed after treatment from a median value of 12.0 kPa (IQR: 10.7 - 14.8) at baseline to a median of 8.1 kPa (IQR: 5.5 – 12.0) 24 weeks after treatment (p<0.0001) (figure s3).

normalisation of plasma Ip-10 levels in ChC patients treated with daasAt baseline, IP-10 levels were significantly increased in CHC patients relative to healthy controls (median 123.3 pg/mL and 39.4 pg/mL respectively, p<0.0001, figure 2a). No significant difference was observed in baseline IP-10 levels between CHC patients who had previously received RG-101 dosing and patients without prior RG-101 dosing (median 87.0 and 189.9 pg/mL respectively, p=0.08, figure 2a). IP-10 levels significantly decreased upon DAA treatment. As early as 1 week after initiation of treatment, IP-10 levels had

table 2. Adverse events

sof+dCv (n=12)

sof+dCv+rBv (n=17)

total (n=29)

Any adverse event No. of events 35 106 141 No. of patients with event 10 16 26 (90)Adverse events occurring in > 10% of patients Agitation 1 4 5 (17) Anemia 0 3 3 (10) Anorexia 1 2 3 (10) Arthralgia 2 1 3 (10) Back pain 1 2 3 (10) Concentration impairment 3 1 4 (14) Dizziness 2 4 6 (21) Dry mouth 1 2 3 (10) Dry skin 1 3 4 (14) Dyspnea 0 3 3 (10) Fatigue 4 8 12 (41) Flu like symptoms 0 4 4 (14) Headache 3 8 11 (40) Insomnia 2 7 9 (31) Myalgia 2 2 4 (14) Nausea 1 5 6 (21) Skin rash 1 4 5 (17) Upper respiratory infection 2 1 3 (10)

Data shown as number (%)

229


13

halved (from median 134.8 to 65.2 pg/mL). At follow-up week 24, IP-10 levels in CHC patients who had cleared the virus were still increased relative to healthy controls (median 54.5 and 39.4 respectively, p=0.02, figure 2B).

no change in hCv-specific t cell function after successful daa treatmentHCV-specific T cell responses at baseline were low in CHC patients (10/29 patients had a positive response to one or more HCV peptide pools in ELISpot assays). After successful treatment and clearance of HCV, there was no significant change in HCV-specific T cell responses (p=0.09) (figure 3a). There were no significant changes in the IFN-γ T cell responses against structural or non-structural HCV proteins (data not shown). Furthermore, there was no difference in IFN-γ T cell response between baseline and follow-up in patients who were previously dosed with RG-101 (p=0.20) nor in patients who were not previously treated with RG-101 (figure 3B).

normalisation of plasma mir-122 levels in successfully treated ChC patientsCHC patients without prior RG-101 dosing had significantly higher relative plasma miR-122 levels at baseline (median 0.761 (IQR: 0.623 - 1.053)) as compared to patients with prior RG-101 dosing and healthy controls (p<0.001, and p<0.0001, respectively) (figure 4a). Relative plasma miR-122 levels at baseline were not different between CHC patients previously dosed with RG-101 and healthy controls (median -0.004 (IQR: -0.431 - 0.286) versus -0.399 (IQR: -0.638 - 0.018), respectively, p=0.175) (figure 4a). At follow-up week 12, plasma miR-122 levels of CHC patients without previous RG-101 dosing decreased 20-fold compared to baseline values (p=0.008), whereas no significant difference between

(a) (B)

figure 2. IP-10 levels of patients with and without previous RG-101 dosing. (2A) Baseline IP-10 levels in patients with (black dots) and without (grey dots) previous RG-101 dosing and healthy controls (HC, white triangles), median and interquartile range are shown. Mann-Whitney U test was used to compare study groups. (1B) IP-10 levels in CHC patient throughout the study period up to follow-up week 24 (FU24), median and interquartile range are shown. Mann-Whitney U test was used to compare study groups. LLOQ, lower limit of quantification. * p<0,05; *** p < 0.001; **** p<0,0001.

230

Chapter 13

baseline and follow-up was observed in patients previously dosed with RG-101 (median fold decline 1.42, p=0.287) (figure 4B). At follow-up week 12, there was no difference in median relative plasma miR-122 levels between healthy controls and patients with and without prior anti-miR-122 dosing (-0.399, -0.344, and -0.275, respectively, p=0.555) (figure 4C).

dIsCussIonIn CHC patients with and without prior anti-miR-122 (RG-101) treatment, DAA treatment with SOF+DCV±RBV was highly effective. Successful antiviral treatment resulted in a reduction of immune activation as evidenced by IP-10 levels, and no increase in the magnitude of ex vivo HCV-specific T cell responses. In CHC patients without prior RG-101 dosing, circulating miR-122 levels were elevated at baseline and normalised when SVR was achieved.

This study enrolled patients with various HCV genotypes, including a substantial number of difficult-to-treat HCV genotype 3 patients with advanced fibrosis or compensated cirrhosis. Previous studies showed that 12 weeks of treatment with SOF+DCV results in SVR12 rates of only 63% in HCV genotype 3 patients with cirrhosis,16 and that addition of RBV to the 12- or 16-week regimen increased SVR rates to 83% and 89% respectively.17 In our study, the SVR12 rate in HCV genotype 3 patients with advanced fibrosis (treated for 12 weeks with SOF+DCV+RBV) and compensated cirrhosis (treated for 24 weeks with SOF+DCV+RBV) was 100%. Although the addition of RBV has been shown to improve SVR12 rates in a 12-week SOF+DAC regimen,17 it is accompanied with more frequent adverse events, and the added benefit of extending the duration of this triple combination treatment beyond 12 weeks is unknown.16–18

As expected, IP-10 levels were significantly increased in CHC patients as compared to healthy controls. We have previously shown that IP-10 levels decreased after dosing with

(a) (B)

figure 3. IFN-γ-ELISpot responses in CHC patients at baseline and follow-up week 12. (3A) Sum of positive IFN-γ-ELISpot responses at baseline and follow-up week 12 (FU12) in CHC patients treated with SOF-DAC. Wilcoxon matched pairs test was used to compare baseline and FU12. (3B) Sum of positive IFN-γ-ELISpot responses in patients with (black dots) and without (grey dots) previous RG-101 dosing. Wilcoxon matched pairs test was used to compare baseline and FU12.

231


13

RG-101.12 However, at baseline there was no significant difference in IP-10 levels between CHC patients with and without previous RG-101 dosing. This suggests that the decline in IP-10 levels was transient after dosing with RG-101, and that IP-10 levels have increased again after rebound in HCV viral load. An increase in IP-10 levels with relapse was previously shown in CHC patients treated with DAAs,2 and suggests reactivation of the interferon pathway. After clearance of HCV, IP-10 levels significantly decreased in all CHC patients in line with the disappearance of the virus. However, the immunological status of the patients does not fully normalise during follow-up, as IP-10 levels were still significantly increased at FU week 24 compared to healthy controls.3

It is unknown to what extent HCV-specific T cells play a role in the clearance of HCV upon DAA treatment. In a chimpanzee successfully treated with DAAs intrahepatic IFN-γ producing T cells were observed, but these were unable to prevent persistence upon reinfection.19 Previous data has suggested an increase in the proliferative potential of HCV-specific T cells in humans after successful treatment with DAAs.3,8,20 However, the extent of

(a)

(C)

(B)

figure 4. Relative plasma miR-122 levels. (4A) Relative plasma miR-122 level at baseline in patients with (black dots) and without (grey dots) previous RG-101 dosing and healthy controls (HC, white triangles), median and interquartile range are shown. Mann-Whitney U test was used to compare study groups. (4B) Relative plasma miR-122 levels in CHC patients with (black dots) and without (grey dots) prior RG-101 dosing treated with SOF+DAC at baseline and follow-up week 12 (FU12). Wilcoxon test was used to compare baseline and FU12. Mann-Whitney U test was used to compare FU12 of different study groups. (4C) Relative plasma miR-122 levels in CHC patient throughout the study period up to FU12, as well as in healthy controls, median and interquartile range are shown. One-way ANOVA was used to compare study groups. ** p<0.01; ***: p<0.001; ****: p<0.0001.

232

Chapter 13

the previously observed reversal of the exhausted phenotype was unknown, as no analyses of function (such as cytokine production) were performed. In patients treated with RG-101, we have shown that the function of HCV-specific T cells does not improve in patients who are long term HCV RNA negative.12 Here we show that similarly, the function of HCV-specific T cells does not improve after successful clearance of HCV. This is in line with the recent observation of persistence of an increased proportion of regulatory T cells even when patients are successfully treated with DAAs.21

The majority of patients in our study were previously dosed with RG-101 in a phase 1b study. A number of these patients had RG-101 associated RAS in 5’UTR miR-122 binding sites (C2G+C3U or C3U) at time of viral rebound and/or at start of DAA treatment.11 Here we showed that DAA treatment was highly effective in CHC patients who previously received RG-101, even if 5’UTR RAS were present. This finding was consistent with a preclinical study showing that 5’UTR HCV variants were fully susceptible to DAAs.22 This is of special importance for testing regimens combining RG-101 and one or more DAAs. Next, we demonstrated that 5’UTR RAS spontaneously disappear within several weeks after identification, suggesting that the mutated viruses were outcompeted by the fitter wild-type variant. Taken together, we expect that the relevance of the mutations in 5’UTR miR-122 binding sites will be limited or non-existent in future clinical practice.

Circulating miR-122 levels are higher in CHC patients as compared to healthy controls.23 In line with previous studies, we showed that circulating miR-122 levels decrease (by approximately 20-fold) and normalise in successfully treated CHC patients.24,25 Moreover, we have previously shown that plasma miR-122 levels rapidly decrease by more than 1000-fold in anti-miR-122 treated CHC patients.26 Here we showed that plasma miR-122 levels were still decreased several months after RG-101 treatment compared to other CHC patients. However, plasma miR-122 levels in RG-101 treated patients were similar to levels of healthy controls, which is favourable from a safety perspective since miR-122 modulates the expression of a large number of hepatocyte proteins, some of which have been implicated in hepatocarcinogenesis.27–32

In conclusion, treatment with SOF+DCV±RBV was well tolerated and resulted in high SVR12 rates in CHC patients with and without prior RG-101 treatment. Our data suggests that there is no restoration of HCV adaptive immunity after SVR in CHC patients.

233


13

referenCes1. Shepard CW, Finelli L, Alter MJ. Global

epidemiology of hepatitis C virus infection. Lancet Infect Dis. 2005 Sep;5(9):558–67.

2. Meissner EG, Wu D, Osinusi A, Bon D, Virtaneva K, Sturdevant D, et al. Endogenous intrahepatic IFNs and association with IFN-free HCV treatment outcome. J Clin Invest. 2014 Aug;124(8):3352–63.

3. Spaan M, van Oord G, Kreefft K, Hou J, Hansen BE, Janssen HLA, et al. Immunological Analysis During Interferon-Free Therapy for Chronic Hepatitis C Virus Infection Reveals Modulation of the Natural Killer Cell Compartment. J Infect Dis. 2016 Jan 15;213(2):216–23.

4. Bengsch B, Seigel B, Ruhl M, Timm J, Kuntz M, Blum HE, et al. Coexpression of PD-1, 2B4, CD160 and KLRG1 on exhausted HCV-specific CD8+ T cells is linked to antigen recognition and T cell differentiation. PLoS Pathog. 2010 Jun 10;6(6):e1000947.

5. Klenerman P, Hill A. T cells and viral persistence: lessons from diverse infections. Nat Immunol. 2005 Sep;6(9):873–9.

6. Urbani S, Amadei B, Fisicaro P, Tola D, Orlandini A, Sacchelli L, et al. Outcome of acute hepatitis C is related to virus-specific CD4 function and maturation of antiviral memory CD8 responses. Hepatology. 2006 Jul;44(1):126–39.

7. Barnes E, Gelderblom HC, Humphreys I, Semmo N, Reesink HW, Beld MGHM, et al. Cellular immune responses during high-dose interferon-alpha induction therapy for hepatitis C virus infection. J Infect Dis. 2009 Mar 15;199(6):819–28.

8. Martin B, Hennecke N, Lohmann V, Kayser A, Neumann-Haefelin C, Kukolj G, et al. Restoration of HCV-specific CD8+ T cell function by interferon-free therapy. J Hepatol. 2014 Sep 4;61(3):538–43.

9. Jopling CL, Yi M, Lancaster AM, Lemon SM, Sarnow P. Modulation of hepatitis C virus

RNA abundance by a liver-specific MicroRNA. Science. 2005 Sep 2;309(5740):1577–81.

10. Lanford RE, Hildebrandt-Eriksen ES, Petri A, Persson R, Lindow M, Munk ME, et al. Therapeutic silencing of microRNA-122 in primates with chronic hepatitis C virus infection. Science. 2010 Jan 8;327(5962):198–201.

11. van der Ree MH, de Vree JM, Stelma F, Willemse S, van der Valk M, Rietdijk S, et al. Safety, tolerability, and antiviral effect of RG-101 in patients with chronic hepatitis C: a phase 1B, double-blind, randomised controlled trial. Lancet (London, England). 2017 Jan 10;

12. Stelma F, van der Ree MH, Sinnige MJ, Brown A, Swadling L, de Vree JML, et al. Immune phenotype and function of NK and T cells in chronic hepatitis C patients who received a single dose of anti-miR-122, RG-101 - Accepted for publication. Hepatology. 2017;

13. Barnes E, Folgori A, Capone S, Swadling L, Aston S, Kurioka A, et al. Novel adenovirus-based vaccines induce broad and sustained T cell responses to HCV in man. Sci Transl Med. 2012 Jan 4;4(115):115ra1.

14. Kelly C, Swadling L, Brown A, Capone S, Folgori A, Salio M, et al. Cross-reactivity of hepatitis C virus specific vaccine-induced T cells at immunodominant epitopes. Eur J Immunol. 2015 Jan;45(1):309–16.

15. Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc. 2008 Jan;3(6):1101–8.

16. Nelson DR, Cooper JN, Lalezari JP, Lawitz E, Pockros PJ, Gitlin N, et al. All-oral 12-week treatment with daclatasvir plus sofosbuvir in patients with hepatitis C virus genotype 3 infection: ALLY-3 phase III study. Hepatology. 2015 Apr 23;61(4):1127–35.

17. Leroy V, Angus P, Bronowicki J-P, Dore GJ, Hezode C, Pianko S, et al. Daclatasvir, sofosbuvir, and ribavirin for hepatitis C virus genotype 3 and advanced liver disease:

234

Chapter 13

A randomized phase III study (ALLY-3+). Hepatology. 2016 May;63(5):1430–41.

18. Welzel TM, Petersen J, Herzer K, Ferenci P, Gschwantler M, Wedemeyer H, et al. Daclatasvir plus sofosbuvir, with or without ribavirin, achieved high sustained virological response rates in patients with HCV infection and advanced liver disease in a real-world cohort. Gut. 2016 Sep 7;65(11):1861–70.

19. Callendret B, Eccleston HB, Hall S, Satterfield W, Capone S, Folgori A, et al. T-cell immunity and hepatitis C virus reinfection after cure of chronic hepatitis C with an interferon-free antiviral regimen in a chimpanzee. Hepatology. 2014 Nov;60(5):1531–40.

20. Burchill MA, Golden-Mason L, Wind-Rotolo M, Rosen HR. Memory re-differentiation and reduced lymphocyte activation in chronic HCV-infected patients receiving direct-acting antivirals. J Viral Hepat. 2015 Dec;22(12):983–91.

21. Langhans B, Dieter Nischalke H, Krämer B, Hausen A, Dold L, van Heteren P, et al. Increased peripheral cd4+ regulatory t cells persist after successful direct-acting antiviral treatment of chronic hepatitis C. J Hepatol. 2016 Dec 28;xxx.

22. Ottosen S, Parsley TB, Yang L, Zeh K, van Doorn L-J, van der Veer E, et al. In vitro antiviral activity and preclinical and clinical resistance profile of miravirsen, a novel anti-hepatitis C virus therapeutic targeting the human factor miR-122. Antimicrob Agents Chemother. 2015 Jan 1;59(1):599–608.

23. van der Meer AJ, Farid WRR, Sonneveld MJ, de Ruiter PE, Boonstra A, van Vuuren AJ, et al. Sensitive detection of hepatocellular injury in chronic hepatitis C patients with circulating hepatocyte-derived microRNA-122. J Viral Hepat. 2013 Mar 27;20(3):158–66.

24. Waring JF, Dumas EO, Abel S, Coakley E, Cohen DE, Davis JW, et al. Serum miR-122 may serve as a biomarker for response to direct acting antivirals: effect of paritaprevir/R with dasabuvir or ombitasvir on miR-122 in HCV-infected subjects. J Viral Hepat. 2016 Feb 5;23(2):96–104.

25. Köberle V, Waidmann O, Kronenberger B, Andrei A, Susser S, Füller C, et al. Serum microRNA-122 kinetics in patients with chronic hepatitis C virus infection during antiviral therapy. J Viral Hepat. 2013 Aug;20(8):530–5.

26. van der Ree MH, van der Meer AJ, van Nuenen AC, de Bruijne J, Ottosen S, Janssen HL, et al. Miravirsen dosing in chronic hepatitis C patients results in decreased microRNA-122 levels without affecting other microRNAs in plasma. Aliment Pharmacol Ther. 2016 Jan;43(1):102–13.

27. Gordon EM, Liu PX, Chen ZH, Liu L, Whitley MD, Gee C, et al. Inhibition of metastatic tumor growth in nude mice by portal vein infusions of matrix-targeted retroviral vectors bearing a cytocidal cyclin G1 construct. Cancer Res. 2000 Jul 1;60(13):3343–7.

28. Fornari F, Gramantieri L, Giovannini C, Veronese A, Ferracin M, Sabbioni S, et al. MiR-122/cyclin G1 interaction modulates p53 activity and affects doxorubicin sensitivity of human hepatocarcinoma cells. Cancer Res. 2009 Jul 15;69(14):5761–7.

29. Gramantieri L, Ferracin M, Fornari F, Veronese A, Sabbioni S, Liu C-G, et al. Cyclin G1 is a target of miR-122a, a microRNA frequently down-regulated in human hepatocellular carcinoma. Cancer Res. 2007 Jul 1;67(13):6092–9.

30. Tsai W-C, Hsu PW-C, Lai T-C, Chau G-Y, Lin C-W, Chen C-M, et al. MicroRNA-122, a tumor suppressor microRNA that regulates intrahepatic metastasis of hepatocellular carcinoma. Hepatology. 2009 May;49(5):1571–82.

31. Lin CJ-F, Gong H-Y, Tseng H-C, Wang W-L, Wu J-L. miR-122 targets an anti-apoptotic gene, Bcl-w, in human hepatocellular carcinoma cell lines. Biochem Biophys Res Commun. 2008 Oct 24;375(3):315–20.

32. Zeng C, Wang R, Li D, Lin X-J, Wei Q-K, Yuan Y, et al. A novel GSK-3 beta-C/EBP alpha-miR-122-insulin-like growth factor 1 receptor regulatory circuitry in human hepatocellular carcinoma. Hepatology. 2010 Nov;52(5):1702–12.

235


13

aCKnowledGementsWe thank Hadassa Heidsieck, Jeltje Helder and Martine Peters for their invaluable support and care in this study.

fundInGThis investigator initiated study was in part funded by Bristol-Myers Squibb. Paul Klenerman received financial support from Wellcome Trust (WT091663MA), the Medical Research Council (STOP HCV, MR/K01532X/1), the NIHR Biomedical Research Centre, Oxford, the Oxford Martin School, NIH (U19AI082630) and an NIHR Senior Fellowship. Eleanor Barnes is supported by the Medical Research Council UK the Oxford NIHR Biomedical Research Centre and the Oxford Martin School.

236

Chapter 13

supplementary data

supplementary figure 1. Flow-chart.

237


13

4

Figure S2 28

HCV RNA levels of patients with 5’UTR RAS in prior study with additional follow-up 29

sample available for sequence analysis. Presence of C3U RAS in miR-122 binding 30

sites (open dots), C2G/C3U RAS (open squares), or wild-type sequences (solid dots) 31

are indicated. 32

33

34

35 supplementary figure 2. HCV RNA levels of patients with 5’UTR RAS in prior study with additional follow-up sample available for sequence analysis. Presence of C3U RAS in miR-122 binding sites (open dots), C2G/C3U RAS (open squares), or wild-type sequences (solid dots) are indicated.

5

Figure S3 36

Successful treatment resulted in a significant decline in ALT level (A) and Fibroscan 37

result (B) in all CHC patients. Wilcoxon signed-rank test was used to compare result 38

at baseline and 24 weeks after treatment , ****: p<0.0001. 39

A. 40

41

B. 42

43

5

Figure S3 36

Successful treatment resulted in a significant decline in ALT level (A) and Fibroscan 37

result (B) in all CHC patients. Wilcoxon signed-rank test was used to compare result 38

at baseline and 24 weeks after treatment , ****: p<0.0001. 39

A. 40

41

B. 42

43

(a)

(B)

supplementary figure 3. Successful treatment resulted in a significant decline in ALT level (A) and Fibroscan result (B) in all CHC patients. Wilcoxon signed-rank test was used to compare result at baseline and 24 weeks after treatment , ****: p<0.0001.

CharaCterIsatIon of IntrahepatIC t Cells In patIents wIth ChronIC hepatItIs B or

C InfeCtIon

human IntrahepatIC Cd69+Cd8+ t Cells have a tIssue resIdent memory t Cell phenotype

wIth reduCed CytolytIC CapaCIty

F. Stelma, A. de Niet, M.J. Sinnige, K.A. van Dort, K.P.J.M. van Gisbergen, J. Verheij, E.M.M. van Leeuwen, N.A. Kootstra*, H.W. Reesink*

Scientific Reports 2017;7:6172.


aBstraCt Tissue resident memory T cells (TRM) have been identified in various tissues, however human liver TRM to date remain unidentified. TRM can be recognized by CD69 and/or CD103 expression and may play a role in the pathology of chronic hepatitis B (CHB) and hepatitis C virus infection (CHC). Liver and paired blood mononuclear cells from 17 patients (including 4 CHB and 6 CHC patients) were isolated and CD8+ T cells were comprehensively analysed by flowcytometry, immunohistochemistry and qPCR. The majority of intrahepatic CD8+ T cells expressed CD69, a marker used to identify TRM, of which a subset co-expressed CD103. CD69+CD8+ T cells expressed low levels of S1PR1 and KLF2 and a large proportion (>90%) was CXCR6+, resembling liver TRM in mice and liver resident NK cells in human. Cytotoxic proteins were only expressed in a small fraction of liver CD69+CD8+ T cells in patients without viral hepatitis, however, in livers from CHB patients more CD69+CD8+ T cells were granzyme B+. In CHC patients, less intrahepatic CD69+CD8+ T cells were Hobit+ as compared to CHB and control patients. Intrahepatic CD69+CD8+ T cells likely TRM which have a reduced cytolytic potential. In patients with chronic viral hepatitis TRM have a distinct phenotype.

243

LIVER RESIDENT T CELLS IN CHRONIC VIRAL HEPATITIS

14

IntroduCtIonThe liver is an organ with unique immunologic properties. Generally, a tolerant milieu is maintained in the liver to prevent broad immune activation in response to gut-derived antigens.1 Hepatotropic viruses, such as hepatitis B (HBV) and C virus (HCV), are therefore thought to specifically target the liver for infection. During an infection, upon encounter with their cognate antigen, antigen-specific T cells undergo clonal expansion and subsequently form a memory T cell population.2 Recently, it became evident that this memory population does not only consist of a recirculating fraction - detectable in the peripheral blood - but also includes an important tissue resident memory T cell (TRM) pool, residing in non-lymphoid organs.3 TRM can be identified by expression of CD69, identifying a broad population of which a subset of cells co-expresses CD103 (integrin alpha E).3,4 TRM reside in human tissues such as lung, skin and gut and have unique functions.3,5–7 For example, resident intrahepatic, but not circulatory CD8+ T cells, are the main effectors in an effective immune response against malaria in mice.8 In viral hepatitis, local bystander TRM may play an important role in the pathology observed in chronic viral infection of the liver.9

Whereas previous data has identified a large CD69+ NK cell population in the liver10, data on CD8+ T cells with a tissue resident phenotype in the human liver is lacking, as is knowledge on their phenotype in patients with viral hepatitis. The aim of this study was to examine the presence of intrahepatic TRM in control patients without viral hepatitis, as well as the presence of these cells in the liver from patients who are chronically infected with HBV or HCV.

patIents and methods patient samplesWe included 17 patients of which 7 did not have viral hepatitis (controls), 4 had chronic HBV (CHB) infection and 6 had chronic HCV (CHC) infection. Patient characteristics are listed in Table 1. The study was approved by the Ethical Review Board of the Academic Medical Center, Amsterdam, The Netherlands and all patients gave written informed consent in accordance with the Declaration of Helsinki. The study was performed in accordance with Good Clinical Practice guidelines, Good Laboratory Practice guidelines and local regulatory requirements. Liver tissue with paired blood samples were obtained from patients undergoing surgical liver resection (indications listed in Table 1). From the resected liver tissue, the non-affected, tumor-free margin surrounding the pathology was obtained. Liver samples were perfused to remove any residual blood, disrupted and treated with collagenase IV (Worthington Biochemical Corporation, CA, USA) and DNAse I (Sigma-Aldrich, MO, USA). Mononuclear cells from blood and liver were isolated using standard density-gradient centrifugation and cryopreserved for later analysis.

244

Chapter 14

table 1. Patient characteristics

Case # viral status gender age reason surgery cirrhosis* genotype alt (u/l)

1 Control m 81 cholangiocarcinoma no 422 Control m 81 liver metastasis CRC no 303 Control f 50 liver metastasis CRC no 334 Control m 68 HCC no 245 Control f 37 multiple adenomas no 506 Control m 68 intrahepatic gallstones no 327 Control m 67 liver metastasis CRC no 188 CHB m 51 HCC yes C 819 CHB m 57 HCC F3/F4 C 3310 CHB m 47 cholangiocarcinoma F1 unknown 4311 CHB m 64 HCC F3 unknown 1112 CHC (HBVc) m 58 HCC yes 1b 5713 CHC (HBVc) m 58 HCC F2/F3 1b 12914 CHC m 59 HCC yes 1a 15415 CHC m 61 HCC yes 1a 11116 CHC m 52 HCC yes 3a 3017 CHC m 55 HCC yes 4 112

*Cirrhosis based on METAVIR score at histology. Abbreviations; ALT, alanine aminotransferase; m, male; f, female; CRC, colorectal carcinoma; HCC, hepatocellular carcinoma; CHB, chronic hepatitis B; CHC, chronic hepatitis C; HBVc, cleared acute hepatitis B virus.

flow cytometryCells were thawed and stained with different combinations of fluorescent label-

conjugated mouse monoclonal antibodies (listed in the Supplementary Materials). For intracellular staining, cells were fixed after surface staining, permeabilized, and stained with fluorescent label-conjugated mouse monoclonal antibodies (Supplementary Materials). All measurements were done on an LSR Fortessa flowcytometer (BD Biosciences, CA, USA) and analyzed by FlowJoMacV 9.7.5 software (Tree Star Inc., OR, USA).

ImmunohistochemistryParaffin embedded formalin-fixed sections (4 μm) of non-affected liver tissue from control patients (n=3) were stained using sequential immunohistochemistry.44 Sections were deparaffinized and rehydrated in up to 96% ethanol. Endogenous peroxidase activity was blocked by incubation with 0.3% H2O2 in methanol. Antigen retrieval was performed at 120°C under pressure in a Tris/EDTA buffer. The sections were stained overnight with anti-CD69 (mouse IgG1, clone CH11, Abcam, Cambridge, UK), followed by incubation with polyHRP-anti-mouse IgG (ImmunoLogic, Duiven, Netherlands). HRP activity was demonstrated with VectorNovaRed (Vector Laboratories, CA, USA). The slides were counterstained with hematoxilin, coverslipped and scanned in a Philips Ultra Fast Scanner 1.6RA (Philips, Eindhoven, Netherlands). After removal of the coverslip, the sections were stripped in a Tris-SDS buffer at 50 °C. Next, they were incubated with anti-CD103

245


14

(rabbit IgG, clone EPR4166, Abcam, Cambridge, UK) and subsequent polyHRP-anti-rabbit IgG (ImmunoLogic, Duiven, Netherlands). HRP activity was demonstrated as above. This procedure was repeated with anti-CD8 (mouse IgG1, clone C3/144B, Dako, Glostrup, Denmark) and anti-CD3 (rabbit IgG1, clone Sp7, Thermo scientific, NH, USA). Images were analysed with Fiji Image J software.43 Pseudocolors were applied to provide co-localization images.

Cell sorting and rna extractionLiver lymphocytes from 6 control patients were sorted using fluorescent label-conjugated mouse monoclonal antibodies for CD3, CD8 and CD69 (Supplementary Materials). Liver CD3+CD8+ lymphocytes were sorted into a CD69+ and CD69- fraction on a Sony SH800 cell sorter (Sony biotechnology Inc., CA, USA). After sorting, RNA was extracted with the AllPrep isolation kit (Qiagen, CA, USA).

reverse transcription quantitative polymerase chain reaction (rt-qpCr)Reverse transcription of mRNA was performed using M-MLV reverse transcriptase and oligo-dT primers. Relative quantification of gene expression was determined with the LightCycler 2.0 or LightCycler 480 Real-Time PCR System (Roche Applied Science, CT, USA) using the SYBR Green PCR Master Mix and gene specific primers (Supplementary Table 1). mRNA expression levels were normalized to the arithmetic mean of the housekeeping gene ACTB using the comparative Ct method, and log10 transformed for analysis.

data availabilityThe datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

statistical analysesStatistical analyses were performed using GraphPad Prism 6 software (GraphPad Software Inc., CA, USA). For paired data, the paired t-test (Gaussion distribution) or Wilcoxon signed rank test (non-Gaussion distribution) was used. Similarly, for non-paired data, the t-test or Mann–Whitney U test was used. Correlations were calculated using the Spearman rank test. Two sided p-values below 0.05 were considered to be statistically significant.

results human intrahepatic Cd69+ t cells express a tissue resident phenotypeIn the liver, CD8+ T cells were enriched as compared to the blood (p=0.0042). Inversely, the liver contained fewer CD4+ T cells than the blood (p=0.0004, Figure 1a). To identify intrahepatic TRM, we analysed the expression of two markers commonly expressed by TRM; CD69 and CD103.5,11 In the blood, few CD69 positive cells were present within the CD8+

246

Chapter 14

figure 1. (a) Frequency of CD8+ and CD4+ T cells as a percentage of total CD3+ lymphocytes was compared in control blood and liver (n=7). Statistical analyses; paired t-test. (b) Frequency of CD69+ and CD103+ as a percentage of total CD8+ T cells in blood and liver. Statistical analyses; paired t-test. (c) Representative flow cytometry plots showing CD69 and CD103 expression, gated on CD8+CD3+ positive lymphocytes. (d) Paraffin embedded formalin-fixed sections of control livers (n=3)(tumor-free tissue was used) were stained using sequential immunohistochemistry. An immunohistochemical staining of CD69 is shown with a portal field (*) and a central vein (#), and a portal field with co-localization of CD69+CD8+CD3+ lymphocytes (filled arrow) and CD69+CD103+CD8+CD3+ lymphocytes (open arrow). Original magnification: 20x. (e) Relative expression of S1PR1 and transcription factor KLF2 in sorted CD69+ and CD69-CD8+ T cells from the liver (n=6). Bars indicate median. Statistical analyses; Wilcoxon signed rank test. (f) Frequency of CXCR6+ T cells as a percentage of total CD8+ T cells in blood, liver CD69+ and liver CD69-CD8+ T cells. Statistical analyses; paired t-test. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

(a) (B) (C)

(d)

(e) (f)

247


14

T cell population (mean 4.3%), while a significant population of CD69+CD8+ T cells was detected in the liver (mean 68.0%, p<0.0001, Figure 1b, gating strategy in Supplementary Fig. 1). In the peripheral blood, mean 2.5% of CD8+ T cells expressed CD103, while mean 12.4% of intrahepatic CD8+ T cells were CD103 positive (p=0.03, Figure 1b,c). In contrast to those in the blood, the CD103+CD8+ T cells detected in the liver co-expressed CD69 (Figure 1c). Immunohistochemistry revealed localization of CD69+CD8+CD3+ positive cells in portal fields, central veins, and parenchymal zones 1-3 (Figure 1d, Supplementary Fig. 2).

The sphingosine 1-phosphate receptor-1 (S1PR1) is crucial for recirculation of (naïve) T cells12, and negatively regulated by CD69 7,11–13. S1PR1 and its regulatory transcription factor Krüppel-like Factor 2 (KLF2) were significantly down-regulated in sorted intrahepatic CD69+CD8+ T cells as compared to CD69-CD8+ T cells (Figure 1e). Chemokine receptor CXCR6, associated with homing to the liver,10,14,15 was expressed by mean 91.2% of CD69+CD8+ T cells in the liver, as compared to mean 25.0% of their CD69- liver counterparts (p<0.0001), and mean 21.9% of peripheral blood CD8+ T cells (Figure 1f). Additionally, 25-35% of CD69+ CD8+ T cells in the liver were positive for the MR1 tetramer, CD161 and IL-18Rα, specific for mucosa associated invariant T (MAIT) cells (supplementary Fig. 3).16

Intrahepatic Cd69+Cd8+ t cells are not recently activated. In intrahepatic CD69+CD8+ T cells, the proportion of HLA-DR/CD38 double positive T cells, which in the peripheral blood represent activated cells, was significantly higher (mean 26.4%) than in peripheral blood T cells (mean 5.4%, p=0.0049, Figure 2a, Supplementary Fig. 4). Whereas alanine aminotransferase (ALT) levels significantly correlated with the proportion of HLA-DR/CD38 expressing CD8+ T cells in the blood (p=0.01), in the liver there was no correlation between the two (p=0.33, Supplementary Fig. 5). The proportion of Ki67+ cells, which represent cells that have recently divided, was significantly higher in intrahepatic CD69-CD8+ T cells as compared to CD69+CD8+ T cells (mean 6.5% and 1.6% respectively, p=0.04, Figure 2b). The proportion of cells expressing apoptosis marker TO-PRO-3, was low in all populations (Supplementary Fig. 6). Programmed death-1 (PD-1), associated with T cell inhibition, was higher expressed on liver CD69+CD8+ T cells as compared to liver CD69- cells and peripheral blood T cells (Figure 2c). Other inhibitory molecules were not differentially expressed on intrahepatic CD69+ and CD69- CD8+ T cells (Supplementary Fig. 6). Nur77, an ‘immediate early gene’, up-regulated in response to TCR stimulation,17 was not differentially expressed in CD69+ and CD69-CD8+ T cells isolated from the liver (Figure 2d). As expected, the proportion of naïve peripheral blood CD8+ T cells expressing activation markers was low (Supplementary Fig. 7).

Intrahepatic Cd69+Cd8+ t cells are not naïve, but have a memory or effector-memory phenotype, similar to the peripheral blood. CD8+ T cells can be divided into different subsets, including naïve and memory cells, the latter comprising cells that have been primed by antigen. Antigen experienced cells can be divided into memory T cells (TMEM) and effector-type memory T cells (TEFF).

248

Chapter 14

We identified these populations by CD27 and CD45RA expression, and analysed naïve (CD45RA+CD27+), TMEM (CD45RA-CD27+) and TEFF (CD45RA±CD27-) T cells (Figure 3a).2,18,19 The proportion of TEFF was comparable in blood and intrahepatic CD69+CD8+ T cells (Figure 3b). In liver CD69-CD8+ T cells, a significantly higher proportion of CD8+ T cells were TEFF phenotype as compared to the blood (p=0.01). The proportion of T cells that had a TMEM phenotype was not different between blood, intrahepatic CD69+ and intrahepatic CD69- CD8+ T cells. Naïve cells were significantly less frequent in the CD8+ T cell populations derived from the liver (CD69+ and CD69- mean 0.7%, 2.2% respectively) as compared to the blood (mean 9.7%, Figure 3a,b).

(a) (B)

(C) (d)

figure 2. Frequency of HLA-DR+CD38+ double positive (a) and Ki67 (b) positive cells as a percentage of CD8+ T cells and gMFI of PD-1 (c) in control blood vs. liver (CD69+ and CD69-) CD8+ T cells. Statistical analyses; paired t-test. **p<0.01. (d) Relative expression of Nur77 in sorted CD69+ and CD69-CD8+ T cells from the liver (n=6, one CD69+ sample was unavailable). Bars indicate median. Statistical analyses; Wilcoxon signed rank test. Abbreviations: ns, non-significant.

249


14

the majority of intrahepatic Cd69+Cd8+ teff cells do not express cytotoxic proteins. As intrahepatic NK cells expressing CD69 and CXCR6 have a reduced cytolytic potential10, we next analysed the expression of these effector molecules in intrahepatic TEFF and TMEM subsets. The proportion of intrahepatic CD69+ TEFF cells expressing perforin was significantly lower (mean 12.1%) that of blood TEFF cells (mean 91.8%, p<0.0001) as was the proportion of granzyme B+ cells (mean 45.7% and 82.4 respectively, p=0.0001, Figure 4a,c). Within the intrahepatic TEFF cell population expressing CD69, significantly less cells expressed cytotoxic proteins as compared to CD69- TEFF cells (p<0.0001 for perforin and p=0.001 for granzyme B). In addition, significantly less intrahepatic TMEM expressed perforin (CD69+; mean 9.3 and CD69-; mean 19.8%) than in the blood (mean 52.6%, p=0.0002, Figure 4b,c), whereas no difference in TMEM granzyme B expression was observed (Figure 4b,c).

(a)

(B)

figure 3. (a) Representative flow cytometry plots showing CD27 and CD45RA expression with gates indicating effector memory (TEFF, CD45RA±CD27-), memory (TMEM, CD45RA-CD27+) and naïve (CD45RA+CD27+) CD8+ T cells in control blood and liver (CD69+ and CD69-). (b) Frequency of TEFF, TMEM and naïve CD8+ T cells as a percentage of total CD8+ T lymphocytes in control blood and liver (CD69+ and CD69-). Bars indicate mean. Statistical analyses; paired t-test *p<0.05, **p<0.01, ****p<0.0001. Abbreviations: ns, non-significant.

250

Chapter 14

the majority of Cd69+Cd8+ teff cells in the liver do not express transcription factors t-bet and hobit.Memory formation and effector function is tightly regulated by transcription factors such as the T-box transcription factors T-bet, “homolog of Blimp1 in T cells” (Hobit) and Eomesodermin (Eomes).20–22 In the liver, a significantly less CD69+ TEFF cells expressed T-bet (mean 16.9%) as compared to blood TEFF (mean 77%, p<0.0001, Figure 4a,c). As in the blood, the majority of intrahepatic CD69- TEFF cells expressed T-bet (mean 63%). Hobit was expressed in CD69+CD8+ TEFF from the liver in significantly less cells (mean 44.2%) than in the blood (mean 90.7%) and CD69- liver CD8+ T cells (mean 88.9%, p<0.0001). In naïve peripheral blood CD8+ T cells expression these transcription factors was only observed in a small proportion of cells (mean T-bet 1.94%, Hobit 5.7% and Eomes 9.96%, Supplementary Fig. 7).

Intrahepatic Cd69+ trm express equal amounts of cytokine mrna as Cd69- liver cells. We next investigated the potential of liver-resident T-cells to produce antiviral and proinflammatory cytokines. Cytokine mRNA content was assessed in intrahepatic CD69+

figure 4. Frequency of perforin+, granzyme B+, Tbethi, Hobit+ and Eomes+ cells as a percentage of CD8+ TEFF (a) and TMEM (b) in control blood and liver (CD69+ and CD69-). Statistical analyses; paired t-test. (c) Representative FACS staining and gating for perforin, granzyme B, Eomes and T-bet in blood and liver CD69+ ad CD69- TEFF and TMEM cells. Statistical analyses; paired t-test. (d) Relative expression of cytokines interferon-γ (IFNγ), tumor necrosis factor-α (TNFα) and macrophage inflammatory protein 1β (MIP-1β) in sorted CD69+ and CD69-CD8+ T cells from control livers. Bars indicate median. Statistical analyses; Wilcoxon signed rank test. *p<0.05, **p<0.01, ***p<0.001 ****p<0.0001. Abbreviations: ns, non-significant.

(a)

(B)

(C)

(d)

251


14

and CD69- cells. No differences were observed in interferon-γ (IFN-γ), tumor necrosis factor-α (TNF-α) and macrophage inflammatory protein-1β (MIP-1β) mRNA expression between CD69+ and CD69-CD8+ T cells (Figure 4d).

Intrahepatic Cd69+ Cd8+ t cells express a distinct phenotype in chronic viral hepatitis.The proportion of CD8+ T cells in the liver of patients with HBV and HCV was similar to that of control patients (Figure 5a). The proportion of CD4+ T cells was significantly higher in the livers of patients with CHC (median 52.8%) as compared to control patients (median 28.5%, p=0.02, Figure 5a). As in control patients, the majority of CD8+ T cells in the liver of patients with CHB or CHC expressed CD69 (median 66.2%, 79.9% and 75.6%). A minority of these cells co-expressed CD103 and this population was significantly expanded in patients with CHC (median 25.7%) as compared to control patients (median 8.4%, p=0.03, Figure 5b). Similar to control patients, in CHB and CHC patients, most intrahepatic CD69+CD8+ T cells were CXCR6 positive (Supplementary Fig. 8a). HLA-DR/CD38, Ki67 and PD-1 were similarly expressed on CD69+CD8+ T cells in CHB, CHC and control patients. In CHB patients, significantly more CD69+CD8+ TEFF cells expressed cytotoxic protein granzyme B (median 74.4%) as compared to controls (median 48.1%, p=0.012, Figure 5d), while this was similar between CHC patients and controls. In CHC patients, significantly less CD69+CD8+ liver T cells expressed Hobit as compared to control CD69+CD8+ T cells (TEFF as well as TMEM, p=0.0025, Figure 5e). These differences between patient groups were not observed in intrahepatic CD8+ T cells that were CD69 negative (data not shown). The percentage of cells expressing perforin and transcription factors T-bet and Eomes did not differ in patients with CHB and CHC and controls (Figure 5c, Supplementary Fig. 8b).

dIsCussIonIn this study we have for the first time in man characterized a population of CD69+CD8+ T cells with a tissue resident-like phenotype which is abundantly present in the liver, but not in the peripheral blood. Expression of CD69 identifies TRM in many different human tissues,3,7 as well as on the majority of intrahepatic TRM in mice and resident-like NK cells in the human liver.10,13,23 A small proportion of intrahepatic CD69+CD8+ T cells expressed CD103, in accordance with data on mouse liver TRM.13 Subsets of human TRM including those in lung, brain, skin and gut3,24 co-express CD103, which forms a heterodimer with integrin β7 and recognizes E-cadherin, which is expressed on hepatocytes and bile duct epithelium. In patients with CHC, this population was enriched, possibly causing portal infiltrates and liver pathology.25 CD69 is known to antagonize the expression of the S1P receptor S1PR1,12 which can detect an S1P gradient leading out of the tissues.11,12,26 Lower expression of S1PR1 and its regulatory transcription factor KLF2 in intrahepatic CD69+CD8+ T cells suggests that these cells are not able to egress from the liver, and are indeed TRM, as described for mouse liver TRM.27 Parabiosis experiments, in which mice

252

Chapter 14

share their circulation, revealed that nearly all CD8+ T cells expressing CD69 are non-circulatory and interestingly, these TRM could be localized in the parenchyma as well as within large vessels in the liver.23 Here, immunohistochemistry localized CD69+CD8+ T cells in the parenchyma as well as within portal fields of the liver. Expression of CXCR6 has been associated with human liver infiltrating lymphocytes,14 intrahepatic NK cells,10 and protective responses against malaria in mice.15 High expression of CXCR6 could facilitate maintenance in the liver as its ligand, CXCL16, is present on hepatocytes and

figure 5. (a) Frequency of CD8+ and CD4+ T cells as a percentage of total CD3+ T lymphocytes in control (ctrl, n=7), CHB infected (HBV, n=3-4) and CHC infected (HCV, n=5-6) livers (number of samples depending on sample availability). Bars indicate median. Statistical analyses; Mann-whitney test. (b) Frequency of CD69+ and CD103+ as a percentage of total CD8+ T lymphocytes in control, HBV and HCV livers. Bars indicate median. Statistical analyses; Mann-whitney test. Frequency of perforin+ (c), granzyme B+ (d) and Hobit+ (e) CD8+ T cells as a percentage of intrahepatic CD69+CD8+ TEFF (upper graphs) and TMEM (lower graphs). Bars indicate median. Statistical analyses; Mann-whitney test. *p<0.05, **p<0.01.

(a) (B)

(C) (e)(d)

253


14

liver sinusoidal endothelial cells (LSEC) in healthy and HCV infected livers.14,28 Few naïve

cells could be identified in the population of CD69+CD8+ T cells from the liver, supporting

the hypothesis that CD69 identifies a distinct non-circulating population of memory

T cells.29,30

CD69 has originally been identified as a marker of recent activation. To investigate

the activation state of CD69+CD8+ T cells, we examined the expression of other markers

of recent and chronic stimulation. Low numbers of blood and liver CD8+ T cells were Ki67

positive, suggesting the absence of persistent T cell activation as observed in malaria

specific intrahepatic TRM.27 Nur77, a factor upregulated after recent TCR activation,17 was

not differentially expressed in CD69+ and CD69-CD8+ T cells from the liver, suggesting

that the expression of CD69 was not promoted by recent TCR cross-linking. CD69+CD8+

T cells in the liver contained a higher fraction of HLA-DR/CD38 double positive cells,

which is consistent with data regarding lung residing CD8+ T cells.31 The proportion of

HLA-DR expressing cells correlated with ALT in the peripheral blood but not in the liver,

suggesting in the liver immune activation is not the primary mechanism of HLA-DR/CD38

upregulation. In addition, CD69+CD8+ T cells expressed PD-1 at higher levels than CD69-

liver CD8+ T cells and blood CD8+ T cells which could play a role in maintaining local

tolerance, similar to what has been reported for CD103+ TRM in the brain.32 Local induction

of apoptosis to induce tolerance did not seem to play a role in the liver.33

A hallmark of effector type CD8+ T cells in the periphery is the presence of intracellular

cytotoxic proteins, such as perforin and granzyme B.2 Significantly less CD69+CD8+ TEFF

cells in the liver expressed cytotoxic proteins, as compared to peripheral effector cells,

similar to what has been observed in TRM in the skin,34 brain,35 and lung36, and in intrahepatic

NK cells.10 A reduced cytolytic activity in the brain has been postulated to prevent immune

pathology in response to low or non-specific stimuli35, a similar mechanism could be

at play in the liver. On the other hand, as part of the intrahepatic CD69+CD8+ T cell

population consisted of MAIT cells, it cannot be excluded that some of these differences

could be caused by this population which is generally low in cytotoxic content.37 In CHB

patients, more liver-resident TEFF as well as TMEM cells expressing granzyme B were observed

than in CHC and control patients, possibly antagonizing tolerance and adding to liver

damage. The infiltration of a dense non-virus specific T cell population has previously been

suggested to drive liver damage in viral hepatitis.9,38

T-bet, Hobit and Eomes are the leading transcription factors thought to govern

the process of T cell differentiation,21,22,39–41 as well as TRM development.13,22 T-bet and

Hobit were expressed by significantly less cells in the intrahepatic CD69+CD8+ T cell

population as compared to peripheral blood CD8+ T cells. As these transcription factors

play an important role in the regulation of cytolytic protein expression,41,42 the distinct

transcriptional profile observed here possibly induced a reduction in perforin and granzyme

B expression. A significantly larger proportion of intrahepatic CD69- CD8+ cells expressed

T-bet and Hobit as compared to CD69+ cells, whereas the IFN-γ mRNA content did not

differ between these populations. This suggests that in the liver, mechanisms additional

to T-bet and Hobit are regulators of IFN-γ production.21,41,43 Hobit expression has been

254

Chapter 14

shown to be important in the development of liver TRM in mice.13 As Hobit expression was observed in liver CD69+ T cells, this transcription factor may also be involved in TRM development in humans. However, significantly less intrahepatic CD69+CD8+ T cells expressed Hobit as compared to intrahepatic CD69- T cells and blood CD8+ TEFF cells, which are well known to express Hobit.21 Furthermore, in CHC livers, the proportion of Hobit expressing cells was reduced, which indicates that an alternative differentiation process occurs in these cells.

Previous data on human intrahepatic NK cells did not show differences between healthy donor tissue obtained from organ donors and resection margins of liver metastases.10 However, it would be interesting to evaluate the phenotype of liver the TRM population in truly healthy subjects in further studies. Other groups have already identified the presence of substantial populations of TRM in many tissues in organ donors.3,7 Similar to their results, in this study we have identified an intrahepatic T cell population with a specific tissue resident phenotype. The presence of a chronic viral infection in the liver seems to infer specific changes to the differentiation status of this intrahepatic CD69+CD8+ T cell population, and suggests an altered cytotoxic potential in CHB and specific regulation of differentiation and homing capabilities in CHC infection. Understanding the role of tissue-resident T cells in chronic viral hepatitis may help the development of future therapeutic interventions enabling the prevention of liver cirrhosis and hepatocellular carcinoma.

255


14

referenCes1. Knolle PA, Thimme R. Hepatic immune regulation

and its involvement in viral hepatitis infection. Gastroenterology 2014;146:1193–207.

2. Hamann D, Baars PA, Rep MH, et al. Phenotypic and functional separation of memory and effector human CD8+ T cells. J Exp Med 1997;186:1407–18.

3. Sathaliyawala T, Kubota M, Yudanin N, et al. Distribution and Compartmentalization of Human Circulating and Tissue-Resident Memory T Cell Subsets. Immunity 2013;38:187–97.

4. Watanabe R, Gehad A, Yang C, et al. Human skin is protected by four functionally and phenotypically discrete populations of resident and recirculating memory T cells. Sci Transl Med 2015;7:279ra39.

5. Park CO, Kupper TS. The emerging role of resident memory T cells in protective immunity and inflammatory disease. Nat Med 2015;21:688–97.

6. Farber DL, Yudanin NA, Restifo NP. Human memory T cells: generation, compartmentalization and homeostasis. Nat Rev Immunol 2014;14:24–35.

7. Thome JJC, Yudanin N, Ohmura Y, et al. Spatial Map of Human T Cell Compartmentalization and Maintenance over Decades of Life. Cell 2014;159:814–28.

8. Morrot A, Hafalla JCR, Cockburn IA, et al. IL-4 receptor expression on CD8+ T cells is required for the development of protective memory responses against liver stages of malaria parasites. J Exp Med 2005;202:551–60.

9. Abrignani S. Bystander activation by cytokines of intrahepatic T cells in chronic viral hepatitis. Semin Liver Dis 1997;17:319–22.

10. Stegmann KA, Robertson F, Hansi N, et al. CXCR6 marks a novel subset of T-bet lo Eomes hi natural killer cells residing in human liver. Sci Rep 2016;6:26157.

11. Skon CN, Lee J-Y, Anderson KG, et al. Transcriptional downregulation of S1pr1 is required for the establishment of resident memory CD8+ T cells. Nat Immunol 2013;14:1285–93.

12. Shiow LR, Rosen DB, Brdicková N, et al. CD69 acts downstream of interferon-alpha/beta to inhibit S1P1 and lymphocyte egress from lymphoid organs. Nature 2006;440:540–4.

13. Mackay LK, Minnich M, Kragten NAM, et al. Hobit and Blimp1 instruct a universal transcriptional program of tissue residency in lymphocytes. Science 2016;352:459–63.

14. Heydtmann M, Lalor PF, Eksteen JA, et al. CXC chemokine ligand 16 promotes integrin-mediated adhesion of liver-infiltrating lymphocytes to cholangiocytes and hepatocytes within the inflamed human liver. J Immunol 2005;174:1055–62.

15. Tse SW, Radtke AJ, Espinosa DA, et al. The chemokine receptor CXCR6 is required for the maintenance of liver memory CD8+ T cells specific for infectious pathogens. J Infect Dis 2014;210:1508–16.

16. Reantragoon R, Corbett AJ, Sakala IG, et al. Antigen-loaded MR1 tetramers define T cell receptor heterogeneity in mucosal-associated invariant T cells. J Exp Med 2013;210:2305–20.

17. Moran AE, Holzapfel KL, Xing Y, et al. T cell receptor signal strength in Treg and iNKT cell development demonstrated by a novel fluorescent reporter mouse. J Exp Med 2011;208:1279–89.

18. Appay V, Van Lier RAW, Sallusto F, et al. Phenotype and function of human T lymphocyte subsets: Consensus and issues. Cytom Part A 2008;73:975–83.

19. Gupta S, Gollapudi S. Effector Memory CD8 + T Cells Are Resistant to Apoptosis. Ann NY Acad Sci 2007;1109:145–50.

256

Chapter 14

20. Smith C, Elhassen D, Gras S, et al. Endogenous antigen presentation impacts on T-box transcription factor expression and functional maturation of CD8+ T cells. Blood 2012;120:3237–45.

21. Vieira Braga FA, Hertoghs KML, Kragten NAM, et al. Blimp-1 homolog Hobit identifies effector-type lymphocytes in humans. Eur J Immunol 2015;45:2945–58.

22. Mackay LK, Wynne-jones E, Freestone D, et al. T-box Transcription Factors Combine with the Cytokines TGF- b and IL-15 to Control Tissue-Resident Memory T Cell Fate. Immunity 2015;43:1101–11.

23. Steinert EM, Schenkel JM, Fraser KA, et al. Quantifying Memory CD8 T Cells Reveals Regionalization of Immunosurveillance. Cell 2015;161:737–49.

24. Casey KA, Fraser KA, Schenkel JM, et al. Antigen-Independent Differentiation and Maintenance of Effector-like Resident Memory T Cells in Tissues. J Immunol 2012;188:4866–75.

25. Shimizu Y, Minemura M, Murata H, et al. Preferential accumulation of CD103+ T cells in human livers; its association with extrathymic T cells. J Hepatol 2003;39:918–24.

26. Ledgerwood LG, Lal G, Zhang N, et al. The sphingosine 1-phosphate receptor 1 causes tissue retention by inhibiting the entry of peripheral tissue T lymphocytes into afferent lymphatics. Nat Immunol 2008;9:42–53.

27. Tse S-W, Cockburn IA, Zhang H, et al. Unique transcriptional profile of liver-resident memory CD8+ T cells induced by immunization with malaria sporozoites. Genes Immun 2013;14:302–9.

28. Geissmann F, Cameron TO, Sidobre S, et al. Intravascular immune surveillance by CXCR6+ NKT cells patrolling liver sinusoids. PLoS Biol 2005;3:e113.

29. Weninger W, Crowley MA, Manjunath N, et al. Migratory properties of naive, effector, and memory CD8(+) T cells. J Exp Med 2001;194:953–66.

30. Mackay CR, Marston WL, Dudler L. Naive and memory T cells show distinct pathways of lymphocyte recirculation. J Exp Med 1990;171:801–17.

31. de Bree GJ, van Leeuwen EMM, Out TA, et al. Selective accumulation of differentiated CD8+ T cells specific for respiratory viruses in the human lung. J Exp Med 2005;202:1433–42.

32. Wakim LM, Woodward-Davis A, Liu R, et al. The molecular signature of tissue resident memory CD8 T cells isolated from the brain. J Immunol 2012;189:3462–71.

33. Crispe IN, Dao T, Klugewitz K, et al. The liver as a site of T-cell apoptosis: graveyard, or killing field? Immunol Rev 2000;174:47–62.

34. Schaerli P, Ebert L, Willimann K, et al. A skin-selective homing mechanism for human immune surveillance T cells. J Exp Med 2004;199:1265–75.

35. Smolders J, Remmerswaal EBM, Schuurman KG, et al. Characteristics of differentiated CD8+ and CD4+ T cells present in the human brain. Acta Neuropathol 2013;126:525–35.

36. Piet B, de Bree GJ, Smids-Dierdorp BS, et al. CD8+ T cells with an intraepithelial phenotype upregulate cytotoxic function upon influenza infection in human lung. J Clin Invest 2011;121:2254–63.

37. Kurioka A, Walker LJ, Klenerman P, et al. MAIT cells: new guardians of the liver. Clin Transl Immunol 2017;6:e132.


39. Hersperger AR, Martin JN, Shin LY, et al. Increased HIV-specific CD8+ T-cell cytotoxic potential in HIV elite controllers is associated withT-bet expression. Blood 2011;117:3799–809.

40. Intlekofer AM, Takemoto N, Wherry EJ, et al. Effector and memory CD8+ T cell fate coupled by T-bet and eomesodermin. Nat Immunol 2005;6:1236–44.

257


14

41. Pearce EL, Mullen AC, Martins GA, et al. Control of effector CD8+ T cell function by the transcription factor Eomesodermin. Science 2003;302:1041–3.

42. van Gisbergen KPJM, Kragten NAM, Hertoghs KML, et al. Mouse Hobit is a homolog of the transcriptional repressor Blimp-1 that regulates NKT cell effector differentiation. Nat Immunol 2012;13:864–71.

43. Hombrink P, Helbig C, Backer RA, et al. Programs for the persistence, vigilance and

control of human CD8+ lung-resident memory T cells. Nat Immunol 2016;17:1467–78.

44. van den Brand M, Hoevenaars BM, Sigmans JHM, et al. Sequential immunohistochemistry: A promising new tool for the pathology laboratory. Histopathology 2014;65:651–7.

45. Schindelin J, Arganda-Carreras I, Frise E, et al. Fiji: an open-source platform for biological-image analysis. Nat Methods 2012;9:676–82.

aCKnowledGementsThe authors would like to thank Dr. L. Swadling, Dr. K.M.H. Heutinck and Prof. Dr. C.P. Engelfriet for critically revising the manuscript. In addition, the authors thank N. Claessen for assistance with immunohistochemistry, Dr. E.B.M. Remmerswaal and Ing. B. Hooibrink for technical assistance with FACS and D.S. Koenis for kindly providing Nur77 primers.


summary and future perspeCtIves

260

Chapter 15

thesIs summaryThe studies in this thesis address different aspects of viral hepatitis B and C infection. We have explored new therapeutic options for the treatment of chronic hepatitis B and C infection, including combination treatment for chronic hepatitis B, and a novel host-targeting agent for chronic hepatitis C infection. Furthermore, we have performed exploratory studies to investigate the effect of these treatment modalities on the antiviral immune response. In addition, we performed studies to clarify the immune response in blood and liver during the natural course of hepatitis B or C virus infection.

part I. Immune responses in patients with an acute hepatitis B infectionIn chapter 2, we have analysed the temporal immune response in nine patients who had an acute hepatitis B infection. In general, over 95% of adult patients who have an acute hepatitis B infection spontaneously clear the virus. Of our nine patients, eight spontaneously resolved the infection whereas one patient did not clear the virus. We therefore had the unique opportunity to analyse immune responses during the transition from acute to chronic infection in this patient. We were able to observe differences in innate and adaptive immune effectors such as NK cells and T cells that possibly contributed to development of chronic infection. Furthermore, we observed a significant population of NK cells expressing activation markers, which did not normalize in the patient with chronic infection. Interestingly, in contrast to patients who cleared the acute infection, the patient who developed chronic infection had no measurable HBV-specific T cell responses at any time point during acute hepatitis B infection.

part II. Immune responses in patients with chronic hepatitis B infection In chapter 3, we have analysed the outcome after five years of follow-up in our cohort of HBeAg positive and –negative CHB patients treated with pegylated interferon alfa-2a (pegIFN) and adefovir combination therapy. The patients in this study had a high viral load and liver inflammation, and thus an indication for treatment. At two years of follow-up, relatively high rates of HBsAg loss (14%) were observed. We analysed whether this outcome was sustained during long term follow-up of five years. Our analysis showed that the high proportion of functional cure in our study is sustained over time and that low HBsAg levels at baseline were significantly associated with HBsAg loss in HBeAg negative patients. However, the majority of all patients (60%) met the criteria for retreatment with a nucleos(t)ide analogue, emphasizing that success of combination therapy is limited and that selection of suitable patients is paramount. In chapter 4, HBV-specific T cells responses were analysed in patients treated with the combination of pegIFN and adefovir. Before therapy, HBV-specific T cells could not be detected direct ex vivo nor after in vitro expansion. However, during follow-up, HBV-specific T cells could be selectively expanded in vitro in patients with therapy induced HBsAg loss. The restoration of the proliferative potential of HBV-specific T cells was not observed in patients without HBsAg loss, implying that HBV-specific T cell function is restored in patients with HBsAg loss after combination therapy. In addition, in chapter 5, we showed that this therapy induced drastic changes in

261

SUMMARY AND FUTURE PERSPECTIVES

15

the NK cell compartment. Significant differences in NK cell characteristics were observed between patients who cleared HBsAg versus patients who did not, at baseline as well as at end of treatment. A significant correlation between ALT and NK cell markers of activation suggested that NK cells are important players in the immune response against hepatitis B. In chapter 6 we further focused on NK cell characteristics. We identified a single-nucleotide polymorphism (SNP) within the human leukocyte antigen class C (HLA-C) region that was associated with response in HBeAg positive CHB patients. This SNP was particularly interesting because it defines the ability of HLA-C to bind to Killer-cell immunoglobulin-like receptors (KIRs), which are expressed on NK cells. After combining HLA-C and KIR genotypes in patients, the functionally relevant combination of HLA-C2 and KIR2DL1 genotype was significantly associated with a combined response (defined as HBeAg loss, HBV DNA<2,000 IU/mL, ALT normalization) after combination therapy. This association was maintained after correction for other known predictors of response in multivariable analyses. Together our data imply that HBV-specific T cells as well as NK cells may play an important role in the process of HBsAg clearance. In chapter 7, we have identified a SNP from a genome wide association study (GWAS) which was associated with response to combination therapy. The SNP was located in the SLC16A9 gene, and was significantly associated with combination treatment induced HBsAg loss. This SNP in SLC16A9, which is a transporter, had previously been associated with lower plasma carnitine levels, which we confirmed in our cohort. Baseline plasma carnitine levels were lower in patients who had HBsAg loss as compared to non-responders. Furthermore, in vitro experiments suggested that high carnitine levels can have an inhibitory effect on HBV-specific as well as non-specific T cells.

In chapter 8, we describe the results of a clinical study in which chronic hepatitis B patients with a low viral load were treated with combination therapy. In this randomised controlled open-label trial, 135 patients were included and randomised into 3 arms: pegIFN plus adefovir (n=46), pegIFN plus adefovir (n=45), and no treatment (n=43). At 24 weeks of treatment-free follow-up, 4 (4%) patients in the combination treatment arms had lost HBsAg, as compared to none (0%) in the control arm, which was not statistically different. Our results indicated that treatment of CHB patients with low viral load with combination therapy does not lead to increased clearance of the hepatitis B virus. We did observe a significant decrease in HBsAg levels in the patients who were treated as compared to untreated patients, however, longer follow-up will have to reveal whether this is associated with eventual HBsAg loss. In chapter 9, we discuss the results of a study by Bourlière et al. in which pegIFN was used as add-on therapy in patients who were treated with nucleotide analogues for at least one year. In this randomised, controlled, open-label trial, 185 patients with HBeAg-negative chronic hepatitis B who were on nucleos(t)ide analogue therapy with undetectable HBV DNA were randomly allocated to receive pegIFN add-on therapy for 48 weeks or no additional therapy. The primary endpoint was HBsAg loss at week 96 by intention-to-treat analysis.

In chapter 10, we focused on a cohort of 12 chronic hepatitis B patients treated with a novel compound, REP 2139. This compound targets HBsAg and inhibits its release from the hepatocyte. We show that in addition to a decline in plasma HBsAg and HBV DNA,

262

Chapter 15

a substantial decline in HBV RNA levels occurs. Interestingly, this result is not obtained by treatment with nucleotide analogues. Furthermore, during REP 2139 treatment, IP-10, IFNγ and IL-21 significantly increased, while IL-7 serum concentration significantly decreased, suggesting that REP 2139 treatment, with the ensuing inhibition of HBsAg, leads to immune modulation.

part III. Immune responses in patients with chronic hepatitis C infectionIn chapter 11 we investigated the safety and efficacy of an antisense oligonucleotide targeting miR-122 called RG-101, in chronic hepatitis C patients. Due to GalNAc conjugation, RG-101 specifically targets the liver. This novel compound binds to the liver specific micro-RNA miR-122 and thereby inhibits its stabilising effect on HCV RNA. We show that RG-101 is safe and well tolerated. Furthermore, of 28 patients who were dosed with a single injection of RG-101, 15 patients had an HCV RNA level below the limit of quantification at week 8 after dosing. In addition, 3 patients were HCV RNA negative for over 76 weeks. Viral rebound at or before week 12 was associated with the appearance of resistance associated substitutions in miR-122 binding regions in the 5’UTR of the HCV genome. The exact mechanism by which RG-101 can lead to HCV RNA reduction and sustained virological response is at present not known. In chapter 12 we have analysed the innate as well as adaptive arm of the immune system in RG-101 dosed patients. We show that dosing with RG-101 leads to a decline in plasma IP-10 levels, as well as an increase in the percentage of NK cells. Markers for NK cell activation and differentiation (including TRAIL) normalised after RG-101 dosing, similar to what has been shown in CHC patients treated with DAA’s. Furthermore, we show that by week 8 post RG-101 injection, there is no significant change in the functional HCV-specific IFN-γ-T cell response. Furthermore, no change in the magnitude of HCV-specific T cell responses was observed at later time points, including the 3 patients who were HCV RNA negative at 76 weeks post dosing. In chapter 13, we show that patients with virological rebound following RG-101 dosing can be successfully treated with direct acting antiviral therapy for 12-24 weeks. We show that 29 patients (with and without prior RG-101 dosing) all have a sustained virological response 12 week (SVR12) after treatment with sofosbuvir and daclatasvir ± ribavirin. Furthermore, in this study we have analysed immune characteristics and plasma miR-122 levels of chronic hepatitis C patients with and without prior RG-101 dosing. We show that successful treatment of chronic hepatitis C patients with direct-acting antivirals results in reduction of broad immune activation, but does not lead to a restoration of HCV-specific T cell function. Plasma miR-122 levels were elevated only in chronic hepatitis C patients without prior RG-101 dosing and normalized when sustained virological response was achieved.

part Iv. Characterisation of intrahepatic t cells in patients with chronic hepatitis B or CIn chapter 14, we show that the liver harbors a T cell population with the phenotypic traits of tissue resident T cells. We show that a substantial population of intrahepatic CD8+ T

263


15

cells express CD69 and is low in transcription factor KLF2 and the S1P1 receptor, indicating tissue residency. These cells are phenotypically distinct from intrahepatic CD8+ T cells without CD69 expression, as well as peripheral blood CD8+ T cells. In accordance with liver resident NK cells CD69+ liver T cells express high levels of the chemokine receptor CXCR6 suggesting altered migratory abilities. Furthermore, these cells have a reduced cytolytic capacity and form a transcriptionally distinct population. These findings suggest that liver CD69+CD8+ T cells are more tolerant and fundamentally different from peripheral blood CD8+ T cells. In patients who have a chronic hepatitis B or C infection, intrahepatic CD69+CD8+ T cells had a distinct phenotype with an increase in cytotoxic protein content.

future perspeCtIvesHaving been immersed in the world of viral hepatitis for four years, many complex issues have been clarified during this time. However, some general questions remain. To some of them, we will perhaps never find a suitable answer. Others might be explained in the near future or sometimes, even in science, the answer will likely be more of an opinion instead of a fact.

do we need a cure for hBv infection?Even though a preventive vaccine with 95% effectivity has been available since 1982, chronic hepatitis B virus infection is still a major global health problem. An estimated 240 million people were living with chronic hepatitis B virus infection in 2015, who were at risk of developing hepatocellular carcinoma and liver cirrhosis.1 When patients have high viremia and inflammation of the liver (increased transaminase values in the blood) or fibrosis, treatment is indicated.2 Most patients who have an indication to start treatment are given nucleos(t)ide analogues (NUCs). These agents inhibit the last step in viral replication by being incorporated into the viral DNA, which terminates replication. Treatment with NUCs is generally well tolerated, safe and leads to efficient suppression of viral replication.3 Successful treatment will improve the outcome with a substantial decrease in the chance of developing cirrhosis or hepatocellular carcinoma. A drawback of treatment with NUCs lies in the duration of treatment, which is generally life-long. Discontinuation of NUCs is associated with ALT flares and viral relapse. Furthermore, treatment with NUCs hardly ever leads to functional cure (HBsAg loss with or without anti-HBs antibodies) which is regarded as the optimal outcome of the infection. In patients with functional cure, there is no more production of viral proteins and viral replication has halted. In some hepatocytes cccDNA will remain present, but it is not transcriptionally active.

In the past, many attempts have been made to improve treatment for chronic hepatitis B virus infection to achieve functional cure in more patients. This includes combinations of peg-interferon and NUCs as presented in this thesis (chapter 3, 8 and 9). Unfortunately, many of these attempts have thus far not lead to improved outcome. In addition, data from clinical trials have been re-analysed to seek predictive markers that can select patients that have a higher change of response to treatment, particularly those that respond with a functional cure (e.g. chapter 6,7). In practice, it is still difficult to find a good marker that

264

Chapter 15

is easily accessible in the clinic and applicable to large groups of patients. Currently, novel compounds that can directly interfere with HBV are being developed for the treatment of chronic hepatitis B virus infection. These include HBV entry inhibitors, HBsAg release inhibitors (e.g. chapter 10), short interfering RNAs (siRNAs) and new immune modulators such as TLR agonists.4 Our research has taught us a lot about the complexity of curing HBV. Furthermore, the question is how much is there to gain with novel therapies? The majority of patients with chronic hepatitis B are in the inactive carrier state. They currently have no treatment indication, and a very favourable outcome.5 In these patients, the viral load is spontaneously suppressed, most likely by the immune system. As described above, the outcome in patients who are successfully treated with NUCs is very good, and the medication is safe. New treatment modalities therefore should have an even better safety profile than the currently widely accepted NUCs, consist of a finite duration or further improve the outcome.6 This is possible by inducing functional cure with clearance of HBsAg. However, elimination of the virus from the liver is very difficult due to the formation of the viral mini-chromosome cccDNA. The question of how to target the cccDNA has currently not yet been answered. Most likely, a combination of compounds directly targeting the virus as well as those targeting the antiviral immune response from the host will be needed. Curing hepatitis B would be the favourable outcome for all patients. For the time being, current management of patients with chronic hepatitis B is capable of reducing risk profiles and identifying cases that need close follow-up, such as patients with cirrhosis.7 Possibly, novel therapeutics will lead the way to a cure for HBV infection.

has hCv been cured?The burden of chronic hepatitis C virus infection will likely decrease, as over the past decade highly efficient antiviral drugs have become available. With these drugs, more than 95% cure rates are achieved in trials as well as in ‘real-world data sets’. However, as estimated recently, a considerable disease burden will remain8 because a significant portion of patients is still unaware that they are infected with HCV. This means the identification of patients at risk for HCV needs to be improved.9 Furthermore, the availability of treatment should be simple and affordable for all patients. For difficult-to-treat populations, such as patients with previous DAA failure and viral resistance, it is paramount that novel treatment options are explored. These can consist of new direct-acting antivirals or host-targeting agents, as explored in this thesis (chapter 11). Lastly, a proportion of patients with liver cirrhosis who have cleared HCV infection will still need to be closely monitored over time, as their risk of developing HCC may still be present. In the USA, in the next 35 years an estimated 320,000 patients will die, 203,000 will develop decompensated liver cirrhosis and 157,000 will develop hepatocellular carcinoma, even with the availability of all-oral direct-acting antivirals.8

what is the role of the immune system in viral hepatitisAfter decades of research, the field of hepatitis B and C immunology has gained a lot of insight in the pathogenesis of acute and chronic HBV and HCV infection. During acute HBV

265


15

and HCV infection, functionally active virus-specific T cells are important for clearance of the virus.10,11 Furthermore, NK cells are highly activated early during infection.12 However, the events that lead to chronic hepatitis B infection are difficult to study as patients are often asymptomatic and acute infection in adulthood often leads to spontaneous resolution. In this thesis, we compared the course of eight patients who cleared acute HBV infection with one patient who developed chronicity (chapter 2). We confirm previous data that showed activation of NK cells early during infection.10,12,13 Furthermore, we show that HBV-specific T cells are functionally active in patients who have cleared the virus, whereas these cells could not be observed in the patient who did not clear HBV. Further research is needed, by increasing the cases of persistent infection, which will allow better understanding of the events that lead to chronicity. Understanding dysfunction of immune cells such as NK and T cells in chronic hepatitis B is important to specify immune profiles responsible for viral persistence.

During chronic infection, persistently high antigen loads lead to a process called T cell exhaustion. Exhausted virus-specific T cells are dysfunctional and play an important role in the persistence of the infection during chronicity. Furthermore, NK cells from patients infected with HBV and HCV display an altered function and phenotype as compared to healthy patients.14 The drawback of most of this research is that it has largely been performed in peripheral blood, which may not always adequately represent the situation at the site of infection; the liver. Recently, new T cell populations have been described, that do not circulate, but are resident in specific organs such as the liver.15 In this thesis, we were able to analyse this intrahepatic T cell population (chapter 14). Even though liver samples are hard to collect, future research would benefit from being more focused on intrahepatic sampling as these cells have a significantly different phenotype from peripheral blood T cells.

role of immune system in therapeutic clearanceAs described before, during chronic infection with hepatitis B or C virus, virus-specific T cells are generally exhausted and functionally inactive. However, in hepatitis B virus infected patients with low viral load (inactive carriers), virus-specific T cell exhaustion is less severe.16 Furthermore, in patients whose viral load has been long term suppressed by NUCs, virus-specific T cell function can be partially restored despite high HBsAg levels.17 In this thesis, we described similar restoration of HBV specific T cells after successful combination therapy with pegIFN and NUCs (chapter 4). Furthermore, successful treatment led to changes in NK cell phenotype and function (chapter 5). However, after successful treatment for HCV, no improvement in T cell function was observed (chapter 12 and 13). The question remains whether the virus specific T cells play an active role in viral clearance upon treatment or whether their function improves passively, as a result of the discontinuation of the suppression by the antigen load. The answer to this important question may help to develop novel therapeutic strategies to reconstitute antiviral functions and enhance viral clearance.

266

Chapter 15

referenCes 1. Schweitzer, A., Horn, J., Mikolajczyk, R. T.,

Krause, G. & Ott, J. J. Estimations of worldwide prevalence of chronic hepatitis B virus infection: A systematic review of data published between 1965 and 2013. Lancet 386, 1546–1555 (2015).

2. Lampertico, P. et al. EASL 2017 Clinical Practice Guidelines on the management of hepatitis B virus infection. J. Hepatol. (2017) doi:10.1016/j.jhep.2017.03.021.

3. Buti, M. et al. Seven-Year Efficacy and Safety of Treatment with Tenofovir Disoproxil Fumarate for Chronic Hepatitis B Virus Infection. Dig. Dis. Sci. 60, 1457–1464 (2015).

4. Brahmania, M., Feld, J., Arif, A. & Janssen, H. L. A. New therapeutic agents for chronic hepatitis B. Lancet Infect. Dis. 16, e10–e21 (2016).

5. Invernizzi, F., Viganò, M., Grossi, G. & Lampertico, P. The prognosis and management of inactive HBV carriers. Liver Int. 36 Suppl 1, 100–4 (2016).

6. Sugarman, J. et al. Ethics and hepatitis B cure research. Gut 0, gutjnl-2016-313009 (2016).

7. Tong, M. J., Nguyen, M. O., Tong, L. T. & Blatt, L. M. Development of Hepatocellular Carcinoma After Seroclearance of Hepatitis B Surface Antigen. Clin. Gastroenterol. Hepatol. 7, 889–893 (2009).

8. Chhatwal, J. et al. Hepatitis C Disease Burden in the United States in the era of oral direct-acting antivirals. Hepatology 64, 1442–1450 (2016).

9. Pawlotsky, J. M. The end of the hepatitis C burden: Really? Hepatology 64, 1404–1407 (2016).

10. Fisicaro, P. et al. Early kinetics of innate and adaptive immune responses during hepatitis B virus infection. Gut 58, 974–82 (2009).

11. Sung, P. S., Racanelli, V. & Shin, E.-C. CD8(+) T-Cell Responses in Acute Hepatitis C Virus Infection. Front. Immunol. 5, 266 (2014).

12. Dunn, C. et al. Temporal analysis of early immune responses in patients with acute hepatitis B virus infection. Gastroenterology 137, 1289–300 (2009).

13. Webster, G. J. M. et al. Incubation Phase of Acute Hepatitis B in Man: Dynamic of Cellular Immune Mechanisms. Hepatology 32, 1117–1124 (2000).

14. Rehermann, B. Pathogenesis of chronic viral hepatitis: differential roles of T cells and NK cells. Nat. Med. 19, 859–868 (2013).

15. Thome, J. J. C. et al. Spatial Map of Human T Cell Compartmentalization and Maintenance over Decades of Life. Cell 159, 814–828 (2014).

16. Boni, C. et al. Characterization of hepatitis B virus (HBV)-specific T-cell dysfunction in chronic HBV infection. J. Virol. 81, 4215–25 (2007).

17. Boni, C. et al. Restored function of HBV-specific T cells after long-term effective therapy with nucleos(t)ide analogues. Gastroenterology 143, 963–73.e9 (2012).

nederlandse samenvattInG

lIst of ContrIButInG authors

lIst of puBlICatIons

phd portfolIo

CurrICulum vItae

270

Appendix

nederlandse samenvattInG hepatitis B virus infectieHet hepatitis B virus (HBV) is een DNA virus dat leverontsteking (hepatitis) kan veroorzaken. Geschat wordt dat wereldwijd 1 op de 3 mensen ooit geïnfecteerd is geweest met het hepatitis B virus. Daarnaast zijn er circa 240 miljoen mensen die chronisch geïnfecteerd zijn met het hepatitis B virus. Hepatitis B virus infectie komt veel voor in endemische gebieden als Afrika en Azië, met een prevalentie van meer dan acht procent. In Nederland ligt de prevalentie veel lager (<1%) en komt hepatitis B infectie vooral voor bij risicogroepen, zoals migranten. Wanneer een persoon op de volwassen leeftijd wordt geïnfecteerd met HBV, is de kans groot (>90%) dat het immuunsysteem zorgt voor een goede afweerrespons, en dat het virus spontaan opgeruimd (geklaard) wordt. Wordt het hepatitis B virus overgedragen van moeder op kind, bijvoorbeeld bij de geboorte of in de eerste levensjaren, dan is de kans groot (>90%) dat een acute infectie overgaat in een chronische infectie. Tijdens deze chronische infectie blijft het virus constant aanwezig in de lever. Deze chronische infectie kan voor leverschade zorgen, bijvoorbeeld in de vorm van verlittekening van de lever (fibrose of cirrose) of leverkanker (hepatocellulair carcinoom). Als gevolg hiervan sterven jaarlijks wereldwijd 800.000 mensen. Bij een hoge virus load en tekenen van leverontsteking en fibrose is het risico op complicaties groter en bestaat er een behandel indicatie. Bij de behandeling kan gekozen worden tussen gepegyleerd interferon alfa, een lichaamseigen stof die het immuunsysteem stimuleert en een antivirale werking heeft, en nucleos(t)ide analogen, die de replicatie van het virus remmen. Nucleos(t)ide analogen moeten vaak levenslang gebruikt worden, omdat na staken van de behandeling het virus zich weer volop kan vermenigvuldigen, en voor leverontsteking kan zorgen. Gepegyleerd interferon alfa hoeft slechts een jaar gebruikt te worden, maar deze behandeling gaat gepaard met hevige bijwerkingen. Beide middelen leiden in slechts enkele gevallen tot klaring van het virus, en zullen vooral het virus onderdrukken en zorgen voor normalisatie van de leverenzymen in het bloed (de transaminasen ASAT en ALAT).

Indien er weinig virus deeltjes in het bloed aanwezig zijn en de leverenzym waarden in het bloed normaal zijn, wordt er ook wel van hepatitis B dragerschap gesproken. Er bestaat dan geen behandelindicatie, omdat het eigen afweersysteem het virus al voldoende onder controle heeft. Er kunnen echter in deze fase ook opvlammingen van het virus voorkomen, waarbij de virus load en leverwaarden stijgen en behandeling nodig is. Daarom moeten hepatitis B virus dragers levenslang gecontroleerd worden.

hepatitis C virus infectieHet hepatitis C virus (HCV) is een RNA virus en veroorzaakt eveneens een leverontsteking. Wereldwijd zijn 170 miljoen mensen chronisch geïnfecteerd met het hepatitis C virus. Chroniciteit is vaak het gevolg van een infectie op volwassen leeftijd en kan tot lever-gerelateerde complicaties als cirrose en hepatocellulair carcinoom leiden. Bij een hepatitis C virus infectie bestaat er in feite een behandelindicatie bij iedere patiënt. De behandeling voor chronische hepatitis C heeft de afgelopen decennia een enorme ontwikkeling

271

Appendix

&

doorgemaakt. Eerder konden patiënten enkel worden behandeld met een combinatie van

gepegyleerd interferon alfa en ribavirine, waarbij veel bijwerkingen optraden en erg lage

HCV klaringspercentages gehaald werden. Inmiddels zijn er de zogenaamde ‘direct acting

antivirals’ (DAA’s), waarbij meer dan 95% van de patiënten het virus volledig kwijtraakt.

Omdat er nog steeds patiënten zijn die niet reageren op de nieuwe DAA’s is de zoektocht

naar nieuwe middelen voor de behandeling van chronische hepatitis C nog steeds in gang.

antivirale afweerresponsHepatitis B virus en hepatitis C virus zijn non-cytopatische virussen. Dat wil zeggen dat

ze zelf geen schade aan de levercel (hepatocyt) veroorzaken. De schade aan de lever

die kan optreden bij hepatitis B en C virus infecties is het gevolg van activatie van het

immuunsysteem, dat het virus probeert op te ruimen. Eiwitten die door de virussen

geproduceerd worden, worden herkend door het immuunsysteem en zorgen voor

stimulatie van een bepaalde immuuncellen, zoals de T en B cel. De T cellen kunnen

virus geïnfecteerde levercellen heel specifiek herkennen en vervolgens uitschakelen, of

signaalstoffen uitscheiden die zorgen voor neutralisatie van het virus. Bij een chronische

infectie kan de constante stimulatie van T cellen leiden tot ‘uitputting’ van deze cellen,

waardoor zij niet meer functioneel zijn.

B cellen kunnen antilichamen produceren die het hepatitis B virus kunnen identificeren en

neutraliseren. Tijdens een (chronische) infectie zijn er meer virusdeeltjes dan antilichamen,

en zijn alle antilichamen gebonden aan virusdeeltjes. Wanneer het virus volledig

opgeruimd is worden anti-HBs antilichamen aantoonbaar in het bloed. Deze antilichamen

beschermen dan tegen een nieuwe infectie met het hepatitis B virus. Antilichamen gericht

tegen het hepatitis C virus zijn vaak niet beschermend tegen een nieuwe infectie omdat

er vele varianten zijn van het virus, en de geproduceerde antilichamen specifiek zijn voor

slechts een type.

Een ander type afweercel, de natural killer (NK) cel, gaat minder specifiek te werk,

en herkent iedere cel die abnormaal is, bijvoorbeeld doordat hij geïnfecteerd is met een

virus. Tijdens een acute infectie worden NK cellen geactiveerd en helpen ze in de initiële

fase van infectie de virus geïnfecteerde cellen op te ruimen. Ook NK cellen laten een

afwijkend patroon zien tijdens chronische infectie met hepatitis B of C virus. De NK cellen

zijn dan over-geactiveerd en voeren hun functie niet meer goed uit.

dit proefschriftDe focus van dit proefschrift was om immuun responsen te analyseren in patiënten

geïnfecteerd met het hepatitis B en C virus, om beter te begrijpen welke rol bepaalde

immuun cellen spelen bij de klaring van het virus.

deel IIn hoofdstuk 2 werd de immuunrespons beschreven tijdens een acute hepatitis B virus

infectie. Negen patiënten met een acute hepatitis B infectie werden een half jaar vervolgd.

272

Appendix

In de tijd werden eerst de NK cellen en later de hepatitis B-specifieke T cellen geactiveerd. Eén van de negen patiënten klaarde het hepatitis B virus niet. Bij deze patiënt werd een afwijkend NK cel activatie patroon geobserveerd. Daarnaast konden in deze patiënt geen geactiveerde hepatitis B-specifieke T cellen worden gedetecteerd, wat mogelijk heeft bijgedragen aan de ontwikkeling van de chronische infectie.

deel IIHet tweede deel van dit proefschrift werden patiënten met een chronische hepatitis B virus infectie bestudeerd en werd de immunologische afweer op verschillende manieren onderzocht. Om te beginnen werden in hoofdstuk 3 de lange-termijn resultaten van een eerdere behandelstudie besproken. Chronische hepatitis B patiënten met een behandelindicatie werden behandeld met een combinatie van gepegyleerd interferon alfa en een nucleotide analoog; adefovir. In de initiële studie had twee jaar na deze therapie een aanzienlijk deel (14%) van de patiënten het hepatitis B virus geklaard, en de huidige studie laat zien dat deze respons ook na 5 jaar follow-up behouden is. Desalniettemin moest de meerderheid van de patiënten (60%) herbehandeld worden met nucleos(t)ide analogen vanwege een hoge virusload en verhoogde leverenzymen na staken van de combinatiebehandeling. hoofdstuk 4 toonde aan dat de hepatitis B-specifieke T cellen een deel van hun activiteit terug kregen in de patiënten die het virus geklaard hadden na bovenstaande therapie. Dit herstel werd niet geobserveerd bij patiënten die niet reageerden op de behandeling. In hoofdstuk 5 werden de veranderingen in het NK cel compartiment beschreven in dezelfde patiëntengroep met chronische hepatitis B infectie. De behandeling met gepegyleerd interferon alfa en adefovir induceerde een activatie van NK cellen, en bepaalde veranderingen werden alleen bij patiënten met klaring van het virus geobserveerd. De data suggereerde dat naast T cellen, ook NK cellen een belangrijke rol spelen bij de klaring van het virus. In hoofdstuk 6 werd er gekeken naar het effect van genetische variatie in relatie tot respons op behandeling met gepegyleerd interferon alfa en adefovir. Genetische variatie in het gen HLA-C is van invloed op de werking van specifieke receptoren die tot expressie worden gebracht door NK cellen, zogenaamde KIR’s. Bij patiënten met een goede respons op behandeling kwam een bepaalde combinatie van receptor en ligand (KIR 2DL1 met HLA-C groep 2) vaker voor dan in de non-responders. In hoofdstuk 7 werd met een genome wide association study (GWAS) een gen variatie of single-nucleotide polymorphism (SNP) geïdentificeerd die geassocieerd is met respons op behandeling met gepegyleerd interferon alfa en adefovir. Deze SNP was gelokaliseerd in het gen SLC16A9 dat betrokken is bij het metabolisme van carnitine. Na meten van de carnitine concentratie in het patiëntencohort, bleek dat patiënten met de SNP een lagere plasma carnitine spiegel hadden en een hogere kans op klaring van het hepatitis B virus dan patiënten zonder de SNP. Bij in vitro experimenten werd vervolgens de remmende werking van carnitine op de deling van hepatitis B-specifieke T cellen aangetoond.

In hoofdstuk 8 worden de resultaten van een studie met inactieve chronische hepatitis B patiënten besproken. In deze studie werden hepatitis B dragers met een lage virusload een jaar behandeld met dezelfde combinatietherapie als waar eerder chronische hepatitis

273

Appendix

&

B patiënten met een hoge virus load en behandelindicatie werden behandeld. Het doel van deze studie was om klaring van het virus te induceren bij hepatitis B virus dragers, echter slechts een minderheid (4%) van de patiënten klaarde het virus en er was geen verschil tussen de behandelde groep en de niet behandelde (controle) groep. Dit onderzoek toonde aan dat behandeling met een combinatie van de huidige antivirale middelen voor chronische hepatitis B infectie niet voldoet om het virus te klaren. In hoofdstuk 9 werd een vergelijkbare studie van Bourlière et al. besproken. In deze studie werden patiënten waarbij de virusreplicatie sinds langere tijd onderdrukt was met nucleos(t)ide analogen behandeld met de toevoeging van gepegyleerd interferon alfa voor 1 jaar. Ook van deze studie waren de klaringspercentages lager dan gehoopt, en werd er geen verschil in respons geobserveerd tussen de patiënten die behandeld werden met een toevoeging van gepegyleerd interferon alfa en patiënten die doorgingen met alleen therapie met nucleotide analogen.

In hoofdstuk 10 werd verder ingegaan op een cohort chronische hepatitis B patiënten met een hoge virusload dat behandeld is met een nieuw middel voor de behandeling van chronische hepatitis B infectie: REP-2139. Dit middel remt het vrijkomen van HBsAg, een eiwit dat wordt aangemaakt door het hepatitis B virus, en in overmaat in de circulatie aanwezig is. De hypothese voor het werkingsmechanisme van dit middel was dat grote hoeveelheden van dit eiwit de afweer cellen remt, en dat door deze rem weg te nemen het immuunsysteem zijn functie weer kan uitvoeren. Naast afname van de virusload werd een daling van het HBV RNA, de voorloper van HBV DNA, vastgesteld. Daarnaast werden, na behandeling met het nieuwe middel, verschillende signaalstoffen van het immuunsysteem in het bloed aangetoond, hetgeen kan duiden op activatie van het immuunsysteem.

deel IIIIn hoofdstuk 11 werd de veiligheid en effectiviteit van een nieuw middel (RG-101) voor de behandeling van chronische hepatitis C virus infectie onderzocht. Het hepatitis C virus is voor de replicatie afhankelijk van een klein stukje gastheer RNA; het micro-RNA-122. RG-101 remt dit micro-RNA waardoor de replicatie van het hepatitis C virus ook geremd wordt. In de studie werden 32 patiënten behandeld met een enkele injectie van RG-101. Het middel werd goed verdragen, en daarnaast was bij 15 patiënten het hepatitis C virus niet meer aantoonbaar 8 weken na de injectie. Echter bij het grootste gedeelte van de patiënten werd het virus weer aantoonbaar. Bij 3 patiënten bleef het virus meer dan anderhalf jaar afwezig, wat liet zien dat dit middel potentie zou kunnen hebben als antiviraal middel tegen chronische hepatitis C virus infectie. In hoofdstuk 12 werd gekeken of het immuunsysteem een leidende rol had in de effecten van RG-101 op hepatitis C virus infectie. Na daling van de virusload werden voornamelijk verschillen geobserveerd in het NK cel compartiment, die vooral geassocieerd leken met afname van de mate van ontsteking. Daarnaast werd er geen herstel van de functie van de hepatitis C virus-specifieke T cellen vastgesteld, hetgeen suggereert dat deze cellen geen rol hebben gespeeld bij het klaren van het virus. In hoofdstuk 13 werd beschreven dat patiënten die eerder behandeld waren met RG-101 alsnog succesvol behandeld kunnen worden met

274

Appendix

DAA’s. In deze studie werden 29 patiënten behandeld met DAA’s, waarvan 18 eerder aan het onderzoek met RG-101 hadden deelgenomen. Alle patiënten klaarden het virus na behandeling. Ook bij deze patiënten werd geen herstel gezien van de hepatitis C virus-specifieke T cellen. Wel nam de algehele immuun activatie af in deze patiënten na klaring van het hepatitis C virus.

deel IvIn hoofdstuk 14 werd een lever-specifieke populatie van T cellen beschreven die phenotypisch veel lijkt op weefsel residente T cellen zoals die recent beschreven zijn in andere organen. Deze T cellen waren fenotypisch anders dan de T cellen die in het bloed gevonden werden en leken een verminderde capaciteit te hebben om andere cellen uit te kunnen schakelen. Daarnaast hadden deze cellen in patiënten met chronische hepatitis B en C virus infectie een specifiek fenotype, wat aangaf dat zijn een rol zouden kunnen spelen in de leverpathologie bij deze virale infecties.

Conclusie Dankzij de studies beschreven in dit proefschrift hebben we meer inzicht gekregen in de klinische resultaten van verschillende behandelingen voor chronische hepatitis B en C infectie. Daarnaast hebben we kennis vergaard over de immunologische achtergrond van de geobserveerde effecten. Deze kennis kan in de toekomst bijdragen om, met name voor hepatitis B, nieuwe aangrijppunten voor therapie te selecteren.

275

Appendix

&



m. al-mahtabBangabandhu Sheikh Mujib Medical

University, Dhaka, Bangladesh

e. BarnesNuffield department of Medicine & Oxford

NIHR BRC, University of Oxford, Oxford, UK

m. BazinetReplicor Inc., Montreal, Canada

u.h.w. Beuers Dept. of Gastroenterology & Hepatology,

Academic Medical Center, Amsterdam,

The Netherlands

J. BlemRegulus Therapeutics, San Diego, CA, USA

a. BrownNuffield department of Medicine & Oxford

NIHR BRC, University of Oxford, Oxford, UK

d.J. ChinHoffman La Roche, Nutley, USA

K.a. van dortDept. of Experimental Immunology,


The Netherlands

n.w. GibsonRegulus Therapeutics, San Diego, CA, USA

K.p.J.m. van GisbergenDept. of Hematopoiesis, Sanquin

Research and Landsteiner Laboratory,


The Netherlands

p. GrintRegulus Therapeutics, San Diego, CA, USA

J. GrundyRegulus Therapeutics, San Diego, CA, USA

s. hadiPRA Health Sciences, Zuidlaren, the Netherlands

m. harbersPRA Health Sciences, Zuidlaren, the Netherlands

m. huangRegulus Therapeutics, San Diego, CA, USA

e.p. van Iperen Dept. of Clinical Epidemiology, Biostatistics, and Bioinformatics, Academic Medical Center & Durrer Center for Cardiogenetic Research, Amsterdam, The Netherlands

l. JansenDept. of Gastroenterology & Hepatology, Experimental Immunology, Academic Medical Center, Amsterdam, The Netherlands

h.l.a. JanssenDept. of Gastroenterology & Hepatology, Erasmus Medical Center, Rotterdam, the Netherlands & Toronto Centre for Liver Disease, Toronto Western & General Hospital, University Health Network, Toronto, Canada

p. KlenermanNuffield department of Medicine & Oxford NIHR BRC, University of Oxford, Oxford, UK

276

Appendix

m. KootDept. of Virus Diagnostic Services, Sanquin Blood Supply Foundation, Amsterdam, The Netherlands

n.a. KootstraDept. of Experimental Immunology, Academic Medical Center, Amsterdam, The Netherlands

s.d. KuikenDept. of Gastroenterology & Hepatology, OLVG West, Amsterdam, The Netherlands

e.m.m. van leeuwenDept. of Experimental Immunology, Academic Medical Center, Amsterdam, The Netherlands

r.a. van lierDept. of Hematopoiesis, Sanquin Research and Landsteiner Laboratory, Academic Medical Center, Amsterdam, The Netherlands

K. liuRegulus Therapeutics, San Diego, CA, USA

u. lopatinAssembly Pharmaceuticals, Bloomington, IN, USA

r. molenkampDept. of Medical Microbiology, Clinical Virology Laboratory, Academic Medical Center, Amsterdam, The Netherlands

s. nebenRegulus Therapeutics, San Diego, CA, USA

a. de nietDept. of Gastroenterology &

Hepatology, Experimental Immunology,


The Netherlands

C.m.J. van nieuwkerkDept. of Gastroenterology & Hepatology,

VU Medical Center, Amsterdam,

The Netherlands

a.C. van nuenenDept. of Experimental Immunology,


The Netherlands

a.K. patickRegulus Therapeutics, San Diego, CA, USA

a. pavlicekRegulus Therapeutics, San Diego, CA, USA

m.h. van der reeDept. of Gastroenterology &



The Netherlands

h.w. reesinkDept. of Gastroenterology &



The Netherlands

e.B. remmerswaal Dept. of Experimental Immunology,


The Netherlands

277

Appendix

&

s. rietdijk Dept. of Gastroenterology & Hepatology, Academic Medical Center & OLVG Oost, Amsterdam, The Netherlands

J. schinkelDept. of Medical Microbiology, Clinical Virology Laboratory, Academic Medical Center, Amsterdam, Netherlands

m.J. sinnigeDept. of Experimental Immunology, Academic Medical Center, Amsterdam, The Netherlands

l. swadlingNuffield department of Medicine & Oxford NIHR BRC, University of Oxford, Oxford, UK

r.B. takkenbergDept. of Gastroenterology & Hepatology, Academic Medical Center, Amsterdam, The Netherlands

a. vaillantReplicor Inc., Montreal, Canada

m. van der valkDept. of Gastroenterology & Hepatology, Academic Medical Center, Amsterdam, The Netherlands

e. van der veerDDL Diagnostic Laboratory,

Rijswijk, Netherlands

J. verheijDept. of Pathology, Academic Medical

Center, Amsterdam, The Netherlands

J.m.l. de vreeDept. of Gastroenterology & Hepatology,

University Medical Center Groningen,

Groningen, Netherlands

s. weijerDept. of Internal Medicine, Medical Center

Zuiderzee, Lelystad, The Netherlands

s.B. willemseDept. of Gastroenterology & Hepatology,


The Netherlands

h.l. ZaaijerDept. of Clinical Virology, Academic Medical

Center, Amsterdam, The Netherlands

a.h. ZwindermanDept. of Clinical Epidemiology,

Biostatistics, and Bioinformatics, Academic

Medical Center, A

278

Appendix

lIst of puBlICatIonspublications in this thesisDynamics of the immune response in acute hepatitis B infection.stelma f, Willemse SB, Erken R, de Niet A, Sinnige MJ, van Dort KA, Zaaijer HL, van Leeuwen EMM, Kootstra, NA, Reesink HWOpen Forum Infectious Diseases 2017; in press.

Immune responses in DAA treated chronic hepatitis C patients with and without prior RG-101 dosing.stelma f & van der Ree MH, Willemse SB, Brown A, Swadling L, van der Valk M, Sinnige MJ, van Nuenen AC, de Vree JML, Klenerman P, Barnes E, Kootstra NA, Reesink HWAntiviral Research 2017;146:139-145.

Peg-interferon plus nucleotide analogue in patients with chronic hepatitis B with low viral load - Authors’ reply.Jansen L, stelma f, Niet A, Reesink HWLancet Gastroenterology and Hepatology 2017;2:629.

Human intrahepatic CD69+CD8+ T cells have a tissue resident memory T cell phenotype with reduced cytolytic capacitystelma f, de Niet A, Sinnige MJ, van Dort KA, van Gisbergen KPJM, Verheij J, van Leeuwen EMM, Kootstra NA, Reesink HWScientific Reports 2017;7:6172.

HBsAg loss after peginterferon and nucleotide combination treatment in chronic hepatitis B patients: 5 years of follow-up.stelma f, van der Ree MH, Jansen L, Peters MW, Janssen HLA, Zaaijer HL, Takkenberg RB, Reesink HWJournal of Viral Hepatitis 2017;Jun 20. doi: 10.1111/jvh.12738.

Immune phenotype and function of NK and T cells in chronic hepatitis C patients who received a single dose of anti-miR-122, RG-101. stelma f, van der Ree MH, Sinnige MJ, Brown A, Swadling L, de Vree JM, Willemse SB, van der Valk M, Grint P, Neben S, Klenerman P, Barnes E, Kootstra NA and Reesink HWHepatology 2017;66:57-68.

Peg-interferon plus nucleotide analogue treatment versus no treatment in patients with chronic hepatitis B with a low viral load: a randomised controlled, open-label trial.de Niet A & Jansen L & stelma f, Willemse SB, Kuiken SD, Weijer S, van Nieuwkerk CMJ, Zaaijer HL, Molenkamp R, Takkenberg RB, Koot M, Verheij J, Beuers U, Reesink HWLancet Gastroenterology and Hepatology 2017;2:576-584.

lIst of puBlICatIons

279

Appendix

&

Comment: Peginterferon add-on: an added benefit? stelma f, Reesink HW Lancet Gastroenterology and Hepatology 2017;2:147-148.

A Single Dose of RG-101, a GalNAc-conjugated Oligonucleotide Antagonizing miR-122, Results in Significant Viral Load Reductions in Chronic Hepatitis C Patients: a Randomised Controlled Trial. Van der Ree MH, de Vree JM, stelma f, Willemse S, van der Valk M, Rietdijk S, Molenkamp R, Schinkel J, van Nuenen AC, Beuers U, Hadi S, Harbers M, van der Veer E, Liu K, Grundy J, Patick AK, Pavlicek A, Blem J, Huang M, Grint P, Neben S, Gibson NW, Kootstra NA, Reesink HW Lancet 2017;389:709-717.

HLA-C and KIR combined genotype as new response marker for HBeAg-positive chronic hepatitis B patients treated with peginterferon based combination therapy. stelma f, Jansen L, Sinnige MJ, van Dort KA, Takkenberg RB, Janssen HL, Reesink HW, Kootstra NAJournal of Viral Hepatitis 2016;23:652-9.

Restoration of T cell function in chronic hepatitis B patients upon treatment with interferon based combination therapy. stelma f & de Niet A, Jansen L, Sinnige MJ, Remmerswaal EB, Takkenberg RB, Kootstra NA, Reesink HW, van Lier RA, van Leeuwen EMJournal of Hepatology 2016;64:539-46.

Natural killer cell characteristics in patients with chronic hepatitis B virus (HBV) infection are associated with HBV surface antigen clearance after combination treatment with pegylated interferon alfa-2a and adefovir. stelma f, de Niet A, Tempelmans Plat-Sinnige MJ, Jansen L, Takkenberg RB, Reesink HW, Kootstra NA, van Leeuwen EMThe Journal of Infectious Diseases 2015;212:1042-51.

HBsAg loss in patients treated with peginterferon alfa-2a and adefovir is associated with SLC16A9 gene variation and lower plasma carnitine levels.Jansen L, stelma f & de Niet A, van Iperen EP, van Dort KA, Plat-Sinnige MJ, Takkenberg RB, Chin DJ, Zwinderman AH, Lopatin U, Kootstra NA, Reesink HWJournal of Hepatology 2014;61:730-7.

280

Appendix

other publicationsNovel gene function revealed by mouse mutagenesis screens for models of age-related disease.Potter PK, Bowl MR, Jeyarajan P, Wisby L, Blease A, Goldsworthy ME, Simon MM, Greenaway S, Michel V, Barnard A, Aguilar C, Agnew T, Banks G, Blake A, Chessum L, Dorning J, Falcone S, Goosey L, Harris S, Haynes A, Heise I, Hillier R, Hough T, Hoslin A, Hutchison M, King R, Kumar S, Lad HV, Law G, MacLaren RE, Morse S, Nicol T, Parker A, Pickford K, Sethi S, Starbuck B, stelma f, Cheeseman M, Cross SH, Foster RG, Jackson IJ, Peirson SN, Thakker RV, Vincent T, Scudamore C, Wells S, El-Amraoui A, Petit C, Acevedo-Arozena A, Nolan PM, Cox R, Mallon AM, Brown SD.Nature Communications 2016;7:12444.

Non-syndromic hereditary sensorineural hearing loss: review of the genes involved. stelma f, Bhutta MF. J Laryngol Otol 2014;128:13-21.

Casuïstiek: Hamartoom uitgaande van de sinus piriformis. stelma f, Dikkers FG, Nieken J. Ned Tijdschr KNO heelkunde 2014;20:12-4.

281

Appendix

&

phd portfolIoName PhD student: Femke StelmaPhD period: June 2013 – June 2017Name PhD supervisors prof. dr. U.H.W. Beuers, prof. dr. T.B.H. Geijtenbeek, dr. H.W. Reesink, dr. N.A. Kootstra

phd training year workload(eCts)

General courses

Basic course in legislation and organisation for clinical researchers (BROK), AMC 2013 1.0Basic Laboratory safety, AMC 2013 0.4Infectious diseases course, AMC 2013 1.3Practical biostatistics, AMC 2014 1.1

seminars, workshops, master classes

Weekly seminars Gastroenterology and Hepatology department, AMC 2013-2017 4.0Weekly seminars Experimental Immunology department, AMC 2013-2017 4.0Yearly PhD retreat Experimental Immunology department, AMC 2013-2017 1.5Gut-club meetings, Amsterdam 2013-2017 0.5Postgraduate course EASL 2014-2017 0.5Postgraduate course AASLD 2013 0.5Hepatitis Masterclass, Virology Education 2013 1.1Advanced Immunology course 2014 1.5EASL Basic school, Lausanne 2015 1.5EASL Basic school, Freiburg 2015 1.5Medical Business Masterclass, Amsterdam 2015 0.5National hepatitis day, Amsterdam 2016 0.3

oral presentations

Dutch Society of Gastroenterology, NVGE voorjaar (1) 2014 0.5Dutch Society of Gastroenterology, NVGE voorjaar (1) 2015 0.5The Asian Pacific Association for the Study of the Liver (2) 2016 1Dutch Society of Gastroenterology, (NVGE) najaar (1) 2016 0.5Dutch Society of Gastroenterology, (NVGE) voorjaar (1) 2017 0.5

Invited oral presentations

Early morning workshop, EASL 2016 0.5

poster presentations

AASLD annual meeting San Francisco, USA (2) 2015 1.0EASL annual meeting Vienna, Austria (3) 2015 1.5AASLD annual meeting Boston, USA (3) 2016 1.5EASL annual meeting Barcelona, Spain (3) 2016 1.5EASL annual meeting Amsterdam, Netherlands (1) 2017 0.5Dutch Society of Immunology, Noordwijkerhout, Netherlands (1) 2013 0.5Dutch Society of Immunology,Kaatsheuvel, Netherlands (1) 2014 0.5Dutch/British Society of Immunology, Liverpool, UK (1) 2016 0.5Dutch Society of Gastroenterology, NVGE voorjaar (2) 2016 1.0

phd portfolIo

282

Appendix

phd portfolIo (continued)

yearworkload(eCts)

attendance (Inter)national conferences

Annual meeting American Association for the Study of Liver Disease (AASLD) 2013-2016 4.0Washington, Boston, San Fransisco, BostonAnnual meeting Dutch/British Society of Immunology (NVVI/ BSI) 2013-2016 4.0Noordwijkerhout, Kaatsheuvel, Liverpool Annual meeting European Association for Study of the Liver (EASL) 2014-2017 4.0London, Vienna, Barcelona, AmsterdamBiannual meeting Dutch Society of Gastroenterology (NVGE) 2014-2016 2.5Veldhoven Annual meeting Asian Pacific Association for the Study of the Liver (APASL) 2016 1.0Tokyo

teaching

lecturing

2nd year medical students: general immunology course 2015 0.5AASLD highlights, Regional hepatology meeting Amsterdam 2015-2016 0.5Post EASL symposium, Amsterdam 2017 0.5Post EASL symposium, Virology Education 2017 0.5

supervising

Bachelor thesis medicine 2014 2.0Master’s thesis biomedical sciences 2015 4.0

parameters of esteem

Grants

EASL full bursary 2015Young investigator award APASL 2016EASL full bursary 2016EFIS international work visit grant 2016EASL registration bursary 2015NVGE travel grant 2016NVVI travel grant 2016EASL registration bursary 2017

283

Appendix

&

danKwoordDit proefschrift is tot stand gekomen door de inzet van vele personen. Allereerst gaat mijn dank uit naar de patiënten die hebben deelgenomen aan het wetenschappelijke onderzoek wat in dit boek staat beschreven.

Henk, met deze promotieplek nam mijn medische carrière een nieuwe wending. Bedankt voor het vertrouwen dat jij in me hebt gesteld. De vrijheid die jij me tijdens mijn promotietraject gaf was in het begin lastig, later leerzaam, en aan het einde geweldig. Ik heb erg veel van je kunnen leren in de afgelopen vier jaar en je bent voor mij op vele vlakken een voorbeeld. Ik heb genoten van je verhalen over vroeger, en hoe jij het allemaal hebt gedaan. Wat ik vooral erg in je waardeer is dat je oprecht waarde hecht aan de mening van je promovendus. Hierdoor was de drempel altijd laag om even bij je binnen te lopen. Dank deze waardevolle, leerzame tijd!

Neeltje, jouw deur staat letterlijk altijd open. Je bent erg betrokken en straalt een grenzeloos vertrouwen uit. Je bent niet alleen hoofd van je eigen groep, maar je maakt er ook onderdeel van uit en je bent altijd bezig met het wel en wee van je mensen. Ik heb genoten van onze wekelijkse discussies over falende experimenten, reviewers commentaren, en wereldreizen!

Ulrich, van een afstandje hield je mijn voortgang in de gaten. Dank voor je vertrouwen in mij als onderzoeker en straks als clinicus. Het wordt tijd om de andere vlakken binnen de hepatologie te exploreren! Het is een eer dat jij mijn promotor bent.

Theo, als hoofd van de afdeling Experimentele Immunologie keek jij van de zijlijn mee. Dank voor je kritische vragen en je gezelligheid.

Geachte dr. Bert Baak, prof. dr. Peter Jansen, prof. dr. René van Lier, dr. Dave Sprengers, prof. dr. Hans Zaaijer, bedankt voor het zitting nemen in mijn promotiecommissie. Dear prof. dr. Barnes, dear Ellie, thank you for a wonderful time in Oxford and for taking place in my thesis committee.

Ester van Leeuwen dank voor je rol als vraagbaak in de beginperiode van mijn promotie.

Marjan, gedurende mijn promotietijd bleek meer en meer hoe belangrijk een goede analist is voor een promovendus. Bedankt dat je me zo veel hebt geleerd. Onze samenwerking verliep gesmeerd en ook een bakkie doen was altijd erg gezellig.

Coauteurs, dank voor jullie inzet en samenwerking!

Collega’s van K0 en M01, bedankt voor de leerzame tijd. Ester, bedankt voor het delen van al je kennis over fluorochromen en hun eigenaardigheden, Machiel, fijn dat ik je altijd even

danKwoord

284

Appendix

aan je labjas kon trekken, Karel & Ad, mijn terugvalbasis voor het moleculaire gedeelte van

dit proefschrift, Berend, dank voor je eerste-hulp-bij-flowcytometrie, journal club, dank

voor al jullie uitleg, Rachel, Daan, Michiel, Si-La, Kirstin, Hanneke, Nienke, Nina, Zita, Thijs,

leden van de borrelcommissie, dank voor de gezelligheid!

Oxford group thanks for an amazing time! Tony, thanks for the hard (team)work, Maddy, I

miss our fluffy coffees, Leo you are the world’s best explainer, Andy, Prabh, Ayako, Chan,

Felicity, Azim, Senthil, Matt, Tianqi, Jelke, Manu, Lian Lee, Annette, football team, Ellie,

Paul, you’re the best!

Crabbers; John, Judith, Sarah, Suzie, staying with you feels like home.

Jeltje en Martine, vanaf dag één was ik onder de indruk van jullie inzet. Bedankt voor al

het werk dat jullie doen voor de onderzoeksgroep virale hepatitis. Hadassa en Monique,

dank voor jullie super-efficiënte ondersteuning.

Kamergenoten van G4-214: Annikki, Louis, Hanke, Meike, Lowiek, Sem, Victorine, Sara, Tim,

Jorrit, Wyts, Robin (en stiekem Sophie en Bart), wat een super tijd hebben we gehad! Onze

kamer mag dan opgedoekt worden, maar de herinneringen aan onze borrels, sinterklazen

en geouwehoer blijft. Annikki, dank voor het uitgebreide voorwerk dat jij hebt gedaan!

Louis, dank voor de vrolijke samenwerking, Hanke, ik mis je woordgrappen iedere dag,

Meik, hopelijk snel weer kamergenoten, Lo, jouw relaxedheid is inspirerend, Sem, altijd

enthousiast, heerlijk! Vic, jouw valentijnsbriefje bewaar ik altijd! Saar, organisatiemachine,

boarden zonder jou wordt lastig, Tim, geen flauwe grappen en chille setjes meer, Jorrit,

gelukkig kom ik je nog op ieder feestje tegen, Wyts, leuk dat jij er ook bij hoort! Robin, ik

kon me geen betere, slimmere, leukere opvolger wensen! Soph en Bart, ik zal het buurten

op jullie kamer missen, tot over een paar jaar!

MDL onderzoekers: Wat een feestje was het om samen met jullie mijn promotietijd door

te ploeteren. Dank voor de gezelligheid op congressen, wintersportvakanties en borrels.

Lieve vrienden en vriendinnen, mijn sociale leven is mijn motor, dank dat jullie er zijn.

Conjo’s, verspreid over het land maar still going strong, Jos, oudste vriendinnetje, wat

hebben we al veel meegemaakt, Tieb, jin en jang, zo is het gewoon, Roos, let’s embrace

our inner nerds! Ies, Rag, Dorijn, gotta stick to ma girls like glue, jullie zijn vriendinnen

voor altijd, Mira, je bent voor mij (en velen met mij) een supermegageweldigfantastische

rots in de branding, Leo, Em, Jesse je maakt wat mee samen! With a twist, wat is zingen

toch heerlijke ontstressend. Zootje ongeregeld, wat een gekke partijtjes zijn het iedere

keer weer!

285

Appendix

&

Izzy, in addition to all the other stuff we do; thanks for running the therapeutic ten km with me every week. I love how we can be each other’s voice of reason. We are both the same and different, thanks for being in my life.

Familie Buiten, zo leuk dat ik er nog een familie bij heb! Dank voor alle gezelligheid!

Paranimfen; lieve Meik, door jou kwam ik terecht bij de virale hepatitis in het AMC. In vier jaar tijd werden we een onafscheidelijk team en een geoliede machine. Ik ben blij met onze reizen samen, onze publicaties, onze gesprekken. We voelen en vullen elkaar goed aan en het was voor ons niet meer dan vanzelfsprekend dat we elkaars paranimf zouden zijn. Ik ben blij dat we vriendinnen en collega’s blijven.

Lieve Jorrit, jouw tomeloze interesse in nagenoeg alles is onnavolgbaar. Afgezien van mijn directe collega’s besprak ik de inhoud van mijn proefschrift met niemand zo uitvoerig. Jouw intelligente vragen en prettig gestoordheid hielpen me regelmatig om de draad weer op te pakken. Dank dat je mijn paranimf wil zijn.

Lieve Tjal, eindelijk een echte titel, misschien kan ik dan ooit nog een discussie winnen van mijn grote broer. Ik ben blij dat jij en Tess er voor me zijn.

Pap en mam, jullie zijn geweldig, door jullie ben ik geworden wie ik nu ben. Jullie keukentafel is een baken in mijn leven waar alles besproken wordt, van wissewasjes tot levenskeuzes. Dank voor jullie creativiteit, wijsheid en onvoorwaardelijke liefde.

Maurits, allerleukste, met jou erbij is het leven te gek, gaaf en geweldig.

286

Appendix

CurrICulum vItaeFemke Stelma was born on the 4th of July 1987 in Ureterp, the Netherlands. In 2005 she graduated from high school at the Drachtster Lyceum in Drachten. After finishing the first year of Life Science and Technology at the Rijksuniversiteit Groningen, she entered medical school at the same university. She obtained her medical degree with honors in 2013. Shortly thereafter, she started her PhD in the Academic Medical Center in Amsterdam under supervision of dr. H.W. Reesink, dr. N.A. Kootstra, prof. dr. T.B.H. Geijtenbeek and prof. dr. U.H.W. Beuers. She spent four years performing experimental and clinical research to investigate immune responses in viral hepatitis B and C at the Department of Gastroenterology and Hepatology as well as the Department of Experimental Immunology. During these four years she spent three months in the UK to initiate a collaboration with the University of Oxford. In May 2018 she will start her training at the Onze Lieve Vrouwe Gasthuis (dr. Y.F.C. Smets), and she will continue her training at the department of Gastroenterology and Hepatology in the Academic Medical Center in Amsterdam (dr. K.M.A.J. Tytgat and prof. dr. U.H.W. Beuers).

CurrICulum vItae

Femke Stelma was born on the 4th of July 1987 in Ureterp, the Netherlands. In 2005 she graduated from high school at the Drachtster Lyceum in Drachten. After finishing the first year of Life Science and Technology at the Rijksuniversiteit Groningen, she entered medical school at the same university. She obtained her medical degree with honors in 2013. Shortly thereafter, she started her PhD in the Academic Medical Center in Amsterdam under supervision of dr. H.W. Reesink, dr. N.A. Kootstra, prof. dr. T.B.H. Geijtenbeek and prof. dr. U.H.W. Beuers. She spent four years performing experimental and clinical research to investigate immune responses in viral hepatitis B and C at the Department of Gastroenterology and Hepatology as well as the Department of Experimental Immunology. During these four years she spent three months in the UK to initiate a collaboration with the University of Oxford. In May 2018 she will start her training at the Onze Lieve Vrouwe Gasthuis (dr. Y.F.C. Smets), and she will continue her training at the department of Gastroenterology and Hepatology in the Academic Medical Center in Amsterdam (dr. K.M.A.J. Tytgat and prof.dr. U.H.W. Beuers).

Immune responses inhepatitis B and C

virus infection

Imm

une responses in hep

atitis B and

C virus infection

FEMKE STELMA

FE

MK

E S

TE

LMA

uva-dare (digital academic repository) immune responses in ... · 10 chapter 1 hepatitis b virology...

Documents