Embed Size (px)
Citation preview
Review
J Biomed Sci 2001;8:237–247
Vaccine- and Hepatitis B ImmuneGlobulin-Induced Escape Mutations ofHepatitis B Virus Surface Antigen
Michael P. Cooremana Geert Leroux-Roelsc Wilma P. Paulijb
aDepartment of Gastroenterology and Hepatology, Academic Medical Center, University of Amsterdam,Amsterdam, bDepartment of Bioscience and Chemistry, Organon Teknika, Boxtel, The Netherlands;cCenter for Vaccinology, Ghent University, Ghent, Belgium
Received: September 25, 2000Accepted: December 5, 2000
Dr. Michael P. CooremanDepartment of Gastroenterology and HepatologyAcademic Medical Center, University of Amsterdam, Meibergdreef 9NL–1105 AZ Amsterdam (The Netherlands)Fax +31 20 6917033, E-Mail [email protected]
ABCFax + 41 61 306 12 34E-Mail [email protected]
© 2001 National Science Council, ROCS. Karger AG, Basel1021–7770/01/0083–0237$17.50/0Accessible online at:www.karger.com/journals/jbs
Key WordsHepatitis B virus W HBsAg W Mutations W Vaccination W
Hepatitis B immune globulin
AbstractHepatitis B virus surface antigen (HBsAg) vaccinationhas been shown to be effective in preventing hepatitis Bvirus (HBV) infection. The protection is based on theinduction of anti-HBs antibodies against a major clusterof antigenic epitopes of HBsAg, defined as the ‘a’ deter-minant region of small HBsAg. Prophylaxis of recurrentHBV infection in patients who have undergone livertransplantation for hepatitis B-related end-stage liver dis-ease is achieved by the administration of hepatitis Bimmune globulins (HBIg) derived from HBsAg-vacci-nated subjects. The anti-HBs-mediated immune pressureon HBV, however, seems to go along with the emer-gence and/or selection of immune escape HBV mutantsthat enable viral persistence in spite of adequate anti-body titers. These HBsAg escape mutants harbor singleor double point mutations that may significantly alter theimmunological characteristics of HBsAg. Most escapemutations that influence HBsAg recognition by anti-HBsantibodies are located in the second ‘a’ determinantloop. Notably, HBsAg with an arginine replacement for
glycine at amino acid 145 is considered the quintessen-tial immune escape mutant because it has been isolatedconsistently in clinical samples of HBIg-treated individu-als and vaccinated infants of chronically infected moth-ers. Direct binding studies with monoclonal antibodiesdemonstrated a more dramatic impact of this mutationon anti-HBs antibody recognition, compared with otherpoint mutations in this antigenic domain. The clinicaland epidemiological significance of these emergingHBsAg mutants will be a matter of research for years tocome, especially as data available so far document thatthese mutants are viable and infectious strains. Strate-gies for vaccination programs and posttransplantationprophylaxis of recurrent hepatitis need to be developedthat may prevent immune escape mutant HBV fromspreading and to prevent these strains from becomingdominant during the next decennia.
Copyright © 2001 National Science Council, ROC and S. Karger AG, Basel
Introduction
Chronic hepatitis B virus (HBV) infection is a majorhealth burden affecting an estimated number of morethan 400 million people worldwide [43]. Hepatitis B-related morbidity and mortality include complications of
238 J Biomed Sci 2001;8:237–247 Cooreman/Leroux-Roels/Paulij
liver cirrhosis and hepatocellular carcinoma [3]. HBV is asmall virus of the Hepadnaviridae family. It has a 3.2-kbgenome with four partially overlapping open readingframes encoding structural envelope or surface proteins[hepatitis B virus surface antigen (HBsAg)], precore/coreproteins (HBeAg and HBcAg), the reverse transcription-DNA polymerase enzyme and a transactivator X protein.Partly double-stranded genomic DNA is encapsidated byan assembly of core protein dimers. This nucleocapsid inturn is surrounded by an envelope that consists of a hostcell-derived lipid bilayer and viral surface proteins [70].
Inflammation and liver damage are primarily causedby cytotoxic T lymphocyte-mediated reactivity againstinfected hepatic cells presenting MHC class I bound T cellpeptide epitopes [17, 18]. Viral clearance from the blood,on the other hand, is mediated by antibodies directedagainst dominant B cell epitope clusters on the surfaceprotein, most notably those located in the ‘a’ determinantregion [36, 37]. Most (99%) protective antibodies thatappear in serum after natural infection are directedtoward this highly immunogenic region of HBsAg [37].Antibodies directed against the ‘a’ determinant have thusbeen the cornerstone of prophylactic strategies againsthepatitis B infection. The success story of the large-scaleapplication of serum-derived and recombinant hepatitisB vaccines and the rationale for the administration ofhepatitis B immune globulins (HBIg) to reduce the risk ofrecurrent hepatitis B after liver transplantation depend onthe neutralizing effect that anti-HBs antibodies displayafter the binding of conformational ‘a’ loop epitopes onthe surface of HBV virions [12, 49, 65]. Thus, adherenceto the policy of increasing the availability of HBV vaccineto the world’s population is expected to further diminishthe health burden of hepatitis B-related liver disease andcancer. With this perspective in mind, recent reports ofthe appearance of HBV with mutant HBsAg that escaperecognition by anti-HBs antibodies are raising concern,with regard to both vaccination and HBIg therapy [6, 22,23, 26, 33, 34, 50, 55, 57, 66, 68, 69]. To date, thesemutants have been isolated in different places around theworld, but we do not know the clinical consequences ofbreakthrough infections in vaccinated subjects, with re-gard to the development of chronic liver disease and can-cer. Data on immunological characteristics are still lim-ited and information on the viral fitness of these mutantsis even more scarce. We will need more epidemiological,immunological and virological data before conclusionscan be drawn as to the consequences of the emergence ofanti-HBs immune escape mutants for vaccine and HBIgdevelopment.
Genetic Variability of HBV
In spite of overlapping genes, HBV has a quasispeciesnature. The viral life cycle includes an intracellular prege-nomic RNA that is reverse transcribed to DNA within theassembled viral nucleocapsid. Because the reverse trans-criptase-DNA polymerase lacks proof-reading ability,HBV exhibits a mutation rate that is more than 10-foldhigher than other DNA viruses; the estimated mutationrate is approximately 1 nucleotide/10,000 bases/infectionyear [4, 54]. HBV variability is reflected by the occurrenceof serologically defined subtypes and by six differentgenotypic groups characterized to date, designated A–F[51].
Different definitions of variants and mutants havebeen suggested by Carman [10]. According to his propo-sal, the term ‘variants’ is used to describe natural subtypesthat occur without a known intervention-induced selec-tion pressure and have a geographical distribution. Sub-type differentiation is based on monospecific antisera andcorresponds to replacements of one or only a few aminoacids (aa). The dominant variants that dictate subtypedefinition are lysine or arginine at aa 122 and 160 ofmajor HBsAg. All serotypes contain the group-specific ‘a’determinant, whereas aa residues at position 122, but alsoS133T and Y134F, define the d or r subtype, and residuesat position 160 define the w or r subtype [2]. HBV sub-types thus can roughly be classified as adr, adw, ayr orayw, which are further divided into sub-subtypes. Becauseserotyping does not necessarily correspond to sequencevariability at the aa level, Ohba et al. [53] hold that geno-typing by S genes gives more accurate information aboutHBV genetic variation. Subtypes are geographically andethnically distributed, for reasons that are not known. Theterm ‘mutants’, on the other hand, is restricted for variantviruses that emerge under selection pressure, as is the casewith human intervention such as vaccination and antivi-ral therapy.
Mutants and variants for several genomic regions havebeen described, some of them thought to be associatedwith different disease courses [9]. HBV mutants associat-ed with viral selection pressure in clinical settings areobserved in the precore/core gene, in the polymerase geneduring administration of nucleoside analog-reverse tran-scription inhibitors and in HBsAg as a result of vaccine-induced or HBIg-mediated anti-HBs antibody immunepressure. Different clinical manifestations have been de-scribed in particular for mutations in precore/core se-quences that encode HBeAg and HBcAg, respectively, buthave recently also been suggested for HBsAg mutations
Vaccine- and HBIg-Induced EscapeMutations of HBsAg
J Biomed Sci 2001;8:237–247 239
[9, 35, 61]. The majority of mutational changes thatemerge under anti-HBs immunological pressure cluster inthe hydrophilic central core of HBsAg, extending from aa99 to 169. Within this region, they are most often found inantigenic region HBs4, indicating that this is the neutral-izing epitope cluster [16, 24, 25, 32, 62, 63].
The concept that HBsAg mutations occurring in thepresence of anti-HBs antibodies are truly escape mutantsfrom immunological pressure follows from several obser-vations. These mutants emerge, persist and become pre-dominant during HBIg administration, but reversal towild type is observed after withdrawal. Nucleotide pointmutations do not occur randomly within HBsAg but areclustered in antigenic epitope domains. Furthermore, thevast majority of these nucleotide replacements is notsilent but induces an aa substitution, suggesting selectionat the protein level [7, 12, 26, 51].
The coding sequence for HBsAg overlaps with that ofaa 343–569 of the HBV polymerase. HBV polymerasecontains four functional domains: (1) primase or ter-minal protein, (2) spacer, (3) reverse transcriptase and(4) RNAse H. The area of clustered HBsAg mutationswithin the ‘a’ loop is located in the reverse transcriptasedomain of HBV polymerase (aa p459, p460, p461, p463,p464, p466, p477, p488) [64]. Because of the frame shiftbetween both genes, HBsAg mutations do not necessarilycause an aa exchange in HBV polymerase, but some doand their impact on polymerase activity and hence viralfitness remains largely unknown. On the other hand,nucleotide analog treatment induces drug escape muta-tions within HBV polymerase, within the overlappingregion with HBsAg. Lamivudine (3TC) has been mostextensively used for the treatment of HBV infection, andinduces mutations in the conserved catalytic YMDDdomain of the polymerase [35]. These mutations cause aareplacements at positions downstream of the major ‘a’determinant epitopes. Recently, however, double HBIgand lamivudine mutants have been isolated in situationsof viral breakthrough in patients undergoing salvage ther-apy for recurrent HBV infection after liver transplanta-tion. These combined mutants harbored the quintessen-tial anti-HBs G145R escape mutant and displayed highviral replication activity in an in vitro assay [5]. Severalother inhibitors of HBV DNA polymerase are beingdeveloped, such as adefovir, penciclovir or famciclovir,and some of them will probably be available for therapy ofchronic hepatitis B in the near future, alone or in combi-nation. Xenobiotic-induced mutations in both reversetranscriptase-DNA polymerase and in corresponding se-quences of envelope proteins will thus likely increase, and
their impact on the pathogenesis of HBV will need furtherstudy.
HBsAg Antigenicity
The envelope of HBV consists of three proteins with acommon carboxy-terminal end, referred to as major orsmall (S), middle and large HBsAg. Middle HBsAg has anamino-terminal extension of 55 aa (pre-S2), and largeHBsAg has an additional amino-terminal extension of119 aa (pre-S1) [70]. These proteins are the products ofthree different codons for initiation of transcription with-in the same open reading frame. S HBsAg is the principalcomponent of the viral envelope.
The three-dimensional structure of HBsAg is not ex-actly known, as there are no X-ray crystallographic dataavailable. A topological model for S predicts four trans-membrane helices with an exposed hydrophilic segmentextending from aa 99 to 169 of S HBsAg, the major hydro-philic region, which contains a highly conformational epi-tope cluster [67]. This structure is stabilized by disulfidebonds, corresponding to eight conserved cysteine residuescommon to all subtypes of HBV [72]. It contains a clusterof major conformational B cell epitopes, defined as the ‘a’determinant region (fig. 1). This domain extends from aa121 to 149 of S HBsAg and induces protective neutraliz-ing anti-HBs antibodies regardless of the virus subtype; itis therefore the primary target for both active and passiveimmune prophylaxis [18, 72, 73].
The majority of anti-HBs antibodies that appear afternatural infection are directed against the ‘a’ determinantepitope cluster. The antigenic epitopes of the ‘a’ regionhave been analyzed with competitive inhibition assaysand by binding studies to synthetic peptides using anti-HBs monoclonal antibody. Studies of mutant proteinshave shown that the conserved cysteines at positions 124,137, 139 and 149 are essential for the antigenicity andpresumably thus for the conformation of the protein [2,25, 45, 63]. Direct binding studies using antigens withpoint-mutated positions within the second ‘a’ loop (aa139–147) demonstrated significant alteration in antigenicproperties in some cases [20]. These data document theimpact of both conformational and physicochemicalproperties of aa at a given position on the antigenicity ofthe epitope cluster. The major hydrophilic region can beseparated into at least five functional areas correspondingto antigenic epitope clusters, indicated as HBs1 (upstreamof aa 120), HBs2 (aa 120–123), HBs3 (aa 124–137), HBs4(aa 139–147) and HBs5 (aa 149–169) [2]. Recently, we
240 J Biomed Sci 2001;8:237–247 Cooreman/Leroux-Roels/Paulij
Fig. 1. Conformational model of the majorhydrophilic region of HBsAg, containing theB cell immunodominant ‘a’ determinant re-gion. Two major loops are proposed in themajor hydrophilic region, defined by disul-fide bridges. Five presumed antigenic re-gions are termed HBs1–5. Vaccination- andHBIg-associated HBsAg mutants are shownin circles. Most escape mutants have beenconsistently found in the second ‘a’ determi-nant loop, extending from aa 139 to 147 or149, and represented by the HBs4 antigenicregion.
identified a sixth antigenic region located outside the clas-sical ‘a’ determinant, extending from residues 178 to 186of the small S protein, with consequences for the deducedstructure of HBsAg [59]. According to the topologicalmodel of Stirk et al. [67], this region is located within thethird passage of HBsAg polypeptide through the lipidmembrane (aa 160–184). Based on the binding data ofmonoclonal antibody against native HBsAg, however, theepitope region is readily accessible in any HBsAg polypep-tide, suggesting that region 160–270, instead of crossingthe lipid membrane twice, is rather projected in its fulllength over the surface of 20-nm HBsAg particles. Othermodels have been proposed that are commonly used [28,46]. In particular, Chen et al. [15] described two loopsfrom residues 124–137 and 139–147, formed by disulfidelinkages. The region aa 139–147 is a particularly immu-nogenic segment. The major hydrophilic region of the Sgene codes for surface-exposed regions of the S protein.Regions facing the inner side of the HBV particle andregions where transmembrane helices are predicted dis-play little or no variations.
HBsAg-Based Active and Passive Immunization
For almost two decades, hepatitis B vaccines have beenused mainly to protect persons at risk. More recently, andspurred on by the World Health Organization’s ExpandedProgram on Immunization (EPI) for universal infant andchildhood vaccination, the use of hepatitis B vaccines hasbeen expanding, and its benefits have been extensivelydocumented [33, 44, 78]. Hepatitis B has clearly become avaccine-preventable disease.
Commercial vaccines contain viral envelope subunits(small HBsAg). With the advent of recombinant manufac-turing facilities, the first-generation plasma-derivedHBsAg has been widely replaced by recombinant HBsAgderived from the yeast Saccharomyces cerevisiae [47].Both types of HBsAg vaccine expose the ‘a’ determinantand were shown to be extremely effective and safe. Theyseemed to provide complete protection against HBVinfection after pre- and postexposure immunization.HBV vaccine efficacy against both HBV infection andchronic HBV carriage is as high as 95% [1, 74]. Therecombinant HBV vaccine was in fact the first successfulrecombinant vaccine for a human infectious disease [47].HBIg derived from fully vaccinated subjects is thereforeused for passive immunization and has proven efficacy in
Vaccine- and HBIg-Induced EscapeMutations of HBsAg
J Biomed Sci 2001;8:237–247 241
the prevention of graft reinfection in patients who havereceived a liver transplant for hepatitis B-related end-stage liver disease [65].
Transmission patterns of hepatitis B differ betweengeographic areas, with perinatal transmission prevailingin Asia and sibling-to-sibling transmission predominatingin Africa. Overall, chronic carriers become less infectiveas they age, as measured by HBeAg positivity and HBVDNA levels. Vaccine programs aimed at newborn babiesin Asia and at the early childhood population in Africahave already caused transmission to fall to very low rates,and the vaccine remains effective even with progressivelydeclining antibodies [74]. In highly endemic countries,where HBV is transmitted mainly from mother to infantat birth, a strict policy of vaccination of all neonates hasled to a decrease in hepatitis B prevalence. In Taiwan, therate of HBsAg carriers is now 1%, compared to 9% in theprevaccination era. Widespread vaccination further led toa dramatic reduction in the incidence of hepatocellularcarcinoma in children between 6 and 15 years, and there-fore the hepatitis B vaccine is also the first vaccine shownto be effective for the prevention of cancer [14]. However,high-risk infants who have been protected from hepatitisB by the administration of HBIg and HBV vaccine soonafter birth still have a carrier rate of 5–10% [42].
In lowly endemic countries, the efficacy of hepatitisvaccination of high-risk populations has not yet beenunequivocally documented. A study from France found a50% decrease in the incidence of the disease between1985 and 1995, but these changes accompanied otherhygienic and prophylactic measures adopted to preventthe spread of HIV infection [60]. A study from the USA,on the other hand, did not show a significant change in theincidence of HBV infection between 1980 and 1994 [19].
Anti-HBs Antibody-Mediated HBsAg ImmuneEscape Mutants
HBV surface proteins contain both B and T cell anti-genic epitopes. Mutations in T cell epitopes may theoreti-cally influence the anti-HBs antibody profile through theinteraction between CD4+ helper T cells and B cells, yetthe role of these immunological pathways with regard tohepatitis B has not been studied. Direct escape from pro-tective anti-HBs antibodies as observed in vaccine andHBIg prophylaxis failure is caused by mutations in B cellepitopes that diminish or abolish recognition by anti-HBsantibodies [8, 20, 48].
Some recent reports suggest an increase in the isolationof HBsAg escape mutants, coinciding with the increasedavailability of protective anti-HBs antibodies generatedagainst recombinant versus native antigen [64]. While it isknown that HBV HBsAg vaccine induces a narrowerimmune response compared to natural infection, thesedata suggest that the spectrum may be further narrowedby the use of yeast-derived recombinant proteins. Thisphenomenon might explain why HBsAg mutants havebeen identified at an overall low rate in patients with nat-ural chronic infection who seroconverted to anti-HBspositivity, whereas the emergence of replicative HBVviruses with HBsAg that harbors point mutations in dom-inant B cell epitopes is substantially higher in HBIg-treated patients and tends to increase in vaccinatedinfants of chronically infected mothers.
Vaccine Escape Mutants
Persistent HBV infection coinciding with an adequateanti-HBs antibody response was already described in the1980s. HBV DNA has also been detected by dot-blothybridization in the serum of HBsAg-negative blood do-nors, but no molecular cloning of viral sequences to inves-tigate the genetic basis of antigenic variations was done inthat study [39].
Sequence variation in antigenic regions is one of themost powerful viral strategies for escaping recognition byboth the B and T cell-mediated immune system of thehost and facilitates viral persistence. In 1990, Carman etal. [13] first reported a case of hepatitis B infection withan ‘a’ determinant mutant virus in an Italian child froman HBV-positive mother, who had received active-pas-sive immunization at birth. A point mutation from gua-nosine to adenosine at nucleotide 587 was found, chang-ing codon 145 in the presumed second loop of the ‘a’determinant from glycine to arginine.
Since then, several other groups from different parts ofthe world have identified many other point mutationswithin the ‘a’ loop associated with breakthrough infec-tions in vaccinated subjects and HBIg-treated patientswho were thought to have protective anti-HBs antibodies.The ‘a’ determinant region is indeed the most heteroge-nous part of the S gene [78]. Mutants that have beenrepeatedly isolated include I/T126A, Q129H, M133L,T143M, D144H/A and G145R. Strikingly, one particularmutant, G145R, was almost invariably found in these sit-uations, while neighboring aa 144 replacements were alsofrequently isolated [22, 24, 64, 69]. Zuckerman et al. [79]
242 J Biomed Sci 2001;8:237–247 Cooreman/Leroux-Roels/Paulij
found a high number of G145R mutants in vaccinatedchildren, and G145R was also isolated from children whoreceived recombinant HBsAg vaccine alone, without con-comitant passive (HBIg) immunization. In general, re-placements in the presumed second ‘a’ loop, including aapositions 144 and 145, are more often associated withanti-HBs immune pressure than mutations in other epi-topic domains of the major hydrophilic region. Vice ver-sa, second ‘a’ loop mutants have only been rarely identi-fied in the absence of anti-HBs immune pressure [76].Point mutations at several other positions distributedover the entire ‘a’ determinant region have been describedmore locally or clustered, like the T/I126N mutant thatwas isolated in a family in Japan [55]. A population vacci-nation study in Gambia revealed that one half of childrenwho became anti-HBc positive had a K141E mutant[38].
Naturally occurring ‘a’ determinant mutations severe-ly but variably affect the structure of major antigenicdeterminants, but do not prevent the synthesis and secre-tion of envelope proteins [20]. The intriguing question ofwhether these mutants are present at a small percentage asa minor strain together with the predominant wild-typestrain before the onset of antibody-mediated immunepressure, and thus are merely selected, or whether theyemerge under conditions unfavorable for wild-type virus,has not been unequivocally resolved. Viral variant se-quences present at a low percentage (!10%) would easilybe missed with the current isolation techniques based onPCR, cloning and sequencing of viral genomes.
Several recent studies have substantiated the conceptthat these ‘a’ determinant mutations are associated withimmune pressure of vaccine-induced anti-HBs anti-bodies. Mutants in the ‘a’ region occur naturally indepen-dently of the vaccination program, including changes atpositions 126, 129, 133 and 145 [55, 77, 79]. For some ofthese mutants, transmission and infection of the vacci-nated population has been described [55]. However, Hsuet al. [34] examined maternal serum and serial samplesfrom their infants who developed HBV infection despiteimmunoprophylaxis and found that the HBV inoculumtransmitted from mother to infant was generally the wild-type strain. Lee et al. [41] studied 27 HBV-infected chil-dren who had received active-passive vaccination, theirmothers and 21 unvaccinated controls, all HBV positive.Six of the vaccinated carrier children had ‘a’ mutations,only one mutation was isolated from one mother (thesame mutant strain as her child) and all controls had wild-type virus.
Although HBsAg mutations in vaccinees mainly mapto the ‘a’ determinant sequence, a panoptic of otherHBsAg mutations associated with anti-HBs antibody im-mune pressure has been described, including pre-S inser-tions and deletions. These mutants are also supposed toaffect HBsAg antigenicity, because the pre-S region con-tains additional relevant B cell epitopes [35, 71].
A case of fulminant hepatitis was described in a vacci-nated patient who was HBV DNA positive but HBsAgnegative, illustrating mutation-induced diagnosis failure[11]. Sequencing of HBV DNA revealed a combinedinsertion between aa 122 and 123 and a G145R mutation[11]. Changes located outside the ‘a’ determinant regionwere reported from immunized infants in Singapore bornto HBV carrier mothers [58]. Interestingly, some of thesemutant proteins had reduced binding to monoclonal anti-bodies against the ‘a’ determinant, and some were mostlikely transmitted vertically as they were isolated bothfrom infant and maternal serum. Grethe et al. [27]reported unusual escape variants in an HBsAg-negativehepatitis B-positive subject. Multiple aa changes in sur-face-exposed regions of HBsAg, including mutational de-letions upstream of aa 124 in the HBs 1 and 2 regions,were found to abolish the reactivity of monoclonal andpolyclonal anti-HBs diagnostic antibodies against the ‘a’loop epitope cluster, and also were not recognized by thevaccinee’s sera [27]. The region between aa 118 and 123was identified as a hot spot for insertions by other investi-gators, illustrating its immunological importance [31].Mutants with insertions in HBs2 that displayed alteredbinding to anti-HBs and were not detected by diagnosticassays have been isolated from chronic carriers [11, 31,76]. Their role in vaccination failure is currently un-known. Another recent study documented mutants out-side the ‘a’ region isolated from immunized infants bornto HBV carrier mothers in Singapore [56]. Some of themwere within the major hydrophilic loop of HBsAg, butmutations at positions 183 and 184 were also found. Assome of these mutations had decreased binding to an ‘a’-specific monoclonal antibody, their functional analysiswill contribute to our understanding of the antigenicstructure of the HBV envelope.
Clinically relevant mutant virus isolates are those thathave altered interactions with protective antibodies, en-abling them to become dominant strains in the presenceof anti-HBs antibodies. The immunological impact of ‘a’loop mutations has been characterized by direct bindingstudies [20]. These results support epidemiological datathat aa 145 mutations, and to a somewhat lesser extent144 mutations, are major escape mechanisms for HBV in
Vaccine- and HBIg-Induced EscapeMutations of HBsAg
J Biomed Sci 2001;8:237–247 243
postvaccination settings. All other mutants found in vac-cinees had a less dramatic effect on antigenicity. Recentobservations suggest that some S mutants may also bemore pathogenic than wild-type virus [11]. An A126 vari-ant was found in one infant with fulminant hepatitis, andA126, A129R and G145R were isolated from childrenwith elevated serum transaminases, whereas mutationswere not isolated from infected infants that had normalliver tests on follow-up. A 145R mutant virus has beenfound that circulated as a stable strain at high titer for 8years of follow-up [34].
HBIg Escape Mutants
Liver allograft reinfection with hepatitis B frequentlycauses rapidly progressive liver disease with a poor prog-nosis. The pathological characteristics of posttransplanta-tion recurrent hepatitis B are extended necrosis of the liv-er parenchyma or fibrosing cholestatic hepatitis [40]. Theadministration of high-dose HBIg preparations for at least6 months after liver transplantation significantly reducesthe high risk of HBV reinfection of the graft. In a retro-spective study of 372 patients with chronic hepatitis B-related end-stage liver disease, the 3-year actuarial risk ofHBV recurrence, defined as the reappearance of HBsAgin serum, was 50% [65]. For patients whose serum waspositive for HBV DNA, however, the risk was 83%. Therecurrence rate was reduced to 36% when passive prophy-laxis with HBIg was given for 6 months or more. Three-year survival was only 44% for patients with recurrenthepatitis B, compared to 83% for patients who remainedHBsAg negative [65].
In spite of the efficacy of HBIg prophylaxis, the risk ofgraft reinfection is still approximately 30%. Sequenceanalysis data of breakthrough HBV recurrences aftertransplantation revealed point mutations in the immuno-dominant ‘a’ loop of HBsAg, corresponding to thosefound in vaccination failure settings. Again, G145R wasfound almost invariably, together with G145K/E/A, aa144 (D144G/A/E/V) replacements and less frequentlymutations, including aa 129 and 133 (M133T) of the firstloop [12, 68, 69]. Hawkins et al. [29] isolated G145R fromthree patients with fibrosing cholestatic hepatitis; in onepatient, the mutant turned out to have been presentbefore transplantation. In another study, the emergence ofmutants was associated with an increased graft failurerate of 44%, compared to 23% if mutants remainedabsent [64]. Fischer et al. [22] reported mutations in the‘a’ domain in 4 of 6 hepatitis B patients treated with poly-
clonal HBIg who had severe recurrent infection. In 2cases, these aa substitutions were also found before trans-plantation as minor viral subpopulations [22]. All HBVpopulations also had mutations resulting in a complete orpartial defect in HBeAg secretion, which were conceivedto contribute to the severe clinical picture.
Immune escape mutants are found during treatmentwith both monoclonal and polyclonal antibodies, but atendency of increased occurrence with antibodies fromsubjects who have received recombinant vaccine has beenreported [64]. S mutants are in fact found in the majorityof patients with liver graft reinfection, and only excep-tionally could these be isolated from pretransplant se-quences. On the withdrawal of HBIg, the mutants areoften lost and the wild type again becomes dominant.Obviously, for unknown reasons, these mutants have asurvival disadvantage compared to wild-type virus. Asmore sensitive techniques to detect minor mutant strainsof !5% of a quasispecies viral population become avail-able, the search for HBsAg mutations may be incorpo-rated in the pretransplantation workup of patients withhepatitis B-related cirrhosis.
The overall similarity of HBsAg mutations arisingunder high-dose HBIg immune pressure and after vacci-nation suggests identical immune escape phenomena. Wefound that a human monoclonal antibody derived from ahigh-responder HBsAg vaccine recipient reacted with an‘a’ determinant epitope in a variant-independent way[21]. However, dose-response curves showed a completeabsence of recognition of both aa 144 and 145 mutants forall antigen concentrations tested. These direct bindingdata illustrate that high-dose antibodies do not compen-sate for the dramatic impact of these single aa mutationson antigenicity [21]. As the risk of escape mutation can bereduced substantially by combining different antiviralstrategies, the combination of HBIg with nucleoside-ana-log reverse transcription inhibitors such as lamivudine isto be recommended [69].
Relevance for Vaccination Practice
The impact of vaccine-induced HBsAg mutants onhealth care depends on their epidemiology, infectivity,pathogenicity and the degree of vaccine cross-immunityto the mutants. These parameters are the basis of an eval-uation of whether HBsAg mutants are a thread for currentvaccination programs.
There is still uncertainty about the current prevalenceof mutant HBV-associated vaccine failure and its tenden-
244 J Biomed Sci 2001;8:237–247 Cooreman/Leroux-Roels/Paulij
cy to rise. Whether there has been a true increase in theprevalence of HBsAg escape mutant HBV infections orwhether their higher isolation rate is relative to an overalldeclining incidence of wild-type HBV infections is con-troversial [30, 33]. More epidemiological and moleculardata from endemic regions are clearly needed, and thesemay also define additional mutants.
Vaccine variants have been shown to be both infec-tious and pathogenic to chimpanzees [52]. In clinical set-tings, liver disease has been associated with the emer-gence of HBsAg mutants, both after vaccination and afterliver transplantation, and it has been shown for a few vari-ants that they are transmitted directly. There are evenmore or less anecdotal data suggesting that some mutants,e.g. G145R, may be more pathogenic than wild-typevirus. Recently, an in vitro model of HBV replication wasdescribed, and the authors of that study found that HBVwith a G145R HBsAg mutation had the same replicationefficacy as wild-type virus [5]. Direct binding data with ahuman anti-HBs monoclonal antibody derived from avaccinee showed a complete lack of recognition of thequintessential 144 and 145 mutants, even with increasingantigen dosages [21], illustrating zero cross-reactivity forthis vaccine-induced antibody. Taken together, these invitro data suggest that at least the 144 and 145 mutantshave the capacity to become dominant strains and tocause chronic hepatitis B in a vaccinated population.
Under the conditions of anti-HBs-mediated immunepressure, a stable and infective mutant strain will beselected in favor of wild-type strains, and therefore maybecome the predominant viral strain as a result of large-scale vaccination. Current vaccines would then becomeless effective. Information on the time frame needed forthis phenomenon to develop is a prerequisite for the dis-cussion of strategic measures. Limited epidemiologicaldata so far do not unequivocally support the assumptionof a relevant increase in clinically manifested vaccinebreakthrough hepatitis B infections in the near future.Based on the epidemiological characteristics of HBV,Wilson et al. [75] developed a theoretical model for thisscenario. The degree of cross-immunity between the cur-rent HBV vaccine and the mutant strain is an importantparameter in this model. If there is a high degree of cross-reactivity, the vaccine escape would not emerge, but thenafter all it is an escape mutant, and thus one might assumethat cross-reactivity should be rather low, in order to offersurvival advantage to the mutant. Under pessimisticassumptions that the current vaccine has no cross-reactiv-ity against the mutants, the model predicts that it wouldbe at least 50 years before the mutant viruses become
dominant. Based on the low contact rate between infectedand susceptible individuals, HBV variants currently havehad insufficient time to increase to the same prevalence aswild type, even if they were equally infectious as wildtype. The overall low prevalence at present of escapemutant HBV may thus be inherent to the epidemiology ofHBV, aside from the effects of these viral variants oninfectiousness or cross-immunity of the vaccine.
Although the decrease in the protective efficacy of cur-rent vaccines according to this model will thus most prob-ably be a slow process over several decennia, it willincreasingly become a problem for individual subjects inthe short term, and it already poses a substantial burdenon the prognosis of hepatitis B-related liver transplantpatients. Chronic infection of children by mutant HBVmay go undetected for decades because of long asymp-tomatic intervals, while imposing the risk of severe liverdisease and hepatocellular carcinoma on affected individ-uals. With the data available on the impact of the quintes-sential immune escape HBsAg mutants, both from clini-cal studies and from direct binding data, timely vaccinemodifications may prevent large-scale selection of escapemutant viruses in the future.
It should be kept in mind that at the moment, no vac-cine or HBIg protects against HBV escape mutants. Fu-ture vaccines should deal with this phenomenon, forexample by including a mixture of wild-type and mutantantigens based on epidemiological and pathogenicitydata, or by including other antigenic regions of HBV. Theinclusion of pre-S2 sequences does not seem to improvethe efficacy of the vaccine against immune escape mu-tants [37].
Conclusions
Recombinant HBsAg-based vaccines have dramatical-ly reduced the number of HBV infections in children inendemic areas. As the number of vaccinated personsincreases, while that of unvaccinated susceptible subjectsdiminishes, the immunological selection pressure onHBV will increase, and presumably mutant HBsAg will bemore readily detected. HBsAg escape mutants have in-creasingly been observed in immunized newborns ofchronically infected mothers and in HBV carriers treatedwith HBIg after liver transplantation.
De novo infection of a successfully vaccinated individ-ual by escape mutant HBV has not been unequivocallyobserved so far, confirming the efficacy of current vac-cines in this population. However, based on epidemiolog-
Vaccine- and HBIg-Induced EscapeMutations of HBsAg
J Biomed Sci 2001;8:237–247 245
ical data, vaccination strategies and antibody treatmentmay have to be adapted in endemic regions where trans-mission in infancy or early childhood predominates, ac-cording to regional variations in the prevalence of mutantviruses. This may imply both the inclusion of mutantHBsAg in vaccine preparations (broad-spectrum vaccinedesign) and/or the combination of monoclonal antibodieswith defined binding characteristics or the construction ofdesigner antibodies, in order to deal with relevant mutantHBV currently identified in clinical samples.
As for HBIg therapy in liver transplantation settings,the selection of monoclonal or polyclonal antibodies that
recognize relevant mutants, in particular aa 144 and 145replacements, may prevent these mutants from becomingdominant strains. In addition, HBIg will be increasinglyused in combination with reverse transcriptase inhibitors,such as lamivudine. Combination strategies based ondefined and complementary antiviral mechanisms areexpected to greatly reduce the risk of mutational viralescape, also because the compact nature of the HBVgenome with overlapping genes supposedly limits the pos-sibilities of mutational escape of this virus.
References
1 Andre FE. Overview of a 5-year clinical experi-ence with a yeast-derived hepatitis B vaccine.Vaccine 8(suppl):S74–S80;1990.
2 Ashton-Rickardt PG, Murray K. Mutationsthat change the immunological subtype of hep-atitis B virus surface antigen and distinguishbetween antigenic and immunogenic determi-nation. J Med Virol 29:204–214;1989.
3 Beasley RP, Hwang LY. Hepatocellular carci-noma and hepatitis B virus. Semin Liver Dis 4:113–121;1984.
4 Blum HE. Variants of hepatitis B, C and Dviruses: Molecular biology and clinical signifi-cance. Digestion 56:85–95;1995.
5 Bock TC, Torresi J, Tillmann HL, LocarniniSA, Manns MP. Replication fitness of hepatitisB virus polymerase mutants after antiviraltreatment following liver transplantation. He-patology 30:199A;1999.
6 Brind A, Jiang J, Samuel D, Gigou M, Feray C,Brechot C, Kremsdorf D. Evidence for selec-tion of hepatitis B mutants after liver trans-plantation through peripheral blood mononu-clear cell infection. J Hepatol 26:228–235;1997.
7 Brown JL, Carman WF, Thomas HC. The clin-ical significance of molecular variation withinthe hepatitis B virus genome. Hepatology 15:144–148;1992.
8 Bruce SA, Murray K. Mutations of some criti-cal amino acid residues in the hepatitis B virussurface antigen. J Med Virol 46:157–161;1995.
9 Carman WF. Molecular variants of hepatitis Bvirus. Clin Lab Med 16:407–428;1996.
10 Carman WF. The clinical significance of sur-face antigen variants of hepatitis B virus. JViral Hepat 4(suppl 1):11–20;1997.
11 Carman WF, Korula J, Wallace L, MacPhee R,Mimms L, Decker R. Fulminant reactivationof hepatitis B due to envelope protein mutantthat escaped detection by monoclonal HBsAgELISA. Lancet 345:1406–1407;1995.
12 Carman WF, Trautwein C, van Deursen FJ,Colman K, Dornan E, McIntyre G, Waters J,Kliem V, Muller R, Thomas HC, Manns MP.Hepatitis B virus envelope variation aftertransplantation with and without hepatitis Bimmune globulin prophylaxis. Hepatology 24:489–493;1996.
13 Carman WF, Zanetti AR, Karayiannis P, Wa-ters J, Manzillo G, Tanzi E, Zuckerman AJ,Thomas HC. Vaccine-induced escape mutantof hepatitis B virus. Lancet 336:325–329;1990.
14 Chang MH, Chen CJ, Lai MS, Hsu HM, WuTC, Kong MS, Liang DC, Shau WY, Chen DS.Universal hepatitis B vaccination in Taiwanand the incidence of hepatocellular carcinomain children. Taiwan Childhood HepatomaStudy Group. N Engl J Med 336:1855–1859;1997.
15 Chen YC, Delbrook K, Dealwis C, Mimms L,Mushahwar IK, Mandecki W. Discontinuousepitopes of hepatitis B surface antigen derivedfrom a filamentous phage peptide library. ProcNatl Acad Sci USA 93:1997–2001;1996.
16 Chiou HL, Lee TS, Kuo J, Mau YC, Ho MS.Altered antigenicity of ‘a’ determinant variantsof hepatitis B virus. J Gen Virol 78:2639–2645;1997.
17 Chisari FV. Cytotoxic T cells and viral hepati-tis. J Clin Invest 99:1472–1477;1997.
18 Chisari FV, Ferrari C. Hepatitis B virus immu-nopathogenesis. Annu Rev Immunol 13:29–60;1995.
19 Coleman PJ, McQuillan GM, Moyer LA, Lam-bert SB, Margolis HS. Incidence of hepatitis Bvirus infection in the United States, 1976–1994: Estimates from the National Health andNutrition Examination Surveys. J Infect Dis178:954–959;1998.
20 Cooreman MP, van Roosmalen MH, teMorsche R, Sunnen CM, de Ven EM, JansenJB, Tytgat GN, de Wit PL, Paulij WP. Charac-terization of the reactivity pattern of murinemonoclonal antibodies against wild-type hepa-titis B surface antigen to G145R and other nat-urally occurring ‘a’ loop escape mutations. He-patology 30:1287–1292;1999.
21 Cooreman MP, Sunnen CM, Verhoye L, vanRoosmalen MH, Paulij WP, Leroux-Roels G.Reactivity pattern of a human monoclonal an-tibody against recombinant hepatitis B vaccineto HBsAg immune escape mutants. Gastroen-terology 118:A1004;2000.
22 Fischer L, Sterneck M, Muller-Ruchholtz C,Gish R, Will H. Hepatitis B virus variants asso-ciated with clinically severe recurrence afterliver transplantation. Transplant Proc 31:492–493;1999.
23 Fortuin M, Karthigesu V, Allison L, Howard C,Hoare S, Mendy M, Whittle HC. Breakthroughinfections and identification of a viral variantin Gambian children immunized with hepatitisB vaccine. J Infect Dis 169:1374–1376;1994.
24 Fujii H, Moriyama K, Sakamoto N, Kondo T,Yasuda K, Hiraizumi Y, Yamazaki M, SakakiY, Okochi K, Nakajima E. Gly145 to Arg sub-stitution in HBs antigen of immune escape mu-tant of hepatitis B virus. Biochem Biophys ResCommun 184:1152–1157;1992.
25 Gerin JL, Alexander H, Shih JW, Purcell RH,Dapolito G, Engle R, Green N, Sutcliffe JG,Shinnick TM, Lerner RA. Chemically synthe-sized peptides of hepatitis B surface antigenduplicate the d/y specificities and induce sub-type-specific antibodies in chimpanzees. ProcNatl Acad Sci USA 80:2365–2369;1983.
26 Ghany MG, Ayola B, Villamil FG, Gish RG,Rojter S, Vierling JM, Lok AS. Hepatitis Bvirus S mutants in liver transplant recipientswho were reinfected despite hepatitis B im-mune globulin prophylaxis. Hepatology 27:213–222;1998.
27 Grethe S, Monazahian M, Bohme I, ThomssenR. Characterization of unusual escape variantsof hepatitis B virus isolated from a hepatitis Bsurface antigen-negative subject. J Virol 72:7692–7696;1998.
28 Guerrero E, Gavilanes F, Peterson DL. Modelfor the protein arrangement in HBsAg particlesbased on physical and chemical studies. In:Zuckerman AJ, ed. Viral Hepatitis and LiverDisease 606–613;1988.
246 J Biomed Sci 2001;8:237–247 Cooreman/Leroux-Roels/Paulij
29 Hawkins AE, Gilson RJ, Gilbert N, WreghittTG, Gray JJ, Ahlers-de Boer I, Tedder RS,Alexander GJ. Hepatitis B virus surface muta-tions associated with infection after liver trans-plantation. J Hepatol 24:8–14;1996.
30 Hino K, Katoh Y, Vardas E, Schoub B, Car-man WF, Okita K. The effect of introduction ofuniversal childhood vaccination on the preva-lence of serologically negative hepatitis B virusinfection and the selection of immune escapevariants in a developing country. Hepatology30:300A;1999.
31 Hou J, Karayiannis P, Waters J, Luo K, LiangC, Thomas HC. A unique insertion in the Sgene of surface antigen-negative hepatitis Bvirus Chinese carriers. Hepatology 21:273–278;1995.
32 Howard CR. The structure of hepatitis B enve-lope and molecular variants of hepatitis B vi-rus. J Viral Hepat 2:165–170;1995.
33 Hsu HY, Chang MH, Liaw SH, Ni YH, ChenHL. Changes of hepatitis B surface antigenvariants in carrier children before and afteruniversal vaccination in Taiwan. Hepatology30:1312–1317;1999.
34 Hsu HY, Chang MH, Ni YH, Lin HH, WangSM, Chen DS. Surface gene mutants of hepati-tis B virus in infants who develop acute orchronic infections despite immunoprophylax-is. Hepatology 26:786–791;1997.
35 Hunt CM, McGill JM, Allen MI, CondreayLD. Clinical relevance of hepatitis B viral mu-tations. Hepatology 31:1037–1044;2000.
36 Iwarson S, Tabor E, Thomas HC, Goodall A,Waters J, Snoy P, Shih JW, Gerety RJ. Neutral-ization of hepatitis B virus infectivity by amurine monoclonal antibody: An experimentalstudy in the chimpanzee. J Med Virol 16:89–96;1985.
37 Jilg W. Novel hepatitis B vaccines. Vaccine16(suppl):S65–S68;1998.
38 Karthigesu VD, Allison LM, Fortuin M, Men-dy M, Whittle HC, Howard CR. A novel hepa-titis B virus variant in the sera of immunizedchildren. J Gen Virol 75:443–448;1994.
39 Lai ME, Farci P, Figus A, Balestrieri A, ArnoneM, Vyas GN. Hepatitis B virus DNA in theserum of Sardinian blood donors negative forthe hepatitis B surface antigen. Blood 73:17–19;1989.
40 Lau JY, Bain VG, Davies SE, O’Grady JG,Alberti A, Alexander GJ, Williams R. High-lev-el expression of hepatitis B viral antigens infibrosing cholestatic hepatitis. Gastroenterol-ogy 102:956–962;1992.
41 Lee PI, Chang LY, Lee CY, Huang LM, ChangMH. Detection of hepatitis B surface gene mu-tation in carrier children with or without im-munoprophylaxis at birth. J Infect Dis 176:427–430;1997.
42 Lee PI, Lee CY, Huang LM, Chen JM, ChangMH. A follow-up study of combined vaccina-tion with plasma-derived and recombinanthepatitis B vaccines in infants. Vaccine 13:1685–1689;1995.
43 Lee WM. Hepatitis B virus infection. N Engl JMed 337:1733–1745;1997.
44 Lemon SM, Thomas DL. Vaccines to preventviral hepatitis. N Engl J Med 336:196–204;1997.
45 Mangold CM, Unckell F, Werr M, Streeck RE.Secretion and antigenicity of hepatitis B virussmall envelope proteins lacking cysteines in themajor antigenic region. Virology 211:535–543;1995.
46 Mangold CM, Unckell F, Werr M, Streeck RE.Analysis of intermolecular disulfide bonds andfree sulfhydryl groups in hepatitis B surfaceantigen particles. Arch Virol 142:2257–2267;1997.
47 McAleer WJ, Buynak EB, Maigetter RZ,Wampler DE, Miller WJ, Hilleman MR. Hu-man hepatitis B vaccine from recombinantyeast. 1984. Biotechnology 24:500–502;1992.
48 McMahon G, Ehrlich PH, Moustafa ZA,McCarthy LA, Dottavio D, Tolpin MD, NadlerPI, Ostberg L. Genetic alterations in the geneencoding the major HBsAg: DNA and immu-nological analysis of recurrent HBsAg derivedfrom monoclonal antibody-treated liver trans-plant patients. Hepatology 15:757–766;1992.
49 Muller R, Gubernatis G, Farle M, Niehoff G,Klein H, Wittekind C, et al. Liver transplanta-tion in HBs antigen (HBsAg) carriers. Pre-vention of hepatitis B virus (HBV) recurrenceby passive immunization. J Hepatol 13:90–96;1991.
50 Ngui SL, O’Connell S, Eglin RP, Heptonstall J,Teo CG. Low detection rate and maternalprovenance of hepatitis B virus S gene mutantsin cases of failed postnatal immunoprophylaxisin England and Wales. J Infect Dis 176:1360–1365;1997.
51 Norder H, Courouce AM, Magnius LO. Com-plete genomes, phylogenetic relatedness, andstructural proteins of six strains of the hepatitisB virus, four of which represent two new geno-types. Virology 198:489–503;1994.
52 Ogata N, Zanetti AR, Yu M, Miller RH, Pur-cell RH. Infectivity and pathogenicity in chim-panzees of a surface gene mutant of hepatitis Bvirus that emerged in a vaccinated infant. JInfect Dis 175:511–523;1997.
53 Ohba K, Mizokami M, Ohno T, Suzuki K, Ori-to E, Lau JY, Ina Y, Ikeo K, Gojobori T. Rela-tionships between serotypes and genotypes ofhepatitis B virus: Genetic classification of HBVby use of surface genes. Virus Res 39:25–34;1995.
54 Okamoto H, Imai M, Kametani M, NakamuraT, Mayumi M. Genomic heterogeneity of hepa-titis B virus in a 54-year-old woman who con-tracted the infection through materno-fetaltransmission. Jpn J Exp Med 57:231–236;1987.
55 Okamoto H, Yano K, Nozaki Y, Matsui A,Miyazaki H, Yamamoto K, Tsuda F, MachidaA, Mishiro S. Mutations within the S gene ofhepatitis B virus transmitted from mothers tobabies immunized with hepatitis B immuneglobulin and vaccine. Pediatr Res 32:264–268;1992.
56 Oon CJ, Chen WN. Current aspects of hepatitisB surface antigen mutants in Singapore. J ViralHepat 5(suppl 2):17–23;1998.
57 Oon CJ, Lim GK, Ye Z, Goh KT, Tan KL, YoSL, Hopes E, Harrison TJ, Zuckerman AJ. Mo-lecular epidemiology of hepatitis B virus vac-cine variants in Singapore. Vaccine 13:699–702;1995.
58 Oon CJ, Ning CW, Shiuan K, Keow LG. Iden-tification of hepatitis B surface antigen variantswith alterations outside the ‘a’ determinant inimmunized Singapore infants. J Infect Dis 179:259–263;1999.
59 Paulij WP, de Wit PL, Sunnen CM, van Roos-malen MH, Petersen-van Ettekoven A, Coore-man MP, Heijtink RA. Localization of aunique hepatitis B virus epitope sheds newlight on the structure of hepatitis B virus sur-face antigen. J Gen Virol 80:2121–2126;1999.
60 Poirriez J. Is the strategy of universal hepatitisB vaccination necessary in low-endemic coun-tries? Scand J Gastroenterol 33:897–899;1998.
61 Pollicino T, Zanetti AR, Cacciola I, Petit MA,Smedile A, Campo S, Sagliocca L, Pasquali M,Tanzi E, Longo G, Raimondo G. Pre-S2 defec-tive hepatitis B virus infection in patients withfulminant hepatitis. Hepatology 26:495–499;1997.
62 Prange R, Mangold CM, Hilfrich R, StreeckRE. Mutational analysis of HBsAg assembly.Intervirology 38:16–23;1995.
63 Pride MW, Thanavala YM, Strick N, Hought-en RA, Neurath AR. Toleration of amino acidsubstitutions within hepatitis B virus envelopeprotein epitopes established by peptide re-placement set analysis. II. Region S(122–136).Pept Res 5:217–226;1992.
64 Protzer-Knolle U, Naumann U, BartenschlagerR, Berg T, Hopf U, Meyer zum BuschenfeldeKH, Neuhaus P, Gerken G. Hepatitis B viruswith antigenically altered hepatitis B surfaceantigen is selected by high-dose hepatitis Bimmune globulin after liver transplantation.Hepatology 27:254–263;1998.
65 Samuel D, Muller R, Alexander G, Fassati L,Ducot B, Benhamou JP, Bismuth H. Livertransplantation in European patients with thehepatitis B surface antigen. N Engl J Med 329:1842–1847;1993.
66 Sterneck M, Gunther S, Gerlach J, NaoumovNV, Santantonio T, Fischer L, Rogiers X, Gret-en H, Williams R, Will H. Hepatitis B virussequence changes evolving in liver transplantrecipients with fulminant hepatitis. J Hepatol26:754–764;1997.
67 Stirk HJ, Thornton JM, Howard CR. A topo-logical model for hepatitis B surface antigen.Intervirology 33:148–158;1992.
68 Terrault NA, Zhou S, McCory RW, Pruett TL,Lake JR, Roberts JP, Ascher NL, Wright TL.Incidence and clinical consequences of surfaceand polymerase gene mutations in liver trans-plant recipients on hepatitis B immunoglobu-lin. Hepatology 28:555–561;1998.
Vaccine- and HBIg-Induced EscapeMutations of HBsAg
J Biomed Sci 2001;8:237–247 247
69 Tillmann HL, Trautwein C, Bock T, BokerKH, Jackel E, Glowienka M, Oldhafer K,Bruns I, Gauthier J, Condreay LD, Raab HR,Manns MP. Mutational pattern of hepatitis Bvirus on sequential therapy with famciclovirand lamivudine in patients with hepatitis Bvirus reinfection occurring under HBIg immu-noglobulin after liver transplantation. Hepa-tology 30:244–256;1999.
70 Tiollais P, Pourcel C, Dejean A. The hepatitisB virus. Nature 317:489–495;1985.
71 Trautwein C, Schrem H, Tillmann HL, Ku-bicka S, Walker D, Boker KH, Maschek HJ,Pichlmayr R, Manns MP. Hepatitis B virusmutations in the pre-S genome before and afterliver transplantation. Hepatology 24:482–488;1996.
72 Waters JA, Brown SE, Steward MW, HowardCR, Thomas HC. Analysis of the antigenic epi-topes of hepatitis B surface antigen involved inthe induction of a protective antibody re-sponse. Virus Res 22:1–12;1992.
73 Waters JA, O’Rourke SM, Richardson SC, Pa-paevangelou G, Thomas HC. Qualitative anal-ysis of the humoral immune response to the ‘a’determinant of HBs antigen after inoculationwith plasma-derived or recombinant vaccine. JMed Virol 21:155–160;1987.
74 Whittle HC, Maine N, Pilkington J, Mendy M,Fortuin M, Bunn J, Allison L, Howard C, HallA. Long-term efficacy of continuing hepatitis Bvaccination in infancy in two Gambian vil-lages. Lancet 345:1089–1092;1995.
75 Wilson JN, Nokes DJ, Carman WF. The pre-dicted pattern of emergence of vaccine-resis-tant hepatitis B: A cause for concern? Vaccine17:973–978;1999.
76 Yamamoto K, Horikita M, Tsuda F, Itoh K,Akahane Y, Yotsumoto S, Okamoto H, Miya-kawa Y, Mayumi M. Naturally occurring es-cape mutants of hepatitis B virus with variousmutations in the S gene in carriers seropositivefor antibody to hepatitis B surface antigen. JVirol 68:2671–2676;1994.
77 Zhang YY, Nordenfelt E, Hansson BG. In-creasing heterogeneity of the ‘a’ determinant ofHBsAg found in the presumed late phase ofchronic hepatitis B virus infection. Scand JInfect Dis 28:9–15;1996.
78 Zuckerman AJ. Effect of hepatitis B virus mu-tants on efficacy of vaccination. Lancet 355:1382–1384;2000.
79 Zuckerman AJ, Harrison TJ, Oon CJ. Muta-tions in S region of hepatitis B virus (letter).Lancet 343:737–738;1994.
