Immune-escape mutations and stop-codons in HBsAg develop in a large proportion of patients with chronic HBV infection exposed to anti-HBV drugs in Europe

Background HBsAg immune-escape mutations can favor HBV-transmission also in vaccinated individuals, promote immunosuppression-driven HBV-reactivation, and increase fitness of drug-resistant strains. Stop-codons can enhance HBV oncogenic-properties. Furthermore, as a consequence of the overlapping structure of HBV genome, some immune-escape mutations or stop-codons in HBsAg can derive from drug-resistance mutations in RT. This study is aimed at gaining insight in prevalence and characteristics of immune-associated escape mutations, and stop-codons in HBsAg in chronically HBV-infected patients experiencing nucleos(t)ide analogues (NA) in Europe. Methods This study analyzed 828 chronically HBV-infected European patients exposed to ≥ 1 NA, with detectable HBV-DNA and with an available HBsAg-sequence. The immune-associated escape mutations and the NA-induced immune-escape mutations sI195M, sI196S, and sE164D (resulting from drug-resistance mutation rtM204 V, rtM204I, and rtV173L) were retrieved from literature and examined. Mutations were defined as an aminoacid substitution with respect to a genotype A or D reference sequence. Results At least one immune-associated escape mutation was detected in 22.1% of patients with rising temporal-trend. By multivariable-analysis, genotype-D correlated with higher selection of ≥ 1 immune-associated escape mutation (OR[95%CI]:2.20[1.32–3.67], P = 0.002). In genotype-D, the presence of ≥ 1 immune-associated escape mutations was significantly higher in drug-exposed patients with drug-resistant strains than with wild-type virus (29.5% vs 20.3% P = 0.012). Result confirmed by analysing drug-naïve patients (29.5% vs 21.2%, P = 0.032). Strong correlation was observed between sP120T and rtM204I/V (P < 0.001), and their co-presence determined an increased HBV-DNA. At least one NA-induced immune-escape mutation occurred in 28.6% of patients, and their selection correlated with genotype-A (OR[95%CI]:2.03[1.32–3.10],P = 0.001). Finally, stop-codons are present in 8.4% of patients also at HBsAg-positions 172 and 182, described to enhance viral oncogenic-properties. Conclusions Immune-escape mutations and stop-codons develop in a large fraction of NA-exposed patients from Europe. This may represent a potential threat for horizontal and vertical HBV transmission also to vaccinated persons, and fuel drug-resistance emergence. Electronic supplementary material The online version of this article (10.1186/s12879-018-3161-2) contains supplementary material, which is available to authorized users.


Background
Worldwide, around 250 million individuals have a chronic hepatitis B virus (HBV) infection. Among them, around 1 million dies as a consequence of end-stage liver disease or hepatocellular carcinoma (HCC) [1].
HBV is a highly evolving pathogen characterized by a high degree of genetic-variability (a unique property among DNA viruses) that is driven by the lack of proof-reading function of HBV reverse transcriptase (RT) and exacerbated by the high speed of the HBV replication cycle [2].
This high degree of HBV genetic-variability allows the virus to react to endogenous (i.e. immune system), and exogenous (i.e. vaccination, hepatitis B immunoglobulin, antiviral drugs) selective pressures by further modulating its genome structure.
Among the different HBV-proteins, HBV surface antigen (HBsAg) contains the major hydrophilic region that is a dominant epitope crucial for binding to neutralizing-antibodies. So far, around 30 immune-escape mutations in HBsAg (hereafter defined as immune-associated escape mutations), have been identified [3][4][5] to evade neutralizing-antibodies, to allow persistent HBV-infection and to promote viral fitness [2,6]. These mutations can have relevant pathobiological implications at the time of immunosuppression-driven HBV-reactivation, thus favoring the reuptake of viral replication during the initial weakening of immune responses [6][7][8][9]. Immune-associated escape mutations can also hamper HBsAg-recognition by antibodies induced by vaccine, thus posing a potential threat for the global vaccination program also in the setting of mother-to-child transmission [2]. In addition, Immune-associated escape mutations can decrease/abrogate HBsAg-binding to antibodies used in diagnostic assays for HBsAg-detection and -quantification [6,10,11], and thus determine a false-negativity or an underestimation of HBsAg levels, that can pose an issue for a proper diagnosis and staging of chronic HBV-infection.
Furthermore, due to the overlapping between the genes encoding reverse transcriptase (RT) and HBsAg, some RT drug-resistance mutations can introduce mutations in the major hydrophilic region of HBsAg that are capable to reduce the binding affinity for neutralizing antibodies, including those induced by HBV-vaccine [16]. Again, these mutations (hereafter defined as NA-induced immune-escape mutations) may pose a public health concern for their pathogenetic potential and possibility of transmission to vaccinated individuals.
Another type of mutation that can be detected in HBsAg is represented by stop-codons. They are associated with the synthesis of truncated forms of HBsAg that remain trapped in the endoplasmic reticulum. This intracellular HBsAg accumulation can induce an oxidative stress that can favour the neoplastic transformation of hepatocytes [17].
Information about the prevalence of the above-mentioned mutations in patients with chronic HBV-infection exposed to NA in Europe is limited. Filling this gap can provide an estimate of the pool for HBV-transmissions also to vaccinated individuals and/or can have a higher risk of disease progression. Thus, this study was designed to estimate the prevalence and characteristics of i) immune-associated escape mutations ii) NA-induced immune-escape mutations and iii) stop-codons in HBsAg in Europe.

Study population
A multicenter survey was performed on genotypic-resistance testing results generated during routine clinical assessments of patients with chronic hepatitis B attending tertiary referral centers in European countries according to Hermans et al., 2016. Inclusion criteria were: chronic hepatitis B with detectable serum HBV-DNA, exposure to ≥ 1 NA, RT/ HBsAg-sequence availability, and age ≥ 18 years [18]. Inclusion of patients exposed to NAs allows to define the prevalence of immune-associated, and also of NA-induced escape mutations.
935 RT/HBsAg-sequences were collected in the time-window between January 1998 and August 2012. Only 1 sequence per patient was included in the analysis. Patient datasets were collected in the framework of the European Society for translational antiviral research (ESAR) from 15 countries. Countries were grouped in geographical regions (http://unstats.un.org/unsd/) as follows: Northern Europe (Denmark/Norway), Western Europe (Austria/France/Germany/Luxembourg/Netherlands), Eastern Europe (Poland/Romania), and Southern Europe (Greece/Italy/Serbia/Slovenia/Spain) [19]. Israel and Turkey were grouped with Southern European countries [18].

Data characteristics
The following information was collected: serum HBV-DNA; HBsAg; hepatitis B e antigen (HBeAg); anti-HBe; serum-alanine aminotransferase (serum-ALT); exposure to ≥ 1 NA (LAM, LdT, ADV, ETV, TDF, LdT). No administrative permissions were required to review patients' records and to use related data.

RT/HBsAg sequencing
RT/HBsAg sequences obtained by well-standardized population-based sequencing procedures during routine clinical practise were collected. Sequence data consisted of FASTA files containing nucleic acid sequence information of the RT/HBsAg region. The ESAR quality control procedure was applied on all submitted sequences. If amino acid substitutions at immune-escape codons were due to ambiguities consisting of > 2 bases per nucleotide position or > 1 ambiguities per codon, or if insertions or deletions were present causing a shift in the HBsAg open-reading frame that affected immune-escape codons, sequences were excluded from the analysis [18]. Furthermore, there was no specific pattern of mutations linked to a specific center.
HBsAg sequences were analyzed using SeqScape-v2.6 software (Thermo-Fisher Scientific), then the sequences were aligned using Bioedit 7.0 software [20]. Sequences having a mixture of wild-type and mutant residues at single positions were considered to have the mutant(s) at that position. The mixed base identification was set at a percentage of 20%.
HBsAg sequences have been submitted to Genbank with the following accession number: MH218870-MH219804.

Mutation prevalence
HBsAg-sequences were analysed to define the prevalence of immune-associated escape mutations, NA-induced escape mutations, and stop-codons.
Mutations were defined as difference from HBV genotype-A reference sequence (Genbank accession number: JN182318) or HBV genotype-D reference sequence (Genbank accession number: GU456636).

Statistical analysis
Statistical analysis was performed using SPSS software (v19.0; SPSS Inc., Chicago, IL) and the statistical environment R (version 3.2.5). Data were expressed as median (interquartile range [IQR]) for quantitative variables and as counts and percentages for qualitative variables. Chi-Squared Test of Independence based on a 2 × 2 contingency table was used for qualitative data, while Mann-Whitney test for continuous data.
Univariable and multivariable logistic regression analysis was performed in order to assess the potential associations between the presence of at least one i) immune-associated escape mutation, ii) NA-induced immune-escape mutation, iii) stop-codon, with several factors, including: gender, age, serum HBV-DNA at the time of genotypic testing, LAM, ADV, ETV, TDF, geographical origin, year of collection, and HBV-genotype.
To provide a more robust characterization of immune-escape mutations and stop-codons circulating in Europe, the analysis was focused on 828 patients infected with HBV genotype-D and A. Table 1 shows demographics, clinical, biochemical, and virological characteristics of these patients.
To corroborate the correlation between immune-associated escape mutations and drug-resistance mutations, the prevalence of ≥ 1 immune-associated escape mutations was also analysed in an independent dataset of drug-naïve patients (cite Additional file 1: Table S1 for demographic and virological characteristics). The percentage of drug-naive patients harbouring drug-resistant strains is 1% (all genotype D). The only primary drug-resistance mutations detected were rtM204I (0.4%, 1/245) and rtN236T (0.4%, 1/245), while the only secondary mutations detected were rtL180M and rtV173L, each present in 0.4% of patients. Again, the presence of ≥ 1 immune-associated escape mutations in genotype D was significantly higher in drug-exposed patients with drug-resistant strains than in drug-naïve patients ( Our results also showed that the distribution of immune-associated escape mutations differed between European regions (Fig. 3). Indeed, the percentage of HBV genotype-D infected patients with ≥ 1 immune-associated escape mutation was significantly higher in Southern Europe than in Western/Northern Europe (36.7% vs 24.2%, P = 0.02). This increase was also observed in Eastern compared to Western/Northern Europe, although not statistically significant (37.5% vs 24.2%, P = 0.17) (Fig. 3).

Discussion
In this largest-to-date European survey of 828 NA-experienced chronically HBV-infected patients, ≥ 1 immune-associated escape and NA-induced mutation was observed in 22.1 and 28.6% of patients, respectively. Furthermore, in 8.5% of patients, ≥ 1 stop-codon in HBsAg was detected. Fig. 2 The histograms report the prevalence of a immune-associated escape mutations, b NA-induced immune-escape mutations, c stop-codons. The prevalence was calculated in the group of 255 patients infected with HBV genotype-A (yellow bars) and in the group of 573 patients infected with HBV genotype-D (green bars). Statistically significant differences were assessed by Chi Squared Test for independence based on a 2 × 2 contingency table. * P < 0.05; ** P < 0.01; *** P < 0.001. In A) a schematic representation of HBsAg functional domains is also reported: N-terminus HBsAg (encompassing amino acids By multivariable analysis, the selection of immune-associated escape mutations (including vaccine-escape mutations) was significantly higher in HBV genotype-D than A. HBV genotype-D is known to be more prone to the onset of HBeAg-negative chronic hepatitis characterized by an extensive accumulation of mutations in the pre-core/basal core promoter of HBV-genome in response to a potent host-based selection pressure [21]. It is conceivable that this selective pressure may also favor the generation and selection of immune-associated escape mutations in HBsAg, further exacerbating HBV-escape from immunological-pressure.
Only the immune-associated escape mutation G130 N was detected more frequently in genotype-A than -D. This difference can be explained considering the fact that the number of nucleotide substitutions necessary to generate G130 N from the wild-type amino acid is lower in genotype-A than -D [22]. This suggests that the Fig. 3 The histogram reports the percentage of patients with at least one immune-associated escape mutations between European regions. The prevalence was calculated in HBV genotype-D and -A infected patients from Western/Northern (black bars), Southern (grey bars), and Eastern Europe (light grey bars). Statistically significant differences were assessed by Chi Squared Test for independence based on a 2 × 2 contingency table. * P = 0.02  [6][7][8][9]23]. It has been proposed that immune-associated escape mutations can favor the re-uptake of HBV-replication during the initial weakening of immune-system, particularly during rituximab-treatment (known to deplete B-lymphocytes) [6]. The substantial circulation of immune-associated escape mutations may thus pose an issue in term of increased risk of HBV-reactivation in immunosuppressed-patients.
Previous in-vitro studies showed that some immune-associated escape mutations can promote the fitness of HBV lamivudine-resistant strains [23,24]. We found an enrichment of immune-associated escape mutations in drug-exposed patients with drug-resistant strains compared to drug-exposed patients with wild-type virus and to drug-naïve patients. This highlights a strict relationship between drug-resistance and immune-associated escape mutations, and suggests the ability of immune-associated escape mutations to stabilize drug-resistance mutations in viral-quasispecies. We observed that sP120T significantly correlated with rtM204V/I, and their co-presence is characterized by elevated serum HBV-DNA. This is consistent with an in-vitro study showing sP120T ability to rescue HBV-replication impaired by rtM204V/I [24].
The ability of immune-associated escape mutations to promote the fitness of HBV lamivudine-resistant strains can raise the issue on lamivudine-use as prophylaxis in immunosuppressed-patients, and highlights the importance to use potent anti-HBV drugs in order to prevent HBV-reactivation. Since the highly potent anti-HBV drugs will soon become generic, this will also allow to reduce the cost related to the management of immunosuppressed-patients at risk of HBV-reactivation.
This has also implications for those European Countries in which lamivudine is still prescribed, again supporting the role of potent anti-HBV drugs for a proper management of patients with chronic HBV-infection.
The circulation of immune-associated escape mutations can have important implications, since they can potentially affect the efficacy of the current vaccination strategy. Indeed, several studies have highlighted the presence of immune-associated escape mutations in individuals who contracted HBV-infection despite completed HBV-vaccination [25][26][27]. In a study led in Taiwan, a positive HBV-DNA was detected in 10 of 60  [27]. Immune-escape mutations can also play a relevant role in the setting of mother-to-child transmission. Currently, HBV-vaccine (in addition to immunoglobulins) is administered to children born to HBV-infected mothers. In a recent study, serum HBV-DNA was detected in 28% children born from HBsAg-positive mothers, and fully responded to HBV-vaccination. Among them, 62% infected children had ≥ 1 immune-associated escape mutation, suggesting the maternal transmission of viral strains with enhanced capability to evade neutralizing antibodies in vaccinated-children [28].
In chronic HBV-infection, recent studies highlighted that the presence of immune-associated escape mutations at baseline was negatively correlated with HBsAg-loss during treatment with potent anti-HBV drugs [29,30]. It is conceivable that the circulation of these mutations can hamper the full immune control of the virus despite potent anti-HBV therapy. This issue should be considered by the recent therapeutic strategies aimed at achieving HBV-cure.
Finally, different studies showed that some immune-associated escape mutations can affect HBsAg-quantification by altering HBsAg-binding to antibodies used in diagnostic assays [6,31,32]. HBsAg-amount is used to provide a more precise definition of the inactive carrier status and to monitor the efficacy of interferon-treatment. The presence of immune-associated escape mutations may cause an underestimation of HBsAg-levels thus hampering the proper management of chronically HBV-infected patients.
Due to the peculiar HBV-genome organization, drug-resistance mutations rtM204 V, rtM204I, and rtV173 L correspond to the NA-induced immune-escape mutations sI195M, sI196S, and sE164D. In our study, HBV genotype-A was associated with a significantly higher prevalence of NA-induced immune-escape mutations. This is in line with previous studies showing that genotype-A is more prone to develop rtM204V than genotype-D at lamivudine failure [32][33][34]. The issue of NA-induced escape mutations is critical considering the ongoing use of lamivudine in some European regions where genotype-A is predominant [18,35].
Finally, ≥ 1 stop-codon was detected in 8.5% of patients. Stop-codons can determine the accumulation of truncated HBsAg in the endoplasmic-reticulum, thus inducing oxidative stress and in turn enhancing hepatocytes proliferation [36,37]. They were detected at 20 HBsAg-positions including 172 and 182, known to promote the carcinogenic transformation of hepatocytes [38,39]. Notably, stop-codon at HBsAg-position 172 derives from the drug-resistance mutation rtA181T selected under ADV-and (in some cases) LAM-treatment [39]. This represents an important issue probably originating from the broad use of first-generation drugs which may have fuelled the circulation of viral strains with an increased oncogenic potential.

Conclusions
"Immune-escape mutations and stop-codons develop in a large proportion of NA-exposed patients in Europe. These mutant isolates may potentially transmit in general population, including vaccinated individuals, and fuel drug-resistance emergence".
In accordance with National Guidelines and/or legislature, approval by Ethic Committee was not necessary since the study was based on a retrospective analysis of anonymized viral sequences obtained for clinical routine practice for the following centers: Italy (as outlined in art. 6