EASL 2017 Clinical Practice Guidelines On The Management of Hepatitis B Virus Infection

Clinical Practice Guidelines
EASL 2017 Clinical Practice Guidelines on the management

of hepatitis B virus infectionq
European Association for the Study of the Liver ⇑
Summary infection require specific focus. Future treatment strategies to

achieve ‘cure’ of disease and new biomarkers are discussed.
Hepatitis B virus (HBV) infection remains a global public health Ó 2017 European Association for the Study of the Liver. Published
problem with changing epidemiology due to several factors by Elsevier B.V. All rights reserved.
including vaccination policies and migration. This Clinical Prac-
tice Guideline presents updated recommendations for the opti-
mal management of HBV infection. Chronic HBV infection can
be classified into five phases: (I) HBeAg-positive chronic infec- Introduction
tion, (II) HBeAg-positive chronic hepatitis, (III) HBeAg-negative
chronic infection, (IV) HBeAg-negative chronic hepatitis and (V) Infection with hepatitis B virus (HBV) remains an important glo-
HBsAg-negative phase. All patients with chronic HBV infection bal public health problem with significant morbidity and mortal-
are at increased risk of progression to cirrhosis and hepatocellular ity.1–3 New information on the pathogenesis and management of
carcinoma (HCC), depending on host and viral factors. The main HBV infection has become available since the previous EASL Clin-
goal of therapy is to improve survival and quality of life by pre- ical Practice Guidelines (CPGs) prepared in 2011 and published in
venting disease progression, and consequently HCC development. 2012.1 The objective of this manuscript is to update the recom-
The induction of long-term suppression of HBV replication repre- mendations for the optimal management of HBV infection. In
sents the main endpoint of current treatment strategies, while order to keep the manuscript and particularly the reference list
HBsAg loss is an optimal endpoint. The typical indication for within a reasonable length, only references published after
treatment requires HBV DNA [2,000 IU/ml, elevated ALT and/or 2012 have been considered, since the readers can find the older
at least moderate histological lesions, while all cirrhotic patients supportive references in the 2012 EASL HBV CPGs.1 The CPGs
with detectable HBV DNA should be treated. Additional indica- do not fully address prevention including vaccination. In addi-
tions include the prevention of mother to child transmission in tion, despite increasing knowledge, areas of uncertainty still exist
pregnant women with high viremia and prevention of HBV reac- and therefore clinicians, patients and public health authorities
tivation in patients requiring immunosuppression or chemother- must continue to make choices based on the evolving evidence.
apy. The long-term administration of a potent nucleos(t)ide
analogue with high barrier to resistance, i.e., entecavir, tenofovir
disoproxil or tenofovir alafenamide, represents the treatment of
choice. Pegylated interferon-alfa treatment can also be consid- Background
ered in mild to moderate chronic hepatitis B patients. Combina-
tion therapies are not generally recommended. All patients Epidemiology and public health burden
should be monitored for risk of disease progression and HCC.
Treated patients should be monitored for therapy response and Approximately 240 million people are chronic HBV surface anti-
adherence. HCC remains the major concern for treated chronic gen (HBsAg) carriers, with a large regional variation of HBsAg-
hepatitis B patients. Several subgroups of patients with HBV positive patients between low (\2%) and high ([8%) endemicity
levels.2,4 The prevalence is decreasing in several highly endemic
countries due to improvements in the socioeconomic status, uni-
versal vaccination programs and perhaps effective antiviral treat-
Keywords: Hepatitis B; EASL guidelines; Treatment; Interferon; Entecavir; ments.5 However, population movements and migration are
Tenofovir; TAF; HBsAg; Hepatocellular carcinoma; HBV DNA; HBV reactivation;
currently changing the prevalence and incidence in several low
Mother to child transmission.
Received 23 March 2017; accepted 23 March 2017
endemic countries in Europe (e.g., Italy, Germany), owing to the
q
Clinical practice guidelines panel: Chair: Pietro Lampertico; Panel members: higher HBsAg prevalence rates in migrants and refugees from
Kosh Agarwal, Thomas Berg, Maria Buti, Harry L.A. Janssen, George Papatheodor- outside Europe compared with the indigenous population.6,7
idis, Fabien Zoulim; EASL Governing Board representative: Frank Tacke. Even with universal vaccination programs, it has been impossible
⇑ Corresponding author. Address: European Association for the Study of the Liver
to substantially prevent acute cases of HBV infection, especially
(EASL), The EASL Building – Home of Hepatology, 7 rue Daubin, CH 1203 Geneva,
Switzerland. Tel.: +41 (0) 22 807 03 60; fax: +41 (0) 22 328 07 24. in high risk populations.8,9 The number of HBV related deaths
E-mail address: easloffice@easloffice.eu due to liver cirrhosis and/or hepatocellular carcinoma (HCC)
Journal of Hepatology 2017 vol. 67 j 370–398

JOURNAL OF HEPATOLOGY
increased between 1990 and 2013 by 33%, relating to [686,000 infection progresses through distinct disease phases that are
cases in 2013 worldwide.10 strongly associated with age. It has been observed that children
and young adults with chronic HBV infection have an immune
Virology and immunopathogenesis profile that is less compromised than that observed in older
patients, challenging the concept of ‘immune tolerance’.16 Several
The viral life cycle studies showed that HBV persists with virus-specific and global T
Human HBV belongs to the Hepadnaviridae family of small, envel- cell dysfunction mediated by multiple regulatory mechanisms,
oped, primarily hepatotropic DNA viruses. In the host, the virus but without distinct T cellbased immune signatures for clinical
replicates and assembles exclusively in hepatocytes, and virions phenotypes (or clinical phase of infection).16,17 Genome-wide
are released non-cytopathically through the cellular secretory association studies recently identified the INTS10 gene at
pathway. The viral genome shows an extremely compact organ- 8p21.3 as a novel locus contributing to the susceptibility to per-
isation. The small (3.2 kb), partially double-stranded, relaxed- sistent HBV infection among Chinese subjects, and being causa-
circular (rc) DNA features 4 open reading frames encoding 7 pro- tive for HBV clearance by activation of IRF3 and then expression
teins: HBeAg (HBV e antigen, secreted dimeric protein), HBcAg of anti-virus interferons hereby highlighting the role of innate
(HBV core antigen, viral capsid protein), HBV Pol/RT (polymerase, immunity in viral clearance.18
reverse transcriptase activity), PreS1/PreS2/HBsAg (large, med-
ium, and small surface envelope glycoproteins), and HBx (HBV Natural history and new nomenclature for the chronic states
x antigen, regulator of transcription required for the initiation
of infection).11,12 Upon viral uptake into hepatocytes, the HBV Chronic HBV infection is a dynamic process reflecting the interac-
nucleocapsid is transported to the nucleus to release the rcDNA tion between HBV replication and the host immune response and
genome. In the nucleoplasm, the rcDNA is converted into a cova- not all patients with chronic HBV infection have chronic hepatitis
lently closed circular DNA (cccDNA), which is wrapped by his- (CHB). The natural history of chronic HBV infection has been
tones to form an episomal chromatinized structure. It then schematically divided into five phases, taking into account the
serves as a transcription template for all viral transcripts that presence of HBeAg, HBV DNA levels, alanine aminotransferase
are translated into the different viral proteins.13 Besides encoding (ALT) values and eventually the presence or absence of liver
the capsid protein and the viral polymerase, the pregenomic RNA inflammation (Fig. 1). The new nomenclature is based on the
is reverse transcribed into new rcDNA within the viral capsid. The description of the two main characteristics of chronicity: infec-
DNA containing nucleocapsids in the cytoplasm are either recy- tion vs. hepatitis. However, despite this nomenclature, in a signif-
cled into the nucleus to maintain cccDNA reservoir, or enveloped icant number of patients, a single determination of HBV
and secreted via the endoplasmic reticulum.11 In addition to replication markers as well as disease activity markers does not
complete infectious virions (diameter of 42 nm), infected cells allow an immediate classification to one of the phases. Serial
produce a large excess of genome-free, non-infectious sub-viral monitoring of serum HBeAg, HBV DNA and ALT levels is required
spherical or filamentous particles of 22 nm.11 Viral genome inte- in most instances but even after a complete assessment, some
gration in the host genome can occur randomly; it is not required subjects fall into an indeterminate grey area and management
for viral replication, but is one of the important mechanisms needs to be individualised. The phases of chronic HBV infection
involved in hepatocyte transformation.14 are not necessarily sequential:
Phase 1: HBeAg-positive chronic HBV infection, previously ter-
Genetic variability of HBV med ‘‘immune tolerant’’ phase; characterised by the presence of
The lack of reverse transcriptase proofreading activity leads to serum HBeAg, very high levels of HBV DNA and ALT persistently
frequent mutations of the viral genome. This results in the coex- within the normal range according to traditional cut-off values
istence of genetically distinct viral species in infected individuals, [upper limit of normal (ULN) approximately 40 IU/L].1 In the
also called viral quasispecies, which evolve depending on the liver, there is minimal or no liver necroinflammation or fibrosis
pressure from the host environment. The interplay between the but a high level of HBV DNA integration and clonal hepatocyte
virus, hepatocyte and the immune response or antiviral treat- expansion suggesting that hepatocarcinogenesis could be already
ment is thought to drive the emergence of HBV mutants that underway in this early phase of the infection.1,19 This phase is
have the capacity to escape immune responses or antiviral treat- more frequent and prolonged in subjects infected perinatally
ments. Analysis of genome-wide nucleotide divergence has and is associated with preserved HBV specific T cell function at
allowed for the identification of nine genotypes (A-I) and several least until young adulthood.20 The rate of spontaneous HBeAg
sub-genotypes.12,15 loss is very low in this phase. These patients are highly conta-
gious due to the high levels of HBV DNA.
Immunopathogenesis Phase 2: HBeAg-positive chronic hepatitis B is characterised by
In acute resolving infections, the response of the innate and adap- the presence of serum HBeAg, high levels of HBV DNA and ele-
tive immune system to HBV is efficient and timely. Viral clear- vated ALT. In the liver, there is moderate or severe liver necroin-
ance involves the induction of a robust adaptive T cell reaction flammation and accelerated progression of fibrosis1. It may occur
inducing both a cytolytic dependent and independent antiviral after several years of the first phase and is more frequently and/
effect via the expression of antiviral cytokines, as well as the or rapidly reached in subjects infected during adulthood. The out-
induction of B cells producing neutralizing antibodies preventing come of this phase is variable. Most patients can achieve HBeAg
the spread of the virus.16,17 Hepatocyte turnover resulting from seroconversion and HBV DNA suppression and enter the
infected cell death leads to cccDNA dilution. HBeAg-negative infection phase. Other patients may fail to con-
When the acute infection becomes chronic, there is a progres- trol HBV and progress to the HBeAg-negative CHB phase for many
sive impairment in HBV specific T cell function. Chronic HBV years.
Journal of Hepatology 2017 vol. 67 j 370–398 371

Natural history and assessment of patients with chronic HBV infection
HBV markers Liver disease

HBsAg Biochemical parameters: ALT
HBeAg/anti-HBe Fibrosis markers: non-invasive markers
HBV DNA of fibrosis (elastography or biomarkers)
or liver biopsy in selected cases
HBeAg positive HBeAg negative

Chronic infection Chronic hepatitis Chronic infection Chronic hepatitis
HBsAg High High/intermediate Low Intermediate
HBeAg Positive Positive Negative Negative
HBV DNA >107 IU/ml 104-107 IU/ml <2,000 IU/ml°° >2,000 IU/ml
ALT Normal Elevated Normal Elevated*
Liver disease None/minimal Moderate/severe None Moderate/severe
Old terminology Immune tolerant Immune reactive HBeAg positive Inactive carrier HBeAg negative chronic hepatitis
Fig. 1. Natural history and assessment of patients with chronic HBV infection based upon HBV and liver disease markers. *Persistently or intermittently. °°HBV DNA
levels can be between 2,000 and 20,000 IU/ml in some patients without sings of chronic hepatitis.
Phase 3: HBeAg-negative chronic HBV infection, previously incidence of cirrhosis ranges from 8% to 20% in untreated CHB
termed ‘inactive carrier’ phase, is characterised by the presence patients and, among those with cirrhosis, the 5-year cumula-
of serum antibodies to HBeAg (anti-HBe), undetectable or low tive risk of hepatic decompensation is 20%.1 The annual risk
(\2,000 IU/ml) HBV DNA levels and normal ALT according to tra- of HCC in patients with cirrhosis has been reported to be
ditional cut-off values (ULN 40 IU/L). Some patients in this 2–5%.23
phase, however, may have HBV DNA levels [2,000 IU/ml (usually HCC is currently the main concern for diagnosed CHB
\20,000 IU/ml) accompanied by persistently normal ALT and patients and may develop even in patients who have been effec-
only minimal hepatic necroinflammatory activity and low fibro- tively treated.24 The risk of developing HCC is higher in patients
sis. These patients have low risk of progression to cirrhosis or with one or more factors that relate to the host (cirrhosis,
HCC if they remain in this phase, but progression to CHB, usually chronic hepatic necroinflammation, older age, male sex, African
in HBeAg-negative patients, may occur.1 HBsAg loss and/or origin, alcohol abuse, chronic co-infections with other hepatitis
seroconversion may occur spontaneously in 1–3% of cases per viruses or human immunodeficiency virus [HIV], diabetes or
year.1 Typically, such patients may have low levels of serum metabolic syndrome, active smoking, positive family history)
HBsAg (\1,000 IU/ml).21 and/or to HBV properties (high HBV DNA and/or HBsAg levels,
Phase 4: HBeAg-negative chronic hepatitis B is characterised by HBV genotype C > B, specific mutations).24 The above factors
the lack of serum HBeAg usually with detectable anti-HBe, and seem to affect the progression to cirrhosis in untreated CHB
persistent or fluctuating moderate to high levels of serum HBV patients.1
DNA (often lower than in HBeAg-positive patients), as well as fluc- Several risk scores have been recently developed for HCC
tuating or persistently elevated ALT values. The liver histology prediction in CHB patients. Most of them, such as GAG-
shows necroinflammation and fibrosis.1 Most of these subjects HCC, CU-HCC and REACH-B, have been developed and vali-
harbour HBV variants in the precore and/or the basal core pro- dated in Asian untreated CHB patients,25 but they do not
moter regions that impair or abolish HBeAg expression. This phase seem to offer good predictability in most studies including
is associated with low rates of spontaneous disease remission.1 Caucasian CHB patients.26,27 A recently developed and vali-
Phase 5: HBsAg-negative phase is characterised by serum neg- dated new score, PAGE-B, offers good predictability for HCC
ative HBsAg and positive antibodies to HBcAg (anti-HBc), with or during the first 5 years of entecavir or tenofovir therapy in
without detectable antibodies to HBsAg (anti-HBs). This phase is Caucasian, mostly European, CHB patients and can be easily
also known as ‘‘occult HBV infection”. In rare cases, the absence applied in clinical practice, as it is based on widely available
of HBsAg could be related to the sensitivity of the assay used parameters (platelets, age, gender).28 The PAGE-B score
for detection.22 Patients in this phase have normal ALT values appears to predict HCC development even in untreated CHB
and usually, but not always, undetectable serum HBV DNA. patients.29,30
HBV DNA (cccDNA) can be detected frequently in the liver.1
HBsAg loss before the onset of cirrhosis is associated with a min- Initial assessment of subjects with chronic HBV infection
imal risk of cirrhosis, decompensation and HCC, and an improve-
ment on survival. However, if cirrhosis has developed before The initial evaluation of a subject with chronic HBV infection
HBsAg loss, patients remain at risk of HCC therefore HCC surveil- should include a complete history, a physical examination,
lance should continue. Immunosuppression may lead to HBV assessment of liver disease activity and severity and markers of
reactivation in these patients.1 HBV infection (Fig. 1). In addition, all first degree relatives and
sexual partners of subjects with chronic HBV infection should
Factors related to progression to cirrhosis and HCC be advised to be tested for HBV serological markers (HBsAg,
The risk of progression to cirrhosis and HCC is variable and is anti-HBs, anti-HBc) and to be vaccinated if they are negative for
affected by the host’s immune response. The 5-year cumulative these markers.
372 Journal of Hepatology 2017 vol. 67 j 370–398

(1) The assessment of the severity of liver disease is important (e.g. 6 and 12 months instead of 24 and 48/52 weeks,
to identify patients for treatment and HCC surveillance. It respectively).
is based on a physical examination and biochemical
parameters (aspartate aminotransferase [AST] and ALT,
Guidelines
gamma-glutamyl transpeptidase [GGT], alkaline phos-
phatase, bilirubin, and serum albumin and gamma globu-
Goals of therapy
lins, full blood count and prothrombin time). An
abdominal hepatic ultrasound is recommended in all
The main goal of therapy for patients with chronic HBV infection
patients. A liver biopsy or a non-invasive test should be
is to improve survival and quality of life by preventing disease
performed to determine disease activity in cases where
progression, and consequently HCC development. Additional
biochemical and HBV markers reveal inconclusive
goals of antiviral therapy are to prevent mother to child transmis-
results.31 Of the non-invasive methods, which include liver
sion, hepatitis B reactivation and the prevention and treatment of
stiffness measurements and serum biomarkers of liver
HBV-associated extrahepatic manifestations.
fibrosis, the use of transient elastography has been mostly
The likelihood of achieving these goals depends on the timing
studied and seems to offer a higher diagnostic accuracy for
of therapy during the natural course of the infection but also on
the detection of cirrhosis. The diagnostic accuracy of all
the stage of the disease and the patients’ age when treatment is
non-invasive methods is better at excluding than confirm-
started. Regression of fibrosis and cirrhosis can be regarded as a
ing advanced fibrosis or cirrhosis.31,32 The results of tran-
further goal of treatment in patients with established advanced
sient elastography may be confounded by severe
fibrosis or cirrhosis, although its impact has not been fully clari-
inflammation associated with high ALT levels.31,32
fied in clinical outcomes. Treatment strategies to prevent HCC
(2) HBeAg and anti-HBe detection are essential for the deter-
development may differ in some ways from those that are needed
mination of the phase of chronic HBV infection.
to prevent fibrosis progression.
(3) Measurement of HBV DNA serum level is essential for the
In patients with HBV-induced HCC, the goals of nucleos(t)ide
diagnosis, establishment of the phase of the infection, the
analogue (NA) therapy are firstly to suppress HBV replication to
decision to treat and subsequent monitoring of patients.
induce the stabilisation of HBV-induced liver disease and to pre-
(4) Serum HBsAg quantification can be useful, particularly in
vent disease progression, and secondly to reduce the risk of HCC
HBeAg-negative chronic HBV infection and in patients to
recurrence after potentially curative HCC therapies. Stabilising
be treated with interferon-alfa (IFNa).
the HBV-induced liver disease can be also regarded as a prerequi-
(5) HBV genotype is not necessary in the initial evaluation,
site for the safe and effective applications of HCC treatments.
although it may be useful for selecting patients to be trea-
In patients with acute hepatitis B, preventing the risk of acute
ted with IFNa offerering prognostic information for the
or subacute liver failure is the main treatment goal. Improving
probability of response to IFNa therapy and the risk of
the quality of life by shortening the duration of the disease asso-
HCC.
ciated symptoms as well as lowering the risk of chronicity may be
(6) Co-morbidities, including alcoholic, autoimmune, meta-
also regarded as relevant goals of treatment.
bolic liver disease with steatosis or steatohepatitis and
other causes of chronic liver disease should be systemati-
Endpoints of therapy
cally excluded including co-infections with hepatitis D
virus (HDV), hepatitis C virus (HCV) and HIV.
Recommendations
(7) Testing for antibodies against hepatitis A virus (anti-HAV)
should be performed, and patients with negative anti-HAV
should be advised to be vaccinated against HAV.
The induction of long-term suppression of HBV DNA
levels represents the main endpoint of all current treat-
ment strategies (Evidence level I, grade of recommenda-
tion 1).
Methodology
The induction of HBeAg loss, with or without anti-HBe
seroconversion, in HBeAg-positive CHB patients is a
These CPGs were developed by a CPG panel of experts chosen by
valuable endpoint, as it often represents a partial
the EASL Governing Board, peer-reviewed by three external
immune control of the chronic HBV infection (Evidence
experts and approved by the EASL Governing Board. The CPGs
level II-1, grade of recommendation 1).
have been based as far as possible on evidence from existing pub-
lications, and, if evidence was unavailable, on the experts’ per- A biochemical response defined as ALT normalisation
sonal experience and opinion. Manuscripts and abstracts of should be considered as an additional endpoint, which
important meetings published since the last CPG and prior to is achieved in most patients with long-term suppression
December 2016 have been evaluated. The evidence and recom- of HBV replication (Evidence level II-1, grade of recom-
mendations in these guidelines have been graded according to mendation 1).
the Grading of Recommendations Assessment Development and
HBsAg loss, with or without anti-HBs seroconversion, is
Evaluation (GRADE) system33 (Table 1). The strength of recom-
an optimal endpoint, as it indicates profound suppres-
mendations (strong: 1, weak: 2) thus reflects the quality (grade)
sion of HBV replication and viral protein expression
of underlying evidence (I, II-1, II-2, II-3, III). Grades are not pro-
(Evidence level II-1, grade of recommendation 1).
vided for statements and definitions. For practical reasons,
months and not weeks were used in parts of the manuscript

Table 1. Grading evidence and recommendations (adapted from GRADE system).
Grade evidence
I Randomised, controlled trials
II-1 Controlled trials without randomisation
II-2 Cohort or case-control analytical studies
II-3 Multiple time series, dramatic uncontrolled experiments
III Opinions of respected authorities, descriptive epidemiology
Grade recommendation
1 Strong recommendation: Factors influencing the strength of the recommendation included the quality of the evidence, presumed
patient-important outcomes, and cost
2 Weaker recommendation: Variability in preferences and values, or more uncertainty: more likely a weak recommendation is
warranted
Recommendation is made with less certainty: higher cost or resource consumption
The level of HBV replication represents the strongest single loss is achieved at a younger age and/or in the absence of sig-
predictive biomarker associated with disease progression and nificant fibrosis.1,54 In an Asian cohort followed for 287
the long-term outcome of chronic HBV infection. The inhibition patient-years after NA treatment induced HBsAg seroclearance,
of viral replication by antiviral treatment has been shown to only two patients with baseline cirrhosis developed HCC or
achieve the elimination of chronic HBV-induced necroinflamma- died (0.7% annual risk), which was a significantly lower rate
tory activity and progressive fibrotic liver processes in the vast compared with propensity score-matched patients without
majority of patients, in turn reducing the risk of HCC. It therefore HBsAg seroclearance (HR 0.09, p \0.01).47
represents the cornerstone endpoint of all our current therapeu-
tic attempts.1,25,34–40 The level of HBV DNA suppression that
should be attained in order to achieve these benefits is not well Indications for treatment
defined, but inferred that the lower, the better.
Treatment-induced HBeAg loss and seroconversion to anti- Recommendations
HBe characterises the induction of a partial immune control often
leading to a low replicative phase of the chronic HBV infection.
Whether this is a durable phase is only proven after treatment All patients with HBeAg-positive or -negative chronic
cessation. After stopping therapy, HBeAg seroreversion, as well hepatitis B, defined by HBV DNA [2,000 IU/ml, ALT
as the development of HBeAg-negative CHB, may also occur (even [ULN and/or at least moderate liver necroinflammation
after NA consolidation treatment), making this endpoint less reli- or fibrosis, should be treated (Evidence level I, grade of
able.41,42 Hence, continuing oral antiviral therapy irrespective of recommendation 1).
the HBeAg response until HBsAg loss has become an alternative Patients with compensated or decompensated cirrhosis
strategy. need treatment, with any detectable HBV DNA level
Suppression of HBV DNA to undetectable levels is normally and regardless of ALT levels (Evidence level I, grade of
associated with normalisation of ALT levels. Persistence of ele- recommendation 1).
vated ALT levels in patients with complete suppression of viral
replication is associated with a lower chance of fibrosis regres- Patients with HBV DNA [20,000 IU/ml and ALT [2xULN
sion and can be a reason for histologic disease progression.43 should start treatment regardless of the degree of fibro-
The most likely explanation for these findings is the presence of sis (Evidence level II-2, grade of recommendation 1).
concomitant liver injury such as alcoholic or non-alcoholic fatty Patients with HBeAg-positive chronic HBV infection,
liver disease.34,44 In contrast, transient ALT flares may indicate defined by persistently normal ALT and high HBV DNA
some level of immune reconstitution and can be associated with levels, may be treated if they are older than 30 years
favourable outcomes.1,45,46 regardless of the severity of liver histological lesions
The loss of HBsAg is regarded as the optimal treatment end- (Evidence level III, grade of recommendation 2).
point, termed ‘functional cure’, but it is only rarely achieved
with our current antiviral armamentarium. Spontaneous HBsAg Patients with HBeAg-positive or HBeAg-negative
seroreversion with reactivation of the inflammatory liver pro- chronic HBV infection and family history of HCC or cir-
cess after HBsAg loss is rare and may occur in patients with rhosis and extrahepatic manifestations can be treated
a significant impairment of their immune function.47–52 The even if typical treatment indications are not fulfilled
main advantage of HBsAg loss is that it allows a safe discontin- (Evidence level III, grade of recommendation 2).
uation of antiviral therapy. As chronic HBV infection cannot be
completely eradicated due to the persistence of cccDNA and
integrated HBV DNA,1 it remains unclear whether HBsAg loss The indications for treatment are generally the same for both
adds to the prevention of the long-term complications of HBeAg-positive and HBeAg-negative CHB (Fig. 2). This is based
chronic HBV infection beyond what can be achieved by the mainly on the combination of three criteria:
suppression of HBV DNA replication alone. HCC may still Serum HBV DNA levels
develop even after spontaneous HBsAg loss (annual rate Serum ALT levels
approximately 0.55%).53 The risk, however, is lower if HBsAg Severity of liver disease

Patients without cirrhosis should be considered for treatment Patients with HBeAg-negative chronic HBV infection and HBV
when they have HBV DNA levels above 2,000 IU/ml, serum ALT DNA \2,000 IU/ml should have ALT determinations every 6–
levels above the traditional ULN (40 IU/L) and severity of liver 12 months and periodical HBV DNA and liver fibrosis assess-
disease assessed traditionally by liver biopsy showing at least ments, perhaps every 2–3 years. A quantitative determination
moderate necroinflammation and/or at least moderate fibrosis. of HBsAg levels can be helpful in the decision on the frequency
Patients with HBV DNA [20,000 IU/ml and ALT [2x ULN can of follow-up in such patients.21 Patients can be followed for
start treatment even without a liver biopsy. Liver biopsy may ALT levels every 12 months and HBV DNA and liver fibrosis
provide additional useful information but it does not usually assessments every 3 years if they have HBsAg levels \1,000 IU/
change the decision for treatment. A non-invasive method for ml, while follow-up with ALT every 6 months and HBV DNA
the estimation of the extent of fibrosis and, most critically from and liver fibrosis assessment at least every 2 years is advised
a monitoring perspective, to confirm or rule out cirrhosis is useful for patients with HBsAg levels P1,000 IU/ml.1,21,55
in patients who start treatment without liver biopsy. Patients with HBeAg-negative chronic HBV infection and HBV
In patients who have HBV DNA [2,000 IU/ml and at least DNA P2,000 IU/ml should be followed with ALT determinations
moderate fibrosis, treatment may be initiated even if ALT levels at least every 3 months for the first year and every 6 months
are normal. In patients who cannot or are reluctant to undergo thereafter, as well as with assessments of HBV DNA and liver
liver biopsy, non-invasive markers of fibrosis may also be used fibrosis by a non-invasive method every year for at least 3 years.
for decisions on treatment indications. If they do not fulfill any treatment indication within the first
As explained in more detail in the EASL-ALEH CPGs on ‘‘non- 3 years of follow-up, they should be consequently followed for
invasive tests for evaluation of liver disease severity and progno- life, like all patients in this phase.55
sis”,32 patients with chronic HBV infection either with normal
ALT and liver stiffness [9 kPa, or with elevated ALT but below Treatment strategies
5x ULN and liver stiffness [12 kPa at a reliable transient elastog-
raphy can be considered to have severe fibrosis or cirrhosis. Currently, there are two main treatment options for CHB
Equivalent cut-offs from other elastographic or serological meth- patients: treatment with a NA or with IFNa, currently pegylated
ods of assessment of liver fibrosis may also be used once vali- (PegIFNa) (Table 2).1,56 The NAs that have been approved in Eur-
dated in chronic HBV patients. ope for HBV treatment include lamivudine (LAM), adefovir dip-
Indications for treatment may also take into account the ivoxil (ADV), entecavir (ETV), telbivudine (TBV), tenofovir
patients’ age, health status, risk of HBV transmission, family his- disoproxil fumarate (TDF) and tenofovir alafenamide (TAF), and
tory of HCC or cirrhosis and extrahepatic manifestations (Fig. 2). can be classified into those associated with low barrier against
HBV resistance (LAM, ADV, TBV) and those with high barrier to
Monitoring of patients currently not treated HBV resistance (ETV, TDF, TAF) (Fig. 3).1,56,57
The main advantage of treatment with a potent NA with high
Recommendations barrier to resistance (i.e., ETV, TDF, TAF) is its predictable high
long-term antiviral efficacy leading to undetectable HBV DNA
levels in the vast majority of compliant patients as well as its
Patients with HBeAg-positive chronic HBV infection favourable safety profile (Table 2).1,56,57 These drugs can be safely
who are younger than 30 years and do not fulfill any used in any HBV infected patient and represent the only treatment
of the above treatment indications should be followed option for several patient subgroups including those with decom-
at least every 3–6 months (Evidence level II-2, grade of pensated liver disease, liver transplants, extrahepatic manifesta-
recommendation 1). tions, acute hepatitis B or severe chronic HBV exacerbation.57–61
NAs are also the only option for prevention of HBV reactivation
Patients with HBeAg-negative chronic HBV infection
in patients under immunosuppression. In addition, preventing
and serum HBV DNA \2,000 IU/ml who do not fulfill HBV transmission in patients with high viremia who do not fulfill
any of the above treatment indications should be fol-
the typical criteria for treatment initiation represents further indi-
lowed every 6–12 months (Evidence level II-2, grade of cations in which only NAs should be used.1,49,50,52,56,57
recommendation 1).
The rationale for a PegIFNa based approach is to induce long-
Patients with HBeAg-negative chronic HBV infection term immunological control with a finite duration treatment. The
and serum HBV DNA P2,000 IU/ml who do not fulfill main disadvantages of PegIFNa treatment is the high variability
any of the above treatment indications should be fol- of response and its unfavourable safety profile making a signifi-
lowed every 3 months for the first year and every cant number of patients ineligible or unwilling for this type of
6 months thereafter (Evidence level III, grade of recom- treatment (Table 2).1,56 Patient selection according to disease
mendation 1). activity, HBV genotype, stage of the disease, as well as levels of
HBV DNA, HBsAg and HBeAg status can be helpful indicators to
predict the individual response probability.1,56 Early on-
treatment predictors are established and can be used as addi-
Patients who are not candidates for antiviral therapy should be
tional tools (e.g. stopping rules) to individualise the treatment
monitored with periodical assessments of serum ALT and HBV
strategy, this helps to discontinue PegIFNa early in those with a
DNA levels as well as for liver fibrosis severity by non-invasive
low likelihood of long-term response.1
markers (Fig. 2). Patients with HBeAg-positive chronic HBV infec-
Theoretically, a combined NA and PegIFNa approach may pro-
tion who remain untreated should ideally have ALT determina-
vide advantages by combining the potent antiviral effect of NA
tions at least every 3 months, HBV DNA determinations every
plus the immune modulation of IFNa.1,56,62,63 The evidence for
6–12 months and assessment of liver fibrosis every 12 months.1

Suspected HBV infection
HBsAg positive HBsAg negative, anti-HBc positive
Chronic HBV infection1 Chronic hepatitis B (CHB) No specialist follow-up

(no signs of chronic hepatitis) ± cirrhosis1 but inform patient and general
practitioner about the potential
risk of HBV reactivation
Monitor
(includes HBsAg HBeAg, HBV DNA,
ALT, fibrosis assessment) Start antiviral treatment
- no
Consider In case of immunosuppression,
Risk of HCC, risk of HBV reactivation, + start oral antiviral prophylaxis
extrahepatic manifestations, or monitor
yes
risk of HBV transmission
Fig. 2. Algorithm for the management of HBV infection. 1see definitions in text and Fig. 1.
superiority of such a combined approach, however, is lacking, and Serological responses for HBeAg are HBeAg loss and HBeAg
there are still many unresolved issues with respect to patient seroconversion, i.e., HBeAg loss and development of anti-HBe
selection, timing, as well as the duration of the combination (only for HBeAg-positive patients).
strategy, which may be addressed in future studies. Serological responses for HBsAg are HBsAg loss and HBsAg sero-
conversion, i.e., HBsAg loss and development of anti-HBs (for all
Definitions of response patients).
Biochemical response is defined as a normalisation of ALT
Responses can be divided into virological, serological, biochemi- levels based on the traditional ULN (40 IU/L). Since ALT activity
cal, and histological. All responses can be estimated at several often fluctuates over time, a minimum follow-up of at least
time points during and after therapy. The definitions of virologi- 1 year post-treatment with ALT determinations at least every
cal responses vary according to the timing (on or after therapy) 3 months is required to confirm sustained off-treatment bio-
and type of therapy.1 chemical response. It should be noted that the rates of sustained
off-treatment biochemical responses may sometimes be difficult
Virological responses to evaluate, as transient ALT elevations before long-term bio-
(1) NA therapy chemical remission may occur in some CHB patients within the
Virological response during NA is defined as undetectable first year after treatment discontinuation. In such cases, addi-
HBV DNA by a sensitive polymerase chain reaction (PCR) tional close ALT follow-up of at least 2 years after ALT elevation
assay with a limit of detection of 10 IU/ml. Primary non- seems to be reasonable in order to confirm sustained off-
response is defined by a less than one log10 decrease of therapy biochemical remission.
serum HBV DNA after 3 months of therapy. Partial virologi- Histological response is defined as a decrease in necroinflam-
cal response is defined as a decrease in HBV DNA of more matory activity (by P2 points in histologic activity index or
than 1 log10 IU/ml but detectable HBV DNA after at least Ishak’s system) without worsening in fibrosis compared to pre-
12 months of therapy in compliant patients. Virological treatment histological findings.
breakthrough is defined as a confirmed increase in HBV
DNA level of more than 1 log10 IU/ml compared to the nadir
(lowest value) HBV DNA level on-therapy; it may precede a
biochemical breakthrough, characterised by an increase in NAs for naïve CHB patients
ALT levels. HBV resistance to NA(s) is characterised by selec-
tion of HBV variants with amino acid substitutions that con- Efficacy
fer reduced susceptibility to the administered NA(s). Recommendations
In patients who discontinue NA, sustained off-therapy
virological response could be defined as serum HBV DNA
levels \2,000 IU/ml for at least 12 months after the end The long-term administration of a potent NA with high
of therapy. barrier to resistance is the treatment of choice regard-
less of the severity of liver disease (Evidence level I,
(2) PegIFNa therapy grade of recommendation 1).
Virological response is defined as serum HBV DNA levels The preferred regimens are ETV, TDF and TAF as monother-
\2,000 IU/ml. It is usually evaluated at 6 months and at apies (Evidence level I, grade of recommendation 1).
the end of therapy.
Sustained off-therapy virological response is defined as LAM, ADV and TBV are not recommended in the treatment
serum HBV DNA levels \2,000 IU/ml for at least of CHB (Evidence level I, grade of recommendation 1).
12 months after the end of therapy.

Table 2. Main concepts and features of current treatment strategies of chronic hepatitis B.
Features PegIFNa ETV, TDF, TAF

Route of Subcutaneous injections Oral
administration
Treatment duration 48 weeks Long-term until HBsAg loss (stopping NA after some years might be
considered in selected cases)1
Tolerability Low High
Long-term safety Very rarely persistence of on-treatment adverse events Probably not (uncertainties regarding kidney function, bone
concerns (psychiatric, neurological, endocrinological) diseases for some NA)
Contraindications Many (i.e., decompensated disease, co-morbidities etc.) None (dose adjustment according to eGFR2)
Strategy Induction of a long-term immune control by finite Stopping hepatitis and disease progression by inhibiting viral
treatment replication
Level of viral Moderate (variable response pattern) Universally high
suppression
Effect on HBeAg loss Moderate, depending on baseline characteristics Low in the first year, increases to moderate during long-term
treatment
Effect on HBsAg levels Variable, depending on baseline characteristics (overall Low: slowly increases with treatment time in HBeAg-positive
higher as compared to NA) patients3; usually very low in HBeAg-negative patients
Risk of relapse after Low for those with sustained response 6–12 months after Moderate if consolidation treatment provided after HBeAg
treatment cessation therapy seroconversion.
High for HBeAg-negative disease
Early stopping rules Yes No
Risk of viral resistance No Minimal to none4
development
PegIFNa, pegylated interferon alfa; ETV, entecavir; TDF, tenofovir disoproxil fumarate; TAF, tenofovir alafenamide; NA, nucleoside/nucleotide analogues; eGFR, estimated
glomerular filtration rate.
1
See section on ‘Treatment strategies’.
2
Dose adjustments in patients with eGFR \50 ml/min are required for all NA, except for TAF (no dose recommendation for TAF in patients with CrCl \15 ml/min who are
not receiving haemodialysis).
3
A plateau in serologic responses has been observed beyond treatment year 4.
4
So far no TDF or TAF resistance development has been detected.
80 1 yr In patients with HBeAg-negative CHB, the 5-year cumulative

70 2 yr
70 67
3 yr probability of virological and biochemical responses on ETV
60 4 yr was 98% and 95%, respectively, while the rate of resistance to
5 yr
50 49
ETV was \1%.26,64–68 After 8 years, 99% of HBeAg-negative CHB
40 38 patients treated with TDF in the registration trial achieved viro-
30 24
29
logical response (HBV DNA \400 copies/ml) without evidence
20 18 17 of TDF resistance and 88% normalised ALT.69 During 3–4 years
11
10 4
1.2 of TDF treatment in HBeAg-negative CHB patients in real practice,
3 0.5
0 0 0 0 0 0 0 0
0 0.2 the virological response rates ranged from 92% to 100% without
LAM ADV TBV ETV TDF TAF the emergence of TDF resistance, while 75% of patients had nor-
Fig. 3. Cumulative incidence of HBV resistance for lamivudine (LAM), adefovir
malised ALT.70–73 No HBeAg-negative CHB patient cleared HBsAg
(ADV), entecavir (ETV), telbivudine (TBV), tenofovir (TDF) and tenofovir wihin the first year of ETV or TDF therapy and very few (1%)
alafenamide (TAF) in pivotal trials in nucleos(t)ide-naïve patients with achieved this endpoint during long-term (8 years) therapy.
chronic hepatitis B. (Collation of currently available data – not from head-to- In patients with HBeAg-positive CHB, the rates of virologic
head studies). No evidence of resistance has been shown after 8 years of TDF
response on TAF were 64% at week 48 and 75% at week 9674,75
treatment.69
HBeAg loss and anti-HBe seroconversion were achieved in 14%
and 10% of patients at week 48 and 22% and 18% at week 96,
respectively. In the same study, ALT normalisation rates at week
The efficacy of all NAs has been assessed in randomised con-
96 by traditional values were higher in patients treated with TAF
trolled phase III clinical trials (Table 3 and 4). The treatment
than TDF (75% vs. 68%), whereas only 1% of patients cleared
strategy for non-cirrhotic and compensated cirrhotic HBV
HBsAg.74–76 In HBeAg-negative CHB patients, TAF achieved viro-
patients is identical given the efficacy and long-term safety pro-
logic response in 94% of patients at week 48,76 which was main-
file in NA therapy. In HBeAg-positive CHB, 5 years of ETV achieve
tained in most cases at week 96 (90%). Only one TAF treated
a 99% cumulative probability of virologic response and 53% prob-
HBeAg-negative CHB patient (\1%) cleared HBsAg by week
ability of HBeAg loss.26 After 5 years of TDF treatment in patients
96.77 These virological and serological results are similar to those
with HBeAg-positive CHB, 97% of those on-treatment had viro-
observed in the TDF arms in both studies. Only 96-week results
logic response and 73% had normal ALT,34 while HBeAg loss
are available for TAF to date, with the studies still ongoing
was present in 49%, HBeAg seroconversion in 40%, HBsAg loss
(Table 3 and 4).
in 10% and HBsAg seroconversion in 8%.

Monitoring of patients treated with ETV, TDF or TAF creatinine clearance \60 ml/min or serum phosphate levels
Recommendations \2 mg/dl.
In the two registrational TAF trials, TAF compared to TDF
demonstrated superiority in the drug effects on several markers
All patients treated with NA should be followed with of renal (both glomerular and tubular) function and bone turnover
periodical assessments including ALT and serum HBV at weeks 48 and 96.74–77,88 In both groups of patients there was a
DNA (Evidence level I, grade of recommendation 1). significant difference in markers reflecting renal and bone func-
Patients at risk of renal disease treated with any NA and tion at week 48. A significant difference was noted in decrease
all patients regardless of renal risk treated with TDF of eGFR in both studies: 0.6 ml/min vs. 5.4 ml/min in HBeAg-
should undergo periodical renal monitoring including positive patients (p \0.0001), 1.8 ml/min vs. 4.8 ml/min in
at least estimated glomerular filtration rate (eGFR) and HBeAg-negative patients (p = 0.004). Similar mean serum crea-
serum phosphate levels (Evidence level II-2, grade of tinine changes were demonstrated between TAF and TDF treated
recommendation 1). subjects: HBeAg-positive TAF treated patient’s 0.01 mg/dl vs.
0.03 mg/dl in TDF (p = 0.02); HBeAg-negative 0.01 mg/dl vs.
Patients on TDF at risk of development and/or with 0.02 mg/dl (p = 0.32). Likewise, a significantly smaller percentage
underlying renal or bone disease should be considered decline in bone mineral density at the hip was reported in
for a switch to ETV or TAF, depending on previous TAF patients over TDF treated patients (0.10% vs. 1.72% in
LAM exposure (Evidence level II-2/I, grade of recom- HBeAg-positive patients [p \0.0001], and 0.29% vs. 2.16%
mendation 1). in HBeAg-negative [p \0.0001]) and spine (0.42% vs. 2.29% in
HBeAg- positive, 0.88% vs. 2.51% HBeAg-negative. Additional
data with biomarkers of renal tubular function and bone turnover
suggest less systemic effects in TAF compared to TDF, with less
All patients considered for treatment with a NA with high progression of chronic kidney disease and bone effects up to week
barrier to resistance (ETV, TDF, TAF) should undergo periodical 96.75,77 Long-term clinical data are lacking, however, similar find-
monitoring. At baseline, full blood count, liver and kidney (eGFR ings of superiority of TAF over TDF have also been found in recent
and serum phosphate levels) function tests, serum HBV DNA studies in HIV infected patients at risk for or with established
levels assessed by a sensitive PCR assay should be performed. renal and bone impairment.89–92
Appropriate dosing adjustments of ETV and TDF are recom- These co-infected data also demonstrate stabilisation in renal
mended for patients with eGFR \50 ml/min. TAF dosage parameters (GFR, creatinine) but improvement in proteinuria,
remains at 25 mg until eGFR is \15 ml/min, with modeling albuminuria and tubular proteinuria, (p \0.001) as well as
pharmacokinetics data suggesting no dose change between increases in hip and spine bone mineral densitometry from base-
\15 ml/min and formal renal support, although this is not line to week 48 (mean percent change +1.47 and +2.29, respec-
within the label (VemlidyÒ SmPC).78 In addition, the baseline tively, p \0.05)91; with similar findings to 96 weeks.92
renal risk should be assessed for all patients. High renal risk Whether these findings translate into improved long-term
includes one or more of the following factors: decompensated clinical outcomes in CHB patients remains to be defined, but an
cirrhosis, creatinine clearance (eGFR) \60 ml/min, poorly con- optimised safety profile of long-term NA therapy might be pre-
trolled hypertension, proteinuria, uncontrolled diabetes, active ferred, particularly in an ageing CHB population, with accruing
glomerulonephritis, concomitant nephrotoxic drugs, or solid co-morbidities. Thus, in CHB patients with deteriorating renal
organ transplantation. function or low eGFR and/or osteopenia/osteoporosis, particularly
During treatment, liver function tests should be performed in older age, the minimisation of progression of the physiological
every 3–4 months during the first year and every six months decline into pathological abnormality should also be considered
thereafter. Serum HBV DNA should be determined every 3– when choosing NA therapy (Table 5). In such subgroups of CHB
4 months during the first year and every 6–12 months thereafter. patients, both ETV and TAF represent suitable choices with TAF
HBsAg should be checked at 12-month intervals if HBV DNA having an advantage in patients with previous exposure to LAM.
remains undetectable, while patients who clear HBsAg should
be tested for anti-HBs. Long-term outcome during NA
Minimal rates of renal function decline have been reported Recommendations
during long-term therapy with ETV and TDF, but the nephro-
toxic potential is higher for TDF. Cases of Fanconi syndrome
associated with TDF therapy and rescued after a switch to ETV Patients under effective long-term NA therapy should
have been reported. In addition, studies using sensitive markers remain under surveillance for HCC (Evidence level II-2,
of glomerular and tubular kidney function and of bone mineral grade of recommendation 1).
density have also reported chronic tubular damage and decline of
eGFR and bone mineral density in TDF treated patients.70,79–87 HCC surveillance is mandatory for all patients with cir-
Therefore, it seems appropriate for now to monitor all CHB rhosis as well as those with moderate or high HCC risk
patients treated with TDF therapy for adverse renal effects with scores at the onset of NA therapy (Evidence level II-2,
serum creatinine (eGFR) and serum phosphate levels. Moreover, grade of recommendation 1).
CHB patients at high renal risk undergoing any NA therapy
should be monitored with serum creatinine (eGFR) levels. The
frequency of renal monitoring can be every 3 months during Long-term ETV or TDF monotherapy has been shown to halt
the first year and every 6 months thereafter, if no deterioration. progression of liver disease, and can also result in a significant
Closer renal monitoring is required in patients who develop improvement of histological necroinflammation and fibrosis,

Table 3. Results of main studies for the treatment of HBeAg-positive chronic hepatitis B at 6 months following 48 or 52 weeks of pegylated interferon alfa (PegIFNa)
and at 48 or 52 weeks of nucleos(t)ide analogue therapy.
PegIFN Nucleoside analogues Nucleotide analogues

PegIFNa2a PegIFNa2b LAM TBV ETV ADV TDF TAF
Dose* 180 mg 100 mg 100 mg 600 mg 0.5 mg 10 mg 245 mg 25 mg
Anti-HBe-seroconversion 32% 29% 16–18% 22% 21% 12–18% 21% 10%
HBV DNA \60–80 IU/ml 14% 7% 36–44% 60% 67% 13–21% 76% 64%
ALT normalisation# 41% 32% 41–72% 77% 68% 48–54% 68% 72%
HBsAg loss 3% 7% 0–1% 0.5% 2% 0% 3% 1%
References: see EASL CPG 20121 for all drugs except for TAF.76
PegIFNa, pegylated interferon alfa; ETV, entecavir; TDF, tenofovir disoproxil fumarate; TAF, tenofovir alafenamide; LAM, lamivudine; TBV, telbivudine; ADV, adefovir; ALT,
alanine aminotransferase.
*
PegIFNa were given as percutaneous injections once weekly and nucleos(t)ide analogues as oral tablets once daily.
#
The definition of ALT normalisation varied among different trials (i.e., decrease of ALT to 61.25-times the upper limit of normal (xULN) in the ETV or 61.3ULN in the TBV
trial). The lower quantification limit of HBV DNA assays was different across studies: \29 IU/ml for TAF studies.
Table 4. Results of main studies for the treatment of HBeAg-negative chronic hepatitis B at 6 months following 48 weeks of pegylated interferon alfa (PegIFNa) and
at 48 or 52 weeks of nucleos(t)ide analogue therapy.
PegIFN Nucleoside analogues Nucleotide analogues

PegIFNa2a LAM TBV ETV ADV TDF TAF
Dose* 180 mg 100 mg 600 mg 0.5 mg 10 mg 245 mg 25 mg
HBV DNA \60–80 IU/ml 19% 72–73% 88% 90% 51–63% 93% 94%
ALT normalisation# 59% 71–79% 74% 78% 72–77% 76% 83%
HBsAg loss 4% 0% 0% 0% 0% 0% 0%
References: EASL CPG 20121 for all drugs except for TAF.74
PegIFNa, pegylated interferon alfa; ETV, entecavir; TDF, tenofovir disoproxil fumarate; TAF, tenofovir alafenamide; LAM, lamivudine; TBV, telbivudine; ADV, adefovir; ALT,
alanine aminotransferase.
*
PegIFNa was given as percutaneous injections once weekly and nucleos(t)ide analogues as oral tablets once daily.
#
The definition of ALT normalisation varied among different trials (i.e., decrease of ALT to 61.25-times the upper limit of normal [ULN] in the ETV or 61.3-times the ULN in
the TBV trial). The lower quantification limit of HBV DNA assays was different across studies: for TAF studies it was \29 IU/ml.
often with a regression of established cirrhosis.1,34 Moreover, therapy.1,39,94–96 Loss of HBsAg during long-term NA therapy
complications of pre-existing decompensated cirrhosis, particu- may occur in a minority of CHB patients who were initially
larly at an early stage of decompensation, improve or even disap- HBeAg-positive (approximately 10–12% after 5–8 years of ther-
pear and the need for liver transplantation is dramatically apy) while is rare in patients with HBeAg-negative CHB (\1–2%
reduced.1 after 5–8 years of therapy).1,83
HCC may still develop and remains the major concern for CHB
patients treated with NAs.24,25 Long-term therapy with NAs
appears to favourably impact HCC incidence when data from ran- NA discontinuation
domised or matched controlled studies are considered.24,25 After Recommendations
the first 5 years of ETV or TDF therapy in CHB patients, recent
data suggest that the HCC incidence is decreasing further, with
the decrease being more evident in patients with baseline cirrho- NAs should be discontinued after confirmed HBsAg loss,
sis.93 In addition, HCC seems to be the only factor affecting long- with or without anti-HBs seroconversion (Evidence level
term survival in ETV or TDF treated CHB patients with or without II-2, grade of recommendation 1).
compensated cirrhosis.94 Since NAs are used in the majority of NAs can be discontinued in non-cirrhotic HBeAg-
CHB patients because of their favourable effects on the overall positive CHB patients who achieve stable HBeAg sero-
long-term outcome, the main clinical challenge is to identify conversion and undetectable HBV DNA and who com-
the patients at risk of HCC who require close surveillance. The plete at least 12 months of consolidation therapy.
Asian HCC risk scores, GAG-HCC, CU-HCC and REACH-B, have Close post-NA monitoring is warranted (Evidence level
been validated in treated Asian CHB patients,25 however the II-2, grade of recommendation 2).
PAGE-B score is the only one that offers good predictability for
HCC in Caucasian treated CHB patients.28 Based on the HCC risk Discontinuation of NAs in selected non-cirrhotic
scores, patients can be classified into those at low, medium and HBeAg-negative patients who have achieved long-
high risk of HCC. Patients in the low HCC risk group have no or term (P3 years) virological suppression under NA(s)
negligible probability of HCC development and therefore may may be considered if close post-NA monitoring can
not require HCC surveillance.25,28 be guaranteed (Evidence level II-2, grade of recom-
Despite the remaining risk of developing HCC, the overall mendation 2).
survival improves in patients under long-term effective NA(s)

Table 5. Indications for selecting ETV or TAF over TDF.* Management of patients with NA failure
1. Age [60 years
2. Bone disease Recommendations
Chronic steroid use or use of other medications that worsen bone
density
History of fragility fracture Prevention of resistance should rely on the use of first
Osteoporosis line therapy with high barrier to resistance NAs
3. Renal alteration** (Evidence level I, grade of recommendation 1).
eGFR \60 ml/min/1.73 m2
Albuminuria [30 mg/24 h or moderate dipstick proteinuria Compliance to NA therapy should be checked in all cases
Low phosphate (\2.5 mg/dl) of treatment failure (Evidence level II-1, grade of recom-
Hemodialysis mendation 1).
*
TAF should be preferred to ETV in patients with previous exposure to
Management of treatment failure should be based on
nucleoside analogues.
**
ETV dose needs to be adjusted if eGFR \50 ml/min; no dose adjustment of NAs cross-resistance data (Evidence level II-2, grade of
TAF is required in adults or adolescents (aged at least 12 years and of at least 35 recommendation 1).
kg body weight) with estimated creatinine clearance (CrCl) P15 ml/min or in
patients with CrCl \15 ml/min who are receiving haemodialysis. Treatment adaptation should be performed as soon as
virologic failure under NAs is confirmed (Evidence level
II-1, grade of recommendation 1).
Since NA therapy does not usually achieve HBV eradication

and rarely results even in HBsAg loss,83 long-term therapeutic Preventing the emergence of resistance is based on the use of
regimens are given in the majority of NA treated CHB patients. NAs with high barrier to resistance and maximal viral suppres-
A widely accepted stopping rule exists only for a proportion of sion as a first line therapy (Fig. 3). The combination of NAs with
patients with HBeAg-positive CHB who can discontinue NAs if low barrier to resistance, such as LAM or TBV with ADV, should
they achieve HBeAg seroconversion and HBV DNA undetectability be avoided, as this may lead to inappropriate viral suppression
and have completed 6 or preferentially 12 months of ensuing and the emergence of multidrug resistant strains. In addition,
consolidation therapy.1,56 According to the existing data, HBeAg sequential monotherapies with agents with a low barrier to resis-
seroconversion will remain in the majority (approximately 90%) tance should be strictly avoided because of the high risk of
and virological remission defined as HBV DNA \2,000– emerging multidrug resistance strains.99
20,000 IU/ml will be maintained in 50% of such patients at Managing NA failures in CHB patients remains a crucial issue
3 years after NAs cessation.41 Alternatively, clinicans may choose in countries in which ETV, TDF and TAF are not available or fully
to continue NA therapy until HBsAg clearance, which represents reimbursed for naïve patients or treatment experienced patients.
the safest current treatment endpoint.1 By contrast, in countries where NAs with a high barrier to resis-
Long-term, perhaps indefinite, NA therapy is usually given in tance have been routinely used for many years, the impact of
HBeAg-negative CHB patients, who are considered to be able to treatment failures has become minimal. Treatment failure can
safely stop NAs only if they achieve HBsAg loss.1 Recent evidence, be defined as primary non-response, partial virological response
accumulating mainly from Asian countries, in which NAs can be and virological breakthrough (see section ‘Definitions of
discontinued in HBeAg-negative CHB patients who achieve serum response’).100
HBV DNA undetectability on three separate occasions 6 months Primary non-response. In patients with primary non-response
apart,97 suggests that the discontinuation of NAs might be also to any NA, it is important to check for compliance. Poor compli-
feasible in this setting. An important factor affecting the probabil- ance is now the main cause of primary non-response. In a compli-
ity of off-NA virological remission appears to be the duration of ant patient with a primary non-response, genotyping of HBV
on-therapy HBV DNA undetectability.41 According to the existing strains for identifying possible resistance mutations may help
data, virological remission defined as HBV DNA \2,000– in formulating a rescue strategy. Primary non-response is almost
20,000 IU/ml will be maintained in approximately 50% of such exclusively seen with ADV because of suboptimal antiviral
patients 3 years after NAs cessation if they have remained for potency and should lead to a rapid switch to TDF or ETV.
more than two years on virological remission during therapy.41 Partial virological response. Partial virological response may be
Since such findings are based on studies with durations of on- encountered with all available NAs. It is always important to
therapy virological remission of [2 to 5 years,41 the optimal dura- check for compliance. If patients receive NA with low barrier to
tion of on NAs remission before discontinuation remains unclear. resistance (LAM, ADV, TBV), it is recommended to change to a
Since overt hepatitis flares and life-threatening episodes have more potent drug without cross-resistance. Most of the time, par-
been rarely reported in patients with pre-existing cirrhosis who tial virological response under ETV or TDF is associated with a
discontinue NAs,98 treatment discontinuation is currently dis- very high pretreatment viral load and not the result of a lack of
couraged in patients with cirrhosis. Moreover, NAs may be discon- efficacy but rather of the potency limit of the antiviral drug. In
tinued only in patients who can be followed closely with ALT and such patients with a partial virological response at week 48, the
HBV DNA determinations at least during the first year following HBV DNA levels at week 48 and their kinetics must be taken into
NAs cessation. Unfortunately, no reliable predictor of post-NAs account. Patients with declining serum HBV DNA levels may con-
remission has been identified to date. Retreatment criteria are also tinue treatment with the same agent given the rise in rates of
important, but have yet to be determined.41 Based on reasonable virological response over time and the very low risk of resistance
clinical judgment, treatment indications for naïve CHB patients with long-term monotherapy with both agents. It can be assumed
may be also applied in patients who discontinue NAs. that the same applies to TAF. In those with plateauing levels of

HBV DNA a switch to the other drug or a combination of ETV HBeAg negativity was sustained 3 years post-treatment in 81%.
+ TDF/TAF can be envisaged especially in patients with advanced Rates of HBsAg loss following 12 months of treatment are 3–7%.
liver disease (see section ‘Long-term outcome during NA’). HBsAg loss rates increase after the end of PegIFNa therapy in ini-
Virological breakthrough. Virological breakthrough in compli- tially HBeAg-positive CHB patients with sustained virological
ant patients is mainly related to the development of HBV drug responses (Table 3). Of the patients with an initial HBeAg loss,
resistance. The rate of virological breakthrough depends on the 30% experienced HBsAg loss after 3 years of follow-up. The sus-
barrier to resistance of the NA. Treatment adaptation should be tainability of HBsAg loss and seroconversion after PegIFNa is
performed as soon as viral breakthrough is identified and con- good although HBsAg seroreversions have been described.62,106
firmed one month apart to prevent a further increase in viral load, In HBeAg-negative CHB patients, the 48-week PegIFNa regis-
subsequent ALT elevation and progression of liver disease includ- trational trial showed sustained biochemical and virological
ing the risk of liver failure.99,100 response rates of 60% and 44% at 6 months and of 31% and 28%
at 3 years after the end of therapy1,107 (Table 4). PegIFNa was less
Management of antiviral drug resistance effective in HBeAg-negative patients with genotype D or E, who
Although a concern, antiviral drug resistance has become a man- had sustained virological responses in the range of 20%. Few real
ageable issue. The risk of resistance is associated with high base- life studies have addressed the efficacy of PegIFNa in HBeAg-
line HBV DNA levels, a slow decline in HBV DNA and a previous negative CHB patients with genotype B or C. In a retrospective
suboptimal NA treatment. Resistance should be identified by Korean study, approximately 30% of genotype C HBeAg-
HBV DNA monitoring and ideally identification of the pattern of negative patients achieved virological response at 1 year after
resistance mutations should be used to adapt treatment strategy. the end of PegIFNa.1,108
In case of resistance, an appropriate rescue therapy should be HBsAg loss rarely occurred during PegIFNa therapy in HBeAg-
initiated with the most effective antiviral agent that does not negative CHB patients, but the rate of HBsAg loss progressively
share cross-resistance to minimise the risk of inducing multiple increased after PegIFNa discontinuation, from 3% at month 6 to
drug-resistant strains. Table 6 shows cross-resistance data for 9% at year 3 to 12% at year 5 in the registrational trial1 (Table 4).
the most frequent resistant variants. Table 7 shows the recom- Similar rates were confirmed by real life studies.1,107 Overall,
mendations for treatment adaptation.99,100 In patients with mul- among sustained responders, approximately 30% clear HBsAg in
tidrug resistance, genotypic resistance testing should be the long-term.
performed by a reference laboratory. Combination of TDF with Two studies assessed the safety and efficacy of extending the
ETV has been evaluated in several clinical studies and appears duration of PegIFNa therapy beyond 48 weeks in HBeAg-negative
to be a safe option as a rescue therapy.101–104 CHB patients. In a European randomised trial with predominantly
genotype D patients, 96 compared to 48 weeks of PegIFNa ther-
PegIFNa monotherapy for CHB patients apy achieved higher rates of sustained virological response (29%
vs. 12%, p = 0.03) and HBsAg loss (6% vs. 0%).109 Similarly, a Chi-
Efficacy nese study including HBeAg-negative CHB patients with geno-
Recommendations type B or C showed that 72 compared to 48 weeks of PegIFNa
resulted in higher rates of sustained virological response (50%
vs. 16%, p = 0.001) and HBsAg loss (36% vs. 10%, p \0.05).110
PegIFNa can be considered as an initial treatment
option for patients with mild to moderate HBeAg- Monitoring of patients treated with PegIFNa
positive or -negative CHB (Evidence level I, grade of Recommendations
recommendation 2).
The standard duration of PegIFNa therapy is 48 weeks All CHB patients treated with PegIFNa should be fol-
(Evidence level I, grade of recommendation 1). lowed with periodical assessments of at least full blood
The extension of the duration of PegIFNa therapy count, ALT, TSH, serum HBV DNA and HBsAg levels (Evi-
beyond week 48 may be beneficial in selected HBeAg- dence level I/II-2, grade of recommendation 1).
negative CHB patients (Evidence level II-1, grade of rec- HBeAg-positive CHB patients treated with PegIFNa
ommendation 2). should be also followed with periodical assessments of
HBeAg and anti-HBe (Evidence level I, grade of recom-
mendation 1).
Only patients with mild to moderate CHB and perhaps
CHB patients with virological response after PegIFNa
selected patients with compensated cirrhosis but no portal
therapy should remain under long-term follow-up
hypertension should be considered for PegIFNa therapy (Table 2).
because of the risk of relapse (Evidence level II-2, grade
In HBeAg-positive CHB patients, response rates at 6 months fol-
of recommendation 1).
lowing 12 months of PegIFNa therapy are 20–30% (Table 3).
Although most patients respond with HBeAg loss or seroconver-
sion during the first 6 months of therapy, a 6 month course of
PegIFNa and/or a lower dose are inferior to the recommended In patients treated with PegIFNa, full blood counts and serum
12 month course. A combined endpoint of HBeAg loss with HBV ALT levels should be monitored monthly and TSH should be
DNA \2,000 IU/ml at 6 months post-treatment was achieved in monitored every 3 months.1 All patients should be monitored for
23% in a meta-analysis of three large trials105 (Table 3). Among safety through 12 months of treatment. Serum HBV DNA and
patients who achieved HBeAg loss at 6 months post-treatment, HBsAg levels in all CHB patients and HBeAg and anti-HBe in

HBeAg-positive CHB patients should be checked at 3, 6 and on liver biopsy. HBV genotypes A and B have been shown to be
12 months of PegIFNa treatment and at 6 and 12 months post- associated with higher rates of HBeAg seroconversion and HBsAg
treatment. Sustained serum HBV DNA\2,000 IU/ml, or even better loss than genotypes C and D.
HBsAg loss, together with ALT normalisation in all CHB patients as In HBeAg-negative CHB patients, high baseline ALT, low base-
well as with HBeAg seroconversion in HBeAg-positive CHB line HBV DNA, younger age, female gender and HBV genotype
patients are the desired treatment endpoints. Sustained virological were independent predictors of response to PegIFNa therapy
response after PegIFNa is usually associated with remission of the but the negative and positive values of these variables are low.
liver disease, but all such patients require long-term follow-up Patients with genotypes B or C had a better chance of response
because of the risk of exacerbation with development of HBeAg- than genotype D patients.1 Using pooled data from several stud-
negative CHB or even of HBeAg seroreversion in initially HBeAg- ies of PegIFNa therapy in HBeAg-negative CHB patients, a base-
positive patients. The risk of HBV reactivation seems to diminish line score system (ranging from 0–7) that combined five
over time. In patients with undetectable HBV DNA (and negative variables (HBV genotype, HBV DNA, ALT, HBsAg levels and age)
HBeAg), HBsAg should be checked at 12 month intervals, as the identified patients with high and low likelihood of response,
rate of HBsAg loss increases over time. Patients who become but this score has not been validated yet.111,112 Baseline host
HBsAg-negative should be tested for anti-HBs. genetic testing to prioritize CHB patients for PegIFNa therapy is
not currently recommended in clinical practice, as the initial
Predictors of PegIFNa response and stopping rules promising results were not confirmed in subsequent
Recommendations studies.113,114
During treatment. The most important on-treatment predictor
of response to PegIFNa is serum HBsAg levels,105 although they
In HBeAg-positive CHB patients, HBsAg levels are influenced by HBV genotype.115 In HBeAg-positive CHB
[20,000 IU/ml for genotype B and C, or no decline of patients, a decline of HBsAg levels below 1,500 IU/ml at 12 weeks
HBsAg levels for genotype A and D, at 12 weeks of is a reasonable predictor of HBeAg seroconversion (positive pre-
PegIFNa therapy are associated with a very low proba- dictive value: 50%), while HBsAg levels [20,000 IU/ml for HBV
bility of subsequent HBeAg seroconversion and can be genotype B and C or no decline of HBsAg levels for HBV genotype
used as PegIFNa stopping rules (Evidence level II-2, A and D are associated with a very low probability of subsequent
grade of recommendation 2). HBeAg seroconversion105 (Fig. 4). At week 24 HBsAg levels
[20,000 IU/ml predict no response regardless of genotype
In HBeAg-positive CHB patients with genotype A-D,
(Fig. 4). A substantial HBV DNA decrease at 12 weeks has been
HBsAg levels [20,000 IU/ml at 24 weeks of PegIFNa
associated with a 50% chance of HBeAg seroconversion. Also,
therapy are associated with a very low probability of
HBeAg levels and immunologically induced ALT flares followed
subsequent HBeAg seroconversion and can be used as
by a HBV DNA decrease are associated with more frequent HBeAg
PegIFNa stopping rules (Evidence level II-2, grade of rec-
seroconversion. However, clinically meaningful cut-offs for HBV
ommendation 2).
DNA and HBeAg levels as a tool for on-treatment response predic-
In HBeAg-negative CHB patients with genotype D, a tion have not been reported, these are based on validated studies
combination of no decrease in HBsAg levels and using adequate outcomes.116
\2 log10 IU/ml reduction in serum HBV DNA levels at In HBeAg-negative CHB patients, a combination of a lack of
12 weeks of PegIFNa therapy predicts no response and decrease in HBsAg levels and \2 log10 IU/ml decline in HBV
should be used as PegIFNa stopping rules (Evidence DNA at 12 weeks of PegIFNa predicts a no response in genotype
level II-2, grade of recommendation 1). D patients (negative predictive value: 100%) (Fig. 4). This stop-
ping rule would allow approximately 20% of patients to discon-
tinue PegIFNa.1,117,118 No robust on-treatment stopping rules
Pretreatment. In HBeAg-positive CHB patients, pretreatment have been developed for HBeAg-negative CHB patients with
predictors of response are low viral load, high serum ALT levels genotype B or C and very few data are available for those
(above 2–5 times ULN), HBV genotype and high activity scores with genotype A and E.118 Some studies have also looked for
Table 6. Cross-resistance data for the most frequent resistant HBV variants.
HBV variant LAM LDT ETV ADV TDF/TAF*

Wild-type S S S S S
M204V R S I I S
M204I R R I I S
L180M + M204V R R I I S
A181T/V I I S R I
N236T S S S R I
L180M + M204V/I ± I169T ± V173L ± M250V R R R S S
L180M + M204V/I ± T184G ± S202I/G R R R S S
The amino acid substitution profiles are shown in the left column and the level of susceptibility is given for each drug: S (sensitive), I (intermediate/reduced susceptibility),
R (resistant).
ETV, entecavir; TDF, tenofovir disoproxil fumarate; TAF, tenofovir alafenamide; LAM, lamivudine; ADV, adefovir.
*
In vitro data for tenofovir, in vivo data for TDF, no clinical data for TAF.

Table 7. Management of patients who develop NA resistance. cirrhosis.120 The benefit from PegIFNa therapy on the HCC
Resistance pattern Recommended rescue strategies incidence seems to be more clear in Asian patients120 and per-
LAM resistance Switch to TDF or TAF haps superior than that of NA therapy.121 In addition, old cohort
TBV resistance Switch to TDF or TAF studies with standard IFNa and systematic reviews show that the
ETV resistance Switch to TDF or TAF incidence of HCC is decreased in IFNa treated compared to
ADV resistance If LAM-naïve: switch to ETV or TDF or TAF untreated CHB patients, with such an effect being clearer in Asian
If LAM-resistance: switch to TDF or TAF patients and those with sustained off-treatment responses and/or
If HBV DNA plateaus: add ETV*** or switch to ETV compensated cirrhosis.1,24 Cohort studies in both HBeAg-positive
TDF or TAF resistance** If LAM-naïve: switch to ETV and HBeAg-negative CHB have shown that courses of standard
If LAM-R: add ETV*
IFNa treatment result in improved overall long-term outcomes
Multidrug resistance Switch to ETV plus TDF or TAF combination
including survival in patients with sustained off-treatment
ETV, entecavir; TDF, tenofovir disoproxil fumarate; TAF, tenofovir alafenamide;
responses.1 Survival data are not available for PegIFNa therapy,
LAM, lamivudine; ADV, adefovir; TBV, telbivudine.
*
The long-term safety of these combinations is unknown. but the same favourable outcomes are expected if sustained
**
Not seen clinically so far; do genotyping and phenotyping in an expert off-treatment responses are achieved. Rates of HBsAg loss in sus-
laboratory to determine the cross-resistance profile. tained responders are gradually increasing approaching 50% at
***
Especially in patients with ADV resistant mutations (rA181T/V and/or 5 years after the end of therapy.119
rN236T) and high viral load, the response to TDF (TAF) can be protracted.
Combination therapy for CHB
NA plus NA
on-treatment positive predictors of sustained response. For Recommendations
HBeAg-negative CHB patients with non-D genotype, a P10%
decline in serum HBsAg from baseline to week 12 of PegIFNa
treatment had a higher probability of achieving a sustained De novo combination therapy with two NAs with high
response than those with a \10% decline (47% vs. 16%, p \0.01) barrier to resistance (ETV, TDF, TAF) is not recom-
but the positive predictive value was low (50%).107 mended (Evidence level I, grade of recommendation 1).
In treatment-adherent patients with incomplete sup-
Safety of PegIFNa
pression of HBV replication reaching a plateau during
PegIFNa therapy is associated with considerable side effects
either ETV or TDF/TAF long-term therapy, a switch to
although patients with HBV infection seem to tolerate it reason-
the other drug or combining both drugs may be consid-
ably well because they are often younger and have less comor-
ered (Evidence level III, grade of recommendation 2).
bidity than patients who were treated with this agent in the
setting of HCV infection.1 The most frequently reported side
effects are flu-like syndrome, myalgia, headache, fatigue, weight
loss, depression, hair loss and local reactions at the site of injec- There have been only a few studies evaluating the role of de
tion. Hepatitis flares may occur which can result in decompensa- novo combination therapy with potent NAs in treatment naïve
tion of liver disease, therefore PegIFNa is contraindicated in chronic HBV infection. In a large prospective multicentre study,
patients with decompensated cirrhosis. PegIFNa treatment is also HBeAg-positive and –negative CHB patients were randomised
associated with mild myelosuppression, but neutropenia and to either ETV or ETV plus TDF.122 The primary endpoint (HBV
thrombocytopenia are usually well managed with dose- DNA \50 IU/ml at week 96) was reached in 76% and 83% of
reduction and only rarely result in clinically significant infection patients treated with mono- or combination therapy, respectively
or bleeding. The combination of PegIFNa with telbivudine is con- (p = 0.088). In the subgroup of HBeAg-positive patients, ETV/TDF
traindicated due to a high risk of neuropathy. combination achieved significantly higher rates of HBV DNA
\50 IU/ml (80% vs. 70%, p = 0.046), which was entirely attributa-
Long-term outcome after PegIFNa ble to the HBeAg-positive subgroup with baseline HBV DNA
Recommendation levels P108 IU/ml (79% vs. 62%). However, no difference was
found in the rate of HBeAg seroconversions. None of the patients
developed resistance, whereas ALT normalisation was observed
Patients with sustained responses after PegIFNa therapy more frequently in the ETV monotherapy group (82% vs. 69%).
and high baseline HCC risk should remain under surveil- This combination did not provide added value in terms of HBsAg
lance for HCC even if they achieve HBsAg loss (Evidence kinetics.123
level III, grade of recommendation 1). In a second double-blind study, HBeAg-positive treatment
naïve patients with high HBV DNA and normal ALT levels were
randomly assigned to either TDF plus placebo or a combination
of TDF plus emtricitabine for 192 weeks.124 At week 192, 55%
The majority of patients who achieve sustained off-treatment
and 76% of patients in the TDF monotherapy and the combination
responses after IFNa or PegIFNa therapy maintains such
group respectively reached the primary endpoint, which was HBV
responses during long-term follow-up of at least 5 years.119 In
DNA \69 IU/ml (p = 0.016). Of those who did not meet the pri-
sustained responders, there is no progression of liver disease
mary endpoint, the majority had low levels of ongoing HBV repli-
and baseline liver histological lesions improve.1 HCC may still
cation, with serum HBV DNA \500 IU/ml. However, HBeAg
develop after PegIFNa therapy, even in patients with sustained
seroconversion occurred in only 5% of patients (all in the
off-treatment responses, particularly in those with pre-existing
monotherapy group), while no patient developed HBV resistance.

Although both of the above studies showed a higher percent- DNA response also showed that adding TDF is superior with
age of complete HBV DNA suppression with NA combination respect to viral suppression and ALT normalisation as compared
therapy in HBeAg-positive patients with high baseline viral load with continuing ETV monotherapy with either 0.5 mg or the
(HBV DNA [108 IU/ml), the differences in terms of on- higher 1.0 mg daily dose.128,129 Switching to another potent NA
treatment HBV DNA levels and clinical/serological endpoints (i.e., from ETV to TDF/TAF or vice versa) may also sometimes lead
observed with both strategies are not strong enough to recom- to an improved response.
mend de novo combination for this group of patients.
The optimal management of patients with incomplete sup- NA plus PegIFNa
pression of HBV replication during long-term treatment with Recommendations
potent NAs (ETV, TDF or TAF) is still a matter of debate. In most
of these patients a continuous decline in HBV DNA levels can
be observed when continuing the same agent. This approach De novo combination of NA and PegIFNa is not recom-
has been shown to be safe, effective and has not been associated mended (Evidence level I, grade of recommendation 1).
with the development of drug resistance in large prospective
In treatment naïve HBeAg-positive patients, short-term
long-term studies.1,125 So far, there are also no convincing data
pretreatment with a NA before PegIFNa is not recom-
demonstrating that the presence of a minimal residual viremia
mended (Evidence level II, grade of recommendation 1).
with HBV DNA levels \69 IU/ml may have any unfavourable
effects with respect to on-treatment disease progression or HCC In long-term NA suppressed CHB patients, adding
risk in patients without cirrhosis.126 We therefore do not recom- PegIFNa or switching to PegIFNa is not recommended
mend changing the initial treatment strategy in patients with low (Evidence level II, grade of recommendation 1).
level and or declining HBV DNA concentrations on a potent NA
monotherapy.
However, in patients with decompensated cirrhosis, not The combination of NA and PegIFNa has been used in treat-
achieving a virologic response defined as HBV DNA \20 IU/ml ment naïve and NA suppressed CHB patients. For treatment naïve
has been shown to be a significant risk factor for developing patients, there is no robust evidence that a de novo combination
HCC (HR = 7.74; 95% CI 1.34–44.78; p = 0.022) but not in those of PegIFNa and NA is superior compared to PegIFNa or NA alone.
with compensated cirrhosis (p = 0.749).127 Previous studies with LAM and or ADV combined with PegIFNa
The long-term consequences of on-treatment HBV DNA levels failed to show an advantage of the combination therapy.1,130 In
plateauing above 69 IU/ml but being below 2,000 IU/ml are a recent randomised controlled trial, the 72-week HBsAg loss
unclear. As the above mentioned studies showed some advan- rates were superior in the PegIFNa and TDF treated patients com-
tages of combining NAs in terms of HBV DNA suppression this pared to those observed in patients receiving PegIFNa alone or
approach can be considered, and is especially recommended for TDF alone (9% vs. 3% vs. 0%) but the overall rates were low and
those with established cirrhosis. A recent retrospective study in mainly confined to genotype A patients.62,131 For treatment naïve
ETV (0.5 mg daily dose) treated patients with incomplete HBV patients, there is also no robust evidence that short-term
HBeAg-positive CHB
Genotype A B C D
WEEK 12 Stop if HBsAg No decline >20,000 >20,000 No decline
WEEK 24 Stop if HBsAg >20,000 >20,000 >20,000 >20,000
HBeAg-negative CHB
(genotype D)
HBsAg levels Any decline No decline
WEEK 12
HBV DNA levels >2 log decline <2 log decline
Continue Continue Stop
Fig. 4. Week 12 and 24 stopping rules for HBeAg-positive and -negative patients treated with PegIFNa. These rules are based upon viral genotype, HBsAg and HBV
levels.

pretreatment with NA improves the rates of sustained response liver disease) once daily. Less potent NAs are not recommended
to PegIFNa.130,132,133 The multicentre ARES study demonstrated as they have been demonstrated to have inferior outcomes as
that a 24-week course of PegIFNa, given to a small group of compared to potent ones.141,146 Despite an overall high safety
HBeAg-positive patients who were on ETV therapy for only profile, concerns remain that development of lactic acidosis in
24 weeks, can improve the kinetics of HBeAg, HBV DNA and patients with decompensated cirrhosis may be a class effect of
HBsAg, compared to those on ETV monotherapy, but a PegIFNa NAs, and close monitoring for adverse events is especially recom-
monotherapy arm was missing.134 mended for patients with a MELD score [22 and impaired kidney
In CHB patients under long-term effective virological remis- function.1,147 All NAs must be adjusted to renal function. Because
sion by NA treatment, PegIFNa can be used as a ‘switch to’ or of its favourable safety profile, TAF might be also an interesting
‘add-on’ strategy. In HBeAg-positive CHB, two recent Chinese treatment option in patients with decompensated disease, espe-
studies assessed the efficacy and safety of switching to PegIFNa cially in those with kidney dysfunction. However, studies con-
for patients on long-term effective NA therapy. Following a 48- cerning the safety and efficacy of TAF in these patient
week course of PegIFNa, 6–20% of patients cleared HBsAg. Base- populations are lacking. PegIFNa is contraindicated in patients
line HBsAg \1,500 IU/ml predicted the serological with decompensated liver disease.
responses.135,136 Two additional Asian studies assessed whether The main goal of NA treatment in patients with decompen-
a 48-week course of add-on PegIFNa was superior to long-term sated liver disease is to achieve clinical recompensation and to
NA therapy in HBeAg-positive patients.137,138 While the HBsAg avoid liver transplantation.1,57 There is strong evidence that
decline was superior in the combination group, the HBsAg loss antiviral therapy significantly modifies the natural history of
rates did not increase significantly. decompensated cirrhosis, improving liver function and increasing
In HBeAg-negative CHB patients under NA treatment, two survival.57,58,148 Meta-analyses demonstrated an overall and
multicentre European studies have assessed the safety and effi- transplant-free survival in NA treated patients of more than
cacy of a 48-week add-on course of PegIFNa.139,140 These two 80%, respectively.58,59 Approximately 35% of treated patients
studies demonstrated that HBsAg kinetics were fostered by the can be delisted for liver transplantation, and an improvement
addition of PegIFNa, but only a few patients cleared HBsAg. of Child-Pugh Score P2 observed in at least 40–50%. Patients
HBsAg levels at baseline and week 12 may predict HBsAg decline with early treatment initiation had better clinical outcomes than
and/or HBsAg loss. There are no studies assessing the safety and those with delayed treatment.148 High baseline Child-Pugh or
efficacy of switching to PegIFNa monotherapy in HBeAg-negative MELD scores are predictors of poor survival meaning that the dis-
CHB patients under long-term NA therapy. ease may have progressed beyond the point of no return.58,148–150
As all these studies of PegIFNa therapy in patients under long- In contrast, an improvement in MELD or Child-Pugh score early
term NA therapy increase cost and side effects, this strategy on-treatment is highly predictive of transplant-free sur-
should be carefully assessed in each individual patient weighing vival.1,148,149 Undetectable HBV DNA levels can be achieved in
up all potential advantages and disadvantages. [80% after 1 year of treatment, and are associated with a lower
risk of HCC development.58,127,148 Lifelong treatment is recom-
Treatment of patients with decompensated cirrhosis mended for all patients with decompensated disease. Even under
effective NA therapy, the risk of developing HCC is high in these
Recommendations patients, and therefore careful long-term HCC surveillance is
mandatory.1
Patients with decompensated cirrhosis should be imme- Prevention of HBV recurrence after liver transplantation
diately treated with a NA with high barrier to resistance,
irrespective of the level of HBV replication, and should Recommendations
be assessed for liver transplantation (Evidence level II-
1, grade of recommendation 1).
All patients on the transplant waiting list with HBV
PegIFNa is contraindicated in patients with decompen-
related liver disease should be treated with NA (Evi-
sated cirrhosis (Evidence level II-1, grade of recommen-
dence level II, grade of recommendation 1).
dation 1).
Combination of hepatitis B immunoglobulin (HBIG) and
Patients should be closely monitored for tolerability of
a potent NA is recommended after liver transplantation
the drugs and the development of rare side effects like
for the prevention of HBV recurrence (Evidence level II-
lactic acidosis or kidney dysfunction (Evidence level II-
Patients with a low risk of recurrence can discontinue
HBIG but need continued monoprophylaxis with a
Patients with decompensated cirrhosis should be referred for potent NA (Evidence level II-1, grade of recommenda-
liver transplantation and treated with NAs as early as possible, tion 2).
with the goal of achieving complete viral suppression in the HBsAg-negative patients receiving livers from donors
shortest time possible. ETV or TDF are the preferred treatment with evidence of past HBV infection (anti-HBc positive)
options and both drugs have been shown to be effective but also are at risk of HBV recurrence and should receive antivi-
generally safe in patients with decompensated disease.1,141–145 ral prophylaxis with a NA (Evidence level II-2, grade of
The licensed ETV dose for patients with HBV decompensated cir- recommendation 1).
rhosis is 1 mg (instead of 0.5 mg for patients with compensated

Before the advent of NA therapy, recurrent HBV infection in density parameters.159 Persons with liver cirrhosis and low CD4
liver transplantation was a major problem.151 All putative count require careful surveillance in the first months after start-
patients for liver transplantation should therefore be treated with ing ART in order not to overlook immune reconstitution syn-
NA therapy with the aim of achieving an undetectable HBV DNA drome and subsequent liver decompensation due to flares of
level.1 NA therapy in combination with HBIG reduces the risk of liver enzymes.158 Because TDF, TAF and possibly also ETV
graft infection to \5%.1,151 Potent NA therapy allows a more effi- monotherapy can cause HIV resistance mutations, all HBsAg-
cient combination treatment with synergistic effects and better positive patients should be screened for HIV before these drugs
tolerability with the aim of achieving anti-HBs levels of P50– are used in the treatment of HBV infection.
100 IU/L. In selected patients (i.e., HBV DNA negative at liver
transplantation), a short course or HBIG free regimens can be HDV co-infected patients
considered.151 In selected patients, ETV prophylaxis without Recommendations
HBIG has been shown to be safe and effective in preventing
HBV recurrence.152 Conversely, lifelong combination therapy
should be given to patients who are at a high risk for HBV recur- PegIFNa for at least 48 weeks is the current treatment of
rence, namely those who are HBV DNA positive at the time of choice in HDV-HBV co-infected patients with compen-
liver transplantation, who are HBeAg-positive, have HCC, and sated liver disease (Evidence level I, grade of recommen-
HDV or HIV co-infection.153–155 In the setting of liver transplan- dation 1).
tation, nephrotoxicity should always be considered and renal
In HDV-HBV co-infected patients with ongoing HBV
function should be carefully monitored because of the concomi-
DNA replication, NA therapy should be considered
tant use of calcineurin inhibitors.
When the immune system is therapeutically suppressed in
the context of liver transplantation, there is potential for HBV PegIFNa treatment can be continued until week 48 irre-
reactivation in HBsAg-negative patients receiving donor organs spective of on-treatment response pattern if well toler-
with evidence of past HBV infection (anti-HBc positive). These ated (Evidence level II-2, grade of recommendation 2).
patients usually receive receive lifelong LAM prophylaxis.156
Treatment in special patient groups with HBV infection At present, PegIFNa is the only available drug that has been
proven to have some antiviral efficacy against chronic HDV infec-
HIV co-infected patients tion.1,160 Studies applying PegIFNa showed on-treatment viro-
Recommendations logic response rates of about 17–47%.1 The rate of HDV RNA
negativity 24 weeks after treatment cessation was, however,
rather low (approximately 25%), and late relapses of HDV replica-
All HIV-positive patients with HBV co-infection should tion beyond week 24 after stopping therapy occurred in more
start antiretroviral therapy (ART) irrespective of CD4 cell than 50% of the responder patients, thus challenging the concept
count (Evidence level II-2, grade of recommendation 1). of sustained virologic response in HDV-HBV co-infection.161
Hence, long-term follow-up HDV RNA monitoring is recom-
HIV-HBV co-infected patients should be treated with a
mended for all treated patients as long as HBsAg is present in
TDF- or TAF-based ART regimen (Evidence level I for
serum. HBsAg loss may develop in the long-term follow-up in
TDF, II-1 for TAF, grade of recommendation 1).
approximately 10% of PegIFNa patients and can be taken as a
marker of cure from HDV infection.161,162
Several studies tried to increase efficacy by increasing treat-
European and American guidelines on the management of HIV ment duration.163,164 However, clear evidence is lacking to confirm
infected patients recommend the initiation of ART in HIV/HBV co- that this approach is beneficial for most chronically HDV infected
infected patients irrespective of CD4 cell count due to the patients. Even after 96 weeks of PegIFNa therapy, alone or in com-
increased risk of fibrosis progression, cirrhosis and HCC.157,158 bination with TDF, 24-week post-therapy relapses occurred in 36–
All persons with HIV/HBV co-infection should receive ART includ- 39% of the patients with on-treatment response.165
ing either TDF or TAF, which have antiviral activity against HIV The likelihood of the long-term response to PegIFNa can be
and HBV. Stopping TDF- or TAF-containing ART should be avoided estimated to some extent by HDV RNA and HBsAg kinetics at
in persons with HIV/HBV co-infection because of the high risk of weeks 12 and 24.164,166–169 However, stopping PegIFNa prema-
severe hepatitis flares and decompensation following HBV reacti- turely at this stage is not recommended, if treatment is well tol-
vation hepatitis. Drug toxicity (renal, bone density, liver) should erated, as the negative predictive values of these markers are not
be closely monitored during ART. ETV represents an alternative very strong, and late responses may occur in patients with early
anti-HBV treatment without a strong activity against HIV.157 At non-response. Furthermore, long-term follow-up studies suggest
present, limited data exists on the use of TAF in HIV/HBV co- that an IFNa based therapy per se can be taken as an independent
infected patients. In 72 HIV/HBV co-infected patients with a factor associated with a lower likelihood of disease progression,
stable suppression of HIV and HBV DNA, switching ART from a and to develop clinical endpoints.162,170
TDF- to a TAF-containing regimen maintained HIV and HBV Neither NAs nor ribavirin showed significant effects on HDV
suppression in [90% of patients, with improved eGFR and bone RNA levels in patients with HDV infection.1 Although HDV is

often the predominant virus in this co-infection, considerable Out of the 327 patients treated with DAA therapy, 124 had
fluctuating activity of HDV and HBV or both viruses, including evidence of anti-HBc with none experiencing any clinical or viro-
alternating predominance can be seen during the natural history logical sequelae.179
of this chronic co-infection.1 NA treatment is recommended for
those patients with HBV DNA levels being persistently above Acute hepatitis B
2,000 IU/ml, and might be considered in order to block residual Recommendations
HBV replication in those with advanced liver disease. In patients
with decompensated liver disease, PegIFNa should not be used
and these patients should be evaluated for liver transplantation. More than 95% of adults with acute HBV hepatitis do not
NA should be considerd in all patients with decompensated dis- require specific treatment, because they will fully
ease if HBV DNA is detectable. recover spontaneously (Evidence level II-2, grade of
recommendation 1).
HCV co-infected patients Only patients with severe acute hepatitis B, charac-
Recommendations terised by coagulopathy or protracted course, should
be treated with NA and considered for liver transplanta-
tion (Evidence level II-2, grade of recommendation 1).
Treatment of HCV with direct-acting antivirals (DAAs)
may cause reactivation of HBV. Patients fulfilling the
standard criteria for HBV treatment should receive In patients with acute hepatitis B, preventing the risk of
NA treatment (Evidence level II, grade of recommenda- acute or subacute liver failure is the main treatment goal.
tion 1). Improving quality of life by shortening the disease associated
HBsAg-positive patients undergoing DAA therapy symptoms as well as lowering the risk of chronicity can be also
should be considered for concomitant NA prophylaxis regarded as relevant goals of treatment. As outlined in the nat-
until week 12 post DAA, and monitored closely ural course of disease, acute HBV infection will recover clinically
(Evidence level II-2, grade of recommendation 2). and virologically including seroconversion to anti-HBs without
antiviral therapy in more than 95% of adults. A potentially
HBsAg-negative, anti-HBc positive patients undergoing life-threatening disease manifestation is severe or fulminant
DAA should be monitored and tested for HBV reactiva- acute hepatitis B. Characteristics of severe acute hepatitis B
tion in case of ALT elevation (Evidence level II, grade are coagulopathy (most studies defined this as international
of recommendation 1). normalised ratio [INR] [1.5), or a protracted course (i.e., persis-
tent symptoms or marked jaundice for [4 weeks), or signs of
acute liver failure.107,180 Although randomised controlled trials
In patients with chronic HBV infection, HCV co-infection are lacking, several cohort studies indicate that the early antivi-
accelerates liver disease progression and increases the risk of ral therapy with highly potent NAs can prevent progression to
HCC.1,171,172 Therefore, all chronic HBV patients should be acute liver failure and subsequently liver transplantation or
screened for HCV as well as for other blood bourne mortality.107,181 This effect, however, is not seen if antiviral
viruses.1,171,173 therapy is initiated late in the course of severe acute hepatitis
With the advent of effective DAA therapy, uptake of antiviral B in patients with already manifested acute liver failure and
therapy for HCV is increasing rapidly. Sustained virological advanced hepatic encephalopathy.182 Data supports the use of
response rates for HCV in HBV and HCV co-infected patients are TDF, ETV or even LAM. One single-center retrospective analysis
comparable with those in HCV mono-infected patients.173 There from Hong Kong reported a higher short-term mortality in
is a potential risk of HBV reactivation during DAAs therapy or patients with acute exacerbation of chronic hepatitis B (not pri-
after clearance of HCV. It must be noted that most patients with mary hepatitis B infection) upon treatment with ETV compared
HCV/HBV co-infection, and advanced disease should be on effec- to historic controls treated with LAM.183 Large case series, how-
tive NA therapy. Following a series of case reports, the US Food ever, support that TDF, ETV or LAM can be safely used in acute
and Drug Administration has now issued a warning about the risk severe hepatitis B.181,184 In principle, TAF should be also effec-
of HBV reactivating in some patients treated with DAAs for HCV. tive in this setting, but no data are currently available on the
They identified 24 cases of HBV reactivation in HCV/HBV co- use of TAF in severe acute hepatitis B. The use of glucocorticoids
infected patients treated with DAAs during a 31-month period in acute severe hepatitis B is supported by older studies, but
from November 2013 to July 2016.174–177 More recent data con- these studies in most cases did not include current antiviral
firms the risk of HBV reactivation associated with DAA therapy. In drugs.185 The management of acute liver failure and the indica-
one publication, out of 103 patients with evidence of previous tion for liver transplantation are discussed in detail in separate
HBV exposure (HBsAg-negative, anti-HBc positive) undergoing EASL CPGs.151,180 Early NA treatment does not increase the risk
DAA therapy, none experienced reactivation.178 In another publi- of chronicity181,186; in fact, observational data from a multicen-
cation, 3 out of 10 patients with HBsAg-positive status demon- tre cohort even indicated reduced rates of chronicity, if NA
strated significant reactivation with two significant clinical treatment was initiated within 8 weeks of acute hepatitis B pre-
events. sentation in genotype A infected individuals.187

Children persons from the practice or study of surgery, dentistry, medi-
Recommendations cine, or allied health fields.193 However, percutaneous injuries
sustained by healthcare personnel during certain surgical, obstet-
rical, and dental procedures provide a potential route of HBV
In children, the course of the disease is generally mild, transmission to patients as well as providers.194 Thus, healthcare
and most of the children do not meet standard workers may require antiviral therapy, even if they do not fulfill
treatment indications. Thus, treatment should be the typical indications for treatment, to reduce direct transmis-
considered with caution (Evidence level II-3, grade of sion during exposure prone procedures to patients. Policies for
recommendation 1). HBsAg-positive healthcare workers vary among countries. No
prospective clinical trials are available to demonstrate the effi-
In children or adolescents who meet treatment criteria,
cacy of antiviral treatment for preventing transmission by health-
ETV, TDF, TAF, and PegIFNa can be used in this popula-
care workers, but no HBV transmissions from healthcare workers
tion (Evidence level II-2, grade of recommendation 2).
to patients have been reported, if the healthcare worker has
serum HBV DNA levels below 200 IU/ml. Thus, healthcare work-
ers, including surgeons, gynaecologists and dentists, who are
Chronic HBV infection runs an asymptomatic course in most HBsAg-positive with HBV DNA [200 IU/ml may be treated with
children, but the lifetime risk of significant clinical complications a potent NA (i.e., ETV, TDF, TAF) to reduce levels of HBV DNA ide-
is not negligible. Since the World Health Organization recom- ally to undetectable or at least to \200 IU/ml (CDC recommenda-
mended global HBV vaccination, the incidence of HBV in children tion: \1,000 IU/ml; recommendation in many countries:
has declined worldwide.188 Issues around the characterisation of \2,000 IU/ml) before resuming exposure prone procedures.194
the phase of chronic HBV infection, and the indications for treat- Monitoring for compliance and efficacy in practicing surgeons
ment, exist in the paediatric population as well, compounded by is required. Healthcare workers performing exposure prone pro-
an enhanced requirement for safety, and thus extrapolation from cedures that are not on antiviral treatment might be more fre-
adult strategies may be unhelpful. Treatment indications should quently retested, especially if they are tested around the HBV
be carefully evaluated,1 and other co-morbidities, such as non- DNA threshold, due to fluctuations in viremia.195 The long-term
alcoholic fatty liver disease considered. Overall, a conservative safety, efficacy, complications and economic implications of such
approach is warranted. A joint EASL-ESPHAGN review provides a policy are unknown.
a detailed relevant review.189,190
Conventional IFNa, LAM, ADV, ETV and TDF have been evalu-
ated for safety and efficacy in children, which were comparable to Pregnancy
adults.1,190 In a study including adolescents 12 to \18 years old Recommendations
with HBeAg-positive and negative chronic HBV hepatitis,
72 weeks of TDF compared to placebo achieved significantly
higher rates of virological response (HBV DNA \400 copies/ml: Screening for HBsAg in the first trimester of pregnancy
89% vs. 0%, p \0.001) and ALT normalisation (74% vs. 31%, is strongly recommended (Evidence level 1, grade of
p \0.001), but similarly low HBeAg clearance rates. TDF proved recommendation 1).
safe and no patient developed resistance. Another recent
In a woman of childbearing age without advanced
study with ETV in adolescents confirmed no improvement in
fibrosis who plans a pregnancy in the near future, it
HBeAg seroconversion rates in cases with normal ALT
may be prudent to delay therapy until the child is born
(\30 IU/L).188,189,191 ETV has also been studied in children
between the ages of 2–12.192
Pregnant women with CHB and advanced fibrosis or cir-
Healthcare workers rhosis, therapy with TDF is recommended (Evidence
Recommendations level II-2, grade of recommendation 1).
In pregnant women already on NA therapy, TDF should
be continued while ETV or other NA should be switched
HBV infection alone should not disqualify infected per-
to TDF (Evidence level II-2, grade of recommendation 1).
sons from the practice or study of surgery, dentistry,
medicine, or allied health fields (Evidence level III, grade In all pregnant women with high HBV DNA levels
of recommendation 1). ([200,000 IU/ml) or HBsAg levels [4 log10 IU/ml,
antiviral prophylaxis with TDF should start at week
Healthcare workers performing exposure prone proce-
24–28 of gestation and continue for up to 12 weeks after
dures with serum HBV DNA [200 IU/ml may be treated
delivery (Evidence level 1, grade of recommendation 1).
with NA to reduce transmission risk (Evidence level II-2,
grade of recommendation 2). Breast feeding is not contraindicated in HBsAg-positive
untreated women or on TDF-based treatment or
prophylaxis (Evidence level III, grade of recommenda-
tion 2).
According to the recommendations of Center for Disease Con-
trol (CDC), HBV infection alone should not disqualify infected

Family planning should always be discussed with women of Patients undergoing immunosuppressive therapy or chemotherapy
childbearing age before initiating HBV therapy. The woman Recommendations
should be informed about the safety data of the HBV drugs on a
possible pregnancy.
PegIFNa is contraindicated during pregnancy. There are no All candidates for chemotherapy and immunosuppres-
adequate and well-controlled studies of LAM, ADV and ETV in sive therapy should be tested for HBV markers prior to
pregnant women. Reproduction studies have been performed in immunosuppression (Evidence level I, grade of recom-
animal and in humans with TDF and TBV and revealed no evi- mendation 1).
dence of harm to the fetus due to these drugs.1 Among the last
All HBsAg-positive patients should receive ETV or TDF or
two agents, TDF should be preferred, because it has a better resis-
TAF as treatment or prophylaxis (Evidence level II-2,
tance profile and more extensive safety data in pregnant HBV
grade of recommendation 1).
positive women.1,196–198
In a woman of childbearing age without advanced fibrosis HBsAg-negative, anti-HBc positive subjects should
who plans a pregnancy in the near future, it may be prudent to receive anti-HBV prophylaxis if they are at high risk of
delay therapy until the child is born. In a woman of childbearing HBV reactivation (Evidence level II-2, grade of recom-
age with advanced fibrosis or cirrhosis who agrees for a ‘‘planned mendation 1).
pregnancy’’ in the future, PegIFNa therapy may be tried as it is
given for a finite duration. It should be noted that effective con-
traception is required during PegIFNa therapy. If PegIFNa is not In HBsAg-positive and HBsAg-negative, anti-HBc positive
possible or has failed, treatment with TDF has to be initiated patients receiving chemotherapy or immunosuppressive therapy,
and maintained even during a future pregnancy. including the established and emerging new biological response
If female patients become unexpectedly pregnant during HBV modifiers, the risk of HBV reactivation can be high, particularly
therapy, treatment indications should be re-evaluated. The same if rituximab is given alone or in combination with steroids.1
treatment indications apply to women who are first diagnosed to The risk of HBV reactivation can be classified as high ([10%),
have CHB during pregnancy. Patients with advanced fibrosis or moderate (1–10%) or low (\1%).52,203 Therefore, all candidates
cirrhosis should definitely continue to be treated, but the treating for chemotherapy and immunosuppressive therapy should be
agent should be TDF. screened for HBsAg, anti-HBs and anti-HBc prior to immunosup-
The prevention of HBV perinatal transmission, which is consid- pression treatment.
ered to occur mainly at delivery, and causes the majority of Vaccination of HBV seronegative patients is recommended.
chronic HBV infection is based on the combination of HBIG and Higher doses or reinforced vaccine may be required to achieve
vaccination given within 12 h of birth. This prophylaxis reduces anti-HBs response in immunocompromised patients.1,107
the rate of perinatal transmission from [90% to \10%.1 HBIG
and vaccine failures occur almost exclusively in HBeAg-positive
HBsAg-positive patients. All HBsAg-positive candidates for
women with high HBV DNA levels ([200,000 IU/ml) and/or
chemotherapy and immunosuppressive therapy should be
HBsAg level above 4–4.5 log10 IU/ml.198–201 NA prophylaxis could
urgently referred to a specialist for further assessment and diag-
be also useful in the few HBeAg-negative women with high levels
nosis of the phase of HBV infection. All these patients should start
of viremia but normal ALT levels.198–201 These mothers should be
potent NA as a treatment or as prophylaxis.
informed that utilising a NA to reduce their viremia levels increase
Patients with chronic hepatits B should be treated with ETV,
the effectiveness to HBIG and vaccination. LAM, TBV or TDF
TDF or TAF, similarly to the immunocompetent patients. Moni-
prophylaxis has been used in this setting during the last trimester
toring and stopping rules for NAs are the same with the immuno-
of pregnancy. Of them, TDF is the preferred agent due to its
competent patients.
characteristics mentioned previously. In a randomised study in
In contrast, the optimal management of patients with
pregnant HBsAg-positive women with high HBV DNA levels
chronic HBV infection, but without chronic hepatitis, remains
([200,000 IU/ml), the rate of mother to child HBV transmission
controversial. Prophylactic administration of LAM has been
at post-partum week 28 was 0% in those treated with TDF
shown to reduce the risk of HBV reactivation and the associated
compared to 7% in the placebo control group per protocol analysis
morbidity and mortality,1 but a residual risk of HBV reactiva-
having a similar safety profile.198 If NA therapy is given as
tion remains approximately in 10% of chronic HBV patients
prophylaxis, i.e., only for the prevention of perinatal transmission,
its duration is not well defined (stopping at delivery or within the with low viremia (HBV DNA \2,000 IU/ml) and in a higher pro-
first 3 months after delivery). The potential advantage of portion of those with higher viremia levels. As recent studies
stopping at delivery is no interference in breast feeding. In suggest that ETV or TDF can be successfully used in such
addition, TDF ameliorated maternal ALT elevations which can patients,204 prophylaxis with ETV, TDF, TAF is recommended
occur during pregnancy or early after delivery in untreated in this setting. Prophylaxis should continue for at least
mothers.202 12 months (18 months for rituximab-based regimens) after ces-
The safety of NA therapy during lactation is uncertain. HBsAg sation of the immunosuppressive treatment and discontinued
can be detected in breast milk, but breast feeding may not be only if the underlying disease is under remission. Liver function
considered a contraindication in HBsAg-positive mothers. In tests and HBV DNA should be tested every 3 to 6 months during
women treated with TDF, tenofovir concentrations in breast milk prophylaxis and for at least 12 months after NA withdrawal as a
have been reported but its oral bioavailability is limited and thus large proportion of HBV reactivations develops after NA discon-
infants are exposed to only small concentrations. tinuation.205–209

HBsAg-negative, anti-HBc positive subjects. The risk of HBV reacti- tive patients should be vaccinated, preferentially with a
vation in this group varies widely according to the virological reinforced vaccine.215,216
profile, underlying disease and the type and duration of immuno- All HBsAg-positive patients should be referred to a specialist
suppressive regimen. These subjects can be tested for serum HBV for further assessment and diagnosis of the phase of HBV
DNA before immunosuppression. If viremic, they should be trea- infection.
ted similarly to HBsAg-positive patients.
In the high risk group ([10%), including anti-HBc positive Dialysis patients. Patients with chronic HBV infection but not
subjects who need to be treated with rituximab in the onco- chronic hepatitis B should be monitored, as there is no strong evi-
hematological setting or those undergoing stem cell transplanta- dence to suggest they have increased mordibity and mortality.1
tion, antiviral prophylaxis is recommended. Prophylaxis should In contrast, all patients with HBeAg-positive or -negative chronic
continue for at least 18 months after stopping immunosuppres- hepatitis B should receive a NA, as the preferred treatment strat-
sion and monitoring should continue for at least 12 months after egy, independently of the transplantation program.205,217,218 ETV
prophylaxis withdrawal. LAM may be used safely in this setting is recommended for NA naïve patients,219 TAF could be used for
although few cases of HBV exacerbation due to LAM resistance both NA naïve and NA experienced/resistant patients but studies
have been reported.210–212 Prophylaxis with ETV or TDF or TAF are still ongoing.74,76 All doses of NAs should be adjusted accord-
can be also considered in HBsAg-negative, anti-HBc positive ing to eGFR values in patients with eGFR \50 ml/min (see insert
patients receiving highly immunosuppressive regimens of packages), except for TAF which does not require dose adjust-
extended duration.213,214 ment if eGFR is [15 ml/min. PegIFNa could be also used in
In HBsAg-negative, anti-HBc positive subjects with moderate selected patients. Given that dialysis may reduce ALT levels, cau-
(\10%) or low (\1%) risk of HBV reactivation, pre-emptive ther- tion must be taken to use this marker to assess treatment
apy, not prophylaxis, is generally recommended.205,206 The main indications.
virological event in these anti-HBc positive patients is HBsAg HBsAg-negative, anti-HBc positive subjects do not require
reappearance (seroreversion), constantly associated with hepati- treatment nor prophylaxis but must be monitored for HBV
tis flare; by converse HBV DNA detection leads to seroreversion markers.
and hepatitis in only 50% of cases.212 Pre-emptive therapy is
based upon monitoring HBsAg and/or HBV DNA every 1– Renal transplant recipients. All HBsAg-positive patients should
3 months during and after immunosuppression, and starting receive anti-HBV prophylaxis or treatment with a NA.1,205,217,218
ETV, TDF or TAF treatment in case of detectable HBV DNA or ETV is the preferred option for NA naïve patients. TDF should
HBsAg seroreversion. As HBsAg seroreversion can lead to a sev- be avoided because of renal safety issues and may be considered
ere, even fatal, acute hepatitis, NA should be started as early as only for patients with NA resistance if TAF is not available.74,76,79
possible, independently of ALT levels. For selected clinical set- TAF could be a good treatment option for both NA naïve and
tings, characterised by long duration of immunosuppression, lim- resistant patients, although its efficacy and safety in this setting
ited compliance to monitoring or unknown risk of viral are currently unknown. Though several studies have used LAM
reactivation for new biologicals, universal prophylaxis, rather in the past, this drug is not recommended because of the high risk
than pre-emptive therapy, is recommended. of resistance. NA prophylaxis and treatment should be continued
long-term. Long-term NA therapy has been shown to reduce liver
complications and improve survival. PegIFNa is contraindicated
Dialysis and renal transplant patients
because of the risk of rejection.
Recommendations
Renal function should be carefully monitored during treat-
ment with a NAs.79 Unexpected deterioration of renal function
All dialysis and renal transplant recipients should be during NA therapy may necessitate a change of treatment or dose
screened for HBV markers (Evidence level II-2, grade adaptation. Arterial hypertension and diabetes mellitus should be
of recommendation 1). optimally controlled in renal transplant recipients.
HBsAg-negative, anti-HBc positive renal transplant recipients
HBsAg-positive dialysis patients who require treatment do not require prophylaxis or treatment. Monitoring of HBsAg is
should receive ETV or TAF (Evidence level II-2, grade of recommended to identify the few cases of HBsAg seroreversion in
recommendation 1). which ETV or TAF should be started immediately, irrespectively
All HBsAg-positive renal transplant recipients should of ALT levels.
receive ETV or TAF as prophylaxis or treatment (Evi-
dence level II-2, grade of recommendation 1). Extrahepatic manifestations
Recommendations
HBsAg-negative, anti-HBc positive subjects should be
monitored for HBV infection after renal transplantation
(Evidence level III, grade of recommendation 1). Patients with replicative HBV infection and extrahepatic
manifestations should receive antiviral treatment with
NA (Evidence level II-2, grade of recommendation 1).
HBV is still prevalent in dialysis and renal transplant patients
PegIFNa should not be administered in patients with
and may cause significant morbidity and mortality.1 All dialysis
immune-related extrahepatic manifestations (Evidence
and renal transplant patients should be screened for HBV mark-
level III, grade of recommendation 1).
ers. Though vaccine responsiveness is impaired, HBV seronega-

HBV related extrahepatic manifestations include vasculitis, also used to monitor NA or PegIFNa based treatments and pre-
skin manifestations (purpura), polyarteritis nodosa, arthralgias, dicting therapeutic efficacy including the risk of relapse after
peripheral neuropathy and glomerulonephritis. Mixed cryoglobu- stopping NAs.233–235 Most of these studies were performed in
linemias, positive rheumatoid factor or inflammatory markers Japan, and large correlation studies derived from Caucasian
(complement factors C3/C4, C-reactive protein, blood sedimenta- patients are lacking. Therefore, further studies are awaited pro-
tion rate) may be found in these patients. HBsAg-positive patients viding clear evidence for a superiority of this marker for clinical
with extrahepatic manifestations and active HBV replication may decision making over established HBV biomarkers like HBsAg
respond to antiviral therapy. PegIFNa can worsen some immune and HBV DNA quantification.
mediated extrahepatic manifestations and should not be admin- Circulating HBV RNA was first described in 1996 in the serum
istered in HBV infected patients with immune-related extrahep- of HBV infected patients and later as a potential new marker for
atic manifestations. Although controlled studies of antiviral monitoring NA therapy. HBV RNA can be released into the serum
therapy in this setting are lacking, case reports suggest that the in the form of enveloped pregenomic RNA containing virions,236
use of NA is safe and effective.61,220 Plasmapheresis, corticos- but the full characterisation of these circulating RNAs is ongoing.
teroids and potentially other immune-suppressive drugs during Because of its strong correlation with intrahepatic cccDNA, serum
the initial phase can be useful in addition to NA therapy in special HBV RNA is an interesting marker to study cccDNA transcrip-
cases. tional activity.236–238 A strong correlation between quantitative
serum HBV RNA dynamics and HBeAg loss in both NA and
PegIFNa treated patients239 was recently demonstrated by using
New biomarkers of HBV infection
a new rapid amplification of cDNA-ends with polymerase chain
reaction (RACE-PCR).240 HBV RNA quantification might be also
Viral cccDNA is the key genomic form that is responsible for the
helpful in predicting viral rebound after discontinuation of
persistence of infection and was shown to persist in the liver of
NAs.236 It should be further explored whether the simultaneous
infected patients even after long-term NA therapy and even after
testing of different replicative, transcriptional and translational
HBsAg loss and seroconversion.221,222 The regulation of the intra-
HBV biomarkers will allow for a better definition of the individual
hepatic pool of cccDNA involves several factors including the
‘‘activity” of the chronic HBV infection helping to better predict
dynamics of infection in the liver and the intrahepatic antiviral
long-term treatment outcomes.
immune response.223 Furthermore, cccDNA transcriptional activ-
ity is controlled by fine epigenetic regulation which can involve
viral and host factors.13,224 Besides the need for standardized Future treatment options
assays, the main limitation of cccDNA studies is the requirement
of liver biopsy; thus surrogate biomarkers are being evaluated Future treatment options for HBV
(see below). It is noteworthy that not all transcripts are expressed
from cccDNA, but can also be expressed from viral sequences Many research programs are ongoing to develop new treatment
integrated in the host genome. Viral genome replication cannot concepts that focus on the clearance of HBsAg in a significant pro-
occur from these integrants, but HBsAg expression can occur portion of patients, with the principle aims of: i) stopping treat-
either from the envelope gene in cccDNA and/or in viral inte- ment with no risk of virological relapse and no risk of liver
grants, explaining at least in part why quantification of HBsAg disease progression and, ii) to further decrease the risk of HCC.
is not a perfect biomarker for intrahepatic cccDNA.225 Quantifica- Several definitions of cure have been proposed following sev-
tion of cccDNA levels and its transcriptional activity will be eral international workshops.241,242 A true cure may not be feasi-
important in clinical trials evaluating novel treatment concepts ble because HBV DNA is integrated into the host genome.
to cure HBV infection. Furthermore, among persons who recovered from acute hepatitis
Hepatitis B core-related antigen (HBcrAg) is a composite bio- B, viral cccDNA can still be detected in the liver explaining the
marker comprising several antigens expressed from the precore/- reactivation of HBV replication when these ‘recovered’ persons
core gene: HBcAg, HBeAg, and prec22 precursor protein.226 The are profoundly immunosuppressed. The ability to ‘cure’ HBV at
HBcrAg associated proteins can also be detected in circulating earlier stages of liver disease would theoretically have a greater
hepatitis B virions (Dane particles) as well as in HBV DNA nega- impact on reducing the risk of HCC.
tive Dane particles containing the 22 kDa precore protein and The novel treatment options under pre-clinical and early clin-
exceeding Dane particles by 100-fold, and probably also in ical evaluation can be categorized into direct antivirals and
pregenomic RNA containing virions.227 The marker does not over- immunotherapeutic agents.
lap with HBsAg quantification, and in contrast to HBsAg, HBcrAg Direct-acting antivirals include HBV entry inhibitors, drugs
quantification might not be influenced by translation from inte- aiming at cccDNA destruction or silencing, approaches targeting
grated viral sequences. Hence, HBcrAg quantification may pro- viral transcripts by siRNA or anti-sense oligonucleotides, nucleo-
vide additional information concerning the translational activity capsid assembly modulators, approaches to decrease HBsAg
of the HBV infection beyond HBsAg quantification. How to best release in serum. This list is not meant to be comprehensive as
use this new assay in the management of patients with chronic many viral targets are currently being screened for drug discov-
HBV infection is still a matter of debate. It has been demonstrated ery. First phase clinical trials are ongoing for several of these
that the serum HBcrAg levels may partly reflect the amount of agents.241,243
intrahepatic DNA and cccDNA in hepatocytes especially in Several potential target mechanisms for immune modulation
HBeAg-positive patients.228,229 It might also be helpful in defin- to engender or restore HBV specific immune responses in con-
ing the phase of chronic HBV infection, especially in HBe- junction with profound inhibition of HBV replication and HBsAg
negative patients, as well as predicting the long-term HCC production to attain immunological control have been sug-
risk.228,230–232 Some studies suggest that this biomarker can be gested.241,243 Several approaches are currently being evaluated

in clinical trials to restore innate immunity in CHB patients. Conflict of interest
Among them, Toll-like receptors 7 (TLR7) agonists have been
the most explored, but other strategies that restore IFNa respon- Kosh Agarwal: Advisor/lecturer: Abbvie, Astellas, Boehringer
siveness or other antiviral innate pathways are under investiga- Ingelheim, Bristol-Myers Squibb, Gilead, GlaxoSmithKline, Inter-
tion. The lack of a T cell mediated response in chronic HBV is cept, Janssen, Merck Sharp & Dohme, Novartis; Research grants:
partly due to the expression of co-inhibitory receptors and by from Abbvie, Bristol- Myers Squibb, Gilead, Roche.
the expression of immunosuppressive cytokines. Recent cancer Thomas Berg: Research support from Abbvie, Roche, BMS,
therapies have indicated the potential of check-point inhibitors Gilead, Novartis, Merck/MSD, Intercept, Janssen, Novartis,
to restore anti-tumor adaptive immunity. Interesting results have Sequana Medical, and Pfizer; provided consultancy, speakers
been obtained for HBV in animal models and in ex vivo studies in bureau and participated in advisory boards for Abbvie, Alexion,
humans. The main concerns of this approach are the potential Bayer, Boehringer Ingelheim, BMS, Gilead, GSK, Intercept, Janssen,
induction of uncontrolled hepatitis flares and autoimmunity. Sev- MSD/Merck, Merz, Novartis, Sequana Medical and Roche.
eral therapeutic vaccines have been evaluated with limited suc- Maria Buti: advisor/lecturer for Abbvie, Boehringer Ingelheim,
cess, but new vaccine formulations are under clinical Bristol-Myers Squibb, Gilead, GlaxoSmithKline, Janssen, Merck
evaluation.241,243 Sharp & Dohme, Novatis; research grants/clinical trials from Abb-
Combinations of antiviral therapy targeting multiple steps in vie, Bristol- Myers Squibb, Gilead, Janssen, Merck Sharp & Dohme.
the HBV lifecycle that suppress viral replication and viral antigen Harry Janssen: Grants from AbbVie, Bristol-Myers Squibb,
production and immune modulatory therapy to restore immune Gilead, Innogenetics, Janssen, Medimmune, Merck, Novartis,
response to HBV will likely be needed to achieve the goal of a Roche; Consultant for AbbVie, Arbutus, Benitec, Bristol-Myers
HBV ‘cure’. Squibb, Spring Bank Pharma, Gilead, GSK, Ionis Pharmaceuticals,
Given the focused drug discovery effort and the potential of a Janssen, Medimmune, Merck, Novartis, Roche.
future ‘cure’ strategy, it is important to be cognisant, when con- Pietro Lampertico: advisor and speaker bureau for Gilead,
sidering the current clinical management of CHB patients, of Roche, BMS, GSK, MSD, Arrowhead, Alnylam.
the potential evolution of therapy in HBV. Patients who are will- George Papatheodoridis: advisor/lecturer for Abbvie, Boehrin-
ing to participate and/or are in phases of the disease that are not ger Ingelheim, Bristol-Myers Squibb, Gilead, GlaxoSmithKline,
eligible for therapy within the current guidelines may be consid- Janssen, Merck Sharp & Dohme, Novartis, Roche; research
ered for clinical trial participation. grants/clinical trials from Abbvie, Bristol- Myers Squibb, Gilead,
Janssen, Merck Sharp & Dohme, Novo Nordisc, Regulus, Roche;
consultant for Roche; Data Safety Management Board for Gilead.
Future treatment options for HDV Frank Tacke: advisor/lecturer for Abbvie, Boehringer, Bristol-
Myers Squibb, Gilead, Merck Sharp & Dohme, Falk, Intercept,
At present, patients co-infected by HBV and HDV have to be trea- Tobira, Galapagos, Silence Therapeutics; research grants from
ted with PegIFNa. The success rate of these treatments is low. Tobira, Galapagos and Noxxon.
Several candidates are being evaluated in clinical trials mainly Fabien Zoulim: advisor and/or research grants for/from Arbu-
in combination with PegIFNa and/or NA including HBV/HDV tus, Assembly, Contravir, Gilead, Janssen, Roche, Sanofi,
entry inhibitors (Myrcludex-B),244,245 drugs inhibiting the release Transgene.
of HBsAg (nucleic acid polymers),246 and inhibitors of the preny-
lation of the large HDV antigen.243,247 Whenever possible, enroll- Acknowledgements
ment in these new clinical trials should be considered, either as a
rescue of PegIFNa failure or to improve treatment success rate in We would like to thank the reviewers of this Clinical Practice
naïve patients. Guideline for their time and critical reviewing: EASL Governing
Board, Maurizia Brunetto, Henry Chan, Markus Cornberg.
Unresolved issues and unmet needs References
[1] European Association for the Study of the Liver. EASL clinical practice
When to start antiviral therapy in patients with HBeAg-
guidelines: Management of chronic hepatitis B virus infection. J Hepatol
positive chronic HBV infection 2012;57:167–185.
Stopping rules for HBeAg-negative patients treated with NA [2] Schweitzer A, Horn J, Mikolajczyk RT, Krause G, Ott JJ. Estimations of
Retreatment criteria after NA discontinuation worldwide prevalence of chronic hepatitis B virus infection: A systematic
How to accelerate HBsAg decline in long-term NA treated review of data published between 1965 and 2013. Lancet
2015;386:1546–1555.
patients? [3] Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K, Aboyans V, et al. Global
Better baseline or on-treatment predictors of sustained treat- and regional mortality from 235 causes of death for 20 age groups in 1990
ment in patients treated with PegIFNa and 2010: A systematic analysis for the Global Burden of Disease Study
Definition of the residual risk of HCC in patients on long-term 2010. Lancet 2012;380:2095–2128.
[4] Ott JJ, Stevens GA, Groeger J, Wiersma ST. Global epidemiology of hepatitis
NA therapy and impact on surveillance
B virus infection: New estimates of age-specific HBsAg seroprevalence and
Unmet need: new treatments with finite duration and high endemicity. Vaccine 2012;30:2212–2219.
cure rates [5] Chen C-L, Yang J-Y, Lin S-F, Sun C-A, Bai C-H, You S-L, et al. Slow decline of
How to define a cure of HBV infection? Definition of novel hepatitis B burden in general population: Results from a population-based
endpoints survey and longitudinal follow-up study in Taiwan. J Hepatol
2015;63:354–363.
Biomarkers for the cure of infection and for the cure of the [6] Coppola N, Alessio L, Gualdieri L, Pisaturo M, Sagnelli C, Caprio N, et al.
liver disease Hepatitis B virus, hepatitis C virus and human immunodeficiency virus

EASL 2017 Clinical Practice Guidelines On The Management of Hepatitis B Virus Infection

Uploaded by

Copyright:

Available Formats

EASL 2017 Clinical Practice Guidelines On The Management of Hepatitis B Virus Infection

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

EASL 2017 Clinical Practice Guidelines On The Management of Hepatitis B Virus Infection

Uploaded by

Copyright:

Available Formats

Clinical Practice Guidelines

EASL 2017 Clinical Practice Guidelines on the management

Summary infection require specific focus. Future treatment strategies to

Journal of Hepatology 2017 vol. 67 j 370–398

Journal of Hepatology 2017 vol. 67 j 370–398 371

HBV markers Liver disease

HBeAg positive HBeAg negative

372 Journal of Hepatology 2017 vol. 67 j 370–398

Journal of Hepatology 2017 vol. 67 j 370–398 373

374 Journal of Hepatology 2017 vol. 67 j 370–398

Journal of Hepatology 2017 vol. 67 j 370–398 375

HBsAg positive HBsAg negative, anti-HBc positive

Chronic HBV infection1 Chronic hepatitis B (CHB) No specialist follow-up

376 Journal of Hepatology 2017 vol. 67 j 370–398

Features PegIFNa ETV, TDF, TAF

80 1 yr In patients with HBeAg-negative CHB, the 5-year cumulative

Journal of Hepatology 2017 vol. 67 j 370–398 377

378 Journal of Hepatology 2017 vol. 67 j 370–398

PegIFN Nucleoside analogues Nucleotide analogues

PegIFN Nucleoside analogues Nucleotide analogues

Journal of Hepatology 2017 vol. 67 j 370–398 379

Since NA therapy does not usually achieve HBV eradication

380 Journal of Hepatology 2017 vol. 67 j 370–398

Journal of Hepatology 2017 vol. 67 j 370–398 381

HBV variant LAM LDT ETV ADV TDF/TAF*

382 Journal of Hepatology 2017 vol. 67 j 370–398

Combination therapy for CHB

Journal of Hepatology 2017 vol. 67 j 370–398 383

WEEK 12 Stop if HBsAg No decline >20,000 >20,000 No decline

WEEK 24 Stop if HBsAg >20,000 >20,000 >20,000 >20,000

HBsAg levels Any decline No decline

HBV DNA levels >2 log decline <2 log decline

Continue Continue Stop

384 Journal of Hepatology 2017 vol. 67 j 370–398

Journal of Hepatology 2017 vol. 67 j 370–398 385

386 Journal of Hepatology 2017 vol. 67 j 370–398

Journal of Hepatology 2017 vol. 67 j 370–398 387

388 Journal of Hepatology 2017 vol. 67 j 370–398

Journal of Hepatology 2017 vol. 67 j 370–398 389

390 Journal of Hepatology 2017 vol. 67 j 370–398

Journal of Hepatology 2017 vol. 67 j 370–398 391

Unresolved issues and unmet needs References

392 Journal of Hepatology 2017 vol. 67 j 370–398

You might also like