KASL clinical practice guidelines for management of chronic hepatitis B

CHB, immune-tolerant phase

The immune-tolerant phase is generally related to vertical transmission and is characterized by HBeAg positivity, very high level of serum HBV DNA (generally ≥10⁷ IU/mL), persistently normal level of ALT, and minimal or absence of hepatic necroinflammation [2,3]. In a follow-up of immune-tolerant CHB patients, serum ALT was elevated in 16% of cases, and the follow-up fibrosis stage was not different from the initial stage in those who remained immune-tolerant for 5 years [2]. In another study from Taiwan, 5% of 240 immune-tolerant CHB patients progressed to cirrhosis and did not develop HCC in 10 years of follow-up [4]. In a Korean multi-center study, 1% of 946 immune-tolerant CHB patients progressed to cirrhosis, and 1.7% developed HCC during 10 years of follow-up [5].

However, a small-scale in vitro study suggested that early hepatocarcinogenesis could be progressing even during the immune-tolerant phase, as was evident by a high level of HBV DNA integration and clonal hepatocyte expansion in liver tissues from immune-tolerant CHB patients as with immune-active CHB patients [6]. Recently, several studies reported that development of HCC was not uncommon in the immune-tolerant phase, although the studies included patients with serum HBV DNA <10⁷ IU/mL, serum ALT >upper limit of normal (ULN), or high fibrosis index [7], necessitating further studies to clarify the criteria for discriminating patients in “true” immune-tolerant phase.

The immune-tolerant phase can last for more than three decades in patients infected with HBV genotype C due to late HBeAg seroconversion [8]. Therefore, many female patients infected with this genotype are in the immune-tolerant phase when they are of childbearing age, which can lead to vertical transmission of HBV to a child [9].

HBeAg-positive CHB, immune-active phase

With increasing age, most patients in the immune-tolerant phase experience immune responses to HBV. Such changes are due to increased response of cytotoxic T lymphocytes to hepatitis B core antigen (HBcAg) and HBeAg [10], resulting in destruction of infected hepatocytes. This phase is characterized by HBeAg positivity and fluctuating courses of serum ALT and HBV DNA levels [11,12]. Histological findings reveal moderate-to-severe necroinflammation [13]. There can be various stages of liver fibrosis according to severity of liver injury.

Once HBeAg seroconversion occurs, the natural course of the disease can have one of three clinical features: (1) repeated HBeAg reversion and seroconversion, (2) immune-inactive phase, or (3) HBeAg-negative, immune-active phase of CHB [14,15]. Typically, 10–40% of patients who experience seroconversion revert to an HBeAg-positive state and then experience recurrence of seroconversion at least once with progression of hepatitis activity [16,17]. In particular, reversion frequently occurs in patients with HBV genotype C, and the rate decreases with age [9]. Hepatic decompensation, which occurs in 5% of patients with acute exacerbation, can be fatal [18].

CHB, immune-inactive phase

Most patients who seroconvert during the immune-active phase progress to the immune-inactive phase, which is characterized by HBeAg negativity, antibody to HBeAg (anti-HBe) positivity, persistent normal ALT level, and HBV DNA level below 2,000 IU/mL [19-21]. Typical histological findings in the third phase are mild liver inflammation [19], and various stages of liver fibrosis can reflect previous liver injury [22].

This phase persists for an extended period in most patients but has a relatively good prognosis. However, an estimated 20% of such patients will revert to the HBeAg-negative or HBeAg-positive immune-active phase and can experience recurring periods of reactivation and inactivation throughout their lives, which can lead to cirrhosis or HCC [23,24].

HBeAg-negative CHB, immune-active phase

Approximately 20% of patients who experience HBeAg seroconversion during their immune-active HBeAg-positive phase progress to the immune-active HBeAg-negative phase, with HBV DNA level ≥2,000 IU/mL, increased ALT level, and active necroinflammation of the liver [14]. These patients show HBeAg-negativity because they harbor HBV variants in the precore (PC) or basal core promoter (BCP) regions of HBV DNA, resulting in failure or reduction of HBeAg production [25-28]. The immune-active HBeAg-negative phase is associated with older age and lower rates of prolonged spontaneous disease remission, and most patients in this phase will experience persistent hepatocellular inflammation and progress to hepatic fibrosis and cirrhosis [27,29,30]. Severe fluctuations of HBV DNA and ALT levels can make it difficult to differentiate these patients from those in the immune-inactive phase [31]. Therefore, HBV DNA and ALT levels should be monitored in every 3 months for at least 1 year in case of immune-inactive phase to find out the HBeAg-negative immune-active phase requiring antiviral treatment [32,33].

HBsAg loss phase

During the natural course of CHB, HBsAg loss is a very rare transition that indicates potential cure of HBV infection (Table 3) [34-37]. Complete cure, or “sterilizing cure,” of HBV infection implies seroclearance of HBsAg and HBV DNA as well as complete clearance of intrahepatic covalently closed circular DNA (cccDNA) and/or integrated HBV DNA. However, it is difficult to achieve these goals at present [37]. Accordingly, the realistic goal suggested is a “functional cure,” which refers to seroclearance of HBsAg and HBV DNA regardless of antibody to HBsAg (anti-HBs) [37]. Despite the presence of intrahepatic cccDNA and/or integrated HBV DNA, it is a successful immunological control state of CHB and therefore can be viewed as a concept similar to idealistic functional cure [36]. In certain circumstances, such as immunosuppression, the risk of HBV reactivation persists [38].

Table 3.

Definitions of HBV cure

Patients in the immune-inactive phase subsequently progress to HBsAg loss or clearance phase at a rate of 1–2% annually [31,39,40]. According to Liaw’s prospective data, HBsAg loss occurs in 0.5% of CHB patients per year and 0.8% of asymptomatic chronic HBV carriers per year [35]. Korean patients reportedly experience a relatively low rate of HBsAg loss (0.4% annually) [41]. In a few patients, serum HBV DNA can be detected at a very low titer during this phase [42,43]. HBsAg loss is the state of functional cure and is associated with a reduced risk of cirrhosis. However, significant risk of HCC development remains even after HBsAg loss in male patients and in settings where HBsAg loss has been achieved late (presence of cirrhosis or age ≥50 years) [42,44].

HBsAg seroclearance is also achieved in a few patients on antiviral treatment, and the long-term clinical outcome compared with that of spontaneous HBsAg seroclearance is unclear [45,46]. Two studies showed no difference in HCC incidence between patients who achieved treatment-induced and spontaneous HBsAg seroclearance [47,48], yet another study reported a 5-year cumulative incidence of HCC of 0.9% and 3.9% in spontaneous and treatment-induced HBsAg seroclearance patients, respectively [49].

Grey zone

The natural course of CHB is divided into five phases with regard to clinical indicators such as ALT and HBV DNA – immune-tolerant, HBeAg-positive immune-active, immune-inactive, HBeAg-negative immune-active, and HBsAg loss. However, patients who do not fit into any of the usual clinical phases are considered to be in the “grey zone.” Approximately 30% of CHB patients are in the grey zone [50-52], and the prognosis and treatment of CHB patients in the grey zone are under study [52-57].

The decision whether to initiate antiviral therapy for patients in the grey zone is a clinically challenging question, and it is often difficult to make an accurate decision without the results of a liver biopsy. Treatment of patients in the grey zone is summarized in the ‘Treatment Indication’ session.

Quantitation of HBsAg

The HBV encodes the three proteins of the HBsAg, which form the viral envelope, small (SHBsAg), middle (MHBsAg), and large (LHBsAg). These proteins are translated from two HBV subgenomic mRNA transcripts, the preS1 mRNA and the preS2/S mRNA. HBsAg is not only generated by transcription and translation of cccDNA, but also can be generated from HBV DNA episomally integrated into the host genome. HBsAg is assembled with core proteins and DNA polymerases during the viral replication process to form a complete virion (Dane particle). The surface proteins also generate non-infectious excess sub-viral particles, approximately 22 nm in diameter, with either a spherical or a long filamentous form. The quantitation of serum HBsAg is a method of detecting all three forms. There is a positive correlation between levels of serum HBsAg and intrahepatic cccDNA in patients with CHB.

Levels of HBsAg helps to distinguish inactive carriers from CHB patients. In HBeAg-negative patients, one-time measurements of serum HBV DNA <2,000 IU/mL and HBsAg <1,000 IU/mL suggest future inactive carriers [105,106]. In contrast, among HBeAg-negative patients with lower viral load (HBV DNA <2,000 IU/mL), HCC risk is higher in those with a high HBsAg titer (>1,000 IU/mL) than in those with a low HBsAg titer [107]. Levels of HBsAg can predict response during pegylated interferon (peginterferon) therapy in HBeAg-positive patients, possibly providing a guide to stopping treatment earlier [108]. For nucleos(t)ide analogue (NA) treatment, levels of HBsAg has emerged as a valuable tool in identifying patients who will maintain or attain inactive status or develop HBsAg loss if they discontinue long-term NA therapy [109-111].

Hepatitis B core-related antigen (HBcrAg)

HBcrAg is a composite biomarker incorporating several viral antigens expressed from the pre-core/core gene: the HBcAg, HBeAg, and p22 core-related antigen (p22Cr) [112]. Although it is not yet commonly used in Korea, HBcrAg can be easily detected using a chemiluminescent immunoassay kit. Significant positive correlations between serum HBcrAg level and HBV DNA level, as well as amount of intrahepatic cccDNA, have been observed in several Asian and European studies [113]. Serum HBcrAg level varies significantly among the phases of HBV infection. HBcrAg is, therefore, also a good virologic marker to differentiate HBeAg-negative CHB patients (active disease) from HBeAg-negative chronic HBV infection (inactive disease) [114]. Level of HBcrAg can potentially predict the possibility of achieving partial cure in patients on antiviral treatment, as defined by a sustained off-therapy virological control. In the study of 130 Hong Kong patients with undetectable serum HBV DNA during NA, HBcrAg was detectable in 101 (78%) samples [115]. After eight years of NA treatment, 21.3% of patients achieved serum HBcrAg <3 log10 U/mL [116]. Based on these findings, monitoring of HBcrAg and HBsAg quantification is recommended by the Japan Society of Hepatology guidelines to identify patients who can discontinue NA [117]. Furthermore, HBcrAg can be an aid for clinicians in identifying patients with a higher risk of HCC development or post-treatment recurrence [118,119]. Recently, the HBcrAg level has been the most emerging noninvasive predictor reflecting intrahepatic viral replication.

HBV RNA

Serum HBV RNA represents partially reverse-transcribed, encapsidated pregenomic RNA (pgRNA) in virus-like particles. Given that pgRNA is transcribed directly from cccDNA, level of serum HBV RNA can potentially serve as a surrogate marker for transcriptionally active cccDNA [120]. Serum HBV RNA may be an indicator for predicting off-treatment response in patients with suppressed HBV under NA treatment [121,122], although it has limited additional value in differentiating CHB phases [123]. Measurement of HBV RNA could be used to determine cccDNA activity, especially with therapies in development that aim for functional cure [119,124]. Levels of HBV RNA is also associated with increased HCC risk and post-resection recurrence in NA-treated patients [125,126]. Before a widespread clinical application of serum HBV RNA for CHB can be accepted and applied, the methodology for detecting and quantifying serum HBV RNA should be standardized.

HBV cccDNA

In the nucleus of hepatocytes, cccDNA is maintained as a stable mini-chromosome and acts as the template for pgRNA transcription and viral protein production [127]. The existence of intrahepatic cccDNA is the molecular basis of disease chronicity and represents a major barrier to HBV eradication. cccDNA is responsible for persistence of the virus in the liver even after HBsAg loss and seroconversion. Occult HBV infection is a state in which HBsAg is not detected in serum, but HBV DNA is present in liver tissue or blood, which is believed to be due to cccDNA remaining in the nucleus. cccDNA also causes hepatitis B reactivation in patients receiving immunosuppressants or chemotherapy [128]. In terms of HBV treatment, current widely used NAs can block the transcription process of RNA production from cccDNA but not remove cccDNA itself. Therefore, cccDNA acts as a major cause of HBV reactivation after stopping the NA, and the development of drugs that can remove cccDNA itself is essential to cure hepatitis B [129]. In order to quantify cccDNA in the liver, invasive liver biopsy is required, making it difficult to use as a marker in clinical settings.

HBV integration into the host genome

HBV can integrate in human DNA and promote hepatocarcinogenesis by insertional mutagenesis, increased genomic instability, or expression of viral oncoproteins such as the protein HBx [130,131]. The application of next-generation sequencing led to characterization of HBV integration events. Recent studies reported that, among patients with primary liver cancer, a higher number of HBV integration events was associated with worse outcome [130,132-134]. In addition, cell-free HBV-integrated tumor DNA was shown to be a circulating biomarker for detecting tumor load in a majority of patients with HBV-related HCC and aided in monitoring residual tumor and recurrence clonality after tumor resection [135,136].

HBV pgRNA

Among HBV mRNAs, 3.5-kb preC and pgRNA originate from HBV cccDNA, and pgRNA is reverse transcribed to enable the synthesis of HBV DNA. Intrahepatic pgRNA is considered a marker of active HBV replication. RNA interference (RNAi) therapeutics such as siRNA and anti-sense molecules targeting pgRNA are being developed. However, the role of intrahepatic pgRNA as a marker for treatment response has recently been replaced by serum HBV RNA measurements.

Capsid assembly modulator

The HBV core protein aids in capsid assembly and imports pgRNAs and polymerases into the capsid for reverse transcription. Then, enveloped HBV DNA is transmitted to infect other hepatocytes or reenter the nucleus for viral cccDNA replenishment. Thus, inhibition of encapsidation could induce reduction of cccDNA expression [211,212].

NVR 3-778, ABI-H0731, and JNJ-6379 are core protein inhibitors developed as oral drugs. Clinical studies of these drugs have reported a significant reduction in serum HBV DNA but a minor reduction in serum HBsAg titer. Therefore, use of core protein inhibitors alone might be inadequate for functional cure of CHB. Thus, combination therapies with interferon or NAs were tested, but the decrease in serum HBsAg titer remained nonsignificant compared to that in the control group [213-217].

A recent study from Korea demonstrated that ciclopirox, a synthetic antifungal agent, strongly inhibits viral replication by blocking the HBV capsid assembly. When combined with NAs, ciclopirox significantly reduced the serum HBV DNA level and HBsAg titer in in vitro and in vivo studies [218,219]. Preclinical studies for the same are currently underway.

RNA targeted therapeutics

Viral RNA forms the backbone of viral antigens and proteins. RNAi inhibits translation of viral transcripts to prevent its replication and HBsAg production, restore HBV-specific immune response, and potentially lead to a functional cure [220]. Currently, two therapeutics are being used for RNAi-based therapy: antisense oligonucleotides (ASO) and small interfering RNAs (siRNAs). ASOs are 15–20-nucleotide-long single-stranded DNA oligomers, which binds to its complementary site on the target viral RNA to form a DNA-RNA duplex. The viral RNA in the DNA-RNA duplex is then cleaved by ribonuclease H, inhibiting the expression of the corresponding gene. On the other hand, siRNAs are 20–25 nucleotide-long double-stranded RNA molecules. The siRNA guide strand directs the RNA-induced silencing complex (RISC) to the complementary target RNA, which is then cleaved by the Argonaute protein present in the RISC [221,222]. Lipid nanoparticles or N-acetylgalactosamine (GalNac/NAG)-linked particles are used to deliver these agents to hepatocytes [222,223].

The main concerns regarding these agents include the risk of off-target toxicity, the potential toxicity of the delivery vehicle, and the risk of post-treatment reactivation by remaining cccDNA [219]. In addition, ALT flare should be monitored during the course of the treatment [224].

In the ASO, a phase 2 clinical study was conducted for GSK3228836, and both treatment-naïve and previously NA-treated groups each receiving 300 mg six times showed a mean reduction of more than 1.5 log10 IU/mL in serum HBsAg titer [224]. Several siRNA drugs are in phase 2 clinical trials, and interim analyses were presented at an international conference. With NA co-administration, JNJ-3989 induced a decrease in HBsAg greater than 1.0 log10 IU/mL in 98% of the participants and an average decrease of 1.93 log10 IU/mL [225]. Similar reduction in HBsAg titer was shown with AB-729, Vir-2218, and RG6346 [226-228].

HBsAg release inhibitor

HBsAg release inhibitors suppress the assembly and secretion of incomplete HBV particles (subviral particles, SVPs) to reduce serum HBsAg titer and mitigate immune exhaustion in the host, enhancing the HBV-specific immune response. Most of these drugs are nucleic acid polymers, of which REP 2139 and REP 2165 have completed phase 2 clinical trials. After 48 weeks of triple combination therapy with peginterferon and tenofovir disoproxil fumarate (tenofovir DF) in patients with CHB, HBsAg loss was observed in 60% of the patients. Among them, 58.3% maintained HBsAg loss for 48 weeks after discontinuation of REP 2139 or REP 2165 [229]. In addition, both these drugs showed a significant therapeutic effect in the treatment of HBV and hepatitis D virus (HDV) co-infection, supporting the potential of HBsAg release inhibitors for the effective treatment of HDV infection [230].

TLR agonist

TLRs recognize viral and bacterial pathogen-associated molecular patterns (PAMPs) and induce downstream signaling pathways that culminate in the transcriptional activation of interferon-stimulating genes. GS-9688 (selgantolimod) is a TLR-8 agonist, and the interim results of the phase 2 clinical study in combination with NAs were recently presented. Weekly administration of 3 mg GS-9668 for 24 weeks decreased HBsAg titer by 0.5 log10 IU/mL or more in 7% of treatment-naïve patients (vs. 0% in the control group) and demonstrated functional cure in 5% of patients who were virologically stable [231,232].

Immune checkpoint inhibitor

HBV evades the HBV-specific immune response by inducing sustained expression of inhibitory receptors on T cells. Therefore, a functional cure for CHB can be achieved by overcoming the mechanisms underlying immune inhibition using an immune checkpoint inhibitor. A pilot study reported that single administration of nivolumab (anti-PD-1), a widely used anticancer drug, to patients with HBeAg-negative CHB significantly reduced serum HBsAg titer for 24 weeks. This showed the possibility of a functional cure using immune checkpoint inhibitors [233]. Envafolimab, developed as an anti-PD-L1, is undergoing a phase 2 clinical trial as a combination treatment with NAs, and the interim results were recently announced. Serum HBsAg titer decreased by 0.49 log10 IU/mL during a 12-week follow-up after a single dose of 2.5 mg Envafolimab [234].

Therapeutic vaccine

A therapeutic vaccine can enhance the host HBV-specific immune response through exposure to various antigens and proteins of HBV.

NASVAC is a nasal spray vaccine currently undergoing phase 3 clinical trials. During a follow-up period of 138 weeks, 6.3% of the previously NA-treated group and 21.7% of the treatment-naïve group showed a reduction in the serum HBsAg titer [235]. GS-4774, expressing HBs, HBc, and HBx antigens, with tenofovir DF did not show a significant decrease in serum HBsAg titer compared to the control group [236,237].

ALT normalization

Normalization of ALT in CHB treatment reflects a decrease in hepatic inflammatory response, mostly associated with undetectable HBV DNA, and reduces clinical deterioration [166]. Normalization of ALT during treatment reflects improvement in cirrhosis and could be considered reflective of treatment goals. However, 14–40% of patients with persistently normal ALT could have significant fibrosis (≥F2), and there is a variety of concurrent liver conditions affecting ALT level, such as non-alcoholic or alcoholic fatty liver [198,256]. As such, ALT normalization alone is insufficient when determining the endpoint of treatment.

Undetectable HBV DNA

HBV DNA level is the strongest indicator of disease progression and long-term outcomes in the natural course of CHB [68,70]. HBV DNA level is associated with histological activity in CHB patients, with low rate of progression to decompensation and high rate of survival in patients with low HBV DNA [257,258]. Antiviral therapy can reduce HBV DNA level, and histological improvement can be achieved in proportion to HBV DNA reduction [198,258-260]. In addition, since it reduces the progression and exacerbation of liver disease and prevents HCC [261,262], HBV DNA is a useful alternative indicator that reflects treatment endpoints. When HBV DNA is not detectable over a long term, and virological response is well maintained, the HBsAg loss rate increases even after cessation of therapy in patients with HBeAg-negative CHB. Therefore, cessation of therapy could be considered in these patients with long-term undetectable HBV DNA [244-246,250]. The lower is the HBV DNA level, the better is the clinical status; however, when the HBV DNA level is 60–2,000 IU/mL, the risk of cirrhosis and HCC is similar to that of patients with undetectable HBV DNA [68,70]. Evidence for HBV DNA level as a surrogate indicator in these patients is lacking. Also, in practice, most patients with undetectable HBV DNA relapsed after cessation of NA therapy [194,244-246]. Hence, undetectable HBV DNA cannot be the sole indicator determining treatment cessation.

HBeAg loss and/or seroconversion

HBeAg seroconversion in patients with HBeAg-positive CHB is accompanied by HBV DNA reduction, ALT normalization, and histological improvement. After HBeAg seroconversion, HBsAg loss increases to 1.15% per year [263,264]. Therefore, HBeAg loss/seroconversion in HBeAg-positive CHB could be an indicator reflecting achievement of the treatment goal. However, after cessation of NA therapy following HBeAg loss/seroconversion, the sustained virological response is reported to be 62.5%, 53.4%, and 51.5% at 1, 2, and 3 years, respectively [265]. Also, in some patients, HBeAg-negative CHB and HBeAg reversion or acute exacerbation accompanied by jaundice can occur in severe cases [266], so there is a limit to the indicator determining treatment cessation with only HBeAg loss/seroconversion. Sustained undetectable HBV DNA after HBeAg loss/seroconversion and the duration of consolidation therapy to maintain ALT normalization are important for maintaining response after cessation of NA therapy [248,267], but studies on the specific period are lacking. Nevertheless, it is recommended to maintain consolidation therapy for at least 12 months after HBeAg loss/seroconversion [144,151,252,265,268].

Quantitative HBsAg level, quantitative HBcrAg level, and HBV RNA

HBsAg is translated from pre-S1 and S2 messenger RNAs that are transcribed from the S gene. HBsAg exists as three protein subtypes (small, middle, and large) that are not differentiated by commercial assays. Additionally, there are two non-infectious subviral particles secreted, in spherical and filamentous forms, with 100-fold to 100,000-fold higher levels than mature virions [104]. The level of HBsAg has been found to be an indicator for off-therapy sustained response in many studies [269-274]. The degree of decline in HBsAg level during peginterferon can predict the efficacy of treatment-induced immune response. In patients whose HBsAg level does not sufficiently decrease in the early phase of peginterferon, additional treatment is unlikely to be successful, so it is used as the stopping rule, reducing unnecessary treatment [269]. Additionally, HBsAg level can be useful to determine if cessation of NA therapy is an option [270-274]. Chen et al. found the HBsAg level at the end-of-treatment (EOT) was the most important predictor for HBsAg loss and sustained response after stopping lamivudine treatment. At EOT, the HBsAg cutoff value of 300 IU/mL could predict 55.6% HBsAg loss in HBeAg-positive patients. In HBeAgnegative patients, the HBsAg cutoff values of 120 and 200 IU/mL could predict 79.2% HBsAg loss and 93.3% post-treatment sustained response, respectively [271]. In a Taiwanese study of patients with EOT HBsAg level <100 IU/mL, the cumulative rates of clinical relapse and sustained virological response were 9.3% and 45.5%, respectively, during a mean 2-year follow-up period after cessation of entecavir treatment [272]. In a large retrospective study of 691 HBeAg-negative patients who had discontinued NAs, during a median follow-up period of 156 (2–614) weeks, virological and clinical relapse rates were reported in 79.2% and 60.6%, respectively. In this study, EOT HBsAg level <100 IU/mL was an independent factor for HBsAg loss [273]. Another study showed EOT HBsAg level to be an independent predictor of virological and clinical relapse in HBeAg-negative patients who had discontinued tenofovir DF treatment, and HBsAg level of 80 IU/mL was the optimal value. The virological and clinical relapse rates at 78 weeks were 19.6% and 15.4%, respectively, in patients who achieved EOT HBsAg ≤80 IU/mL. Also, the cumulative rate of HBsAg loss at 104 weeks were 45.5% in these patients [274]. As such, low EOT HBsAg level (10–200 IU/mL) has been reported to be a good predictor of sustained virological response and HBsAg loss after cessation of NA therapy. Therefore, monitoring of quantitative HBsAg level during NA therapy could be helpful in practice [243,270,273].

HBcrAg contains three products encoded by the PC/core gene, HBcAg, p22cr, and HBeAg, all of which can be measured by serological testing. Serum HBcrAg level is closely associated with intrahepatic total HBV DNA and cccDNA, as well as serum HBV DNA [104,113,118,241]. HBcrAg level is useful for predicting off-treatment sustained response after cessation of NA therapy. Recently, HBcrAg level alone or a combination of HBsAg levels has been proposed [247,275]. SCALE-B score (HBsAg level [S], HBcrAg [C], age [A], ALT [L], and tenofovir [E] used for HBV [B]) including HBsAg and HBcrAg levels at NA discontinuation was suggested for prediction of post-NA relapse and HBsAg loss in Asian CHB patients [276-278]. The scoring criteria for including HBsAg and HBcrAg levels for cessation of NA therapy have been proposed and incorporated into the Japan Society of Hepatology Guidelines for the Management of HBV infection [117].

HBV RNA has been introduced as a novel biomarker for cessation of NA therapy [104,242]. There are also studies that predict off-treatment response with a combination of other indicators, such as HBsAg level or HBcrAg level [119,124,279]. However, the value of this indicator has not yet been evaluated as a standardized test, and it is necessary to explore the role of HBV RNA in predicting outcomes after cessation of NA therapy.

HBsAg loss

HBsAg loss is defined as a sustained loss of HBsAg and HBV DNA from serum, with or without anti-HBs seroconversion [37]. HBsAg loss is considered “functional cure” as the optimal treatment endpoint from a virological and clinical point of view. The prognosis following spontaneous HBsAg loss is excellent, except in patients with cirrhosis or those with concurrent other viral infection [23]. The incidence of HCC is significantly reduced when HBsAg loss occurs before ages 45–50 years [280]. Some patients with HBsAg loss during antiviral therapy showed HBsAg reversion or low but detectable HBV DNA, but most patients maintain HBsAg loss and undetectable HBV DNA, and their incidence of liver complications, including HCC, is significantly lower compared to that of patients without HBsAg loss [46,47,49,281]. Therefore, HBsAg loss is generally regarded as the optimal outcome of antiviral therapy in CHB, reflecting the treatment goal [37], at which point NAs can be discontinued [46,281,282]. Recently, it has been reported that HBsAg reversion and serum HBV DNA detection are rare if antiviral therapy is discontinued after HBsAg loss, defined as testing HBsAg negative on two separate occasions at least 6 months apart [281,282].

HBV treatment in patients undergoing curative treatment for HCC

A recent meta-analysis demonstrated that the pooled rates of HBV reactivation and biochemical reactivation were 20% and 9%, respectively, in patients who underwent surgical resection for HCC without NA treatment [283]. In contrast, patients receiving antiviral prophylaxis had a pooled HBV reactivation rate of 2–4%. In terms of local ablation therapies such as radiofrequency ablation or percutaneous ethanol injection, the HBV reactivation rate in patients who received no prophylactic NA was 5–9%, whereas HBV reactivation was rarely diagnosed in those who received prophylactic NA therapy [283-286]. In HBsAg-positive HCC patients with undetectable serum HBV DNA level at HCC diagnosis, some retrospective studies in Korea and China reported that 22–33% of patients experienced HBV reactivation after surgical resection [287-290]. The risk of HBV reactivation was lower for those receiving NA.

Several studies have reported that antiviral treatment was associated with lower risk of tumor recurrence after curative treatment for HBV-related HCC. A Taiwanese large-scale retrospective study showed that patients who underwent NA treatment with entecavir, lamivudine, telbivudine, etc. showed a significantly lower risk of tumor recurrence or overall death after surgical resection for HCC compared to those not treated with these drugs, although the prevalence of cirrhosis was significantly higher [291]. A Chinese randomized controlled trial (RCT) to evaluate the effect of NA treatment on postoperative prognosis of HBV-HCC showed that antiviral treatment significantly decreased HCC recurrence (hazard ratio [HR], 0.48; 95% confidence interval [CI], 0.32–0.70) and HCC-related death (HR, 0.26; 95% CI, 0.14–0.50) [292]. Even in RCTs including patients with low-level viremia (HBV DNA <2,000 IU/mL), antiviral treatment was associated with longer recurrence-free survival and overall survival [293]. NA therapy was also associated with a decreased risk of HCC recurrence among patients with HBV-related HCC post-RFA (HR, 0.69; 95% CI, 0.50–0.95) [294]. A meta-analysis showed that the antiviral treatment group had significantly lower risk of tumor recurrence (odds ratio [OR], 0.59; 95% CI, 0.35–0.97), liver-related mortality (OR, 0.13; 95% CI, 0.02–0.69), and overall mortality (OR, 0.27; 95% CI, 0.14–0.50) than the no antiviral treatment group [295].

Regarding the choice of oral antivirals, a Korean retrospective study demonstrated that antiviral treatment with high-potency NAs (i.e., entecavir and tenofovir) showed significantly longer recurrence-free survival than both antiviral treatment with low-potency NAs (i.e., lamivudine, clevudine, and telbivudine; HR, 0.47; 95% CI, 0.34–0.65) and no antiviral treatment (HR, 0.39; 95% CI, 0.30–0.51) [296]. However, it is not clear whether tenofovir and entecavir have different effects on HCC recurrence in patients receiving curative hepatectomy for HBV-related HCC. Recent Korean retrospective study reported that tenofovir DF therapy was associated with a significantly lower risk of HCC recurrence after curative resection and better overall patient survival compared with entecavir [297]. In contrast, no difference in the risk of death or liver transplantation was observed in a Taiwanese retrospective cohort with hepatectomy for early-stage HCC, despite tenofovir DF therapy being associated with a lower risk of HCC recurrence than entecavir [298]. The prognoses in terms of recurrence and death after curative treatment of HBV-related HCC were not statistically different between the entecavir and tenofovir DF groups in a multi-center retrospective study in Korea [299].

HBV treatment in patients undergoing locoregional therapy for HCC

After transarterial chemoembolization (TACE), approximately 4–40% of patients with HBV-related HCC developed HBV reactivation [283,300-304]. Post-TACE risks of HBV reactivation, flare-up hepatitis, and liver failure due to HBV reactivation were 2.8%, 2.8%, and 0%, respectively, in the prophylactic lamivudine treatment group and 40.5%, 29.7%, and 8.1% in the control group. There was a significant difference between the two groups [302]. After hepatic artery infusion chemotherapy (HAIC), HBV reactivation was reported in 24–67% of patients, which was relatively higher than that after TACE. This finding can potentially be explained by a larger total amount of cytotoxic chemotherapeutic agents due to shorter treatment intervals than TACE [305-307]. Furthermore, prophylactic antiviral treatment is significantly associated with improved long-term survival among patients undergoing transarterial chemotherapy, including TACE and HAIC (10-year overall survival, 26.5% vs. 12.8%) [203]. The prospective Chinese study involved 98 patients with HBsAg-positive/HBV DNA-negative HCC reported that HBV reactivation occurred in 23.4% of patients in the nonantiviral group (11/47, 23.4%), but only 5.9% of patients in the antiviral group (P<0.05) [308].

In patients who underwent external beam radiation therapy (EBRT) for HCC, reactivation and ALT elevation were reported in 0% and 2.3%, respectively, of the lamivudine-prophylaxis group and 21.8% and 12.5% of the control group. The control group had a significantly higher risk of HBV reactivation [283,309]. A recent Korean retrospective study also showed that the absence of antiviral treatment was an independent risk factor for HBV reactivation in EBRT for HCC (OR, 8.34; 95% CI, 2.53–27.47) [310]. Therefore, preventive antiviral therapy is recommended for patients with HCC who are scheduled to receive EBRT. The combination treatment with TACE and EBRT had twice the risk of HBV reactivation compared to TACE treatment alone [311]. In addition, either TACE or combination treatment of TACE with EBRT can reactivate HBV replication in HBsAg-negative/anti-HBc-positive individuals [312,313]. Patients with prior chronic HBV infection who undergo intensive TACE should be closely monitored, with an alternative approach of antiviral prophylaxis against HBV reactivation.

HBV treatment in patients undergoing systemic therapy/immune checkpoint inhibitors for HCC

HBV reactivation was diagnosed in 6.2% of patients treated with sorafenib based on the data of a recently published meta-analysis [283]. A few retrospective studies showed that underlying HBV DNA >2,000 IU/mL was associated with poor prognosis in sorafenib-treated patients, whereas antiviral prophylaxis was significantly related to the improvement of overall survival [314,315].

Immune checkpoint inhibitors can enhance host immunity and consequently have a lower risk of HBV reactivation. However, immune checkpoints can result in severe acute aggravation of hepatitis since it can upregulate antiviral immunity against HBV. Therefore, suppression of HBV replication with antiviral treatment is necessary before use of immune checkpoint inhibitors [316]. The overall pooled HBV reactivation rate was 7.8% in HCC patients receiving immune checkpoint inhibitors, and most of them did not receive antivirals [283]. A Korean retrospective study including 398 HBV-HCC patients receiving both immune checkpoint inhibitors and antiviral prophylaxis demonstrated that two (0.5%) showed HBV reactivation, both of whom showed poor adherence to antivirals [317]. In addition, retrospective studies reported that antiviral prophylaxis is less likely to reactivate HBV regardless of baseline HBV DNA level [318,319]. Thus, antiviral prophylaxis is recommended in HCC patients receiving immune checkpoint inhibitors.

In atezolizumab and bevacizumab combination treatment or lenvatinib, which have recently been approved as the first-line treatment for liver cancer, further studies are needed due to the lack of data on HBV reactivation.

Patients with renal dysfunction or metabolic disease prior to starting treatment

Patients with chronic kidney disease have relatively higher rate of exposure to HBV infection [324]. In Korea, a 4–7% HBsAg-positive rate has been reported among hemodialysis patients [325-329]. When starting NA treatment in patients with chronic kidney disease, the dose must be adjusted according to the creatinine clearance (Table 8). Tenofovir alafenamide fumarate (tenofovir AF) and tenofovir DF are not recommended in patients with creatinine clearance below 15 and 10 mL/min, respectively, without renal replacement. This is also true for besifovir in cases of creatinine clearance below 15 mL/min.

Table 8.

Dose adjustment of NAs for adult patients with altered creatinine clearance

Because NA treatment itself can affect renal function or bone density, it is necessary to select an appropriate drug if there is any risk factor such as long-term steroid use. In addition, patients treated with tenofovir DF should be monitored for bone mineral density during treatment. In a large phase 3 trial comparing tenofovir AF and tenofovir DF over a 96-week treatment period, among those with any risk factor for renal dysfunction (age over 50 years, hypertension, cardiovascular disease, diabetes, or hyperlipidemia), patients treated with tenofovir DF, compared to patients treated with tenofovir AF, had worsening renal function (compared to baseline, median changes in estimated glomerular filtration rate [eGFR] were -5.0 mL/min and -0.3 mL/min, respectively) [330-332]. Additionally, among those with a risk factor for decreasing bone density (female, age over 50 years, Asian, and baseline eGFR <90 mL/min) patients treated with tenofovir DF showed a greater decrease in bone mineral density compared to patients treated with tenofovir AF (-3.29% and 1.23%, respectively) [330-332]. Therefore, it is recommended to avoid the use of tenofovir DF among patients with risk factors for renal dysfunction such as baseline eGFR <60 mL/min, proteinuria, albuminuria (urine albumin: creatinine ratio >30 mg/g), hypophosphatemia (<2.5 mg/dL), uncontrolled diabetes, or hypertension. If patients have a diagnosis of osteopenia or osteoporosis, need to be on chronic steroid treatment, or take other medications that can lower the bone density, other NAs that more weakly affect bone density should be considered over tenofovir DF (Fig. 5) [151].

Figure 5.

Indications for selecting entecavir, tenofovir alafenamide fumarate, or besifovir over tenofovir disoproxil fumarate. AF, alafenamide fumarate. ^*Dose was adjusted if creatinine clearance <50 mL/min, refer to Table 8. ^†Not indicated if creatinine clearance <15 mL/min without dialysis. ^‡Not indicated if creatinine clearance <15 mL/min.

In addition to tenofovir AF, entecavir and besifovir have less of an effect on renal function and bone metabolism. Besifovir, recently approved for use, had been evaluated in a clinical trial for safety in reduction of renal function and bone density. In a multicenter phase 3 trial from Korea, the median decrease from baseline in eGFR was -0.5 mL/min in patients with besifovir and -7.8 mL/min in patients with tenofovir DF. In terms of bone mineral density, patients treated with besifovir showed median decreases from baseline in bone mineral density in spine and hip of 0.33% and 0.44%, respectively, whereas patients treated with tenofovir DF had median decreases of 0.85% and 1.29% [333]. Therefore, besifovir is thought to have a better effect on renal function or bone density compared to tenofovir DF. However, besifovir is not indicated in cases of severely decreased renal dysfunction (eGFR <15 mL/min) due to lack of clinical data.

Patients who developed renal dysfunction or decrease in bone density on treatment with NAs

If patients develop renal dysfunction or decrease in bone density while on NAs, the causative factors must be identified and corrected. Then, the dose must be modified accordingly (Table 8) or a review should be performed for drug change (Fig. 5).

A phase 3 clinical trial compared the safety of tenofovir AF to that of tenofovir AF during 240 weeks. In this trial, patients were given either treatment for 96 weeks each, after which some patients treated with tenofovir DF switched to tenofovir AF, while the remaining switched to tenofovir AF at 144 weeks. Patients who switched from tenofovir DF to tenofovir AF at 96 weeks (tenofovir AF for 48 weeks afterward) showed improved renal function at 144 weeks compared to patients who continued tenofovir DF at 144 weeks (compared to baseline, mean change in eGFR: 4.2 mL/min and -0.9 mL/min, respectively). In terms of bone mineral density in spine and hip, patients who were treated with tenofovir DF showed greater decreases than patients treated with tenofovir AF until week 96. However, after switching to tenofovir AF at week 96 or week 144, bone mineral density recovered to baseline at week 240 in these patients, resulting no significant difference in patients initially treated and then switched and those who maintained tenofovir AF for 240 weeks [334].

In another phase 3 non-inferiority switching trial, patients who switched to tenofovir AF from tenofovir DF had improvements in renal function (compared to baseline, median change 0.94 mL/min vs. -2.74 mL/min, respectively) and bone mineral density (compared to baseline, median change 0.66% vs. -0.51% [hip]; 1.74% vs. -0.11% [spine]) compared to patients who continued tenofovir AF at week 48 [335]. Additionally, in a randomized non-inferiority trial of patients with multidrug-resistant HBV, patients who switched to tenofovir AF from tenofovir DF, compared to patients continuing tenofovir AF, showed a greater increase in median eGFR (compared to baseline, 7.3% vs. 1.9%, respectively) and an improvement in bone mineral density (compared to baseline, 1.8% vs. 0.2%, respectively) [336]. Therefore, it is thought that reduction in renal function and/or bone density while on tenofovir DF can be improved by switching to tenofovir AF.

In a phase 3 trial comparing besifovir with tenofovir D, patients receiving tenofovir DF for 48 weeks showed a decrease in eGFR by -7.8 mL/min. However, renal function as measured by eGFR was recovered to baseline after switching to besifovir and continuing to 144 weeks. In terms of bone mineral density in spine and hip, patients receiving tenofovir DF showed decrease of 1.12% and 0.62%, respectively. After switching to besifovir, bone density was improved and recovered to baseline, showing no significant difference from patients who were initially treated with besifovir. These improvement in renal function and bone density were maintained after 144 weeks of besifovir treatment [337,338]. Therefore, during the NA treatment of CHB, if patients develop renal dysfunction or metabolic bone disease, and/or carry risk factors, appropriate drug change can be an option for overcoming the side effects (Fig. 5).

Reactivation of CHB

Reactivation of hepatitis B indicates recurrence of active necrotizing inflammatory disease in patients in the inactive phase of CHB or those who recovered from previous active infection. Such reactivation can be largely divided into two categories, “exacerbation of chronic HBV infection” for those with positive HBsAg and “relapse of past HBV infection” for those with negative HBsAg and positive anti-HBc [340]. In the latter category, patients who remained in an “occult HBV infection” status can show viral replication triggered by immunosuppression, leading to reverse seroconversion or seroreversion, with redetection of HBsAg [341-343]. Exacerbation of chronic HBV infection is defined in those with seropositive HBsAg as an increase of serum HBV DNA by more than 100 times the baseline level. Relapse of past HBV infection is defined as seroconversion of HBsAg-negative to positive or detection of serum HBV DNA from undetectable level to greater than 100 IU/mL. Hepatitis flare due to HBV reactivation is defined as a ≥3-fold increase in serum ALT level, to a total exceeding 100 IU/L [344,345].

Various rates of reactivation have been reported but are about 20–50% in HBsAg-positive patients receiving anti-cancer chemotherapy. For an accurate diagnosis, liver damage related to chemotherapy, tumor metastasis, or hepatitis secondary to other viruses should be excluded. In many cases, patients are asymptomatic but occasionally present with jaundice or are in various stages such as decompensated liver disease or death [343,346-348]. Typical reactivation is noted by detection of serum HBV DNA during immunosuppression or chemotherapy or elevation of serum ALT after stopping the immunosuppressive therapy. If the reactivation occurs during chemotherapy, it can lead to reduction or discontinuation of chemotherapy, adversely affecting treatment success [349-351]. There are risk factors of hepatitis B reactivation related to the virus, the host, and treatment. Virus factors include serum HBV DNA, HBeAg seropositivity, hepatocyte cccDNA, and PC/BCP mutation prior to treatment; host factors include type of malignant tumor, male gender, young age, and high serum ALT level; and treatment factors include the type and intensity of immunosuppressant or chemotherapy regimen, hematopoietic stem-cell transplantaton (HSCT), and/or solid organ transplantation [352]. The type and intensity of chemotherapy regimen related to the risk of hepatitis B reactivation can be classified into three categories: high risk group (reactivation risk of 10% or more), moderate risk group (reactivation risk between 1–10%), and low risk group (reactivation risk less than 1%) (Table 9) [316,353,354].

Table 9.

Risk of hepatitis B reactivation associated with immune-related therapies

Reactivation of hepatitis B during chemotherapy for lymphoma and other hematologic malignancies

During chemotherapy for lymphoma, hepatitis B reactivation is reported to be frequent, with the rate up to 24–67%. This not only implies that the chemotherapy used for lymphoma is strong enough to cause bone marrow suppression, but also that patients with lymphoma have higher rates of seropositive HBsAg than those without lymphoma [347,355-357]. Rituximab, commonly used in combination with steroids for the treatment of lymphoma, is known to increase the risk of reactivation [358,359]. Rituximab therapy increased the risk of hepatitis B reactivation in patients with non-Hodgkin’s lymphoma who had seropositive HBsAg or seronegative HBsAg/seropositive anti-HBc combination (relative risk [RR], 2.14; 95% CI, 1.42–3.22; P<0.001). Even in patients with seronegative HBsAg/seropositive anti-HBc combination, the use of rituximab therapy was associated with higher risk of reactivation (RR, 5.52) [360]. In a retrospective observational study of lymphoma patients treated with rituximab, hepatitis B reactivation rate was 27.8% in patients with seropositive HBsAg, and the risk could be lowered by the prophylactic antiviral treatment (22.9% [32/140] vs. 59.1% [13/22]; P<0.001). However, hepatitis B reactivation developed in more than 20% of patients who received prophylactic antiviral treatment. In a retrospective study, hepatitis B reactivation rate was as low as 2.4% in patients with seronegative HBsAg/seropositive anti-HBc combination [361]. According to a recent meta-analysis, this rate was 17% in prospective studies and 7% in retrospective studies [362]. In addition, in a prospective observational study of lymphoma patients with seronegative HBsAg/seropositive anti-HBc combination who were treated with rituximabCHOP (R-CHOP) chemotherapy, hepatitis B reactivation and subsequent worsening of hepatitis were common (10.4 vs. 6.4 per 100 person-year). In that study, there were cases where hepatitis worsened during periodic monitoring of HBsAg and serum HBV DNA, and immediate antiviral treatment was safe for reactivation of hepatitis B. In particular, seroreversion of HBsAg was the most important indicator of hepatitis B-related liver damage (100% vs. 28%) [363]. There was a significant difference in the reactivation of hepatitis B in patients with and without prophylactic antiviral therapy (13.3% vs. 60%) during treatment with rituximab [364]. Furthermore, prior to receiving chemotherapy (R-CHOP), screening for hepatitis B in all patients, rather than limiting to high-risk groups, resulted in a 10-fold reduction in hepatitis B reactivation rate and economic and survival benefits [365]. Lamivudine was the most commonly used antiviral agent, but the hepatitis B reactivation rate was significantly lower in the entecavir group than in the lamivudine group (6.3% vs. 39.3%; P<0.05). In a recent, prospective, randomized control study, hepatitis B reactivation was reported to be 0% in the group treated with prophylactic tenofovir DF during rituximab treatment but 10.7% (P=0.09) in the group not receiving prophylactic antiviral therapy [366].

With other hematologic malignancies, if patients are receiving high-intensity chemotherapy prior to HSCT, the risk of reactivation is high [366,367]. In particular, in patients with seropositive HBsAg or seronegative HBsAg/seropositive anti-HBc awaiting high intensity chemotherapy prior to HSCT, antiviral therapy with a high barrier to resistance is recommended [368]. During immunosuppressive therapy or chemotherapy for hematologic disorders, for patients with evidence of hepatitis B infection, prophylactic treatment with lamivudine or entecavir can significantly lower the reactivation rate [369,370].

Meanwhile, in a recent Korean study evaluating 93 multiple myeloma patients with seronegative HBsAg who were treated with daratumumab (anti-CD38), hepatitis B reactivation occurred in 6 (6.5%, 6/93). Subgroup analysis for patients with seronegative HBsAg/seropositive anti-HBc combination showed a higher risk of 12.5% (3/24). Furthermore, special attention is needed because one patient eventually died due to liver failure even though all six hepatitis B-reactivated patients were treated with prompt antiviral therapy [371].

Reactivation of hepatitis B during chemotherapy for solid tumors

Reactivation of hepatitis B in patients with solid tumors occurs in 14–21%, while that of breast cancer is higher at about 41– 70%, which is thought to be related to the high doses of treatment agents and the use of anthracycline-based chemotherapy and steroids [351,372,373]. Steroids not only suppress the immune system, but directly stimulate the replication of HBV, increasing the risk of reactivation. Use of prophylactic antiviral agents in most solid tumors, not only breast cancer, but also lung cancer in which immunotherapy is being used as a first-line treatment, has significantly reduced the rates of hepatitis B reactivation and discontinuation of chemotherapy treatment [374,375].

Reactivation of hepatitis B during immune-related treatments for inflammatory bowel disease (IBD) or rheumatoid arthritis (RA)

The reactivation of hepatitis B also might be associated with the use of tumor necrosis factor-alpha (TNFα) inhibitors (infliximab, etanercept, adalimumab, etc.) for the treatment of IBD or RA [375-380]. For TNF α inhibitors used for the treatment of IBD, hepatitis B reactivation occurred in 22.5–39% of patients with seropositive HBsAg and 5% of seronegative HBsAg/seropositive anti-HBc combination [381]. In cases treated with TNFα inhibitors and disease-modifying antirheumatic drugs (DMARDs) for RA treatment, the rate of reactivation of hepatitis B was around 12.3% in patients with seropositive HBsAg [382]. In another study, reactivation was reported in 39% of HBsAg-positive patients and 5% of antiHBc-positive patients; among those given antiviral prophylaxis, the reactivation rate was significantly lower (23% vs. 62%; P=0.003) [383]. In a meta-analysis of patients treated for RA, hepatitis B reactivation occurred in 14.6% of HBsAg-positive patients and was lower at 9.0% in those treated with prophylactic antiviral therapy [384]. However, the reactivation rate tended to decrease in the TNFα inhibitor treatment group (4.4% vs. 15.6%; P=0.05), but no significant difference was observed from the DMARD treatment group (27.1% vs. 22.4%) [384]. This finding might be related with the high heterogeneity of studies included in this metaanalysis, and further studies are needed. In the same study, hepatitis B reactivation occurred in 1.6% of cases where only anti-HBc was positive [384].

Acute exacerbation of hepatitis B during immunotherapy treatment

Recently, immune checkpoint inhibitors such as anti-PD-1 (nivolumab), anti-PD-L1 (atezolizumab), and anti-CTLA4 (ipilimumab) have been used in various carcinomas including liver cancer. According to a recent retrospective study analyzing patients with multiple solid cancers and some lymphomas, the incidence rates of hepatitis B reactivation of the total patients, HBsAg-positive patients, and HBsAg-negative patients were 0.14% (5/3,465), 1.0% (5/511), and 0.0% (0/2,954), respectively. The hepatitis B reactivation rates were 0.4% (2/464) and 6.4% (3/47) in patients with and without antiviral prophylaxis, respectively [317]. Hepatitis B reactivation is rare in immune checkpoint inhibitor chemotherapy, but prophylactic antiviral therapy can be considered in HBsAgpositive patients. Pembrolizumab among various immune checkpoint inhibitors showed a significant correlation with hepatitis B reactivation [385]. In a retrospective study evaluating 89 patients with hematologic malignancies treated with chimeric antigen receptor (CAR)-T cells, all 19 HBsAg-positive patients received prophylactic antiviral therapy with entecavir, but hepatitis B reactivation occurred in one (5.3%). However, hepatitis B reactivation was not observed in patients with seronegative HBsAg/seropositive anti-HBc combination despite only two of 37 patients receiving prophylactic entecavir therapy. Furthermore, no hepatitis B reactivation occurred during prophylactic antiviral therapy in the 33 patients with seronegative HBsAg/seronegative anti-HBc combination. Therefore, prophylactic antiviral therapy might be beneficial to HBsAg-positive patients treated with CAR-T cells, who are thought to have higher than intermediate risk of hepatitis B reactivation [386]. There are concerns about the possibility of acute exacerbation of hepatitis B in relation to these treatments, but there is insufficient data, and further consideration of prophylactic antiviral therapy is needed.

Start and end points of prophylactic antiviral therapy

When the reactivation of hepatitis B occurs, there is a risk of liver failure or even death. Therefore, prevention is most important. Prior to starting an immunosuppressive therapy or chemotherapy, screening for HBsAg and anti-HBc is necessary. If there is no evidence of HBV infection in the past (with negative HBsAg and anti-HBc), HBV vaccination can be considered. In HBsAg-positive cases, regardless of serum HBV DNA level, antiviral prophylaxis is recommended. Instead of waiting for the serum HBV DNA level to rise, administering an antiviral agent at the start of the immunosuppressive therapy or chemotherapy or 7 days prior to the treatment start date is reported to be more effective [377,387-389]. The end point of the prophylactic antiviral treatment should theoretically be continued until the immune system is adequately recovered, but there is lack of sufficient evidence to suggest a specific end point. It has been reported that the risk of HBV reactivation is high when prophylactic lamivudine is discontinued about 3 months after the end of chemotherapy. The risk is especially higher when the serum HBV DNA prior to the treatment is elevated (≥2,000 IU/mL) [390]. Therefore, when HBV is actively replicating before prophylactic antiviral therapy, following the present CHB treatment guidelines for the discontinuation of antiviral agents might prevent virus reactivation after treatment. However, regardless of the serum HBV DNA level prior to the treatment, reactivation is reported more than 6 months after the completion of chemotherapy, so caution is required. Therefore, antiviral prophylaxis should be maintained for at least 6 months after the chemotherapy is completed, and extension should be considered according to the chemotherapy risk. Especially, for patients receiving chemotherapy involving rituximab, it is recommended to extend the antiviral prophylaxis to at least 12 months after the completion of chemotherapy [391-393]. Even after the end of prophylactic antiviral therapy, it is necessary to closely monitor for relapse for more than at least 12 months.

Meanwhile, as described above, hepatitis B reactivation requires attention because it can occur not only when HBsAg is positive, but also when HBsAg is negative and anti-HBc is positive. In particular, under immunosuppressive conditions, patients who were only positive for anti-HBc had a higher risk of hepatitis B reactivation than patients who were positive for both anti-HBc and anti-HBs [316,394]. Therefore, when HBsAg-negative and anti-HBc-positive patients are treated with a rituximab-containing regimen or HSCT for leukemia, it is necessary to consider prophylactic antiviral therapy because the risk of hepatitis B reactivation is greater than 10%. Meanwhile, in a Korean study analyzing the risk of hepatitis B reactivation in patients who received chemotherapy including rituximab for lymphoma, the risk of hepatitis B reactivation was low in patients with resolved HBV (HBsAg-negative and HBcAb-positive) when anti-HBs titer was high (>100 IU/mL). For these patients, meticulous follow-up for ALT and HBV DNA and prompt commencement of antiviral therapy upon hepatitis B reactivation are a possible strategy [395]. For patients receiving intermediate- or low-risk groups of chemotherapy, HBsAg and HBV DNA should be monitored periodically (every 1–3 months) during and after the chemotherapy, and antiviral treatment should be commenced upon hepatitis B reactivation [316].

Treatment medications

Lamivudine is the most widely studied drug for prophylactic antiviral therapy. It is well known to significantly reduce reactivation, liver failure, and death according to randomized controlled trials of lymphoma patients in Hong Kong and Taiwan [356,367,387,396]. However, lamivudine has been reported to be resistant even during prophylaxis. If the treatment duration is expected to be long, it is necessary to select a therapeutic agent with a high barrier to resistance considering the resistance rate [356]. In a retrospective study of lymphoma patients, the incidence of hepatitis and chemotherapy disruption due to HBV reactivation was significantly lower in the entecavir group than in the lamivudine group [397]. In a randomized controlled study comparing 121 patients treated with RCHOP chemotherapy for lymphoma as entecavir-treated (n=61) and lamivudine-treated groups (n=60), the entecavir group performed better in terms of the risk of hepatitis B reactivation (4% vs. 18%; P=0.001) and the rate of discontinuation of chemotherapy due to exacerbation of hepatitis B (1% vs. 11%; P=0.002) [398]. In a meta-analysis, entecavir prophylaxis was shown to prevent reactivation of hepatitis B more effectively than lamivudine prophylaxis [399]. In a retrospective comparative analysis of 419 CHB patients who received chemotherapy for solid cancers and lymphoma in Korea, HBeAg positivity, serum HBV DNA, and cancer types were identified as risk factors related to hepatitis B reactivation, and a risk stratification tool was developed based these factors. This study also confirmed that entecavir had a better preventive effect than lamivudine or telbivudine in the high-risk group [400]. However, since most prophylactic therapy studies were conducted on lymphoma patients, prospective studies evaluating various malignancies including solid cancers on appropriate antiviral agents and treatment duration for each cancer type and anticancer drug are needed. Considering the failure rate and tolerance rate, entecavir and tenofovir are expected to be relatively safe options. In a retrospective study evaluating patients receiving chemotherapy or immunosuppressant treatment for solid cancers, lymphoma, or rheumatic diseases, the tenofovir AF group (n=11) had similar effects compared to the entecavir group (n=66) in terms of HBV DNA reduction effect (-2.83±1.45 log IU/mL vs. -3.05±2.47 log IU/mL; P=0.86), HBV non-detection rate (78.8 vs. 90.9%; P=0.68), and renal function decline rate (-0.62±11.2 mL/min/1.73 m2 vs. -3.67±13.2 mL/min/1.73 m2; P=0.29), respectively [401]. Therefore, tenofovir AF can also be considered as a safe drug with good preventive effects. Such a conclusion regarding besifovir requires additional data. However, considering the characteristics of the drugs, entecavir, tenofovir AF, and besifovir are expected to be relatively safe choices as a prophylactic antiviral treatment for patients at risk or with renal and/or bone diseases.

Hepatitis C virus (HCV) coinfection

In patients with CHB, the anti-HCV-positivity rate varies from 1.5% to 2.4% in Korea [454,455]. Patients with HBV/HCV coinfection have more severe necroinflammation and fibrosis as well as high risk of cirrhosis and HCC compared to those with a monoinfection [456-459]. Patients with HBV/HCV coinfection should be tested for HBV DNA and HCV RNA before antiviral treatment for each virus. If HCV RNA is detectable, antiviral treatment for HCV (direct acting antiviral [DAA] therapy) should be started. If HBV DNA is detectable, check for ALT, HBV DNA, and presence of liver cirrhosis and treat according to the HBV guidelines, as with HBV monoinfection (Fig. 3).

There is a potential risk of HBV reactivation during DAA therapy or after clearance of HCV. In a meta-analysis, HBV DNA was newly detected or the level increased in 14.1% of patients after administration of DAAs for 4–12 weeks, whereas active hepatitis accompanied by ALT elevation was noted in 12.2% of patients [460]. Therefore, patients with HBV/HCV coinfection should be monitored for the risk of HBV reactivation by monitoring ALT and HBV DNA level during and after DAA therapy. In addition, if HBV is indicated for treatment at the time of starting DAA therapy for HCV, NA therapy for HBV should be started. In patients with a history of cirrhosis or HCC, simultaneous NA therapy for HBV can be considered along with DAA therapy for HCV to reduce the risk of liver failure risk resulting from HBV reactivation [461,462]. Entecavir, tenofovir DF, and tenofovir AF are preferred choices. There was no significant drug-drug interaction between NAs for HBV (entecavir, tenofovir DF, tenofovir AF) and DAAs or HCV. However, when selecting antiviral drugs in patients with HBV/HCV/HIV triple infection, drug-drug interactions should be considered [463,464].

In HBsAg-negative and anti-HBc-positive patients with HCV, HBV reactivation rate (HBsAg positivity) during DAA therapy is very low at 0–0.1% [465-467]. Although HBV reactivation rate during DAA therapy is low, HBsAg and HBV DNA should be tested to check for HBV reactivation in cases of ALT elevation.