Abstract
There is a strong association between the distribution of Hepatitis B virus (HBV) carriers and the incidence of hepatocellular carcinoma (HCC). About 60% of HCC in Japan is caused by viral hepatitis. Ten to 15 percent of hepatitis virus-related HCCs derive from HBV. Recently, antiviral therapy against HBV has developed, and interferon therapy and nucleos(t)ide analogues (NAs) are currently the standard of care. NAs exhibit antiviral activity by inhibiting DNA polymerase and suppressing HBV replication. NAs are highly effective in suppressing HBV-DNA and improving alanine aminotransferase. The long-term treatment goal for chronic hepatitis B is HB surface antigen (HBsAg) loss. However, the number of patients who achieve HBsAg loss by NA (i.e., functional cure) is low and there have been cases of HCC incidence during (or after) NA therapy. In this article, we review the efficacy of NA therapy in suppressing HBV-derived carcinogenesis.
Hepatitis B virus (HBV) is a hepatitis virus first discovered in 1965 and belongs to the family Hepadnaviridae (1). The HBV gene is a circular, incomplete double-stranded DNA of approximately 3.2 kb. During replication, the gene is replicated using RNA as an intermediate and HBV’s own reverse transcriptase (1). HBV can only infect hepatocytes, and sodium taurocholate cotransporting peptide (NTCP), which was reported as an HBV receptor in 2012, is specifically expressed in hepatocytes, which may determine hepatocyte targeting by HBV (2). HBV enters the cell by binding to NTCP on the cell surface. The HBV genome, which has an incomplete double-stranded DNA structure of 3.2 kb, is subsequently transported to the nucleus, where it is converted into covalently closed circular DNA (cccDNA) by nuclear enzymes. The cccDNA, also called minichromosome, is thought to be almost as stable as the host DNA in the chromosome in the absence of cell division, and remains in the hepatocyte in an extremely stable state for a long time, which is the main reason why complete HBV elimination is difficult (3).
There is a strong association between the distribution of HBV carriers and the distribution of hepatocellular carcinoma (HCC) incidence. In a study in Taiwan, the incidence of HCC in HBV carriers showed a relative risk ratio more than 200 times higher than that in non-carriers (4, 5). In addition, HB vaccination significantly reduced the incidence of HCC in children (6). Furthermore, a prospective study of 3653 cases in Taiwan showed that the patient’s serum HBV-DNA level determines the risk of future liver carcinogenesis (7). In other words, HBV infection and the HBV virus itself are potent risk factors for liver carcinogenesis.
Approximately 60% of HCC in Japan is said to be caused by viral hepatitis (8). Although the proportion of viral hepatitis has decreased, it is still high. Ten to 15% of virus-related HCCs derive from HBV (8). In addition, the development of HCC is the greatest poor prognostic factor in cases of chronic hepatitis B (CHB) (1). Antiviral therapy against HBV has recently developed and is now the standard of care. This antiviral therapy can be divided into two main types: interferon (IFN) therapy and nucleos(t)ide analogues (NAs) therapy. Compared to NAs, IFNs are administered for a limited period and are said to have a high HB surface antigen (HBsAg)-lowering effect and a high HBsAg negative conversion rate, making them an excellent treatment if the patients who are expected to benefit from IFN therapy are selected (9). Pegylated IFN monotherapy is the first choice for patients who wish to avoid long-term continuous administration of NAs, such as young patients and those who wish to have a baby (9). The enzyme DNA polymerase is involved in the replication of HBV in hepatocytes, and NA has an antiviral effect by inhibiting DNA polymerase and suppressing the replication of HBV. NAs lead to the quiescence of hepatitis by strongly inhibiting viral replication, and thus NAs are highly effective in suppressing HBV-DNA and improving serum alanine aminotransferase (ALT) levels. It has been shown that the earlier ALT normalizes after initiation of NA therapy, the lower the rate of HCC incidence (10). NAs are an easy-to-administer formulations because they are oral and have few side effects, but the problem is that the duration of treatment is often prolonged. The negativity ratio of HBsAg level with NA therapy is low, at approximately 5-10% over 10 years (11, 12).
In Japan, lamivudine (LAM) was approved in 2000, but the high frequency of drug-resistant viruses has been a problem since its introduction (13). Drug-specific mutations in the polymerase region changes the steric structure of the polymerase, making it difficult for NAs to bind to the polymerase, thereby weakening the effect of LAM. In 2004, adefovir (ADV) was approved for the treatment of CHB patients with LAM resistance. In 2006, entecavir (ETV) was approved, and has been used as the standard treatment for NA in clinical practice for more than 10 years because of its very low frequency of drug-resistant virus emergence and good antiviral efficacy. In 2014, tenofovir disoproxil fumarate (TDF), which, like ETV, has a very low frequency of drug-resistant virus emergence and a strong antiviral effect, was also approved. In late 2016, tenofovir alafenamide (TAF), a prodrug of TDF, was approved (14). Since then, no new drugs for CHB have been approved in Japan. Current Japanese hepatitis B guidelines recommend Peg-IFN or NAs (ETV, TDF, and TAF) for CHB patients with HBV-DNA ≥2,000 IU/ml (3.3 LogIU/ml) and ALT ≥31 U/l (regardless of whether the patient is HBeAg positive or negative). In the case of cirrhosis, NAs (ETV, TDF, and TAF) are recommended if HBV-DNA is positive, regardless of ALT level and HBeAg positivity/negativity (15). These NAs have high antiviral efficacy and low resistance mutation, but they are not effective against HBV-cccDNA, the origin of replication, and require long-term administration. The long-term treatment goal for CHB is HBsAg loss (16). However, the number of patients who achieve HBsAg loss by NAs is low, so the more realistic short-term goals are the sustained normalization of ALT levels, HBeAg loss, and suppression of HBV-DNA levels.
However, cases of hepatocarcinogenesis during (or after) NA therapy in patients with CHB have been observed (17). In this review of the literature, we describe the efficacy of NA therapy in suppressing HBV-derived hepatocarcinogenesis and the characteristics of cases in which hepatocarcinogenesis occurred during NA treatment.
Inhibitory Effect of NAs on Carcinogenesis
Unlike IFN, NAs have almost constant antiviral and hepatic function-improving effects. In addition, since the oral administration of NAs is basically continued for a long period after the start of administration, it is easier to examine their hepatocarcinogenesis inhibitory effect than that of IFN. Although many studies have examined the inhibitory effect of NAs on hepatocarcinogenesis, randomized controlled trials (RCTs) are scarce (18). Liaw et al. conducted a 2:1 RCT of LAM (n=436) versus placebo (n=215) and found that 3.9% of patients in the LAM group and 7.4% in the placebo group developed HCC, with a significantly lower cumulative HCC incidence rate in the LAM group [p=0.047, hazard ratio (HR)=0.49] (18). Drug resistance of LAM is a problem, and in this study, the YMDD drug-resistant mutation of LAM was observed in 49% of LAM patients after the start of treatment. However, as the antiviral and liver function improving effects of NAs became clear, it became ethically problematic to conduct a long-term RCT, and the discussion on the inhibitory effect of NAs on liver carcinogenesis was inconclusive.
Subsequently, ETV, which has an extremely low rate of drug-resistant virus emergence, came into use, and although not an RCT, the results of a study comparing HCC incidence rates between ETV-treated and control groups matched by background using propensity scores (PSs) were reported (19). In this study, among 316 cases in each PS matched group, 1.9% in the ETV group and 22.7% in the control group developed HCC (p<0.001, HR=0.37). Comparison of cumulative HCC incidence rates by background liver stage showed that the ETV group was not clearly effective in suppressing HCC incidence in patients with chronic hepatitis, but the ETV group had a significantly lower rate of HCC incidence in patients with cirrhosis (p<0.001). In addition, a comparison of HCC incidence rates stratified patients between three groups of hepatocarcinogenesis risk classified using a risk scoring system (Chinese University of Hong Kong-HCC, CU-HCC): scored by age, albumin, bilirubin, HBV-DNA, and cirrhosis (20), showed a clear inhibitory effect on HCC development as the risk score worsened (19).
The differences between ETV and TDF are controversial (21). TDF has antiviral effects comparable to those of ETV, with higher IFN-λ3 production capacity and stronger HBsAg lowering activity than ETV (22). IFN-λ exhibits potent antitumor activity in mouse cancer models, including liver cancer, and it is speculated that this antitumor activity contributes to the differential risk of developing HCC (23). The first large-scale study comparing the difference in hepatocarcinogenesis inhibition between ETV and TDF was that by Choi et al. in Korea using a national medical database, in which PS matched analysis showed that the TDF group had a significantly lower risk of HCC incidence than the ETV group with an HR of 0.68 (24). However, Kim et al. analyzed the cumulative HCC incidence rate in a multicenter hospital cohort consisting of Korean patients who started ETV or TDF as first-line treatment from 2012 to 2014 and showed that there was no statistically significant difference between the ETV and TDF groups in both PS matched analyses and inverse probability of treatment weighting (IPTW) analyses (25). Since the report by Choi and Kim et al., numerous cohort studies have compared the hepatocarcinogenesis inhibitory effects of ETV and TDF. Some of them showed the superiority of TDF and many others found the two to be equivalent, and no definitive conclusion has been reached. A recent meta-analysis of 15 studies (16,101 patients given TDF and 45,686 given ETV) reported that TDF reduces the incidence of HCC by approximately 20% compared to ETV, but also noted the need for RCTs (26). Another meta-analysis (6,979 patients given TDF and 35,960 patients given ETV) also reported that TDF significantly suppressed hepatocarcinogenesis compared to ETV, especially in HBe antigen positive cases (27). Since most previous studies support the equivalence of ETV and TDF or the superiority of TDF, and none show the superiority of ETV, further studies are needed to identify subsets of patients that may benefit from TDF or TAF.
In a study comparing the carcinogenic inhibition effect of ETV (n=1,525) and TAF (n=285) in NA treatment-naïve CHB patients, there was no significant difference in the cumulative HCC incidence rate between the two groups (p=0.255), and no significant difference was found by PS matching (p=0.953) or IPTW analyses (p=0.743) (28). In a study comparing the carcinogenic inhibitory effect of TAF (n=502) and TDF (n=2,245) in patients with CHB without prior NA treatment, there was no significant difference in the cumulative HCC incidence rate between the two groups (p=0.30), nor was there any significant difference using PS matching analysis (p=0.60) (29). In this study, there was also no significant difference in virological response or ALT normalization rate between the two groups. Because TAF is more efficiently taken up into hepatocytes than TDF, the therapeutic dose of TAF 25 mg was less than that of TDF 300 mg. Therefore, the blood concentration of tenofovir in TAF is reduced to approximately 10% of that in TDF, and renal impairment and bone mineral density loss are reduced, and safety is considered to be high. Therefore, in clinical practice, the number of cases switching from TDF to TAF is increasing in Japan (30).
However, adherence of patients to NA medication is also clinically important. Medication adherence has been shown to be significantly associated with HCC development, HCC-related death, and liver disease-related events (31).
HCC Incidence Under NA Therapy
Papatheodoridis et al. reported on the risk of hepatocarcinogenesis during NA therapy (32). In 1,815 patients on ETV or TDF for CHB, platelet counts, age, and sex, were risk factors, and the concordance index value of the PAGE-B score (platelet, age, and sex combined risk score) was 0.81, which is a high and useful score. However, since this risk score was based on a study of Caucasians, and the risk of HCC incidence in CHB varies widely by race (33), it was not clear whether it could be used for races in Asia, where the majority of worldwide CHB patients are located. Therefore, Kim and colleagues reported a modified PAGE-B score that can be used with the highest accuracy in Asian racial groups undergoing NA therapy for CHB (34). In addition to age, sex, and platelet counts in PAGE-B, albumin level was included as a risk factor in this study, which was also validated (35). In addition, Risk Estimation for Hepatocellular Carcinoma in Chronic Hepatitis B (REACH-B), Individual Prediction Model (IPM), CU-HCC, Guide with Age, Gender, HBV DNA, Core Promoter Mutations and Cirrhosis-HCC (GAG-HCC), and Nomogram-HCC (NGM-HCC) are also used (36-38). Recently, artificial intelligence has also been vigorously used to predict liver carcinogenesis during NA treatment (39).
Significance of HB Core-related Antigen for Carcinogenesis During NA Treatment
NAs have been shown to suppress hepatocarcinogenesis, but they do not eliminate carcinogenesis. Currently, it has been shown that NAs can maintain low levels of HBV-DNA in many CHB patients. But why hasn’t liver carcinogenesis been eliminated in these patients? Focusing on serum HB core-related antigen (HBcrAg), which is said to reflect the amount of HBV in the liver, a study on the relationship between HBcrAg and hepatocarcinogenesis during NA therapy has been reported (40). The study included 601 HBeAg-negative patients, and examined the HCC incidence rate based on HBcrAg levels at the start of NA therapy and one year after, and classified the patients into three groups according to changes in HBcrAg levels during the course of NA treatment, and the group with persistently high HBcrAg level had a significantly higher rate of HCC incidence than the other groups (40). The results suggest that even if HBV-DNA levels are apparently low upon NA therapy, the carcinogenic effect of NAs may not be exerted if HBcrAg levels are high. Another report showed that in HBeAg-negative cases, HBcrAg levels were associated with HCC incidence in cases with HBV-DNA levels between 2000-19999 IU/mL and normal ALT levels (41). HBcrAg is an HBV-related marker developed in Japan and is said to reflect the amount of cccDNA in hepatocytes, the template for HBV transcription (42-45). HBcrAg is also said to correlate with the transcriptional activity from cccDNA (42-45). Even if HBV replication is inhibited by NAs, it is presumed that carcinogenic potential may remain if the amount of cccDNA or transcriptional activity in hepatocytes remains. Therefore, to achieve further carcinogenesis inhibition, novel therapies that significantly reduce cccDNA in the liver are desirable and are currently underway. Recently, a highly sensitive HBcrAg test (iTACT method) has been developed, which may be useful for monitoring HBV reactivation (46) and as an indicator of HCC incidence risk during NA therapy (47). Measurement using the iTACT method, a more sensitive assay, showed a significant difference in HBcrAg levels between the HCC- and non-HCC groups after HBsAg seroclearance (47).
A recent report proposed the concept that by discontinuation of NA after 4 or more years of continuous therapy and when HBsAg has decreased to some extent, HBsAg disappears along with ALT flare (48). This means that HBV-specific T cells were induced by NA discontinuation, and the elimination of HBV-infected cells proceeded. In fact, in patients with HBsAg-positive HCC, the use of immune checkpoint inhibitors is unlikely to cause ALT flare under NA therapy. Therefore, in the absence of NA therapy, induction of T cells by immune checkpoint inhibitors or therapeutic vaccines or activation of innate immunity by TLR7/8 agonists may promote HBcrAg reduction and HBV elimination, which await further investigation in clinical trials. However, since there is a risk of developing severe hepatitis after NA discontinuation, close monitoring is necessary (48). The Tanaka Group of the Ministry of Health, Labour and Welfare in Japan reported that the risk of hepatitis relapse is expected to be 10-20% after NA discontinuation and discontinuation can be considered if: 1) NA therapy is used for at least 2 years; 2) HBeAg is negative and HBV-DNA is not detected in the blood; and 3) HBcrAg <3.0 logU/ml and HBsAg <80 IU/ml (49).
Direct Carcinogenic Effects of Hepatitis B Virus
Most hepatitis B carcinogenesis occurs in patients with cirrhosis, but occasionally carcinogenesis is seen in young patients with chronic hepatitis or inactive carriers who have not yet developed cirrhosis, suggesting the existence of a direct mechanism of carcinogenesis by HBV itself that is not caused by chronic inflammation.
It has been reported that HBx protein, one of the proteins produced by HBV, contributes to carcinogenesis by activating transcription factors, such as NF-
B, stimulating cytokine production, such as TNFα and IL-6, and affecting host cell growth and apoptosis (50). It has also been reported that pre-S2, a part of the open reading frame encoding for HBsAg, is deleted at the onset of HCC, suggesting that the pre-S deletion protein is involved in carcinogenesis by inducing endoplasmic reticulum stress (51). In addition, incorporation of the HBV genome into the host genome is also thought to cause genetic abnormalities that may be associated with carcinogenesis, although cases of carcinogenesis despite a good course of NA therapy are extremely rare. However, it is currently difficult to predict such carcinogenesis.
Liver Fibrosis and Carcinogenesis During NA Therapy, Insulin Resistance and Carcinogenesis, and Prevention of Recurrence After Complete Remission for HCC
Mac-2-binding protein glycosylation isomer (M2BPGi) was developed in Japan as a serum biomarker to diagnose liver fibrosis progression by blood test alone (52). M2BPGi well reflects the degree of liver fibrosis in CHB and is also associated with seroconversion of HBeAg (53, 54). M2BPGi is associated with the progression of liver fibrosis during NA therapy in CHB and can predict the development of HCC, although the presence of cirrhosis is an established predictor for HCC incidence. Regarding the risk of HCC incidence, Ichikawa et al. reported a significantly higher cumulative HCC incidence rate at M2BPGi values of 0.71 cutoff index (COI) or higher in NA-naive CHB patients (55) and Cheung et al. at 0.69 COI or higher in CHB patients during NA therapy (56). Furthermore, when combined with HBcrAg, cases with high risk for HCC incidence are more clearly identified (57). M2BPGi can also be a predictor for HCC recurrence after complete resection in patients with HBV-related HCC (58, 59).
Metabolic syndrome (MS) is associated with a higher rate of insulin resistance (60); MS is a poor prognostic factor in CHB patients receiving NA (61). Furthermore, insulin resistance is a risk factor for hepatic carcinogenesis (62). In our study, fasting blood sugar (FBS, p=0.0227), HBeAg positivity (p=0.0067) and HBsAg (p=0.0497) were extracted by multivariate analysis as factors correlated with intrahepatic cccDNA, with HBcrAg levels showing a significant trend (p=0.0562) (45). Elevated FBS due to insulin resistance may induce hepatocarcinogenesis because of increased levels of cccDNA in the liver (45).
In addition, the impact of NA is also significant in the prevention of recurrence after complete remission of HCC, and the effect of NA in preventing recurrence when compared with NA-naive patients has been demonstrated (63, 64). A meta-analysis (16101 patients given TDF and 45686 given ETV) comparing the impact of TDF and ETV on recurrence after complete remission of HCC showed that TDF significantly suppressed late recurrence compared to ETV, but not early recurrence (late recurrence: HR=0.58, 95%CI=0.45-0.76, early recurrence: HR=0.88, 95%CI=0.76-1.02) (65).
Development of New Drugs and Inhibition of Liver Carcinogenesis
The short-term goal of antiviral therapy for persistently HBV-infected patients is, as mentioned above, sustained normalization of ALT (<30 U/l), HBeAg-negativity and HBeAb-positivity, and inhibition of HBV-DNA replication, while the long-term goal is HBsAg loss, or functional cure (16). HBsAg loss has been shown to inhibit the progression to cirrhosis and hepatocarcinogenesis (66).
The rate of complete HBV elimination by conventional NA and PEG-IFNα therapy alone remains low. Intense basic research has elucidated the mechanisms of HBV replication and the role of the host immune system in HBV elimination, and the development of novel anti-HBV drugs (drugs that block each pathway of the HBV life cycle) targeting processes different from those of existing drugs is underway worldwide. In hepatitis C, direct-acting antiviral agents (DAAs) that inhibit NS3, NS5A, and NS5B viral proteins involved in viral replication have been developed, and their combination therapy has shown dramatic efficacy (67, 68). In hepatitis B, however, only polymerase and X protein are involved in viral replication. Although NA is a potent inhibitor of polymerase and thus significantly reduces the amount of HBV-DNA in the blood, it has no direct effect on cccDNA in the nucleus of hepatocytes, which is the starting point of HBV replication. There are two strategies for new drug development in CHB: DAAs, which target the virus itself, and host-targeting antiviral agents (HTAs), which target host factors, including immunity, to eliminate residual cccDNA in hepatocytes, the ultimate goal in curing HBV infection (16, 69). Elimination of residual cccDNA in hepatocytes is associated with HBsAg loss and is expected to suppress carcinogenesis beyond NA. However, most of the new HBV therapeutics currently under development are up to phase II, and the long-term prognosis of carcinogenesis and other outcomes with new HBV therapeutic intervention is unknown. Future studies are awaited.
Closing Remarks
The article outlines NA therapy and hepatocarcinogenesis in CHB. It is a fact that advances in NA therapy have drastically reduced the number of cases of hepatocarcinogenesis and greatly improved the prognosis. However, it is also true that there are cases of hepatocarcinogenesis during NA therapy, and we must strive for early detection of these cases. In addition, new HBV drugs are expected to be introduced into clinical practice in the near future, but their efficacy in inhibiting carcinogenesis is unknown, and further studies in the clinical practice are awaited. Schematic explanation of HBV-related HCC incidence is shown in Figure 1.
Factors contributing to Hepatitis B virus-related hepatocellular carcinoma.
Acknowledgements
The Authors gratefully thank all medical staff in our department for their significant help.
Footnotes
Authors’ Contributions
Writing – original draft: H.N. Writing – review & editing: all Authors. Final approval: all Authors.
Conflicts of Interest
The Authors have no conflicts of interest to declare in relation to this study.
- Received August 5, 2023.
- Revision received September 8, 2023.
- Accepted September 11, 2023.
- Copyright © 2024 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).







