Optimizing care of HBV infection and HBV-related HCC

INTRODUCTION

Despite the availability of vaccines and effective antivirals for the prevention and treatment of chronic hepatitis B (CHB), the global burden remains enormous, with around 330,000 deaths related to HBV cirrhosis and 192,000 deaths due to HBV-related HCC in 2019.^1,2

There are 3 major postulated mechanisms for hepatocarcinogenesis in HBV-related HCC, even before the onset of cirrhosis³: (1) cis-mediated insertional mutagenesis of HCC—HCC-associated genes, (2) induction of chromosomal instability by integrated DNA, (3) expression of mutant HBV genes from the persistent integrated form (Figure 1). The continued suppression of the innate and adaptive host immune response also adds to the persistent risk of hepatocarcinogenesis. Antiviral therapy suppresses HBV replication, prevents progression to cirrhosis, and substantially reduces the risk of HCC. However, it does not eliminate the covalently closed circular DNA from infected hepatocytes, necessitating lifelong antiviral treatment in many patients. Studies have also demonstrated that even after more than a decade of successful HBV suppression, the risk of developing HBV-related HCC persists,^4–7 emphasizing the need for the development of a drug that could result in complete eradication of the virus.

FIGURE 1.

Hepatocellular carcinogenesis in patients with CHB. Created by Biorender. com.

MAJOR SOCIETY RECOMMENDATIONS FOR THE INITIATION OF ANTIVIRAL THERAPY IN CHB

In patients with HBeAg-positive CHB without cirrhosis, both the American Association for the Study of Liver Diseases (AASLD)⁸ and the Asian Pacific Association for the Study of the Liver (APASL)⁹ suggest initiating treatment in patients with HBV DNA levels of >20,0000 IU/mL with higher alanine aminotransferase (ALT) values of >2× upper limit of normal (ULN). On the other hand, the European Association for the Study of the Liver (EASL)¹⁰ suggests initiating treatment when the HBV DNA >2000 IU/mL and ALT>ULN and/or at least moderate liver necroinflammation or fibrosis. In patients with HBeAg-negative CHB, AASLD, APASL, and EASL recommend initiating treatment when HBV DNA levels exceed 2000 IU/mL but vary concerning the thresholds for ALT. For patients with a histological examination showing moderate inflammation and fibrosis or significant fibrosis, the AASLD, APASL, and EASL also recommend starting treatment.¹¹ The WHO recommends that patients without cirrhosis above the age of 30 with persistently abnormal ALT and HBV DNA>20,000 IU/mL are at risk of disease progression and should be considered for antiviral therapy.¹²

There are several approved treatment options for CHB. Conventional interferon alpha-2a was the first treatment approved and was used widely in the 1990s. Subsequently, the first nucleoside analog reverse transcriptase inhibitor lamivudine was introduced in 1998, with entecavir and tenofovir introduced around a decade later.¹³ Multiple studies and systematic reviews have demonstrated the efficacy of antiviral therapy in reducing the risk of HCC, both for primary prevention and secondary prevention. A cohort study conducted in South Korea by Choi et al¹⁴ and a retrospective study conducted in China by Yip et al¹⁵ suggested that treatment with tenofovir may be associated with a lower risk of HCC than entecavir. Multiple studies and meta-analysis have since examined this issue, with some concluding that tenofovir was superior to entecavir for preventing HCC¹⁶ and others finding no significant difference between tenofovir and entecavir.^17–19 The difference in findings is likely related to the heterogenous inclusion criteria utilized by these retrospective, observational studies and the potential for residual confounding. However, both tenofovir and entecavir appear to be effective in preventing HCC.

THE INDETERMINATE PHASE OF CHB

Emerging data suggest that a substantial proportion of patients beyond standard treatment criteria for CHB may be at higher risk for liver-related complications. A provocative study demonstrated a higher risk of developing cirrhosis or HCC among patients with CHB who never met AASLD treatment criteria and who were not treated, compared with treatment-eligible patients who received treatment.²⁰ The AASLD classifies the phases of CHB into the immune active phase, the inactive phase, and the immune-tolerant phase.⁸ However, many patients do not fall into any of these well-defined phases. For example, patients with a high HBV DNA but ALT 1–2× ULN or<ULN or high ALT but low HBV DNA do not fall into any of these categories and are classified as being in the indeterminate phase. It was previously unclear if these patients would transition to another phase or were at increased risk of HCC. A retrospective cohort study of 3366 adult treatment-naive adults with CHB determined that 38.7% were in the indeterminate phase at baseline.²¹ After 10 years of follow-up evaluation, up to 53% of patients remained in the indeterminate phase, and 22% transitioned to the immune active phase. This study showed that the majority of patients in the indeterminate phase would either remain indeterminate or progress to the active phase, rather than become inactive eventually. When compared with the patients who remained in the inactive phase, those who remained in the indeterminate phase had a significantly higher 10-year cumulative HCC incidence (4.6% vs. 0.5%; p<0.0001) and a 14 times higher risk of developing HCC (adjusted HR 14.1 (95% CI: 1.30–153.3), p=0.03). Another study of 242 patients in the “gray” or indeterminate phase of CHB who underwent liver biopsy determined that 73% of these “gray zone” patients had significant histological disease,²² lending further weight to the notion that these patients were at risk of liver-related outcomes. Table 1 summarizes recent studies describing the natural history of untreated patients with CHB in the indeterminate phase.

TABLE 1.

Studies on the natural history of untreated patients with chronic HBV in the “indeterminate phase”/“gray zone”

References	Year	Study population	Main findings
Oliveri et al23	2017	153 untreated anti-HbE–positive HbsAg carriers with serum HBV-DNA<20,000 IU/mL and normal ALT values Classified into 3 main groups after follow-up of 1 y:  ○ Inactive carrier [IC]: HBV DNA <2000 IU/mL and normal ALT)  ○ Low viremic active carrier [LV-AC]: HBV DNA>2000 IU/mL at least once but always<20,000 IU/mL and normal ALT  ○ Chronic hepatitis B [CHB]: HBV DNA>20,000 and normal or elevated ALT Thereafter IC and LV-AC were followed up for additional 57.2 months	After the first year diagnostic follow-up:  ○ 13.1% of inactive carriers became classified as patients with CHB, 30% LV-AC and 56.9% IC Thereafter, during the long-term follow-up:  ○ 0% of IC reactivated, 21.8% cleared HBsAg  ○ 54.3% of LV-AC remained stable, 43.5% became IC and 2.2% developed CHB The best single-point CHB and IC diagnostic-accuracies were total-anti-HBc (84.2%, NPV-98.2%) and HBV-DNA/total-anti-HBc/HBcrAg combination (89.5%, 93%-sensitivity, 84.8%-specificity,) respectively.
Bonacci et al.24	2018	287 untreated HbeAg-negative chronic HBV infection with HBV DNA<20,000 IU/mL and ALT<80 U/L 4 main groups:  ○ Inactive carrier (HBV DNA undetectable or<2000 IU/mL and ALT<40 U/L)  ○ GZ-1 (HBV DNA<2000 IU/mL and ALT 40–80 U/L)  ○ GZ-2 (HBV DNA 2000–20,000 IU/mL and ALT<40 U/L)  ○ GZ-3 (HBV DNA 2000–20,000 IU/mL and abnormal ALT 40–80 U/L) Median follow-up of 8.2 y	HbsAg clearance was seen in 44 (15.3%) patients, 25 (18.3%) from the IC group, and 19 (12.6%) from the GZ group Transition to IC state occurred in 40% of patients with GZ DNA fluctuations >2000 IU/mL correlated inversely with transition into IC and HbsAg loss. HbsAg levels were significantly lower in ICs than in patients with GZ (338 IU/mL vs. 5763 IU/mL) HbeAg-negative CHB occurred in only 18 (6.3%) patients with GZ. No patient developed cirrhosis or advanced fibrosis ALT/HBV-DNA fluctuations, and HbeAg-negative CHB development were more frequent in genotype B/C patients, whereas HbsAg loss occurred only in genotype A/D patients
Huang et al25	2021	Retrospective study 3366 adult untreated patients without cirrhosis with CHB with 1 or more years of follow-up evaluation and more than 2 sets of HbeAg, ALT, and HBV DNA within 12 mo Mean follow-up period 12.5 y	38.7% of patients were in the indeterminate phase 52.7% remained indeterminate while 21.7% became immune active after a 10-year follow-up Patients who remained in the indeterminate phase had a higher 10-y cumulative HCC incidence (4.6% vs. 0.5%) and an adjusted hazard ratio for HCC of 14.1, compared to those who remained in the inactive phase Among persistently indeterminate CHB patients, age 45 y and older was associated with an 18 times higher risk for HCC development
Yao et al26	2021	4759 treatment naive patients with CHB Gray zone patients with CHB classified into 4 groups:  ○ GZ-A: HbeAg-positive, normal ALT and HBV DNA<106 IU/mL  ○ GZ-B: HbeAg positive, elevated ALT levels and HBV DNA<2×104 IU/mL  ○ GZ-C: HbeAg negative, elevated ALT and HBV DNA<2 ×10^3 IU/ml  ○ GZ-D: Hbeag negative, elevated ALT, HBV DNA<2×103 IU/mL	1322 patients (27.78%) were in the gray zone (GZ). 33% of patients in the GZ-B had advanced fibrosis while 25.8% had cirrhosis Older age, HbeAg-positive status, and higher ALT levels were independent risk factors of advanced disease in patients with GZ CHB
Chen et al27	2023	Retrospective study 1226 patients with chronic HBV infection (HbsAg positive for 6 mo) All patients underwent US-guided percutaneous liver biopsy Histological necroinflammation grade (G0–4), and fibrosis stage (S0-4) were evaluated according to the Scheuer classification system G0–1 was considered as no or mild inflammation, G2–4 was considered as significant inflammation, and significant fibrosis was defined as S2–4	441 (36%) were in the indeterminate phase Among patients in the indeterminate phase, 202 (45.8%) had mild inflammation (G0-1) and 239 (54.2%) had moderate to severe inflammation (G2-4) 11.6% of patients had a degree of inflammation of G≥3, which was higher than that in the immune-tolerant phase, inactive phase, and HBeAg-negative immune active phases Severity of liver inflammation was significantly associated with liver fibrosis in patients in the indeterminate phase PT, PLT, ALT, and HBV DNA are independent risk factors for significant liver inflammation for patients in indeterminate phase, with AUC of 0.717 (sensitivity 63%, specificity 71%)

Oliveri et al.²³

Bonacci et al.²⁴

Huang et al.²⁵

Yao et al.²⁶

Chen et al.²⁷

Abbreviations: ALT, alanine aminotransferase; CHB, chronic hepatitis B; GZ, grey zone; HBcrAg, hepatitis B core-related antigen; IC, inactive carrier; LV-AC, low viremic active carrier; PLT, platelet count; PT, prothrombin time.

ANTIVIRAL THERAPY FOR THE INDETERMINATE PHASE OF CHB

A multicenter study of 14 sites in the United States, Europe, and Asia analyzed 855 patients to determine the association between antiviral therapy and HCC development in patients with CHB in the “indeterminate” phase.²⁵ The 5-, 10- and 15-year HCC incidence was 2.5%, 3.9%, and 9.4%, among antiviral-treated patients versus 2.7%, 14.7%, and 19.1% among untreated patients (p=0.02). Reduction in HCC risk occurred mainly after 5 years. After adjusting for age, sex, diabetes, platelet count, ALT, HBeAg, and HBV DNA levels, antiviral therapy was also associated with a 70% reduction in HCC risk (adjusted hazards ratio 0.3, 95% CI: 0.2–0.7, p=0.003). Additionally, lower HCC incidence was seen in several patient subgroups that received antiviral therapy: male patients, those older than 35 years, those with positive HbeAg, and those with HBV DNA>1000 IU/mL. In a multicenter, randomized controlled trial in participants aged 25–70 years with substantial viremia (viral DNA>2000 IU/mL) and ALT 1–2×, the ULN randomized participants to oral tenofovir disoproxil fumarate or placebo once daily for 3 years. These participants underwent paired liver biopsies. Liver fibrosis progressed (an increase of ≥1 stage) in 26% of patients in the tenofovir disoproxil fumarate group and 47% of patients in the placebo group (relative risk 0·56, 95% CI: 0·35–0·88; p=0.013), although there was no significant change in necroinflammation.²⁸

These findings highlight an important gap in the current treatment paradigm for CHB, as a significant proportion of patients with CHB fall in the “indeterminate” phase. These patients are potentially at an increased risk of HCC, which may be mitigated by early initiation of well-tolerated antiviral therapy. Another study of treatment-naive patients with CHB showed that the proportion of patients who eventually developed HCC outside the treatment recommendations according to the APASL, AASLD, and EASL criteria were 64.0%, 46%, and 33.5%, respectively.²⁹ Therefore, treatment indications could be potentially expanded to include high-risk patients in the indeterminate phase.

UTILIZATION OF ANTIVIRAL THERAPY FOR HBV AND HBV-RELATED HCC

While existing data suggest that treating patients with CHB with antiviral therapy according to the indications described above (and potentially some in the indeterminate phase as well) can help lower the risk of HCC development and stabilize the underlying liver disease, it remains very poorly utilized. A nationwide US study showed that only 1 in 3 patients with HBV-related HCC in general received antiviral therapy despite having insurance coverage. Lower treatment rates were seen in females, non-Asian patients, patients without cirrhosis, and patients without gastroenterologist/hepatologist or infectious disease specialist care.³⁰ Another cohort study of 1906 patients with HBV or HCV-related HCC who underwent curative surgical resection showed that antiviral therapy initiated before or within 6 months of HCC diagnosis was independently associated with lower mortality in both HBV and HCV-related HCC (adjusted hazards ratio 0.6 and 0.18, respectively). However, the utilization of antivirals for patients with HBV-related HCC in this cohort was only 57%.³¹

One of the reasons for the low uptake rates of antivirals in HBV treatment could be the fact that there is no finite duration of treatment, poor access, and social stigma. Asian countries/territories had lower HBV utilization compared to non-Asian countries, which may be related to local reimbursement barriers.³¹ This emphasizes the need for multidisciplinary collaboration between health care policymakers, primary physicians, hepatologists, and surgeons to improve the access and availability of antiviral therapy in the treatment of patients with HBV.

SURVEILLANCE FOR HBV-RELATED HCC

HCC surveillance is associated with earlier detection, curative therapy, and improved survival.³² The AASLD recommends surveillance with ultrasound examination and alpha-fetoprotein (AFP) every 6 months in people at risk of HCC.³³ In patients with CHB with cirrhosis, both AASLD and EASL recommend surveillance in patients with Childs-Pugh stages A or B and in patients with Childs-Pugh C eligible for liver transplantation, while APASL generally recommends surveillance in all individuals with cirrhosis. In patients with CHB without cirrhosis, AASLD recommends surveillance for patients in endemic countries (men>40 y old, women>50 y old), patients of African descent, family history of HCC, or those with a platelet-age-gender–HBV risk (PAGE-B) score of >/=10. APASL has similar recommendations, while EASL recommends surveillance in individuals at intermediate at high risk of HCC (page-B score ≥ 10).

Risk stratifying people with HCC may help identify high-risk patients who should be carefully surveyed or those who are at low risk of HCC who may forego HCC surveillance. For patients who are antiviral treatment naive, the Chinese University-HCC score comprising 5 factors (age, albumin, bilirubin, HBV DNA, and cirrhosis) divides HCC risk into low, medium, and high risk with the 5-year prediction of HCC development at 98.3%, 90.5% and 78.9%, respectively. The risk estimation for hepatocellular carcinoma in chronic hepatitis score, on the other hand, consists of variables sex, age, HBeAg, ALT, and HBV DNA levels. Patients with the highest risk of HCC had 81% risk at 10 years.³⁴

In patients who are currently on antiviral treatment (mainly entecavir and Tenofovir disoproxil fumarate), the cirrhosis, age, male sex, and diabetes mellitus model was unique in identifying the presence or absence of diabetes as a variable. Scores <8, 8–13, and >13 stratified patients into low, medium, or high-risk subgroups, which correlated with the 3-year cumulative incidences of HCC of 0.27%, 2.4%, and 10.75%, respectively.³⁵ Additionally, the Real-world Effectiveness from the Asia Pacific Rim Liver Consortium for HBV (REAL-B) model developed by Yang et al³⁶ also predicts the overall HCC risk of patients with CHB on antiviral therapy. It consists of 7 variables (male sex, age, alcohol use, diabetes, baseline cirrhosis, platelet count, and AFP), with scores categorized into 3 groups: 0–3 low risk (<1% risk of HCC), 4–7 moderate risk (1%–5% risk of HCC), and 8–13 high risk (>5% risk of HCC). AUROC was >0.80 for HCC risk at 3, 5, and 10 years, and these were significantly higher than other risk models (p<0.001) such as the PAGE-B scores (across all prediction years), and other risk scores developed for untreated patients (Risk Estimation for Hepatocellular Carcinoma in Chronic Hepatitis B, GAG-HCC) or partially treated patients (Chinese University-HCC score). Lastly, a retrospective cohort study³⁷ compared the performance of 10 risk prediction models (Risk Estimation for Hepatocellular Carcinoma in Chronic Hepatitis -B, PAGE-B, m-PAGE-B, Chinese University-HCC, HCCRESCUE, cirrhosis, age, male sex, and diabetes mellitus, APA-B, REAL-B, AASL-HCC, RWS-HCC) for HBV-related HCC in HBV-treated patients and concluded that AUC values for 3-year HCC risk were the highest for risk scores RWS-HCC, APA-B, REAL-B, and AASL-HCC (all >0.80). Five of the models (PAGE-B, m-PAGE-B, cirrhosis, age, male sex, and diabetes mellitus, AASL-HCC, and REAL-B) identified the low-risk group, in which no cases of HCC occurred within the 3-year timeframe of the study period.

Moving forward, physicians may consider individualizing HCC surveillance based on individual patient risk factors (eg, those with multiple risk factors, such as family history, type 2 diabetes, and heavy alcohol consumption). With the rise in the prevalence of obesity and steatotic liver disease, new risk scores may be helpful to reflect the interplay between chronic hepatitis B and host risk factors in determining overall HCC risk.³⁸

BARRIERS TO EFFECTIVE HCC SURVEILLANCE

The actual proportion of HCC surveillance among patients with CHB without cirrhosis remains low at about 32%,^38,39 and only about 39% of these patients with CHB without cirrhosis received abdominal imaging after 6 months of follow-up.⁴⁰ Another study determined that only a quarter or less of patients with cirrhosis (in general) received HCC surveillance.⁴¹

Several factors may contribute to barriers to effective HCC surveillance in patients with CHB. Patient-related factors include the lack of access to medical care, misconceptions regarding screening and medical costs, or insurance-related issues. Physician-related factors include the general lack of knowledge regarding screening and failure to recognize patients at risk. To address patient-related barriers to HCC surveillance, health care policymakers can consider improving general population education on HBV and HCC surveillance through outreach strategies via mail or social media platforms. Physicians must also be kept abreast and updated with the latest guidelines. Reminders for HCC surveillance can be set on electronic medical records with alert systems if surveillance appointments have been missed.³⁸

Effective HCC surveillance may also be compromised by the limitations of ultrasonography. The sensitivity of ultrasonography in the detection of early-stage HCC (defined by Milan criteria) is only 47%, and when combined with AFP, only up to about 63%.⁴² This has prompted several groups to identify other surrogate markers in the prediction of HCC development. HBcrAg is a good surrogate marker of covalently closed circular DNA levels and transcriptional activity, and a study showed that HBcrAg >2.9 logU/mL was independently associated with HCC development and was better than HBV DNA in terms of predictive power for HCC development.⁴³ Another biomarker such as protein induced by vitamin K absence-II, has been described in relation to HCC,⁴⁴ with a recent consensus expert panel from Asia-Pacific agreeing that protein induced by vitamin K absence-II was beneficial in the detection of early HCC when combined with AFP,⁴⁵ although a widely accepted cutoff value has not been established. qHbsAg was also studied in the Taiwanese population for its utility as a biomarker to predict HCC risk, and it was found that in patients with HBV DNA level<2000 IU/mL who had a similar risk of HCC, HBsAg level ≥1000 IU/mL was identified as a new independent risk factor for HCC development.⁴⁶ However, more data are required before these emerging biomarkers can be used to risk-stratify patients with CHB in routine clinical practice.

CONCLUSIONS

In summary, antiviral therapy is associated with substantially improved outcomes for people with CHB but is underutilized. Given the increasing evidence that some patients with CHB in the indeterminate phase have a higher risk of HCC development, treatment indications may be expanded to include select higher-risk patients. Increased efforts are required to improve the utilization of antiviral therapy and HCC surveillance in people with CHB.

FUNDING INFORMATION

Daniel Q. Huang receives funding support from the Singapore Ministry of Health’s National Medical Research Council (MOH-001370).

CONFLICTS OF INTEREST

Daniel Q. Huang advises Gilead. The remaining authors have no conflicts to report.