Introduction
Chronic hepatitis B virus (CHB) infection continues to be a substantial public health burden accounting for 30% of all deaths from cirrhosis and 40% of all deaths related to hepatocellular carcinoma (HCC) globally.1 Antiviral treatment in time is effective for reducing the complications associated with chronic hepatitis B virus (HBV) infection.2 Generally, the necessity of antiviral treatment for patients with CHB is based on comprehensive consideration of the levels of alanine aminotransferase (ALT) and serum HBV DNA, the severity of liver disease, age, and family history of liver cirrhosis/HCC.3–5 Other factors, such as the effectiveness of antiviral drugs, patient willingness, and compliance, also influence the final decision on antiviral therapy.
For CHB patients with normal ALT levels, especially those with negative hepatitis B e antigen (HBeAg), it has long been considered that treatment can be postponed for no or slight liver injury and a benign prognosis.6–8 However, new evidence has changed that traditional stereotype. Significant necroinflammation and/or liver fibrosis can exist in CHB patients with normal ALT, indicating the necessity of treatment.3,4 If untreated, HBeAg-negative CHB patients with normal ALT and a high viral load (≥2,000 IU/mL) are more likely to have an increased risk of HCC and death/transplantation than treated patients with abnormal ALT.9 A previous study has demonstrated that HBV DNA integration and clonal hepatocyte expansion were present across the disease phases, underscoring that the potential risk of HCC development persists, even in patients with early-stage chronic HBV infection, a stage previously considered benign.10
Recently, the indications for antiviral therapy have been further expanded in China with the aim of reducing the risk of HBV-related advanced liver events. Improvement of the effectiveness and accessibility of antiviral drugs, decreased treatment costs and the early completion of the World Health Organization plan to eliminate the global harm of viral hepatitis also promote changes in the management of CHB patients. In the latest Chinese Guidelines for the prevention and treatment of chronic hepatitis B (version 2022), the threshold of ALT for initiating treatment was lowered to 30 U/L for men and 19 U/L for women; meanwhile, patients with normal ALT and positive DNA should be treated if they are over 30 years of age, have a family history of cirrhosis/HCC, or have significant liver necroinflammation/fibrosis by invasive or noninvasive examination.11
However, there is still a lack of clinical data on the benefits of antiviral therapy in HBeAg-negative CHB patients with normal ALT levels, as those patients have been discouraged from treatment for quite some time. In this study, we aimed to determine the treatment efficacy and prognostic improvement after a long course of nucleos(t)ide analog (NA) therapy in noncirrhotic CHB patients with normal ALT levels, negative HBeAg, and positive HBV DNA, so as to provide direct and reliable evidence for expanding the indication for antiviral treatment in such a controversial population.
Methods
Patient selection
This was a single-center, retrospective cohort study. CHB patients were enrolled at the hepatitis clinic of West China Hospital of Sichuan University between February 2017 and March 2020. Enrollment criteria were: (1) detectable hepatitis B surface antigen (HBsAg) for at least 24 weeks, with negative HBeAg and positive hepatitis B e antibody (HBeAb); (2) at least three values of ALT below the upper limit of normal (ULN) within 1 year and at least 3 months apart. The upper limit of normal (ULN) was 50 U/L for men and 40 U/L for women; (3) detectable serum HBV DNA (≥20 IU/mL); (4) no signs of cirrhosis or liver nodules; and (5) HBV treatment-naïve patients, without previous NA or interferon treatment. Exclusion criteria were (1) <18 years of age; (2) pregnant; (3) coinfection with human immunodeficiency virus (HIV), hepatitis C virus (HCV), or hepatitis D virus (HDV); and (4) confirmed liver cirrhosis, HCC or other liver diseases, such as alcoholic, autoimmune, or biliary liver disease. The design of this study was consistent with the principles of the Declaration of Helsinki and Good Clinical Practice Guidelines and was approved by the Institutional Review Board of the West China Hospital of Sichuan University. The clinical trial registration number is ChiCTR2100050064.
Treatment and monitoring during follow-up
Grouping was based on whether patients had initiated antiviral treatment. Patients in the treated group self-administered entecavir (ETV) 0.5 mg, tenofovir disoproxil fumarate (TDF) 300 mg, or telbivudine (LDT) 600 mg per day. Patients in the untreated group received no treatment, including NAs or interferon alpha. Abdominal ultrasonography and serum HBV DNA, aspartate aminotransferase (AST) and ALT levels, platelet (PLT) count and alpha-fetoprotein (AFP) were measured at baseline and every 6 months. All data was sourced from the laboratory information management system of West China Hospital of Sichuan University.
Outcomes
The primary outcome was the change in serum HBV DNA, including virological response, breakthrough, and relapse. The virological response was an undetectable serum HBV DNA during therapy. Virological relapse was a serum HBV DNA>2,000 IU/mL after stopping treatment in patients who achieved virological response. Virological breakthrough was >1 log10 (10-fold) increase in serum HBV DNA from the nadir during therapy in patients adherent to treatment and with an initial virological response.12
Secondary outcomes included change in aspartate aminotransferase to platelet ratio index (APRI) and fibrosis-4 index (FIB-4) scores, and the incidence of liver nodules, liver cirrhosis, and HCC. Liver cirrhosis was suspected when platelet (PLT) <100×109/L (excluding other causes), APRI>2.00, FIB-4>1.45, or ultrasound-suggested signs of cirrhosis and confirmed by liver biopsy, digestive endoscopy, computed tomography (CT), or magnetic resonance imaging (MRI). Patients received regular HCC surveillance by abdominal ultrasonography and serum alpha-fetoprotein (AFP) tests. HCC diagnosis was based on histological examination and/or typical features by dynamic CT and/or MRI.
APRI and FIB-4 are noninvasive serological indices to assess liver fibrosis and cirrhosis. APRI=AST (ULN)/PLT (109/L)×100. The ULN of AST was 34 IU/L for women and 36 IU/L for men, as previously reported.13 FIB-4=(Age [years]×AST [U/L])/(PLT [109/L]×ALT[U/L]1/2). Liver fibrosis staging: the predefined thresholds APRI<1.0 and FIB-4<1.45 were used to reject advanced fibrosis, and APRI>2 and FIB-4>3.25 to rule in cirrhosis. Both indices have indeterminate zones (i.e. APRI 1.00–2.00, and FIB-4 1.45–3.25) where the stage of liver fibrosis is unclassifiable.14
Serum assays
Serum biochemical indexes were measured by standard procedures (Olympus AU5400, Olympus Corporation, Tokyo, Japan). Serum HBsAg levels were measured with an Elecsys HBsAg II Quant Assay (Roche Diagnostics, Penzberg, Germany). Serum HBeAg status was determined by an electrochemiluminescence immunoassay (Roche Diagnostics, Indianapolis, IN, USA). Serum HBV DNA concentration was quantitatively determined with Cobas TaqMan assay kits (Roche Diagnostics, Branchburg, NJ, USA) with a lower limit of detection of 100 IU/mL.
Statistical analysis
Continuous variables were reported as median and interquartile range (IQR) and categorical variables were reported as absolute and relative frequencies. The frequencies and distributions of categorical variables were compared with chi squared and Fisher’s exact tests. Continuous variables were compared with Student’s t-tests. The Wilcoxon signed-rank test was used to compare paired samples. The cumulative incidence of liver nodules, cirrhosis, and HCC were estimated by the Kaplan-Meier method, and statistical significance was determined using log-rank tests. Univariable Cox proportional hazards models was used to determine the hazard ratio (HR) of liver nodules, liver cirrhosis, and HCC. A two-sided p<0.05 was considered statistically significant. The analyses were performed with SPSS for Windows, version 26.0 (IBM Corp., Armonk, NY, USA).
Results
Patient characteristics
From February 2017 to December 2020, 529 CHB patients with persistently normal ALT were screened at West China Hospital of Sichuan University. A total of 194 patients were included and divided into a treated group (n=127) and an untreated group (n=67). During follow-up, treatment was interrupted in 40 patients (31.5%) in the treated group, i.e. the treatment-discontinued group. The remaining 87 patients were the treatment-continued group. Figure 1 is a flow diagram of patient selection and grouping. Of the 194 patients, the median duration of follow-up was 54 months. Their median age was 43 years, female patients accounted for a higher proportion (52.6%), and most (80.9%) had HBV DNA concentrations ≥2,000 IU/mL (approximately 3.3 log10 IU/mL). About two-thirds of the patients (67.8%) received ETV treatment. The untreated group was the youngest, with only about 50% of patients ≥40 years of age. The treatment-discontinued group had the lowest AST level and APRI score. The treatment-continued group had a higher median FIB-4 score than either of the other two groups. There were no significant differences in sex, serum HBV DNA load, ALT level, or PLT count among the three subgroups. Table 1 gives the patient baseline characteristics.
Table 1Baseline characteristics of the enrolled patients
| Characteristic | Entire, n=194 | Treated, n=127
| Untreated, n=67 | p# |
|---|
| Continued, n=87 | Discontinued, n=40 |
|---|
| Duration of follow-up*, month | 54 (24, 72) | 60 (36, 72) | 57 (48,72) | 48 (24, 72) | 0.895 |
| Duration of treatment*, month | – | 60 (48, 72) | 12 (6, 24) | – | <0.001 |
| Age, years | 43 (36, 50) | 45 (39, 50) | 43 (38, 54) | 39 (20, 49) | <0.001 |
| ≥40 | 122 (62.9%) | 62 (71.3%) | 27 (67.5%) | 33 (49.3%) | 0.015 |
| Male | 92 (47.4%) | 44 (50.6%) | 20 (50.0%) | 28 (41.8%) | 0.220 |
| NA(s) regime | | | | | 0.241 |
| ETV | 82 (42.3%) | 59 (67.8%) | 23 (57.5%) | – | – |
| TDF | 44 (22.7%) | 28 (32.2%) | 16 (40.0%) | – | – |
| LDT | 1 (0.5%) | 0 (0.0%) | 1 (2.5%) | – | – |
| HBV DNA, log10 IU/mL | 3.88 (3.41, 4.56) | 3.86 (3.43, 4.44) | 4.09 (3.61, 4.62) | 3.64 (3.08, 4.56) | 0.117 |
| ≥3.3 | 157 (80.9%) | 74 (85.1%) | 34 (85.0%) | 49 (73.1%) | 0.146 |
| ALT, U/L | 22 (17, 27) | 21 (17, 25) | 22 (18, 28) | 24 (17, 30) | 0.101 |
| ALT ≤20 | 78 (40.2%) | 39 (44.8%) | 13 (32.5%) | 26 (38.8%) | 0.402 |
| AST, U/L | 24 (20, 28) | 25 (22, 28) | 21 (17, 25) | 24 (20, 28) | <0.001 |
| PLT, ×109/L | 183±51 | 177±55 | 186±53 | 195±50 | 0.118 |
| APRI | 0.39(0.29, 0.51) | 0.41 (0.32, 0.57) | 0.31 (0.24, 0.46) | 0.37 (0.27, 0.44) | <0.001 |
| <1.00 | 191 (98.5%) | 84 (96.6%) | 40 (100%) | 67 (100%) | 0.311 |
| 1.00–2.00 | 3 (1.5%) | 3 (3.4%) | 0 (0.0%) | 0 (0.0%) | 0.311 |
| >2.00 | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | – |
| FIB-4 | 1.22 (0.87, 1.67) | 1.47 (1.04, 1.99) | 0.98 (0.81, 1.57) | 1.04 (0.71, 1.44) | <0.001 |
| <1.45 | 124 (63.9%) | 43 (49.4%) | 29 (72.5%) | 52 (77.6%) | 0.001 |
| 1.45–3.25 | 64 (33.0%) | 39 (44.8%) | 11 (27.5%) | 14 (20.9%) | 0.005 |
| >3.25 | 6 (3.1%) | 5 (5.7%) | 0 (0.0%) | 1 (1.5%) | 0.186 |
Changes in serum HBV DNA level
In the untreated group, serum HBV DNA loads declined from 3.64 log10 IU/mL to 3.27 log10 IU/mL (Fig. 2A). No subject had spontaneously achieved undetectable HBV DNA at the end of the follow-up (Fig. 2D). In the treatment-continued group, the cumulative virologic response rates at 6, 12, and 18 months were 92.0% (80/87, 95% CI: 83.6–96.4), 97.7% (85/87, 95% CI: 91.2–99.6), and 100% (87/87, 95% CI: 94.7–100.0), respectively. No patient experienced virological breakthrough following virological response. Of the 87 patients, 82.8% (72/87, 95% CI: 72.8–89.7) maintained virological response (HBV DNA<100 IU/mL and 16.1% (14/87, 95% CI, 9.4–25.9) had an intermittent fluctuation of low-level of serum HBV DNA (<2,000 IU/mL; Fig. 2B, E). In the treatment-discontinued group, 97.5% of the patients (39/40, 95% CI: 85.3–99.9) achieved virologic response at the end of treatment (EOT). At the end of follow-up (EOF), the median load of HBV DNA rebounded to 3.92 log10 IU/mL, a decrease of 0.25 log10 IU/mL (median) from baseline (p=0.035; Fig. 2C). The cumulative incidence of HBV DNA detectable after virological response within 6, 12, and 24 months were 59.0% (23/39, 95% CI: 42.2–74.0), 87.2% (34/39, 95% CI: 71.8–95.2), and 92.3% (36/39, 95% CI: 78.0–98.0), among whom 84.6% (32/39, 95% CI: 65.9–91.9) had a virological relapse (HBV DNA>2,000 IU/mL). The cumulative incidence of virological relapse at 6, 12, and 24 months were 17.9% (7/39, 95% CI: 8.1–34.1), 48.7% (19/39, 95% CI: 32.7–65.0), and 76.9% (30/39, 95% CI: 60.3–88.3%), respectively (Fig. 2F).
Changes in APRI and FIB-4
In the untreated group, the APRI decreased from 0.37 at the baseline to 0.35 (p=0.135). The proportion of patients with an APRI<1.0 significantly decreased from 100% to 89.5% (p=0.020). The median score of FIB-4 increased from 1.04 to 1.20 (p<0.001). The proportion of patients with FIB-4<1.45 decreased from 77.6% to 64.2% (p=0.128) and the proportion of those with an FIB-4 of >3.25 increased from 1.5% to 7.5% (Fig. 3). In the treatment-continued group, the APRI decreased significantly from 0.41 to 0.33 (p<0.001). The proportion of patients with an APRI of <1.0 increased to 100% (p=0.244). The median FIB-4 also decreased significantly from 1.47 to 1.10 (p<0.001), with the proportion of FIB-4<1.45 increasing significantly to 70.1% (p=0.005), and that with an FIB-4>3.25 decreased to 3.5% (Fig. 3). In the treatment-discontinued group, the median duration of treatment was 12 (6–18) months. At the EOT, the ALT level had increased significantly vs. baseline (22 vs. 25 U/L, p=0.019), but the APRI score (0.31 vs. 0.31, p=0.627) and FIB-4 score (0.98 vs. 1.11, p=0.572) had not significantly changed. At the EOF, the APRI and FIB-4 scores had significantly increased from those at the EOT (APRI, 0.31 vs. 0.67, p<0.001; FIB-4, 1.11 vs. 2.34, p<0.001) and baseline (APRI, 0.31 vs. 0.67, p<0.001; FIB-4, 0.98 vs. 2.34, p<0.001; Supplementary Table 1 and Supplementary Fig. 1). The proportion of patients with an APRI of <1.0 decreased significantly from 100% to 85.0% (p=0.034). The proportion of patients with an FIB-4 of <1.45 decreased significantly from 72.5% to 27.5% (p<0.001), and that of patients with an FIB-4 of >3.25 increased significantly from 0.0% to 10.0% (Fig. 3). The changes in scores of APRI and FIB-4 in relapsed and nonrelapsed patients had the same tendencies (Supplementary Table 1 and Supplementary Fig. 1). There were no significant differences in the increase in APRI and FIB-4 at EOF compared with the scores at EOT in the two subgroups. The APRIEOF-EOT scores in relapsed vs. nonrelapsed patients were 0.35 vs. 0.30, p=0.453. The FIB-4EOF-EOT scores in relapsed vs. nonrelapsed patients were 1.21 vs. 1.53, p=0.798 (Supplementary Table 2 and Supplementary Fig. 2).
Liver nodules, liver cirrhosis, and HCC
Patients at all HBV DNA levels
Adverse event of liver nodules was reported in 53 patients (27.3%, 95% CI: 21.3–34.3) of the entire cohort (n=194), with a mean follow-up of 55.5 months (95% CI: 51.4–59.5), 26 in the untreated, 10 in the treatment-continued, and 17 in the treatment-discontinued groups. Of the 53 patients with liver nodules, 16 (30.2%, 95% CI: 18.7–44.5) progressed to cirrhosis, with a mean follow-up of 60.8 months (95% CI: 53.2–68.3), one, eight, and seven in the three subgroups. Of the 16 patients with liver cirrhosis, two (12.5%, 95% CI: 2.2–39.6) developed HCC, with a mean follow-up of 72 months, one in the untreated group and the other in the treatment-discontinued group. The incidence of adverse events reported in each subgroup during follow-up are shown in Supplementary Figure 3A, B.
The incidence of liver nodules was 1.82 per 100 patient-years (PYs) in the treatment-continued group, significantly lower than that of the untreated group (10.4 per 100 PYs, p<0.001, hazard ratio (HR) 0.24, 95% CI: 0.11–0.49, p<0.001) and the treatment-discontinued group (1.82 vs. 3.11 per 100 PYs, HR: 0.23, 95% CI: 0.10–0.49, p<0.001). Although the estimated incidence of liver nodules in the untreated group was higher than that in the treatment-discontinued group, the difference was not significant (10.4 vs. 3.11 per 100 PYs, HR: 1.06, 95% CI: 0.57–1.98, p=0.845; Fig. 4A and Table 2). The cumulative incidence of liver cirrhosis was also significantly reduced in the treatment-continued group compared with the untreated group (0.17 vs. 2.49 per 100 PYs, HR: 0.11, 95% CI: 0.01–0.91, p=0.013) and treatment-discontinued group (0.17 vs. 1.37 per 100 PYs, HR: 0.05, 95% CI: 0.01–0.44, p=0.006). Difference in the incidence of liver cirrhosis in the untreated and treatment-discontinued groups was not significant (p=0.121; Fig. 4B and Table 2). Differences in the incidence of HCC among the three subgroups were not significant (p=0.228; Fig. 4C and Table 2).
Table 2Incidence of liver nodules, liver cirrhosis, and HCC in patients with HBV DNA at all levels and with HBV DNA≥2,000 IU/mL of the three subgroups¶
| Subgroup | HBV DNA at all levels, n=194
| HBV DNA≥2,000 IU/mL, n=157
|
|---|
| Patient-years | No. of events | No./100 patient-years (95% CI) | HR (95% CI) | Patient-years | No. of events | No./100 patient-years (95% CI) | HR (95% CI) |
|---|
| Liver nodules | | | | | | | |
| Untreated | 249 | 26 | 10.4 (7.06–15.01) | 0.24 (0.11–0.49) | 246 | 20 | 8.13 (5.16–12.46) | 0.22 (0.10–0.48) |
| Discontinued | 547 | 17 | 3.11 (1.88–5.03) | 0.23 (0.10–0.49) | 541 | 16 | 2.96 (1.76–4.87) | 0.21 (0.09–0.47) |
| Continued | 550 | 10 | 1.82 (0.93–3.43) | – | 545 | 9 | 1.65 (0.81–3.22) | – |
| Liver cirrhosis | | | | | | | |
| Untreated | 281 | 7 | 2.49 (1.10–5.28) | 0.11 (0.14–0.91) | 279 | 7 | 2.51 (1.10–5.32) | 0.08 (0.01–0.67) |
| Discontinued | 586 | 8 | 1.37 (0.64–2.79) | 0.05 (0.01–0.44) | 586 | 7 | 1.19 (0.52–2.55) | 0.07 (0.01–0.53) |
| Continued | 594 | 1 | 0.17 (0.01–1.09) | – | 589 | 1 | 0.17 (0.01–1.10) | – |
| HCC§ | | | | | | | |
| Untreated | 283 | 1 | 0.35 (0.02–1.09) | – | 281 | 1 | 0.36 (0.02–2.92) | – |
| Discontinued | 591 | 1 | 0.17 (0.01–1.09) | – | 591 | 0 | 0.0 (0.0–0.81) | – |
| Continued | 600 | 0 | 0.0 (0.0–0.79) | – | 595 | 0 | 0.0 (0.0–0.80) | – |
Patients with HBV DNA≥2,000 IU/mL
Of the 194 patients, about four-fifths (n=157) had HBV DNA≥2,000 IU/mL, 49 in the untreated group, 74 in the treatment-continued group, and 34 in the treatment-discontinued group. Except for the shortest follow-up duration in the untreated group, the lowest AST and APRI levels in the treatment-discontinued group, and the highest FIB-4 score in the treatment-continued group, there were no significant differences in the baseline characteristics in the three new subgroups (Table 3).
Table 3Baseline characteristics of patients with HBV DNA≥2,000 IU/mL in the three subgroups¶
| Characteristic | Entire, n=157 | Treatment group, n=108
| Untreated, n=49 | p |
|---|
| Continued, n=74 | Discontinued, n=34 |
|---|
| Duration of follow-up*, months | 54 (24, 72) | 57 (48, 72) | 57 (48, 72) | 48 (24, 72) | 0.006 |
| Duration of treatment*, months | – | 60 (36, 72) | 12 (6, 24) | – | <0.001 |
| Age, years | 43±12.5 | 44±8.6 | 44±9.6 | 41±11.1 | 0.194 |
| ≥40 | 96 (61.1%) | 50 (67.6%) | 21 (61.8%) | 25 (51.0%) | 0.186 |
| Male | 75 (47.8%) | 38 (48.6%) | 16 (47.1%) | 30 (61.2%) | 0.311 |
| NA (s) regimen | | | | 0.061# |
| ETV | 68 (63.0%) | 51 (68.9%) | 17 (50.0%) | – | |
| TDF | 39 (36.1%) | 23 (31.1%) | 16 (47.1%) | – | |
| LDT | 1 (0.9%) | 0 (0.0%) | 1 (2.9%) | – | |
| HBV DNA, log10 IU/mL | 4.11 (3.61, 4.71) | 4.08 (3.59, 4.64) | 4.25 (3.87, 4.72) | 4.11 (3.58, 4.74) | 0.368 |
| ALT, U/L | 23±7 | 22±6 | 22±6 | 24±9 | 0.178 |
| ALT≤20 | 62 (39.5%) | 32 (43.2%) | 12 (35.3%) | 18 (36.7%) | 0.687 |
| AST, U/L | 24±6 | 26±6 | 21±6 | 24±5 | <0.001 |
| PLT, ×109/L | 187±53 | 179±57 | 187±51 | 193±47 | 0.361 |
| APRI | 0.38 (0.29, 0.51) | 0.41 (0.32, 0.58) | 0.31 (0.23, 0.44) | 0.37 (0.27, 0.46) | 0.004 |
| <1.00 | 154 (98.1%) | 71 (95.9%) | 34 (100.0%) | 49 (100.0%) | 0.313 |
| 1.00–2.00 | 3 (1.9%) | 3 (4.1%) | 0 (0.0%) | 0 (0.0%) | 0.313 |
| >2.00 | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | – |
| FIB-4 | 1.18 (0.87, 1.70) | 1.38 (1.00, 1.99) | 0.98 (0.68, 1.52) | 1.04 (0.75, 1.52) | 0.003 |
| <1.45 | 99 (63.1%) | 38 (51.4%) | 25 (73.5%) | 36 (73.5%) | 0.016 |
| 1.45–3.25 | 52 (33.1%) | 31 (41.9%) | 9 (26.5%) | 12 (24.5%) | 0.089 |
| >3.25 | 6 (3.8%) | 5 (56.8%) | 0 (0.0%) | 1 (2.0%) | 0.306 |
Liver nodules developed in 45 (28.6%, 95% CI: 21.9–36.5%) of the 157 patients, with a median follow-up of 55.3 months (95% CI: 50.9–59.8), 20 in the untreated, nine in the treatment-continued, and 16 in the treatment-discontinued groups. Of the 45 patients with liver nodules, 15 (33.3%, 95% CI: 20.4–49.1) progressed to cirrhosis, with a mean follow-up of 60.0 months (95% CI: 52.1–67.9), seven, one, and seven in the three subgroups, respectively. Of the 15 patients with liver cirrhosis, 1 (6.7%, 95% CI: 0.35–34.0%) in the untreated group developed HCC at a follow-up of 72 months. The cumulative incidences of adverse events are shown in Supplementary Figure 3C, D.
Patients with HBV DNA≥2,000 IU/mL in the treatment-continued group had a significantly lower incidence of liver nodules (1.65 vs. 2.96 vs. 8.13 per 100 PYs) and liver cirrhosis (0.17 vs. 2.51 vs. 1.19 per 100 PYs) than those who were untreated and discontinued treatment (Fig. 4D, E). There were no significant differences in the incidence of liver nodules (p=0.887) and liver cirrhosis (p=0.551) in the untreated and the treatment-discontinued groups (Fig. 4D, E). Compared with untreated patients, the risks of liver nodules and liver cirrhosis in patients remaining on antiviral therapy decreased by 78% (HR: 0.22, 95% CI: 0.10–0.48) and 92% (HR: 0.08, 95% CI: 0.01–0.67), respectively (Table 2). Compared with those interrupting therapy, the risks of liver nodules and liver cirrhosis decreased by 79% (HR: 0.21, 95% CI: 0.09–0.47) and 93% (HR: 0.07, 95% CI: 0.01–0.53), respectively (Table 2). No patients in the treatment-continued or the treatment-discontinued groups developed HCC, but one in the untreated group did. Differences in the HCC incidence among the three subgroups were not significant (p=0.462; Table 2 and Fig. 4F).
Patients with HBV DNA of <2,000 IU/mL
There were 37 patients with HBV DNA<2,000 IU/mL, 18 in the untreated, 13 in the treatment-continued, and six in the treatment-discontinued groups. Eight patients developed liver nodules, six in the untreated group and one each in the other two groups. Patients of the treatment-continued group had a lower incidence of liver nodules than those untreated (0.20 vs. 2.47 per 100 PYs, p=0.105), and the treatment-discontinued group (0.20 vs. 0.26 per 100 PYs, p=0.619), but the differences were not significant. Compared with the untreated group, patients with interrupted therapy had a significantly higher incidence of liver cirrhosis (0.25 vs. 0.0 per 100 PYs, p=0.019) and HCC (0.25 vs. 0.0 per 100 PYs, p=0.003), but neither incidence was significantly higher than those in the treatment-continued group (0.25 vs. 0.0 per 100 PYs, p=0.061 for liver cirrhosis; 0.25 vs. 0.0 per 100 PYs, p=0.317 for HCC) (Table 4).
Table 4Incidence of liver nodules, liver cirrhosis, and HCC in patients with HBV DNA<2,000 IU/mL of the three subgroups¶
| Subgroup | Patient-years | No. of events | No./100 patient-years (95% CI) | p |
|---|
| Liver nodules | | | | |
| Untreated (n=18) | 243 | 6 | 2.47 (1.01–5.56) | 0.105* |
| Continued (n=13) | 494 | 1 | 0.20 (0.001–1.30) | – |
| Discontinued (n=6) | 380 | 1 | 0.26 (0.01–1.69) | 0.619* |
| Liver cirrhosis | | | |
| Untreated (n=18) | 273 | 0 | 0.0 (0.0–1.73) | 0.019# |
| Continued (n=13) | 534 | 0 | 0.0 (0.0–0.89) | 0.061# |
| Discontinued (n=6) | 402 | 1 | 0.25 (0.01–1.60) | – |
| HCC | | | | |
| Untreated (n=18) | 274 | 0 | 0.0 (0.0–1.72) | 0.003# |
| Continued (n=13) | 540 | 0 | 0.0 (0.0–0.88) | 0.317# |
| Discontinued (n=6) | 406 | 1 | 0.25 (0.01–1.59) | – |
Discussion
In this single-center retrospective study, we analyzed the efficacy of NAs treatment and the improvement of long-term outcomes in HBeAg-negative patients with persistent normal ALT and detectable HBV DNA. The results showed those patients had an excellent virological response to NAs therapy but also a high rate of virological relapse if treatment was interrupted. Compared with those untreated and those stopping treatment, we observed significant declines of APRI and FIB-4 scores, significantly reduced risk of liver nodules and liver cirrhosis, and a low incidence of HCC in patients who continued antiviral therapy, especially in those with HBV DNA≥2,000 IU/mL, who accounted for 80.9% of the whole cohort. To our knowledge, this is the first study to investigate the long-term prognosis of antiviral therapy in HBeAg-negative CHB patients with persistently normal ALT and positive HBV DNA, providing direct evidence for expanding the indications of antiviral treatment in such a population.
Continuous inhibition of virological replication and favorable compliance are foundations of reduced risk of liver cirrhosis. In this study, a high rate of virological response was achieved in both the treated and treatment-discontinued groups. However, poor off-treatment sustainability of NAs effectiveness was observed, with more than 80% of patients experiencing virological relapse (HBV DNA>2,000 IU/mL) in the treatment-discontinued group. Worse still, those with virological relapse had a significantly higher incidence of liver nodules (2.77 vs. 1.82 per 100 PYs, p<0.001) and cirrhosis (1.19 vs. 0.17 per 100 PYs, p<0.001) than those adherent to treatment. Additionally, the median APRI (0.28 vs. 0.66, p<0.001) and FIB-4 scores (0.94 vs. 2.24, p<0.001) increased significantly from the baseline in patients with virological relapse. This is consistent with the results of a 10-year longitudinal observational study that included 894 treatment-naïve CHB patients who received ETV therapy and revealed poor medication adherence was associated with increased mortality and risk of HCC and cirrhotic complications.15 Moreover, inappropriate discontinuation or intermittent treatment failed to effectively inhibit virological replication and led to virological resistance and low-level viremia.13,15
Several reasons were responsible for treatment interruption. By making phone calls and reviewing electronic medical records, we identified some reasons for stopping medication, including lack of awareness of long-term treatment, conflict between the time for oral medicines and daily dietary habits, inability to adhere to long-term therapy, mistaken belief that continuously undetectable serum HBV DNA and normal ALT are signs of medication withdrawal, fear of drug side effects and virological resistance during long-term medication, and concerns of drug interactions because of treatment for other diseases. Data from a previous study revealed that adherence to NAs treatment by CHB patients was about 75% in low-, middle-, and high-income countries.16 A Chinese study based on a questionnaire survey (Morisky Medication Adherence Scale-8) reported that over 50% of patients had low medication compliance (a score of <6 points) and only 16.5% patients, those with a score of eight points, had high adherence, much lower than the proportions reported by similar studies conducted in other countries such as Australia (74.1%), the USA (66%), and New England (52%).17 Therefore, increased attention should be paid to patient compliance management. Given the limited effectiveness and increased risk of adverse liver events after drug withdrawal, sustained treatment is recommended.
Continuous antiviral therapy helps delay the progression of liver fibrosis. Liver biopsy is the most accurate method to evaluate liver inflammation and fibrosis, but is challenging to carry out in clinical practice. Therefore, this study used noninvasive methods, APRI and FIB-4, which are recommended by major international guidelines, to assess liver fibrosis.12,18–20 APRI and FIB-4 have good specificity in excluding significant liver fibrosis and cirrhosis.14 In this study, the APRI and FIB-4 scores in the continuous treatment group were significantly decreased from baseline, and the proportion of APRI<1.00 (p=0.244) and FIB-4<1.45 (p=0.005) increased, but changes in the untreated and treatment-discontinued groups were almost opposite. That suggests that the process of liver fibrosis did not stop in patients with normal ALT, and long-term antiviral therapy helped to slow down hidden changes in the liver even in patients with normal ALT. Although the mechanism of the reversal of liver fibrosis at an early stage and delay of cirrhosis by antiviral treatment has not been fully explained, published results based on liver pathology also support the use of long-term standardized antiviral therapy to reverse liver fibrosis and early-stage cirrhosis in some patients.21–23
Serum HBV DNA has an important role in the antiviral treatment decision. In a study by Lim et al.,9 the untreated replicative phase group (ALT<ULN, HBV DNA≥2,000 IU/mL) had significantly increased risks of HCC (HR: 1.76; 95% CI: 1.00–3.10, p=0.05) and death/transplantation (HR: 2.14; 95% CI: 1.09–4.21, p=0.03) compared with the treated active-phase group (ALT≥2ULN, HBV DNA≥2,000 IU/mL). Therefore, early antiviral intervention before ALT elevation was recommended for selected HBeAg-negative CHB patients with high viral loads. Limitations of that study were that active-phase patients were the control group and replicative phase patients were not treated. Hence, the concrete benefit of antiviral therapy is not known if replicative phase patients receive therapy. Our study results fill the gap, finding that for a median follow-up of 60 months, cirrhosis in patients adhering to antiviral treatment was reduced by 91%, with a 92% reduction in patients with an HBV DNA≥2,000 IU/mL compared with their counterparts and no patients developed HCC.
Significant necroinflammation and fibrosis may exist even with a normal ALT, indicating the necessity of antiviral treatment. Zhuang et al.4 assessed changes of liver histology in 327 HBeAg-negative CHB patients with normal ALT levels, finding that an ALT>20 U/L and a HBV DNA≥2,000 IU/mL were independently predictive of significant histopathology. Therefore, we used 20 U/L as the critical stratification value. However, there were no significant difference in the incidence of liver nodules (3.76 vs. 6.91 per 100 PYs, p=0.815) or liver cirrhosis (0.36 vs. 1.43 per 100 PYs, p=0.330) in patients with an ALT<20 U/L (n=26) vs. ALT≥20 U/L (n=43) in the untreated group, indicating the limitation in predicting disease progression by ALT level. The results are not consistent with a previous study that found patients with ALT levels of 0.5–1×ULN had an increased risk of liver-related complications and HCC than those with ALT levels of <0.5×ULN (both p<0.001).24 Therefore, the results need validation by future studies.
However, patients in the untreated group with HBV DNA≥2,000 IU/mL had a higher incidence of liver cirrhosis than those with HBV DNA<2,000 IU/mL (2.51 vs. 0.0 per 100 PYs, p=0.036), indicating a certain correlation underlying the risk of liver cirrhosis and serum viral load. The correlation between HBV DNA level and disease outcomes has been reported in previous studies.25,26 Later studies have also shown a higher risk of HCC in patients with higher serum HBV DNA levels.1,25,27,28 In this study, antiviral treatment, significantly lowered the risks of liver nodules and liver cirrhosis in treated patients with HBV DNA levels ≥2,000 IU/mL compared with untreated patients.
We suggest increasing the weight of serum HBV DNA in antiviral treatment decisions rather than rigidly adhering to ALT level, age, or family history of cirrhosis/HCC, considering HBeAg-negative patients with high viral loads are common in clinical practice, and antiviral therapy decreases the long-term risk of adverse liver events.5,29,30 Recent Chinese hepatitis B guidelines (version 2022) have lowered the ALT level for initiating treatment to 30 U/L for men and 19 U/L for women.31 That will enable more patients to meet treatment standards but still fails to include some patients with a high risk of disease progression, such as those with ALT levels below the ULN but with high viral loads ≥2,000 IU/mL.
This study has some limitations. (1) It was a retrospective study and the sample size was relatively small if HCC is taken as endpoint event, thus, the benefits of antiviral therapy may be underestimated. However, when we started this study, only a small number of patients had received long-term (at least 5 years) antiviral therapy following the recommended regular examination and no treatment in most clinical settings. In the beginning, 529 patients with persistent normal ALT were screened, but only 194 cases met the inclusion criteria. Thus, multicenter, prospective studies with a larger sample size and prolonged follow-up time are urgently needed. (2) APRI and FIB-4 replaced liver biopsy to estimate liver fibrosis because many patients were reluctant to accept this invasive examination. In the future, hepatitis B treatment indications will be less dependent on liver biopsy, with more emphasis on high-risk factors, such as viral load. (3) The number of patients with HBV DNA loads <2,000 IU/mL and younger than 30 years of age is small and the benefits of antiviral therapy for those individuals need further confirmation. (4) Our subjects were all Chinese, consequently the results need to be verified in non-Chinese patients.