Patient selection: Between January 2008 and December 2020, a total of 1617 consecutive adult patients (age, ≥ 18 years) diagnosed with CHB (hepatitis B surface antigen positive > 6 mo) who had undergone liver biopsy at the Third Affiliated Hospital of Sun Yat-Sen University were included in this retrospective analysis. The exclusion criteria were as follows: (1) Viral coinfection (hepatitis C virus, hepatitis D virus, or HIV); (2) Alcohol abuse (≥ 30 g of alcohol per day for men, ≥ 20 g of alcohol per day for women), nonalcoholic fatty liver disease (diagnosed by liver biopsy) and autoimmune liver disease; (3) Decompensated cirrhosis, HCC, and nonliver cancer; (4) Liver transplantation; and (5) Prior or current antiviral treatment (Supplementary Figure 1). The study was approved by the Ethics Committee of The Third Affiliated Hospital of Sun Yat-sen University. Written informed consent was obtained from each participant prior to the liver biopsy.

Definitions: The clinical phases of patients with CHB were determined in accordance with the EASL 2017 clinical practice guidelines, taking into consideration the highest HBV-DNA levels and ALT levels observed in at least two determinations within the 12 mo preceding enrolment (Supplementary Table 1)[4]. Patients who did not meet the criteria for any of the five phases were classified as GZ, with subcategories including GZ-A (HBeAg positive, normal ALT levels, and HBV DNA ≤ 10⁷ IU/mL), GZ-B (HBeAg positive, elevated ALT levels, and HBV DNA < 10⁴ or > 10⁷ IU/mL), GZ-C (HBeAg negative, normal ALT levels, and HBV DNA ≥ 2000 IU/mL), and GZ-D (HBeAg negative, elevated ALT levels, and HBV DNA ≤ 2000 IU/mL)[9,12]. We used ALT and gamma-glutamyl transpeptidase (GGT) levels of 40 U/L and 60 U/L as the ULN, respectively[4]. The calculations were formulated as follows: Aspartate aminotransferase (AST)-to-platelet ratio index (APRI) = [AST (U/L)/ULN]/[PLT (10⁹/L)] × 100; fibrosis score based on four factors (FIB-4) = [age (years) × AST (U/L)]/[PLT (10⁹/L) × √ALT (U/L)]; GGT-to-platelet ratio (GPR) = [GGT (U/L)/ULN]/[PLT (10⁹/L)] × 100.

Histological assessment: Ultrasonography-guided percutaneous liver biopsy was conducted using a 16-gauge disposable needle. The minimum sample length required was 15 mm, with a minimum inclusion of 6 portal tracts. Inflammation grade (G0-G4) and fibrosis stage (S0-S4) were estimated according to Scheuer's classification[14]. In accordance with the pathological staging system, significant liver inflammation and significant fibrosis were defined as ≥ G2 and ≥ S2, respectively. Significant hepatic injury (SHI) was defined as ≥ G2 and/or ≥ S2[12,15]. The biopsy samples were subjected to blind and independent observation and interpretation by two proficient pathologists. In cases where discordance arose between the two pathologists, a third pathologist, Jianning Chen, was consulted for additional evaluation, leading to a consensus being achieved through subsequent discussion.

SPSS version 25.0 software (SPSS Inc., Chicago, IL, United States) was used for statistical analyses. Continuous variables are expressed as the median and interquartile range, and categorical data are expressed as counts and percentages. Kruskal-Wallis tests and Pearson’s chi-squared tests were applied to compare variables that were significantly different between groups. The independent predictors of SHI were determined by univariate and multiple logistic regression analyses. Areas under the receiver operating characteristic curve (AUROC) were calculated to investigate the diagnostic performance of the noninvasive scores and were compared by the Delong test. All P values were two-sided, and P < 0.05 was deemed statistically significant.

The baseline characteristics of the 1617 treatment-naive patients are presented in Table 1. Based on the defined criteria, 161 (9.96%) patients were classified as having HBeAg-positive chronic HBV infection, 203 (12.55%) patients were categorized as HBeAg-positive CHB, 171 (10.58%) patients were identified as having HBeAg-negative chronic HBV infection, and 270 (16.70%) patients were classified as HBeAg-negative CHB. Interestingly, 812 (50.22%) patients did not meet the criteria for any of the aforementioned phases and were therefore designated as GZ. Notably, there were significant variations in clinical characteristics across the different phases. The average age of patients with GZ was 36.0 years, with 74.1% being male. These patients had a mean HBV DNA level of 5.24 Log₁₀ IU/mL and an intermediate ALT level of 37.0 U/L.

The clinical and liver histological characteristics are presented in Table 1 and Figure 1. In the GZ group, the proportion of significant liver inflammation (≥ G2) was 56.8%, and the proportion of significant fibrosis (≥ S2) was 53.4%. Among GZ patients, 63.7% had SHI. Higher proportions of SHI were observed in HBeAg-positive (84.2%) and HBeAg-negative (76.7%) CHB patients. However, HBeAg-positive (41.0%) and HBeAg-negative chronic HBV infection (53.2%) had lower but still relatively high proportions of SHI.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Distribution and liver histological features of chronic hepatitis B patients in different immune states. A and D: Proportions of liver inflammation grades in different immune status groups; B and E: Proportions of fibrosis stages in different immune status groups; C and F: Proportions of significant hepatic injury in different immune status groups. A: HBeAg-positive chronic hepatitis B virus (HBV) infection; B: HBeAg-positive chronic hepatitis B; C: HBeAg-negative chronic HBV infection; D: HBeAg-negative chronic hepatitis B; GZ: Grey zone; EHI: Evidenced hepatic injury. ^aP < 0.001.

The clinical characteristics of 812 GZ patients are shown in Supplementary Table 2. Among these patients, the proportion of GZ-C was the highest (41.1%), followed by GZ-B (34.6%), GZ-A (15.8%), and GZ-D (8.5%). Notably, patients in the GZ-A (35.0 years) and GZ-B (31.0 years) subgroups were younger than those in the GZ-C (40.0 years) and GZ-D (36.0 years) subgroups. Furthermore, higher HBV DNA levels were observed in GZ-B (7.86 Log10 IU/mL), followed by GZ-A (5.39 Log10 IU/mL), GZ-C (4.40 Log10 IU/mL), and GZ-D (2.40 Log10 IU/mL).

As shown in Supplementary Figure 2, HBeAg-positive GZ patients exhibited a significantly higher rate of significant liver inflammation (≥ G2) (66.1% vs 47.4%, P < 0.001) and SHI (70.4% vs 56.8%, P < 0.001) than HBeAg-negative GZ patients. However, there was no statistically significant difference in fibrosis stages between HBeAg-positive and HBeAg-negative GZ patients (55.8% vs 51.2%, P = 0.186). The distributions of liver inflammation grades, fibrosis stages, and SHI are shown in Supplementary Table 2 and Supplementary Figure 2. The highest prevalence of significant liver inflammation (≥ G2) was observed in GZ-B (66.5%), while GZ-D (62.3%) had the highest proportion of significant fibrosis (≥ S2). In terms of SHI, the highest proportion was found in patients from GZ-B (70.5%), followed by GZ-A (70.3%), GA-D (69.6%), and GZ-C (54.2%).

SHI was observed in 517 (63.7%) GZ patients. Of those, 288 were HBeAg-positive GZ patients, and 229 were HBeAg-negative GZ patients. In the HBeAg-positive cohort, univariate analysis indicated that PLT, ALT, AST, GGT, total bilirubin (Tbil), albumin (ALB), and prothrombin time(PT) were associated with SHI. However, multiple logistic regression analysis indicated that only PLT, AST, GGT, and ALB remained significantly associated with SHI. For the HBeAg-negative cohort, female sex, HBV-DNA, ALT, AST, GGT, Tbil, and PT were associated with higher SHI, whereas PLT and ALB were negatively associated with this event by univariate analysis. By multiple logistic regression analysis, female sex, HBV-DNA, GGT, and PLT were associated with SHI (Supplementary Table 3 and Figure 2).

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Multiple logistic regression analysis. A: Multiple logistic regression analysis of clinical parameters of chronic hepatitis B patients in HBeAg-positive grey zones associated with significant hepatic injury; B: Multiple logistic regression analysis of clinical parameters of chronic hepatitis B patients in HBeAg-negative grey zones associated with significant hepatic injury; C: Receiver operating characteristic (ROC) curves of aspartate aminotransferase-to-platelet ratio index (APRI), fibrosis score based on four factors (FIB-4), and gamma-glutamyl transpeptidase-to-platelet ratio (GPR) in the prediction of significant hepatic injury (SHI) in HBeAg-positive grey zones; D: ROC curves of APRI, FIB-4, and GPR in the prediction of SHI in HBeAg-negative grey zones. PLT: Platelet; AST: Aspartate aminotransferase; GGT: Gamma-glutamyl transpeptidase; ALB: Albumin; APRI: Aspartate aminotransferase-to-platelet ratio index; FIB-4: Fibrosis score based on four factors; GPR: Gamma-glutamyl transpeptidase-to-platelet ratio.

Further investigation was conducted to assess the diagnostic efficacy of established scoring systems such as APRI, FIB-4, and GPR in predicting SHI (Figure 2). The AUROCs for these three tests showed no significant difference in the HBeAg-positive GZ (P > 0.05). However, in HBeAg-negative GZ, APRI demonstrated superior performance compared to FIB-4 (P < 0.001) and was comparable to GPR (P = 0.30).

To examine the significance of lowering the treatment threshold for ALT, a total of 794 patients with normal ALT levels (ULN ≤ 40 U/L) were chosen for further investigation. The distribution of their immune states was as follows: 161 (20.28%) patients had HBeAg-positive chronic HBV infection, 128 (16.12%) patients fell within GZ-A, 171 (21.54%) patients had HBeAg-negative chronic HBV infection, and 334 (42.07%) patients were categorized under GZ-C.

Based on the AASLD 2018 criteria of the ALT antiviral treatment threshold (35 U/L for males and 25 U/L for females), we subsequently evaluated the ratio of SHI in different groups. Among the 794 chronic HBV infections, more than one-quarter of them (221/794, 27.8%) were above the AASLD criteria (Figure 3A). Of these 221 individuals, 29.2% (47/161) with HBeAg-positive chronic HBV infection, 36.7% (47/128) with GZ-A, 21.1% (36/171) with HBeAg-negative chronic HBV infection, and 27.2% (91/334) with GZ-C were above this ALT threshold (Figure 3B). It is worth noting that 54.8% (121/221) of patients had significant liver inflammation (≥ G2), which was significantly higher than that of patients below the ALT threshold (43.1%, 247/573) (P = 0.003) (Figure 3C). In addition, the proportion of SHI in the high ALT group was significantly higher than that in the low ALT group (60.6% vs 51.3%, P = 0.018). In total, the proportion of patients with ALT ≤ 40 U/L who required antiviral therapy was 64.86% [(221 + 294)/794] according to the AASLD 2018 Clinical Practice Guidelines.

Open in New Tab   Full Size Figure   Download Figure

Figure 3 The alanine aminotransferase treatment threshold was lowered to the American Association for the Study of Liver Diseases 2018 treatment criteria (35 U/L for males and 25 U/L for females). A: Among the 794 chronic hepatitis B virus infections, 27.8% (221/794) of patients were above the American Association for the Study of Liver Diseases criteria; B: Comparison of the proportions of patients exceeding the alanine aminotransferase (ALT) threshold in different groups; C: A total of 43.1% (247/573) of patients below the ALT threshold had significant liver inflammation (≥ G2); D: The proportion of significant hepatic injury (SHI) in patients below the ALT threshold was 51.3%. Comparison of the proportions of SHI in each group. ALT: Alanine aminotransferase; NS: Not significant. ^aP < 0.05, ^bP < 0.01, ^cP < 0.001.

Obviously, the SHI value in patients with HBeAg-positive chronic HBV infection below the new ALT threshold was substantially lower than that of GZ-A patients above the ALT threshold (36.8% vs 57.4%, P = 0.016). The former was in the “truly” HBeAg-positive chronic HBV infection group due to high HBV DNA levels and low ALT levels. Similarly, GZ-C patients had a significantly higher SHI ratio than the “truly” HBeAg-negative chronic HBV infection patients (62.6% vs 48.1%, P = 0.032) (Figure 3D).

According to the new recommendations for the treatment threshold of ALT (30 U/L for males and 19 U/L for females), we investigated the rate of SHI in different groups separately. Among the 794 chronic HBV infections, nearly half of them (393/794, 49.5%) were below the new criteria (Figure 4A). Among these 393 patients, 46.0% (74/161) with HBeAg-positive chronic HBV infection, 36.7% (47/128) with GZ-A, 64.9% (111/171) with HBeAg-negative chronic HBV infection and 48.2% (161/334) with GZ-C were below this ALT threshold (Figure 4B). Notably, the proportions of significant liver inflammation (≥ G2) and SHI in patients above the ALT threshold were significantly higher than those in patients below the ALT threshold (50.9% vs 41.7%, P = 0.01; 57.4% vs 50.4%, P = 0.049) (Figure 4C). In total, the proportion of patients with ALT ≤ 40 U/L who required antiviral therapy was 75.44% [(401 + 198)/794] according to the “expert opinion on expanding anti-HBV treatment for chronic hepatitis B” in China.

Open in New Tab   Full Size Figure   Download Figure

Figure 4 Lowering the alanine aminotransferase treatment threshold to new criteria (30 U/L for men and 19 U/L for women). A: Among the 794 chronic hepatitis B virus infections, 50.5% (401/794) of patients were above the new criteria; B: Comparison of the proportions of patients exceeding the alanine aminotransferase (ALT) threshold in different groups; C: A total of 41.7% (164/393) of patients below the ALT threshold had significant liver inflammation (≥ G2); D: The proportion of significant hepatic injury (SHI) in patients below the ALT threshold was 50.4%. Comparison of the proportions of SHI in each group. NS: Not significant. ^aP < 0.05, ^bP < 0.01, ^cP < 0.001.

The SHI values in patients with HBeAg-positive chronic HBV infection below the new ALT threshold was 36.4%. However, higher SHI values of 67.9% were seen in the GZ-A patients above the new ALT threshold, and the difference was statistically significant (P < 0.001). However, there was no significant difference in the proportion of SHI between patients with HBeAg-negative chronic HBV infection below the new ALT threshold and GZ-C patients above the new ALT threshold (45.0% vs 54.9%, P = 0.105) (Figure 4D).

The median age was 31, 35, 37, and 40 years for HBeAg-positive chronic HBV infection, GZ-A, HBeAg-negative chronic HBV infection, and GZ-C patients, respectively. There was an increasing trend of age in these states (Figure 5A). Among 794 patients, 76.7% (609/794) were > 30 years old, and almost 70% were HBeAg-negative patients, including 77.8% with HBeAg-negative chronic HBV infection and 88.3% with GZ-C (Figure 5B). The ratio of SHI in patients ≤ 30 years old was 49.1%, and it increased to 55.3% for those patients > 30 years old. However, there was no significant difference (P = 0.136). A similar result was observed in all states regardless of whether they were older or younger than 30 years old (P > 0.05) (Figure 5C).

Open in New Tab   Full Size Figure   Download Figure

Figure 5 Comparison in different groups. A: Comparison of age in different groups; B: Comparison of the proportions of patients older than 30 years in different groups; C: The proportion of significant hepatic injury (SHI) in patients ≤ 30 years of age was 49.1%. Comparison of the proportions of SHI in each group. NS: Not significant. ^aP < 0.05.

In clinical practice, a considerable proportion of chronic hepatitis B (CHB) patients (27.8%-55%) fall into the “grey zone (GZ)” or “indeterminate phase”. Additionally, there is still debate regarding how best to treat these GZ patients and the advantages of antiviral therapy. Moreover, an issue that complicates the management of CHB is the disagreement regarding the appropriate treatment threshold for alanine aminotransferase (ALT) levels.

To explore the impact of varying the threshold of ALT levels in identifying significant hepatic injury (SHI) among GZ patients.

Our research evaluated the clinical and histological characteristics and additionally explored the impact of adjusting the threshold of ALT in identifying significant liver injury among GZ patients.

This retrospective analysis involved a cohort of 1617 adult patients diagnosed with CHB who underwent liver biopsy. Significant hepatic injury was defined as the presence of notable liver inflammation (≥ G2) and/or significant fibrosis (≥ S2). Kruskal-Wallis tests and Pearson’s chi-squared tests were applied to compare variables that were significantly different between groups.

The study showed that 50.22% of the patients with HBV infection fell into the GZ category, and more than half of the patients (63.7%) in the GZ category exhibited SHI. The areas under the receiver operating characteristic curves of Aspartate aminotransferase-to-platelet ratio index, fibrosis score based on four factors, and gamma-glutamyl transpeptidase-to-platelet ratio in predicting SHI were 0.717, 0.713, and 0.727, respectively, in the HBeAg-positive GZ phases and 0.717, 0.645, and 0.692, respectively, in the HBeAg-negative GZ phases. Lowering the ALT treatment thresholds to the American Association for the Study of Liver Diseases 2018 treatment criteria can more accurately identify patients with significant liver damage in the GZ phases. When we lowered the ALT treatment threshold to the new criteria, the same outcome was revealed.

This study showed that 50.22% of CHB patients were in the GZ, and over half of GZ patients (63.7%) had SHI. Lowering ALT thresholds can more accurately identify patients with significant liver damage at an earlier stage and reduce the need for some unnecessary liver biopsies. Furthermore, age may not be a limitation for initiating antiviral therapy in patients with CHB who have normal ALT levels.

Further investigation is needed to determine the assessment and treatment strategy for CHB patients in the GZ phases.