This is a cross-sectional analysis of a prospectively collected dataset for patients receiving percutaneous liver biopsy in Shanghai Ruijin Hospital. Details of the biopsy procedure, data collection, and biobank maintenance were described previously (Cao et al., 2017a; Cao et al., 2017b). All the patients provided written informed consent for liver biopsy. The indications for liver biopsies are as follows: 1) staging of known parenchymal liver disease to predict prognosis; 2) instructing on whether to start antiviral therapy based on the histologic staging of inflammation and fibrosis, especially in patients with normal or mildly elevated alanine aminotransferase (ALT); and 3) diagnosis of other liver diseases other than HBV infection. Currently, there is no established scoring system for assessing the activity of fatty liver in patients with CHB. For the purpose of analysis, a robust score, “steatosis, activity, and fibrosis (SAF) score” established (Bedossa and Consortium, 2014) and validated (Nascimbeni et al., 2020) in patients with NAFLD was used to evaluate steatosis, lobular inflammation, and ballooning degeneration. Based on the SAF score, patients were classified into three groups—NASH, NAFL, and no NAFLD—according to the fatty liver inhibition of progression (FLIP) algorithm. All the liver specimens from eligible patients were retrospectively assessed for the SAF score and FLIP classification. The current study complied with the Declaration of Helsinki and was approved by the ethics review committee of Shanghai Ruijin Hospital. Patient informed consent was waived by the committee considering the retrospective design.

All patients who received liver biopsy between January 2009 and December 2018 at our unit were assessed for eligibility. Inclusion criteria i) aged older than 16 years and ii) with positive serum hepatitis B surface antigen for at least 6 months. Exclusion criteria were i) with evidence of other chronic liver diseases including other viral hepatitis, autoimmune hepatitis, drug-induced liver disease, and primary biliary cholangitis; ii) with evidence of HCC; iii) with previous history of antiviral therapy; iv) with unreliable scoring results derived from unqualified liver samples, defined as less than 10 mm or containing less than six portal triads; and v) with excessive alcohol assumption, defined as alcohol intake of ≥20 g per day for men and ≥10 g per day for women. A flowchart of the study design is illustrated in Supplementary Figure 1 .

All patients received comprehensive clinical and laboratory assessments within 1 month before liver biopsy. Medical history of hypertension, diabetes mellitus (DM), alcohol consumption, and anthropometric parameters including body weight and body height were recorded. Overweight was defined as body mass index (BMI) ≥23 kg/m², and obesity as BMI ≥ 25 kg/m², following Asian-specific recommendations (Consultation, 2004). The upper limit of normal (ULN) for ALT was defined as ≤40 U/L according to the European Association for the Study of the Liver (EASL) criteria (European Association for the Study of the Liver. Electronic address and European Association for the Study of the, 2017).

Biopsy specimens were routinely stained with H&E and Masson trichrome stains and assessed by two liver pathologists immediately after liver biopsies. Scheuer’s scoring system was used for staging fibrosis (S0–S4) and inflammation (G0–G4) (Scheuer, 1991). Significant fibrosis was defined as S ≥ 2 and severe fibrosis as S ≥ 3. Steatosis, lobular inflammation, and cytological ballooning degeneration were reevaluated according to the SAF score by a liver pathologist blinded to clinical data/biological data. According to the SAF scoring system, steatosis was scored on a 0–3 scale, lobular inflammation on a 0–2 scale, and ballooning on a 0–2 scale (Bedossa and Consortium, 2014). Histopathological lesions in NAFLD were assessed separately in the SAF score so that the SAF score was recommended and adopted to assess NAFLD in this article (Nascimbeni et al., 2020). Moderate-to-severe steatosis was defined as intrahepatic steatosis ≥33%. Liver specimens with a length ≥10 mm or containing at least six portal triads were considered qualified for analysis, and the rest were excluded.

Continuous variables were expressed in mean ± SD or median (interquartile range [IQR]), as appropriate, whereas categorical variables were presented as number (percentage). Differences in continuous variables were examined for statistical significance using Student’s t-test or Kruskal–Wallis rank-sum test depending on data distribution. Categorical variables were analyzed with the chi-squared test or Fisher’s exact test. Correlation analyses were performed according to Spearman’s method. Logistic regression analyses were used to identify risk factors for significant fibrosis or severe fibrosis. All the variables with a p-value less than 0.1 under univariate analysis entered the stepwise selection process, and those with a p-value less than 0.05 were retained. Variables were expressed as odds ratio (OR) and 95% CI. p-Values <0.05 were considered statistically significant. We classified our patients into lean or non-lean groups according to the ethnic-specific BMI cutoff of 23 kg/m² in our current study with the Asian population. Subgroup analyses were performed by logistic regression. All data were analyzed using SAS (Version 9.4; SAS Institute Inc., Cary, NC, USA).

A total of 1,081 patients with treatment-naïve CHB were identified from 1,701 CHB patients who underwent a liver biopsy between 2009 and 2018 ( Supplementary Figure 1 ), and patient characteristics are shown in Table 1 . Among 1,081 treatment-naïve CHB patients, the median age was 37 years (IQR, 31–47), and 63.1% of them were males. Their median viral load was 5.33 log10 IU/ml (IQR, 3.74–7.18 log10 IU/ml), and 44.4% of them were HBeAg-negative patients. The overall prevalence of diabetes, hypertension, and overweight was 1.5%, 3.1%, and 50.8%, respectively. ALT elevation in this population was ubiquitous and accounted for 52.0% of all patients. The percentage of significant fibrosis (S ≥ 2) and severe fibrosis (S ≥ 3) according to Scheuer’s scoring system was 50.1% and 22.9%, respectively. The histopathological findings are illustrated in detail in Table 1 .

The FLIP algorithmic tree shows the number of patients in the different algorithmic pathways ( Figure 1A ). According to the FLIP algorithm, 62.6% (677/1,081), 28.4% (307/1,081), and 9.0% (97/1,081) CHB patients were classified as without NAFLD, with NAFL, and with NASH, respectively. Among all the patients with NAFLD, about a quarter (24.0%, n = 97) of patients were diagnosed with NASH, among whom 17.5% (17 of 97) patients were diagnosed with severe NASH ( Figure 1B ).

Compared with patients without NAFLD, patients with NAFL or NASH showed significantly higher median age, BMI, and ALT (all p < 0.01) ( Table 1 ). A higher proportion of male, diabetic, and arterial hypertensive patients were observed in patients with NAFLD than in those without (all p < 0.01) ( Table 1 ).

The proportion of different fibrosis stages between NAFL and NASH subgroups was different ( Table 1 ). Percentage of patients with significant fibrosis or severe fibrosis was significantly higher in NASH subgroups than those in NAFL subgroups (74.2% vs. 38.8%, p < 0.001, or 35.1% vs. 24.8%, p = 0.047, respectively) ( Table 1 ).

Factors associated with significant fibrosis or severe fibrosis are shown in Table 2 . The presence of NASH, overweight, age, blood platelets count, ALT, hepatitis B surface antigen level, HBeAg status, and serum HBV DNA level were significantly associated with the presence of significant fibrosis ( Table 2 ). Among these factors, the presence of NASH, overweight, age, and blood platelets count were independent predictors for significant fibrosis under a multivariable analysis. Using the same analytic strategy, we identified the presence of NASH, diabetes, overweight, blood platelets count, albumin, and HBV DNA level as independent factors associated with severe fibrosis ( Table 2 ).

Based on the significant association of significant/severe fibrosis with the presence of NASH, we further compared the distribution of fibrosis stages based on Scheuer’s scoring system between patients with and without steatosis, lobular inflammation, or cytological ballooning and found that the proportion of significant fibrosis was significantly different between patient with or without lobular inflammation (57.1% vs. 34.3%, p < 0.001) and patients with or without cytological ballooning (75.1% vs. 44.6%, p < 0.001) ( Figure 2 ). The proportion of significant fibrosis showed no difference between patients with or without steatosis (47.3% vs. 51.9%, p = 0.146) ( Figure 2 ). According to spearman correlation analyses, both the degree of lobular inflammation and degree of cytological ballooning were significantly associated with fibrosis stage (both p < 0.001) ( Table 3 ). The presence of steatosis, presence of moderate-to-severe steatosis, and degree of steatosis were not related to the fibrosis stage (all p > 0.05).

Correlation analyses of metabolic and viral factors and NASH-related histological changes were performed according to Spearman’s method ( Table 4 ). It was demonstrated that metabolic factors (including BMI, diabetes, and hypertension) were mildly correlated with the degree of steatosis or cytological ballooning but not lobular inflammation. Among viral factors, HBV DNA but not HBsAg or HBeAg status was associated with lobular inflammation. Clinical factors related to lobular inflammation only included ALT and AST.

To further analyze the association between NASH and significant fibrosis in different populations, we performed univariable analyses and multivariable analyses in subgroups (HBeAg-positive, HBeAg-negative, ALT ≤ ULN, ALT > ULN, and BMI < 23 kg/m²). The results showed that NASH was independently related to significant fibrosis in four subgroups (p = 0.004 for HBeAg-positive; p = 0.008 for HBeAg-negative; p = 0.007 for ALT ≤ ULN; p = 0.038 for ALT > ULN) ( Tables 5 and 6 ).

Moreover, 512 (47.4%) patients were classified as lean CHB, among whom 122 (23.8%) and 18 (3.5%) patients had concomitant NAFLD or NASH, respectively according to the FLIP algorithm ( Supplementary Figure 2 ). Patient characteristics and pathological features according to the presence of NAFLD or NASH are enumerated in Supplementary Table 3 . Similar to the results in the entire cohort, significant fibrosis was more frequently presented in lean patients with lobular inflammation (52.7% vs. 27.9%, p < 0.01), or with cytological ballooning (68.9% vs. 41.5%, p < 0.01) but not with steatosis (47.3% vs. 53.6%, p = 0.171) than those without ( Supplementary Figure 3 ). The OR of significant fibrosis and severe fibrosis in this specific population was 2.55 (95% CI 0.94–6.91, p = 0.065) and 3.06 (95% CI 1.15–8.13, p = 0.025), respectively.