Background
With the dramatic rise in the prevalence of non-alcoholic fatty liver disease (NAFLD), it frequently coexists with other conditions such as alcohol consumption and viral hepatitis. In the meantime, chronic hepatitis B (CHB) caused by hepatitis B virus (HBV) infection is still one of the most common causes of chronic liver diseases in China [1]. Consequently, CHB and NAFLD are frequently observed together with an estimated 30% having hepatic steatosis among those with CHB [2].
Until now, the interplay between the two diseases has not been thoroughly evaluated. Several studies revealed that hepatic steatosis in chronic HBV infection did not appear to affect the severity of liver histology [3, 4]. Conversely, some recent researches concluded steatosis was associated with advanced fibrosis in CHB [5‐8]. Therefore, the effect of fatty liver on the natural history of chronic HBV infection still remains controversial [9].
Anzeige
With this background, we aim to compare the histologic differences between simple chronic HBV infection and HBV infection with concomitant NAFLD, and to study whether fatty liver predict severe liver histology like significant liver inflammation or fibrosis in chronic HBV infection. Furthermore, noninvasive models were developed to accurately identify significant liver inflammation and significant fibrosis in this study.
Methods
Study design and patients
A retrospective cross-sectional study was conducted at Zhejiang Provincial People’s Hospital. We included patients with chronic HBV infection over 18 years-old who underwent a liver biopsy between 2016 and 2021. The exclusion criteria were as follows: (1) excessive alcohol consumption (ethanol consumption more than 140 g in men and 70 g in women per week); (2) other types of viral hepatitis (e.g., chronic hepatitis C virus infection); (3) other causes of liver injury (e.g., drug-induced liver disease, autoimmune liver disease, or hereditary disorders); (4) pregnancy, malignancy, severe cardiopulmonary disorders, or renal dysfunction.
The natural history of chronic HBV infection has been divided into four clinical phases as follows, taking into account the clinical data of patients including presence of hepatitis B e antigen (HBeAg), Hepatitis B virus DNA (HBV DNA) levels and alanine aminotransferase (ALT) values [10]. (1) HBeAg-positive chronic HBV infection, i.e. “immune tolerant” phase: positive serum HBeAg, very high levels of HBV DNA and ALT persistently within the normal range (upper limit of normal (ULN) of 40 IU/ml). (2) HBeAg-positive CHB: positive serum HBeAg, high levels of HBV DNA and abnormal ALT values. (3) HBeAg-negative chronic HBV infection, i.e. “inactive carrier” phase: negative serum HBeAg, undetectable or low (< 2000 IU/ml) HBV DNA levels and normal ALT; Some patients in this phase may have serum HBV DNA levels > 2000 IU/ml (usually < 20,000 IU/ml) accompanied by persistently normal ALT. (4) HBeAg-negative CHB: negative serum HBeAg, moderate to high levels of HBV DNA, and elevated or fluctuating ALT levels.
This study was approved by the Ethics Committee of People’s Hospital of Hangzhou Medical College and followed the guidelines for studies in humans. Informed consents were obtained from all subjects.
Anzeige
Clinical data and liver biopsy
Clinical, demographic and laboratory data were collected from the medical records of patients. Complete blood counts, biochemical and virological (HBV DNA, the positivity of HBeAg) data were recorded as the closest results to the date on which liver biopsy was performed.
Skilled doctors performed percutaneous liver biopsy using the MAX-CORE Disposable Core Biopsy Instrument (Bard Peripheral Vascular, Inc., Mexico). The specimens were fixed, paraffin-embedded, and stained by haematoxylin and eosin (H&E) and Masson’s trichrome for further pathological evaluation by an experienced liver pathologist. All liver biopsy slides should be qualified for scoring of histologic features.
A threshold of 5% macrovesicular steatosis made a diagnosis of NAFLD [11]. Steatosis was graded as the percentage of liver parenchyma replaced by fat: (1) 5 − 33%, (2) 34 − 66%, or (3) more than 66% [12]. Lobular inflammation was scored on a scale of 0–3: (0) none, (1) mild, (2) moderate, and (3) many. The degree of portal inflammation and hepatocellular ballooning were divided as: (0) none, (1) mild inflammation or few balloon cells, and (2) prominent inflammation or ballooning. Liver inflammation (G0 to G4) and fibrosis (S0 to S4) were assessed according to the Scheuer scoring system [13]. The grades of liver inflammation were classified into the following 5 stages: G0, no inflammation; G1, inflammatory but no necrosis; G2, focal necrosis or acidophil bodies; G3, severe focal cell damage; and G4, widely bridging necrosis and piecemeal necrosis. Significant liver inflammation was defined as inflammation grade of G2 to G4. Liver fibrosis was scored as follows: S0, no fibrosis; S1, portal fibrosis without septa; S2, portal fibrosis with rare septa; S3, numerous septa without cirrhosis; and S4, cirrhosis. Fibrotic stage of S2 to S4 was considered as significant fibrosis, while fibrotic stage of S3 to S4 was defined as advanced fibrosis [14].
Statistical analysis
Statistical analyses were performed by SPSS (version 23) and Python 3.7. Continuous variables are described as the mean ± standard deviation (SD), and categorical variables are presented as numbers (percentages). Propensity score-matching (PSM) was used to adjust the potential confounding factors including age, gender, HBV DNA and the positivity of HBeAg. When evaluating differences between groups, the t-test or the chi-squared test was applied. A two-sided P value < 0.05 was considered statistically significant.
Logistic analysis was applied to assess the risk factors for significant liver inflammation, significant fibrosis and advanced fibrosis. The associated factors observed in the univariate analysis were utilized for model training. The establishments of models were based on the Scikit-Learn package of Scientific Python 3.7 libraries. A binary logistic regression was performed to predict the probability of significant liver inflammation and significant fibrosis. We used the LIBLINEAR library to carry out the computation. Area under receiver operating characteristic curve (AUROC) was used to evaluate the predictive accuracy of the model.
Results
Study Population
A total of 871 patients with chronic HBV infection were included in this study. The mean age of the study population was 39.14 (± 9.93) years, and 586 (67.28%) were male patients. Among them, 352 patients (40.41%) were HBeAg-positive. Majority of the patients (93.92%) did not receive any antiviral therapy. Hepatic steatosis was present in 255 patients (29.28%): 219 grade 1 steatosis, 33 grade 2 steatosis, and 3 grade 3 steatosis. Significant hepatic inflammation was present in 461 patients (52.62%). Significant fibrosis was observed in 527 patients (60.51%), while advanced liver fibrosis was observed in 171 patients (19.63%).
Comparisons between HBV-infected patients with and without hepatic steatosis
The characteristics of HBV-infected patients with and without hepatic steatosis are exhibited in Table 1. Compared with patients without fatty liver, chronic HBV infection with concurrent NAFLD were more likely to be male, HBeAg-negative, and to have hypertension and type 2 diabetes mellitus (T2DM). Furthermore, they had significantly higher height, weight, body mass index (BMI) and higher levels of white blood cells (WBC), albumin, total bilirubin, ALT, GGT and ALP.
Table 1
Clinical characteristics of HBV-infected patients with and without hepatic steatosis
HBV infection (N = 616) | HBV infection + NAFLD (N = 255) | P value | |
|---|---|---|---|
Age | 38.87 ± 10.19 | 39.78 ± 9.25 | 0.222 |
Gender (M/F) | 366/250 | 220/35 | < 0.001 |
Height | 164.63 ± 8.23 | 167.63 ± 7.97 | < 0.001 |
Weight | 60.69 ± 10.14 | 73.00 ± 10.71 | < 0.001 |
BMI | 22.33 ± 2.98 | 25.98 ± 3.55 | < 0.001 |
WBC | 5.61 ± 1.51 | 6.00 ± 1.37 | < 0.001 |
PLT | 184.83 ± 52.47 | 191.60 ± 52.89 | 0.084 |
ALB | 43.77 ± 3.57 | 44.54 ± 3.40 | 0.003 |
GLB | 29.22 ± 3.93 | 29.02 ± 4.00 | 0.482 |
Tbil | 15.23 ± 6.22 | 16.51 ± 6.29 | 0.006 |
ALT | 38.62 ± 32.75 | 52.02 ± 37.48 | < 0.001 |
AST | 33.30 ± 19.40 | 35.81 ± 17.58 | 0.075 |
GGT | 27.24 ± 26.64 | 42.89 ± 36.26 | < 0.001 |
ALP | 81.10 ± 24.27 | 89.68 ± 25.51 | < 0.001 |
HBV DNA (Log10 IU/mL) | 4.55 ± 2.66 | 4.23 ± 2.52 | 0.093 |
HbeAg (+) | 264 (42.86%) | 88 (34.51%) | 0.022 |
Anti-viral therapy | 36 (5.84%) | 17 (6.67%) | 0.644 |
Hypertension | 29 (4.71%) | 32 (12.55%) | < 0.001 |
T2DM | 7 (1.14%) | 13 (5.10%) | < 0.001 |
Anzeige
Compared to HBV-infected patients without fatty liver, chronic HBV infection with concurrent NAFLD had a higher severity of hepatic steatosis, hepatic inflammation especially portal inflammation, and hepatic ballooning (Table 2). There were no significant differences based on the grades of lobular inflammation or fibrosis. The prevalence of significant fibrosis was similar between the two groups. However, there was borderline difference in the presence of advanced liver fibrosis (P = 0.049).
Table 2
Histologic characteristics of HBV-infected patients with and without hepatic steatosis
HBV infection (N = 616) | HBV infection + NAFLD (N = 255) | P value | |
|---|---|---|---|
Steatosis | < 0.001 | ||
0 | 616 (100%) | 0 | |
1 | 0 | 219 (85.88%) | |
2 | 0 | 33 (12.94%) | |
3 | 0 | 3 (1.18%) | |
Lobular inflammation | 0.323 | ||
0 | 1 (0.16%) | 0 | |
1 | 389 (63.15%) | 150 (58.82%) | |
2 | 210 (34.09%) | 93 (36.47%) | |
3 | 15 (2.44%) | 12 (4.71%) | |
4 | 1 (0.16%) | 0 | |
Portal inflammation | < 0.001 | ||
1 | 360 (58.44%) | 101 (39.61%) | |
2 | 227 (36.85%) | 127 (49.80%) | |
3 | 26 (4.22%) | 26 (10.20%) | |
4 | 3 (0.49%) | 1 (0.39%) | |
Hepatic ballooning | |||
0 | 608 (98.70%) | 239 (93.73%) | |
1 | 8 (1.30%) | 9 (3.53%) | < 0.001 |
2 | 0 | 7 (2.75%) | |
3 | 0 | 0 | |
G | < 0.001 | ||
1 | 318 (51.62%) | 92 (36.08%) | |
2 | 268 (43.51%) | 135 (52.94%) | |
3 | 28 (4.55%) | 27 (10.59%) | |
4 | 2 (0.32%) | 1 (0.39%) | |
S | 0.071 | ||
0 | 11 (1.79%) | 0 | |
1 | 236 (38.31%) | 97 (38.04%) | |
2 | 259 (42.05%) | 97 (38.04%) | |
3 | 54 (8.77%) | 30 (11.76%) | |
4 | 56 (9.09%) | 31 (12.16%) | |
Significant liver fibrosis | 369 (59.90%) | 158 (61.96%) | 0.572 |
Advanced liver fibrosis | 110 (17.857%) | 61 (23.922%) | 0.049 |
Significant liver inflammation | 298 (48.38%) | 163 (63.92%) | < 0.001 |
To alleviate the effects of confounders, PSM was performed at a 1:1 ratio, yielding 254 matched pairs. Compared with HBV-infected patients without fatty liver, chronic HBV infection with concurrent NAFLD had higher possibility of having hypertension, T2DM, and had greater body weight, BMI, and greater concentrations of ALT, GGT and ALP (Table 3).
Table 3
Clinical characteristics of HBV-infected patients with and without hepatic steatosis after PSM
HBV infection (N = 254) | HBV infection + NAFLD (N = 254) | P value | |
|---|---|---|---|
Age | 39.71 ± 9.33 | 39.79 ± 9.27 | 0.332 |
Gender (M/F) | 219/35 | 219/35 | / |
Height | 167.76 ± 6.61 | 167.62 ± 7.98 | 0.796 |
Weight | 64.19 ± 9.27 | 73.01 ± 10.73 | < 0.001 |
BMI | 22.79 ± 2.94 | 25.98 ± 3.56 | < 0.001 |
WBC | 5.77 ± 1.60 | 6.01 ± 1.37 | 0.066 |
PLT | 183.13 ± 53.90 | 191.55 ± 52.99 | 0.068 |
ALB | 44.09 ± 3.45 | 44.53 ± 3.40 | 0.093 |
GLB | 28.61 ± 3.89 | 29.03 ± 4.01 | 0.212 |
Tbil | 16.10 ± 6.50 | 16.51 ± 6.31 | 0.456 |
ALT | 41.76 ± 35.98 | 52.08 ± 37.54 | 0.001 |
AST | 33.46 ± 19.05 | 35.84 ± 17.61 | 0.132 |
GGT | 31.25 ± 30.39 | 42.90 ± 36.33 | < 0.001 |
ALP | 84.67 ± 21.86 | 89.56 ± 25.49 | 0.015 |
HBV DNA (Log10 IU/mL) | 4.22 ± 2.51 | 4.25 ± 2.51 | 0.436 |
HbeAg (+) | 88 (34.65%) | 88 (34.65%) | / |
Anti-viral therapy | 16 (6.30%) | 16 (6.30%) | / |
Hypertension | 12 (4.72%) | 32 (12.60%) | 0.002 |
T2DM | 3 (1.18%) | 13 (5.12%) | 0.011 |
Anzeige
The histologic characteristics of the patients after PSM were exhibited in Table 4. It revealed that chronic HBV infection with concurrent NAFLD had a greater severity of hepatic steatosis, lobular and portal inflammation, hepatic ballooning and fibrosis. Chronic HBV infection with NAFLD is more prone to have significant inflammation and advanced liver fibrosis, while but the probability of having significant fibrosis between the two groups was similar.
Table 4
Histologic characteristics of HBV-infected patients with and without hepatic steatosis after PSM
HBV infection (N = 254) | HBV infection + NAFLD (N = 254) | P value | |
|---|---|---|---|
Steatosis | < 0.001 | ||
0 | 254 (100%) | 0 | |
1 | 0 | 218 (85.83%) | |
2 | 0 | 33 (12.99%) | |
3 | 0 | 3 (1.18%) | |
Lobular inflammation | < 0.001 | ||
0 | 1 (0.39%) | 0 | |
1 | 193 (75.98%) | 150 (59.06%) | |
2 | 59 (23.23%) | 92 (36.22%) | |
3 | 1 (0.39%) | 12 (4.72%) | |
4 | 0 | 0 | |
Portal inflammation | < 0.001 | ||
1 | 165 (64.96%) | 100 (39.37%) | |
2 | 81 (31.89%) | 127 (50.00%) | |
3 | 8 (3.15%) | 26 (10.24%) | |
4 | 0 | 1 (0.39%) | |
Hepatic ballooning | 0.001 | ||
0 | 253 (99.61%) | 238 (93.70%) | |
1 | 1 (0.39%) | 9 (3.54%) | |
2 | 0 | 7 (2.76%) | |
3 | 0 | 0 | |
G | < 0.001 | ||
1 | 148 (58.27%) | 92 (36.22%) | |
2 | 98 (38.58%) | 134 (52.76%) | |
3 | 8 (3.15%) | 27 (10.63%) | |
4 | 0 | 1 (0.39%) | |
S | 0.010 | ||
0 | 9 (3.54%) | 0 | |
1 | 105 (41.34%) | 97 (38.19%) | |
2 | 100 (39.37%) | 97 (38.19%) | |
3 | 20 (7.87%) | 29 (11.42%) | |
4 | 20 (7.87%) | 31 (12.20%) | |
Significant liver fibrosis | 140 (55.12%) | 157 (61.81%) | 0.126 |
Advanced liver fibrosis | 40 (15.75%) | 60 (23.62%) | 0.034 |
Significant liver inflammation | 106 (41.73%) | 162 (63.78%) | < 0.001 |
Factors associated with significant liver inflammation, significant fibrosis and advanced liver fibrosis
Univariate analysis showed hepatic steatosis, ALT, AST, GGT, ALB, GLB, PLT and the positivity of HBeAg were predictors of significant liver inflammation. The above characteristics entered into the subsequent multivariate analysis, revealing that the following seven characteristics were associated with higher odds of having significant inflammation in chronic HBV infection: steatosis, decreased ALT and ALB, elevated AST, GGT and GLB, and the positivity of HBeAg (Table 5).
Table 5
Factors associated with significant liver inflammation
Variables | Univariate | Multivariate | |||
|---|---|---|---|---|---|
P value | β ± SE | Wald | OR (95% CI) | P value | |
Age | 0.544 | ||||
Gender | 0.690 | ||||
HBV DNA (Log10 IU/mL) | 0.863 | ||||
HbeAg (+) | 0.031 | 0.436 ± 0.157 | 7.760 | 1.547 (1.138–2.103) | 0.005 |
BMI | 0.242 | ||||
Hypertension | 0.377 | ||||
T2DM | 0.756 | ||||
WBC | 0.105 | ||||
PLT | 0.053 | -0.003 ± 0.001 | 3.644 | 0.997 (0.994-1.000) | 0.056 |
ALB | 0.003 | -0.080 ± 0.023 | 11.827 | 0.923 (0.881–0.966) | 0.001 |
GLB | 0.013 | 0.057 ± 0.021 | 7.602 | 1.059 (1.017–1.103) | 0.006 |
Tbil | 0.082 | ||||
ALT | 0.038 | -0.013 ± 0.005 | 5.500 | 0.987 (0.977–0.998) | 0.019 |
AST | < 0.001 | 0.054 ± 0.012 | 21.319 | 1.055 (1.031–1.079) | < 0.001 |
GGT | 0.041 | 0.008 ± 0.004 | 4.606 | 1.008 (1.001–1.015) | 0.032 |
ALP | 0.178 | ||||
Steatosis | 0.001 | 0.750 ± 0.176 | 18.210 | 2.117 (1.500-2.988) | < 0.001 |
Anzeige
Both univariate and multivariate analyses demonstrated associated factors for higher probability of significant fibrosis in chronic HBV infection were as follows: decreased levels of HBV DNA, hypertension, decreased ALT, elevated AST, GGT and GLB (Table 6).
Table 6
Factors associated with significant liver fibrosis
Variables | Univariate | Multivariate | |||
|---|---|---|---|---|---|
P value | β ± SE | Wald | OR (95% CI) | P value | |
Age | 0.196 | ||||
Gender | 0.211 | ||||
HBV DNA (Log10 IU/mL) | 0.004 | -0.127 ± 0.029 | 19.317 | 0.881 (0.832–0.932) | < 0.001 |
HbeAg (+) | 0.706 | ||||
BMI | 0.769 | ||||
Hypertension | 0.028 | 0.890 ± 0.341 | 6.802 | 2.434 (1.247–4.751) | 0.009 |
T2DM | 0.727 | ||||
WBC | 0.194 | ||||
PLT | 0.086 | ||||
ALB | 0.712 | ||||
GLB | 0.037 | 0.040 ± 0.020 | 3.977 | 1.041 (1.001–1.083) | 0.046 |
Tbil | 0.952 | ||||
ALT | 0.002 | -0.017 ± 0.005 | 12.163 | 0.983 (0.974–0.993) | < 0.001 |
AST | < 0.001 | 0.045 ± 0.010 | 19.461 | 1.046 (1.025–1.067) | < 0.001 |
GGT | 0.059 | 0.007 ± 0.003 | 4.711 | 1.007 (1.001–1.083) | 0.030 |
ALP | 0.952 | ||||
Steatosis | 0.784 | ||||
Table 7 showed factors related to higher risk of advanced liver fibrosis in chronic HBV infection included male gender, decreased levels of HBV DNA, decreased levels of PLT and ALT, and elevated levels of AST, GGT and GLB.
Table 7
Factors associated with advanced liver fibrosis
Variables | Univariate | Multivariate | |||
|---|---|---|---|---|---|
P value | β ± SE | Wald | OR (95% CI) | P value | |
Age | 0.067 | ||||
Gender (if male) | 0.036 | 0.644 ± 0.232 | 7.720 | 1.904 (1.209–2.998) | 0.005 |
HBV DNA (Log10 IU/mL) | 0.001 | -0.139 ± 0.038 | 13.350 | 0.871 (0.808–0.938) | < 0.001 |
HbeAg (+) | 0.099 | ||||
BMI | 0.465 | ||||
Hypertension | 0.560 | ||||
T2DM | 0.576 | ||||
WBC | 0.554 | ||||
PLT | < 0.001 | -0.010 ± 0.002 | 24.919 | 0.990 (0.987–0.994) | < 0.001 |
ALB | 0.903 | ||||
GLB | < 0.001 | 0.105 ± 0.024 | 19.018 | 1.111 (1.059–1.164) | < 0.001 |
Tbil | 0.338 | ||||
ALT | 0.031 | -0.011 ± 0.005 | 5.384 | 0.989 (0.980–0.998) | 0.020 |
AST | 0.007 | 0.023 ± 0.008 | 7.211 | 1.023 (1.006–1.040) | 0.007 |
GGT | 0.005 | 0.010 ± 0.003 | 9.924 | 1.010 (1.004–1.016) | 0.002 |
ALP | 0.533 | ||||
Steatosis | 0.229 | ||||
Liver histology based on the natural history of chronic HBV infection
The 818 patients who haven’t receive any antiviral therapy were divided into four disease phases of chronic HBV infection according to their clinical data. Among them, 238 patients had concurrent NAFLD, while the other 580 did not. We compared the presence of significant liver inflammation, significant fibrosis and advanced liver fibrosis between the two groups with or without NAFLD according to the natural history of chronic HBV infection. Patients with normal ALT (chronic HBV infection with no evidence of CHB) and elevated ALT values (CHB) were also analyzed separately.
As shown in Table 8, in patients with HBeAg-positive and HBeAg-negative chronic HBV infection (immune tolerant and inactive carrier phase), that is, in those HBV-infected patients with persistent normal ALT value, the presence of significant liver inflammation was higher in NAFLD group than that in the non-NAFLD group (68.750% vs. 43.571%, P = 0.011; 59.302% vs. 35.685%, P < 0.001; 61.864% vs. 38.583, P < 0.001 in chronic HBV infection with no evidence of CHB). Furthermore, in patients with HBeAg-positive chronic HBV infection (immune tolerant phase), the presence of advanced liver fibrosis was higher in NAFLD group than that in the non-NAFLD group (25% vs. 10.714%, P = 0.044).
Table 8
Liver histology based on the natural history of chronic HBV infection
Phase of chronic HBV infection | NAFLD or not | Total | Significant liver inflammation (%) | P value | Significant liver fibrosis (%) | P value | Advanced liver fibrosis (%) | P value |
|---|---|---|---|---|---|---|---|---|
HBeAg-positive chronic HBV infection | Non-NAFLD | 140 | 61 (43.571) | 0.011 | 66 (47.143) | 0.434 | 15 (10.714) | 0.044 |
NAFLD | 32 | 22 (68.750) | 18 (56.25) | 8 (25) | ||||
HBeAg-positive CHB | Non-NAFLD | 111 | 78 (70.270) | 0.455 | 64 (57.658) | 1.000 | 18 (16.216) | 0.508 |
NAFLD | 51 | 39 (76.471) | 29 (56.863) | 11 (21.569) | ||||
HBeAg-negative chronic HBV infection | Non-NAFLD | 241 | 86 (35.685) | < 0.001 | 150 (62.241) | 1.000 | 44 (18.257) | 0.749 |
NAFLD | 86 | 51 (59.302) | 53 (61.628) | 17 (19.767) | ||||
HBeAg-negative CHB | Non-NAFLD | 88 | 57 (64.773) | 0.325 | 61 (69.318) | 0.397 | 19 (21.591) | 0.848 |
NAFLD | 69 | 39 (56.522) | 43 (62.319) | 16 (23.188) | ||||
Chronic HBV infection with no evidence of CHB | Non-NAFLD | 381 | 147 (38.583) | < 0.001 | 216 (56.693) | 0.524 | 59 (15.486) | 0.160 |
NAFLD | 118 | 73 (61.864) | 71 (60.169) | 25 (21.186) | ||||
CHB | Non-NAFLD | 199 | 135 (67.839) | 0.625 | 125 (62.814) | 0.636 | 37 (18.593) | 0.471 |
NAFLD | 120 | 78 (65) | 72 (60) | 27 (22.5) |
Subgroup analysis based on metabolic factors
To increase the strength of the correlation between the virus and NAFLD, we make a list of subgroup analyses (Table 9). Firstly, patients with T2DM, hypertension and overweight (defined as BMI ≥ 23 kg/m2) were excluded (Subgroup 1). In this subgroup of a total of 409 patients, 41 had concurrent NAFLD, while the other 368 did not. The presence of significant liver inflammation in HBV-infected patients with NAFLD but no metabolic factors is borderline higher than those without NAFLD (60.976% vs. 45.109%, P = 0.069), but the prevalence of significant or advanced liver fibrosis between the two groups was comparable.
Table 9
Subgroup analysis based on metabolic factors
Subgroups | NAFLD or not | Total | Significant liver inflammation (%) | P value | Significant liver fibrosis (%) | P value | Advanced liver fibrosis (%) | P value |
|---|---|---|---|---|---|---|---|---|
Subgroup 1 | Non-NAFLD | 368 | 166 (45.109) | 0.069 | 212 (57.609) | 0.623 | 59 (16.033) | 1.000 |
NAFLD | 41 | 25 (60.976) | 22 (53.659) | 6 (14.634) | ||||
Subgroup 2 | Non-NAFLD | 377 | 170 (45.093) | 0.035 | 218 (57.825) | 0.871 | 62 (16.446) | 0.670 |
NAFLD | 43 | 27 (62.791) | 24 (55.814) | 8 (18.605) |
We observed the presence of steatosis in lean individuals (BMI < 23 kg/m2) with HBV infection in this study was approximately 10%, similar with the overall prevalence of steatosis in lean individuals according to literature [15]. There were 420 lean HBV-infected subjects totally, and liver histology was compared in the NAFLD and non-NAFLD group (Subgroup 2). NAFLD group had higher prevalence of significant liver inflammation than the non-NAFLD group (62.791% vs. 45.093%, P = 0.035), but the presence of significant or advanced liver fibrosis between the two groups was similar.
Noninvasive models for significant liver inflammation and significant fibrosis
All included patients with and without significant liver inflammation or significant fibrosis were randomly selected with a ratio of 1:1. 80% of the patients were included in the derivation cohort, while the other 20% were included in the validation cohort.
The derivation cohort for significant inflammation was composed of 656 patients. Model score = 0.00592 + 0.86997 * AST − 0.40814 * ALT + 0.19083 * GGT − 0.32590 * ALB + 0.29954 * GLB + 0.31760 * (if steatosis) + 0.17445 * (if HbeAg positive) − 0.16393 * PLT. The AUROC of the model was 0.735. In the validation cohort of 164 patients, the AUROC was 0.825 (Fig. 1). At a cut-off value of 0.478, the model could determine significant hepatic inflammation with a sensitivity of 0.705, a specificity of 0.810, and a diagnostic accuracy of 0.758.
Fig. 1
Receiver-operating characteristic (ROC) curves of the predictive model for significant liver inflammation, with the area under the ROC curve of 0.825
×
![]()
The derivation cohort for significant fibrosis was consisted of 555 patients. A model was developed with an AUROC of 0.702. Model score = 0.08888 + 0.20617 * (if hypertension) − 0.16412 * ALT + 0.32559 * AST + 0.45615 * GGT + 0.24697 * GLB − 0.41932 * Log10 HBV DNA. In the validation cohort of 110 patients, the AUROC was 0.760 (Fig. 2). At a cut-off value of 0.500, the model could diagnose significant fibrosis in CHB with a sensitivity of 0.725, a specificity of 0.714, and a diagnostic accuracy of 0.719.
Fig. 2
Receiver-operating characteristic (ROC) curves of the predictive model for significant fibrosis, with the area under the ROC curve of 0.760
×
![]()
Discussion
The current study highlighted the pathological findings in HBV-infected patients with and without concurrent NAFLD. HBV-infected patients with fatty liver had a higher severity of hepatic steatosis, hepatic inflammation, hepatic ballooning and a higher probability of advanced liver fibrosis than patients with simple chronic HBV infection. Fatty liver was not a risk factor for significant or advanced fibrosis, but it could independently predict significant liver inflammation in chronic HBV infection. Especially, in patients with either HBeAg-positive or HBeAg-negative chronic HBV infection, that is, in HBV-infected patients with persistent normal ALT, the presence of significant liver inflammation was higher in NAFLD than those without NAFLD. The prevalence of advanced liver fibrosis was higher in NAFLD than non-NAFLD group only in HBeAg-positive chronic HBV infection, while NAFLD did not increase fibrosis burden in other stages of HBV infection. Furthermore, we developed noninvasive models for significant liver inflammation and significant fibrosis with good diagnostic performance.
In 2020, a group of experts suggested the nomenclature of NAFLD should be updated to metabolic associated fatty liver disease (MAFLD) [16, 17], because metabolic liver disease usually coexists with other conditions such as viral hepatitis, and should not be described as a condition of “exclusion” [18]. We need to pay attention to patients with dual aetiology fatty liver disease who meet the criteria of NAFLD and who also have other concomitant condition. It is generally recognized that both NAFLD and HBV infection are common types of chronic hepatitis, and both can cause cirrhosis, hepatic failure and hepatocellular carcinoma [19, 20]. The coexistence of NAFLD and chronic HBV infection happens frequently. Therefore, it’s of great importance to explore the relationship between hepatic steatosis and HBV infection.
The interaction between NAFLD and HBV infection is complex and unclear. HBV infection might be related to decreased risk of NAFLD [2, 21], but the mechanisms whereby HBV influences steatosis has not been well understood. On the other hand, how hepatic steatosis influences the clinical outcomes of chronic HBV infection is not entirely clear. Wong et al. has found fatty liver, measured by controlled attenuation parameter (CAP), is associated with advanced fibrosis [6]. Mak’s study demonstrated hepatic steatosis promoted fibrosis progression in virologically quiescent CHB [22]. Another study by Seto et al. revealed that severe steatosis was related to severe fibrosis in both treatment-naive and on-treatment patients with HBV infection [5]. However, the definition of liver fibrosis in the above studies was determined by liver stiffness measurement (LSM) under transient elastography. As we know, moderate to severe hepatic steatosis might result in overestimation of LSM in HBV-infected patients, which should be considered seriously to avoid misdiagnosing fibrosis [23]. Therefore, histology by liver biopsy is always the golden standard. Even though in studies reporting the relationship between concomitant NAFLD and the severity of liver histology in patients based on liver biopsy, the results were not consistent [3, 4, 7].
In this study, we found that HBV-infected patients with concomitant NAFLD had a higher severity of portal inflammation and hepatic ballooning, and were more likely to develop significant liver inflammation and advanced fibrosis (borderline difference). To lighten the effects of confounding factors including age, gender and virological profiles, we performed further evaluation in 254 matched pairs. The results showed that chronic HBV infection with concurrent NAFLD had a greater severity of lobular and portal inflammation, hepatic ballooning and fibrosis, and were more likely to have significant inflammation and advanced fibrosis than simple HBV infection. But the probability of having significant fibrosis between the two groups was similar. Multivariate logistic analyses identified that steatosis was an independent predictor for only significant liver inflammation but not significant or advanced fibrosis in chronic HBV infection. While histologic alterations represent a continuous process, and persistent liver inflammation will definitely lead to the progression of fibrosis [24], it’s better to carry out a longitudinal cohort study to explore whether concomitant NAFLD in HBV-infection could promote liver fibrosis.
As we know, the natural history of HBV infection follows four disease phases with different levels of viral replication and dynamics in liver disease progression. As reported in this study, the influence of concurrent NAFLD on liver histology is different in distinct clinical phases in chronic HBV infection. Briefly, NAFLD could aggravate liver inflammation in HBV-infected patients with persistent normal ALT (immune tolerant and inactive carrier phase), and NAFLD increased the burden of advanced fibrosis only in the immune tolerant phase. This is a pretty novel and interesting finding. We should further explore the inside mechanism relating the interaction between NAFLD and distinct clinical phases in chronic HBV infection.
HBV replication and subsequent liver inflammation and fibrosis account for disease progression in CHB. Significant liver necroinflammation (grade G ≥ 2) and significant fibrosis (stage S ≥ 2) by liver histology greatly increase the risk of cirrhosis, hepatocellular carcinoma and end-stage liver diseases. As a consequence, early and timely antiviral therapy is recommended in these situations. Nevertheless, as the only golden standard for evaluating significant liver inflammation and fibrosis, liver biopsy is limited in clinical practice due to its invasive nature. Current noninvasive predictions of liver fibrosis mainly include LSM and biomarkers like Chitinase 3-like 1 (CHI3L1) [25, 26], Golgi protein 73 [27], et al. The detection of the newly-developed noninvasive markers has not yet been applied widely and their diagnostic accuracies are still limited. LSM could be impacted by a variety of factors like BMI, waist circumference, fatty liver, skin capsular distance and so on [28, 29]. Central obesity and fatty liver will lead to an overestimation of LSM in viral hepatitis including CHB [23] and chronic hepatitis C [30]. Given the gradually increasing prevalence and severity of NAFLD in HBV-infected patients, it’s necessary and has always been a research hotspot to develop noninvasive diagnostic models for significant liver inflammation and fibrosis. Our models were based on routine widely available clinical and laboratory parameters, and have presented with good diagnostic performance.
Despite our best effort, there were still some limitations in this study. Firstly, it was limited to a cross-sectional study, which was unable to evaluate the clinical outcomes of patients with concurrent NAFLD with HBV infection. Long-period follow-up and repeated liver biopsy are of great necessity to understand the influence of progression or regression of fatty liver on the histologic characteristics and clinical outcomes of the disease. Secondly, lack of data on more metabolic factors such as waist circumference and insulin resistance could also affect the interpretation of results. Last but not least, the mechanism behind interactions between NAFLD and HBV infection have not been explored in this study. As literatures report that specific diet and drugs can prevent the progression of chronic HBV infection and NAFLD [31, 32], further mechanism research is warranted in the purpose of reducing the synergistic negative effects of HBV and NAFLD, alleviating liver injury, and preventing decompensation events.
Conclusions
In conclusion, HBV-infected patients with concomitant NAFLD had a higher severity of liver inflammation, and a higher probability of significant liver inflammation and advanced fibrosis than simple chronic HBV infection. Hepatic steatosis could independently predict significant liver inflammation in nondrinker patients with chronic HBV infection, but it was not a risk factor for significant or advanced fibrosis. NAFLD could aggravate liver inflammation in the immune tolerant and inactive carrier phase of HBV infection, and increase the burden of advanced fibrosis only in the immune tolerant phase. In addition, noninvasive models established in this study could efficiently predict significant liver inflammation and significant fibrosis in patients with chronic HBV infection.
Acknowledgements
Not applicable.
Declarations
Competing interests
The authors declare no competing interests.
Ethics approval and consent to participate
This study was approved by the Ethics Committee of People’s Hospital of Hangzhou Medical College (2021QT257) and followed the guidelines for studies in humans. Informed consents were obtained from all subjects.
Consent for publication
Not applicable.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.