| IntroductionHepatitis B virus (HBV) infection is epidemic worldwide, and one of the major risk factors for end-stage liver disease including liver failure, liver cirrhosis and hepatocellular carcinoma (HCC). Liver fibrosis is the common, pathological repair response to different chronic injuries in hepatic tissue, which is characterized by over-precipitation of extracellular matrix and may progress to cirrhosis finally. The five-year accumulation morbidity of liver cirrhosis in patients with chronic HBV infection is8%-20%. Additionally, liver cirrhosis is an independent risk factor for the development of HCC. The incidence of HCC in patients with liver cirrhosis is3%-6%. Evolution of liver fibrosis has been demonstrated to be correlated with the long-term risk of chronic hepatitis B (CHB) related liver disease patients. However, liver is such a silent organ that most patients may already have developed into the stage of liver cirrhosis or even liver failure when they first notice the typical clinical symptoms and signs.There are two methods of diagnosing liver fibrosis. One is non-invasive method such as liver stiffness measurement (LSM) by transient elastography (TE), which is a new ultrasound-based diagnostic method for liver fibrosis assessment. LSM by TE has obvious advantages such as being non-invasive, simple and reproducible operation, satisfying distinguishibity for advanced fibrosis and cirrhosis. LSM by TE has been validated extensively by amounting clinical studies and recently been approved for clinical use by FDA consequently. Although much progress has been made in assessment of liver fibrosis by LSM, disadvantages should not be ignored for it can not read inflammatory activity, and the mild to moderate fibrosis (F0-F2) precisely, is significantly influenced by hepatic inflammatory activity and the other conditions such as BMI, obesity and so on. Another is invasive method such as liver biopsy. However, the disadvantages of liver biopsy are low compliance and difficulty in reproducible operation for its invasive. Additionally, the small amounts of liver sampling and its error may cause inaccurate diagnosis. So liver biopsy is restrictly used, but it remains as the "gold standard" reference method for liver fibrosis assessment.So evaluating the liver histological abnormalities in CHB patients accurately play an important role in clinical diagnosis, management and prognosis of CHB related liver disease. Pathological abnormalites including fibrosis, necroinflammation and so on by liver histological analysis still could not be replaced by other clinical examinations. Regarding this, the2012Asian-Pacific Association for the Study of the Liver (APASL) guideline recommends liver biopsies should be considered to be performed in viremic CHB patients over40years old, especially those with high normal or minimally raised alanine aminotransferase (ALT), while the2009American Association for the Study of Liver Diseases (AASLD) guideline proposes the same opinion. However, the2012European Association for the Study of the Liver (EASL) guideline suggests that besides patients with fluctuated ALT, liver biopsies or even therapy should be considered to be performed in immune tolerant patients over30 years old. Thus regarding what kinds of patients should be considered for taking liver biopsy, especially among those with persistent normal ALT (PNALT), there are obviously some differences between the relevant suggestions in various guidelines. In this context,2012EASL guideline suggests that in the future research, it is necessary to consider about how to solve the issues of uncertain indications for treatment in immune tolerant patients, and HBeAg-negative patients with the level of HBV DNA below20,000IU/ml (100,000copies/ml).Serum level of ALT is the most commonly used clinical biochemical marker for evaluating the extent of hepatic tissue injury. The current standard ’normal’ value of ALT level has been set to be40U/L. The standard value was established statistically based on the population parameter of the cohort that was supposed to be liver-disease free. However, it was laterly realized that some subpopulation of it might have included the individuals with subclinical liver disease (such as chronic hepatitis C and nonalcoholic fatty liver disease) due to the screening technological limitation at that time. Additionally, many studies complaint that such a standard value failed to identify patients with mild to moderate hepatic injuries from the ones without hepatic injury. Therefore it was suggested that the current standard reference ranges for ALT level could probably underestimate the frequency of chronic liver disease and the upper limit normal (ULN) of ALT should better be revised. In this context, a European research revealed the ULN of ALT could be adjusted to30U/L for male and19U/L for female, but another study from Korea showed it might be changed to33U/L for male and25U/L for female. It would be more accurate theoretically if the ULN of ALT is revised according to the liver histological abnormalities. However there were few reports exploring the possible association between the ULN of ALT and liver histological abnormalities in CHB patients.End-stage liver disease has high risk of mortality and HBV infection is a major risk factor for it, but the underlying mechanism remains obscure. Among the possible cues, viral mutation is believed to be involved in the pathogenesis of end-stage liver disease. Relationships between the factors of HBV genotype and mutations in the precore (PC)/core promoter region and the development of end-stage liver disease have attracted a lot of interest. For example, it has been showed that genotype B was associated with the development of HBV related-acute-on-chronic liver failure (HB-ACLF), and genotype C was associated with the development of HCC. Our previous study found CHB patients with genotype B, G1896A mutation in the PC region and A1762T/G1764A mutations in the basal core promoter (BCP) region had a higher tendency to develop HB-ACLF in comparison with patients with wild type. In another study comparing the full-length HBV sequences in longitudinal patients before and after ACLF, results showed that C53T, A1846T and G1896A mutations were associated with the occurrence of HB-ACLF. In the context of liver cirrhosis development, A1762T/G1764A mutations in the BCP region may result in patients evolving from asymptomatic HBV carriers or chronic hepatitis developing into liver cirrhosis. A domestic study revealed that A1762T/G1764A, T1768A and A1846T mutations all were independent factors of liver cirrhosis. Furthermore, research has indicated that the A1762T/G1764A mutations were associated with HCC development, and the A1762T/G1764A mutations in liver tissue independently predicted postoperative survival rate in HCC. However, a recent study reported only T1674C/G mutation was specific to predict HCC. To explore the biomarkers for the severity of liver disease in CHB patients from HBV genotype and mutation, it should compare the HBV sequences from CHB patients with different liver histological abnormalities.However there were few reports on it.AimTo first observe the characteristics of liver histological abnormalities in Chinese CHB patients in a large cohort, and then explore the relationship between the ULN of ALT, HBV genotypes and BCP/PC mutations and liver histological abnormalities, and to identify the viral factors influencing histological changes in CHB paitents, which might provide experimental evidence for the evolving of CHB.MethodsThe research was divided into three independent parts and all the subjects were treatment-naive CHB patients with liver biopsy:the first cohort included675CHB patients to investigate the characteristics of liver histological changes; the second cohort included167CHB patients with PNALT to explore the association between the ULN of ALT level and liver histological abnormalities; the third cohort included219CHB patients with serum sample to investigate the relationships between HBV genotype, BCP/PC mutations and liver histological abnormalities. The CHB was diagnosed on the basis of serological markers and liver function test. HBV serological markers were detected by chemiluminescent enzyme immunoassay. Serum aminotransferase were determined by commercial kits. All patients were seronegative for hepatitis C, D or E virus. PNALT was defined by having at least three ALT values equal to or less than ULN every6-12months apart with the observation periods from18-36months and no elevated ALT at any time points prior to the liver biopsy. Liver biopsies were implemented after hospitalization and were scored using the Metavir scoring system for both necroinflammation grade and fibrosis stage. Significant histological abnormality was defined as necroinflammation grade≥A2and/or fibrosis stage≥F2. HBV DNA was extracted from200u1serum using the QIAamp DNA Blood Mini Kit according to the manufacturer’s instructions. The HBV S gene was amplified for genotyping/subgenotyping with primers BS1and Pol2for the first round PCR, and if necessary, primers BS1and P29were used for semi-nested PCR. The PCR products were analyzed by electrophoresis on2%agarose gel,stained with ethidium bromide, and sequenced by an automated DNA sequencer. The HBV BCP/PC regions were amplified using primers Is2-2and P30A for the first round PCR and primers Is2-2and HC24R for the second round. PCR products were separated on2%agarose gel and sequenced by an automated DNA sequencer. The sequence data were analyzed using Lasergene software suite V6.0, DNASTAR. HBV DNA sequences were aligned using CLUSTALW software version1.8along with HBV A-G genotype reference sequences retrieved from GenBank/EMBL/DDBJ. Phylogenetic trees were constructed by MEGA software V4. All data were analyzed using the statistical package SPSS13.0. Results were given as mean±SD or no.(%) of patients. Mann-Whitney was used for similar comparison of nonparametric data. Chisquare was used for categorical variable analysis. Logistic regression analysis was used to determine the association between variables and significant liver histological abnormalities. Two tailed P-value of<0.05was considered statistically significant.Results1) In HBeAg-positive patients (n=516), the frequency of significant necroinflammation stratified by ALT levels was1.2%(1/85) in PNALT,23.8%(30/126) in ALT1-2×ULN and51.1%(156/305) in ALT>2xULN group, while significant fibrosis was49.4%(42/85) in PNALT,69.8%(88/126) in ALT1-2xULN and81.6%(249/305) in ALT>2xULN group, respectively. Frequencies of significant histological abnormalities in patients with ALT>2xULN were much higher than those in patients with PNALT and ALT1-2xULN(both P<0.001).2) HBeAg-positive patients with PNALT (n=85) were further stratified into ALT <0.5xULN (n=14) and ALT0.5-1×ULN (n=71) subgroups. No differences were found in the frequencies of significant histological necroinflammation (0.0%vs1.4%, P=1.000) and significant fibrosis (57.1%vs47.9%, P=0.527) between low normal and high normal ALT subgroups. Patients with PNALT were also stratified into age≤30(n=77) and age>30(n=8) subgroups. Results indicated the frequency of significant necroinflammation was similar between subgroups (1.3%vs0.0%, P=1.000), but older patients had a higher frequency of significant fibrosis than younger patients (87.5%vs45.5%, P=0.058) with an almost significant statistical difference.3) In HBeAg-negative patients (n=159), the frequency of significant necroinflammation stratified by ALT levels was9.1%(5/55) in PNALT,17.8%(8/45) in ALT1-2×ULN and57.6%(34/59) in ALT>2×ULN group, while significant fibrosis was30.9%(17/55) in PNALT,73.3%(33/45) in ALT1-2×ULN and94.9%(56/59) in ALT>2×ULN group, respectively. Frequencies of significant histological abnormalities in patients with ALT>2×ULN were much higher than in patients with PNALT and ALT1-2×ULN (both P<0.001).4) HBeAg-negative patients with PNALT (n=55) were stratified into ALT <0.5xULN (n=23) and ALT0.5-1×ULN (n=32) subgroups, but no differences were found in the frequency of liver histological significant necroinflammation (0.0%vs15.6%,P=0.130) and significant fibrosis (17.4%vs40.6%, P=0.066) between low normal and high normal ALT subgroups. While patients with PNALT were stratified into age<40(n=46) and age>40(n=9) subgroups, the frequencies of significant necroinflammation (10.9%vs0.0%, P=0.578) and significant fibrosis (30.4%vs33.3%, P=1.000) were also comparable. When patients were stratified by HBV DNA levels into≤41og10copies/ml (n=50) and4-5log10copies/ml (n=5) subgroups, no differences were found in the frequency of liver significant necroinflammation and significant fibrosis between the subgroups either (6.0%vs40.0%, P=0.060;32.0%vs20.0%,P=0.963, respectively). 5) In HBeAg-positive patients with PNALT (n=85), univariate analysis indicated that no parameter was associated with significant histological abnormality. In patients with elevated ALT (n=431), univariate analysis indicated age, levels of platelet (PLT), prothrombin activity (PTA), albumin (ALB), ALT, aspartate aminotransferase (AST) and HBV DNA were associated with significant necroinflammation, while age, levels of PLT, ALB, ALT, AST and HBV DNA were associated with significant fibrosis. The multivariate analysis indicated increasing age (OR=1.042[1.009-1.076], P=0.012), higher AST (OR=1.015[1.008-1.022], P<0.001) and lower HBV DNA (OR=0.671[0.546-0.825], P<0.001) were associated with significant necroinflammation, while higher AST (OR=1.024[1.013-1.036], P<0.001), lower HBV DNA (OR=0.685[0.530-0.885], P=0.004) and ALB (OR=0.933[0.877-0.992], P=0.027) were associated with significant fibrosis in patients with elevated ALT.6) In HBeAg-negative patients with PNALT (n=55), univariate analysis indicated that no parameter was associated with significant histological abnormality. In patients with elevated ALT (n=104), levels of PTA, ALB, ALT, AST and HBV DNA were associated with significant necroinflammation, while levels of ALT and AST were associated with significant fibrosis. The multivariate analysis indicated only higher AST was associated with significant necroinflammation in patients with elevated ALT (OR=1.021[1.005-1.038], P=0.009).7) HBeAg-positive patients with PNALT (n=96) were stratified by the ULN of ALT according to Prati’s suggestion, the frequencies of significant necroinflammation (5.0%vs3.6%, P=1.000) and significant fibrosis (47.5%vs50.0%, P=0.809) were similar between the ALT≤ULN group (n=40) and the ALT>ULN group (n=56). When patients were stratified by the ULN of ALT found in Korea, no differences were found in the frequency of liver significant necroinflammation (4.7%vs3.1%, P=1.000) and significant fibrosis (51.6%vs43.7%, P=0.470) either between the ALT≤ULN group (n=64) and the ALT>ULN group (n=32).8) HBeAg-negative patients with PNALT (n=71) were stratified by the ULN of ALT according to Prati’s suggestion, the frequencies of significant necroinflammation (6.5%vs16.0%, P=0.388) and significant fibrosis (39.1%vs40.0%, P=0.943) were similar between the ALT≤ULN group (n=46) and the ALT>ULN group (n=25). When patients were stratified by the ULN of ALT found in Korea, no differences were found in the frequency of liver significant necroinflammation (8.9%vs13.3%, P=0.984) and significant fibrosis (41.1%vs33.3%, P=0,586) either between the ALT<ULN group (n=56) and the ALT>ULN group (n=15).9) There were significant differences in the frequencies of mutations in the BCP/PC regions between HBeAg-positive (n=148) and HBeAg-negative (n=71) patients. HBeAg-positive patients had a significantly higher frequency of T1753V, A1846T and G1896A compared with HBeAg-negative patients (10.1%vs19.7%, P=0.050;18.9%vs53.5%, P<0.001;20.3%vs36.6%, P=0.009, respectively), but no differences in other mutation sites. Differences in the frequencies of mutations in the BCP/PC regions were also found between genotypes B and C. Genotype C had a significantly higher frequency of T1753V, A1762T/G1764A and C1766T/T1768A (22.4%vs4.5%, P<0.001;68.2%vs24.1%, P<0.001;10.3%vs0.0%, P<0.001, respectively), but a lower frequency of G1896A (15.0%vs35.7%, P<0.001) compared with genotype B.10) In HBeAg-positive patients (n=148), univariate analysis indicated that genotype C, A1762T/G1764A and A1846T mutations were associated with significant necroinflammation, while A1762T/G1764A and G1896A mutations were associated with significant fibrosis. The multivariate analysis indicated A1846T mutation was associated with significant necroinflammation independently (OR=5.360[2.084-13.787], P<0.001), and A1762T/G1764A (OR=7.098[2.497-20.177], P<0.001) and G1896A (OR=16.816[2.150-131.553], P=0.007) mutations were associated with significant fibrosis independently. In HBeAg-negative patients (n=71), univariate analysis indicated only A1762T/G1764A mutations were associated with significant necroinflammation (OR=4.296[1.092-16.898], P=0.037), but age, sex, genotype and other mutation sites were not associated with significant histological abnormality.11) CHB patients were divided into four groups on basis of the mutation patterns in the BCP and PC regions. Group1had no mutations in either region (BCP-/PC-, n=83), group2only had a BCP mutation (BCP+/PC-, n=80), group3only had a PC mutation (BCP-/PC+, n=36), and group4had mutations in both regions (BCP+/PC+, n=20). BCP+subjects had the A1762T/G1764A double mutations, and PC+subjects had the G1896A mutation. The AST levels among different BCP/PC mutation patterns were different, with both BCP+/PC-group and BCP-/PC+group having higher level of it in contrast to the BCP-/PC-group (P=0.002and P=0.020, respectively). There were no differences in the HBV DNA levels among different BCP/PC mutation patterns, but the BCP+/PC+group had significantly lower level of HBV DNA than the BCP-/PC-group (6.13±1.091og10copies/ml vs6.69±1.80log10copies/ml, P=0.027). The frequency of HBeAg-negative patients increased in a stepwise manner in patients with the pattern sequence of BCP-/PC-(24.1%), BCP+/PC-(31.3%), BCP-/PC+(33.3%) and BCP+/PC+(70.0%), respectively (P=0.001).12) The frequency of significant fibrosis was statistically different among the four BCP/PC mutation patterns in HBeAg-positive CHB patients (n=148)(P<0.001). Almost70%patients with significant fibrosis (n=116) had BCP and/or PC mutations, while over80%patients without significant fibrosis (n=32) had neither BCP nor PC mutations. There were17patients with ALT<ULN among148HBeAg-positive patients, using BCP and/or PC mutations for detecting significant fibrosis in them had the sensitivity, specificity, positive predictive value and negative predictive value of70%,100%,100%and70%respectively.Conclusion1) Significant fibrosis is not rare in Chinese patients with PNALT. Liver fibrosis assessment should be strongly considered in these patients, especially HBeAg-positive patients over30years old.2) Significant fibrosis is very common in CHB patients with minimally raised ALT.3) The ULN of ALT recommended by Prati and found in Korea could not predict the significant histological abnormality in CHB patients with PNALT. The appropriate ULN in Chinese CHB patients needs to be specifically established.4) In HBeAg-positive CHB patients, A1846T mutation in BCP/PC regions might be associated with significant necroinflammation, while A1762T/G1764A and Gl896A mutations might be associated with significant fibrosis.5) In HBeAg-negative CHB patients, A1762T/G1764A mutations in BCP/PC regions might be associated with significant necroinflammation.6) A1762T/G1764A and/or G1896A mutations might be reliable for detection of significant fibrosis in HBeAg-positive CHB patients with normal ALT. |
PDF Full Text Request