Introduction

Hepatitis B (HBV) infection is an important worldwide health problem, and it infects about 350–400 million people in the world. It is estimated that a million people chronically infected with HBV die of cirrhosis and its complications every year [1]. Therefore, it is most important to stop the development of early-stage fibrosis before the onset of cirrhosis [2].

Although the role of liver biopsy as a diagnostic test in chronic hepatitis B (CHB) has declined, it still has increasing importance in predicting treatment decisions. Currently, the major indication for liver biopsy in CHB is to evaluate the intensity of the disease, especially in respect to both necro-inflammation (grade) and fibrosis (stage) which is quite essential for prognosis and therapeutic management [3].

Liver biopsy is the gold standard for assessment of fibrosis in patients with HBV. However, it has some disadvantages, including inter-observer and intra-observer variability in biopsy interpretation and specimen variation [4]. A standard biopsy specimen represents only about 0.0002 % of the whole liver. It has been shown that two biopsy samples taken during the same procedure have significant influence on the diagnostic performance of interpretation in patients with chronic hepatitis C (CHC) or non-alcoholic steatohepatitis (NASH) [5, 6]. However, no study evaluating the sampling variability for liver biopsy in CHB is available in the literature. Therefore, we aimed to assess the influence of collecting two liver biopsy samples during the same procedure on staging and grading of CHB.

Patients and methods

Twenty-seven patients with CHB were enrolled consecutively into the study between June 2011 and April 2012. Laboratory and clinical data were obtained from our hospital's computerized patient registry database: HBV–DNA level, hepatitis B e-antigen (Hbe-Ag) seropositivity, alanine transaminotransferase (ALT), aspartate transaminotransferase (AST), gamma–glutamyl transpeptidase, alkaline phosphatase, bilirubin, international numeric ratio, white blood cell count, platelet count, hemoglobin. Patients co-infected with other viral hepatitis and HIV or liver transplantation, clinically overt cirrhosis, Wilson disease, hemochromatosis, alcohol abuse, and/or autoimmune or cholestatic liver diseases were excluded from the study.

Liver biopsies were performed on treatment-naive CHB patients. Paired liver biopsy samples were obtained from the right and left hepatic lobes using ultrasound guidance with 17-gauge biopsy needles. Samples were fixed with 10-% formalin and embedded in paraffin. The tissue sections were processed with hematoxylin and eosin, Masson’s trichrome, and reticular fiber staining. An expert pathologist of our institution reviewed the biopsy samples. He was also blinded to the two separate biopsy specimens of the same patient to show discrepancy in staging of fibrosis between the two biopsy specimens, if present. The biopsy samples more than 18 mm long or including at least four portal tracts were accepted for pathological evaluation.

The Knodell scoring system, modified by Ishak, was used in evaluating the fibrosis staging and grading. According to the scoring system, the histological activity index (HAI) score was counted from 0 to 18 for minimal to maximal points (periportal or periseptal interface hepatitis, confluent necrosis, focal lytic necrosis, apoptosis and focal inflammation, portal inflammation), and the fibrosis score was counted from 0 to 6 for minimal to maximal points. Minimal/mild/moderate/severe HAI grading scores were defined 1–3/4–8/9–12/13–18, respectively. F0–F1 was termed “no/minimal fibrosis” and F2–F6 was termed ‘‘significant fibrosis’’.

The Statistical Package for Social Sciences (SPSS, Illinois, Chicago) 15.0 for windows was used to analyze the data. The values were reported as mean ± standard deviation (SD). For continuous variables, paired-samples tests were used to analyze the variance among groups, if appropriate. Categorical variables were compared with the Chi square test.

Results

Paired biopsy samples from 27 patients with CHB were evaluated. Median age of the patients was 43.51 ± 11.69. Fifteen patients were female, 12 patients were male. Demographic features and laboratory findings of patient and control groups are reported in Table 1.

Table 1 Demographic features and laboratory findings of patients
Full size table

All of the patients were HBs Ag positive, anti-HBc Ig G positive, anti-HBe positive and HBV DNA levels were higher than 2000 UI/ml. Serological tests for hepatitis D and hepatitis C were negative in all the patients.

Numbers of portal tract and biopsy size were equal in the two samples (right/left lobe portal tract: 7.92 ± 4.21/6.74 ± 3.73). HAI score in the right/left lobe was 4.85 ± 1.99/4.25 ± 1.85 and the difference was significant (p value = 0.04). 4/27 (14.8 %) patients had 1-point, 2/27 (7.4 %) had 2-point, and 3/27 (11.1 %) had 3-point differences in HAI score between the two samples of the same patient. On the other hand, fibrosis stages in the right/left lobe were 1.74 ± 1.12/1.81 ± 1.24 (p). Neither the difference in grading nor in fibrosis staging between the two samples was significant. 7/27 patients (25.9 %) had a 1-point difference and only 1 patient (3.7 %) had a 2-point difference in liver fibrosis stage between the two samples.

When we evaluated the pre-cirrhotic and cirrhotic patients, two patients had F3 fibrosis, which remained the same after the second examination. Three patients with F4 fibrosis had a one-point discrepancy in fibrosis score. On the other hand, 5/27 patients with either F1 or F2 had a 1-point difference. When we compared patients with F1–F2 fibrosis and patients with F3–F4 fibrosis, the difference in fibrosis score of the two samples was also not significant.

A significant HAI score (Knodell ≥ 8) rate was observed in right lobe biopsy samples (3.7 %, 1/27). On the other hand, no patient had a significant HAI score (Knodell ≥ 8) in the left-lobe biopsy samples. The rate of significant fibrosis stage (stage ≥ 2) was 66.7 % (18/27) in right-lobe biopsy samples; the same rate was observed (63.0 %, 17/27) in left-lobe biopsy samples, but the difference was not significant. The mean degree of steatosis was detected as 9.11 ± 16.07 (min: 0; max: 70). Portal tract numbers, stage and grade scores are given in Table 2.

Table 2 Number of portal tracts, length of samples, stage and grade scores
Full size table

Discussion

This study showed that the numbers of portal tracts and biopsy sizes were equal in right- and left-lobe samples. The right-lobe sample’s HAI score was higher from that of the left-lobe sample. However, the difference in terms of grading was not significant. Besides, we did find no significant difference in fibrosis staging between the two samples.

HBV infection is an important worldwide health problem, especially in developing countries. CHB is associated with a wide range of clinical conditions, from a healthy asymptomatic carrier status with a normal liver histology to severe and chronic liver diseases, including cirrhosis and hepatocellular carcinoma (HCC) [7]. There is a particular concern in the Asian region, where chronic HBV infection is common, with a chronic HBV carrier rate of approximately 10 %. About one-third of the CHB patients will die due to chronic liver disease [8]. The goal of treatment in patients with early stages is to stop the progression to cirrhosis; as such, liver biopsy is the first step of treatment.

Liver biopsy continues to have a main role as the ‘‘gold standard’’ in diagnosis and treatment of patients with CHB. In daily clinical practice, an important problem of liver biopsy is sampling variability. a liver biopsy sample is an only small amount of the liver, representing a 1:50,000 sample/whole organ ratio [9, 10]. Many studies have shown the clinical significance of sampling variability in chronic liver diseases [11–14]. In some studies, when paired samples were obtained from right and left lobes of the liver in patients with CHC, a difference of grades and/or fibrosis stages could be observed between the two samples in 20–30 % of patients [5, 15]. Similar differences were reported for non-alcoholic fatty liver disease where significant sampling variability was shown for diagnosis of NASH [16]. However, our study is the first to show liver biopsy sampling variability in CHB patients. In contrast to other research, we found no significant difference between two CHB samples in terms of grading and fibrosis staging.

According to pathological criteria, the optimal liver biopsy sample size is recommended to be at least 20 mm long, 1.4 mm wide or including 10 (at least 4–6) portal tracts [17–20]. In our study, numbers of portal tracts were appropriate to provide the basal criteria.

The most important limitation of our study is that only one pathologist evaluated the samples. The second limitation is the limited number of patients.

In conclusion, based on the findings of our study, we believe there is no need for a second biopsy in optimal sample collection conditions. This study showed that if the liver biopsy taken is the proper size, it is sufficient to predict treatment decisions in patients with hepatitis B. However, further studies with a larger number of patients are needed to compare the paired samples for grading and staging score in patients with CHB.