| |
Significance of HBV DNA Levels in Liver Histology of HBeAg and Anti-HBe Positive Patients with Chronic Hepatitis B
|
| |
| |
American Journal of Gastroenterology
Volume 99 Issue 10 Page 2032 - October 2004
Man-Fung Yuen, M.D., Irene Oi-Lin Ng, M.D., Sheung-Tat Fan, M.D., He-Jun Yuan, Ph.D., Danny Ka-Ho Wong, M.Sc., John Chi-Hang Yuen, B.Sc., Simon Siu-Man Sum, B.Sc., Annie On-On Chan, M.D., and Ching-Lung Lai, M.D.
Department of Medicine; Department of Pathology; Department of Surgery, The University of Hong Kong, Queen Mary Hospital, Hong Kong
ABSTRACT
OBJECTIVE: To determine the relationship between hepatitis B virus (HBV) DNA levels and total histologic activity index (HAI), necroinflammation (HAI-NI), and fibrosis (HAI-F) scores.
PATIENTS AND METHODS: Liver histology and HBV DNA levels were determined in 94 patients with chronic hepatitis B.
RESULTS: There was no association between HBV DNA levels and liver histology in hepatitis-B-e antigen-positive patients (n = 43). In anti-HBe-positive patients (n = 51), HBV DNA levels correlated positively with HAI-NI (r = 0.31, p= 0.014) and HAI-F (r = 0.33, p= 0.017) scores. Though the majority of anti-HBe-positive patients with HBV DNA levels <105 copies/ml had mild necroinflammation and no fibrosis, 14.3% had established fibrosis. Anti-HBe-positive patients with core promoter mutations had a poorer histology compared to those without. There was no difference in the histology between anti-HBe-positive patients with and without precore mutations. Alanine aminotransferase (ALT) level correlated positively with HAI-NI score. Patients with persistently normal ALT levels had a significantly lower median HAI-NI score compared to patients with either persistently or intermittently elevated ALT levels.
CONCLUSIONS: In anti-HBe-positive patients, though HBV DNA level <105 copies/ml (100,000 copies/ml) was associated with better histology, 14.3% patients had established fibrosis. Further studies to define a better cut-off HBV DNA level to differentiate low- and high-risk patients for disease progression are required.
BACKGROUND
Chronic hepatitis B (CHB) infection causes a spectrum of different diseases ranging from clinically asymptomatic carrier state to the development of cirrhosis-related complications and hepatocellular carcinoma. Factors determining the clinical outcome in CHB patients still remain unknown. Hepatitis-B-e antigen (HBeAg) seroconversion was previously believed to be associated with the favorable long-term prognosis. However, there is emerging evidence indicating that in the Asian and Mediterranean population, disease progression is still possible in patients after HBeAg seroconversion, i.e., patients with antibody to HBeAg (anti-HBe). Measuring hepatitis B virus (HBV) DNA level becomes the most important serologic marker to study the natural history of CHB and assess treatment efficacy. Niitsuma et al. suggest HBV DNA levels of <10 copies/ml to be the goal for successful treatment of CHB. However, with the currently licensed and other novel nucleoside analogues, this theoretical level may not be achievable in the near future. The HBV DNA level that is associated with a better prognosis for patients with CHB remains an unsettled issue.
In the National Institutes of Health Workshop on "Management of Hepatitis B" held in September 2000, HBV DNA level of <105 copies/ml was arbitrarily chosen as the cut-off level to differentiate patients with active disease and inactive disease. The reason this level was chosen at that time was based on the fact that most of the non-PCR assays could not detect HBV DNA levels <105 copies/ml (100,000 copies/ml). With the advancement in molecular technology, the lower limit of detection of HBV DNA level improves from levels above 105 as measured by hybridization assays to 50 copies/ml as measured by quantitative PCR assay. Studies adopting these sensitive assays would help to determine whether the cut-off HBV DNA level should be set at 105 copies/ml. A more practical approach to determine the cut-off HBV DNA level, is to study the relationship between HBV DNA levels and liver histology.
Measuring the alanine aminotransferase (ALT) levels remains the most common and convenient way to reflect the liver inflammation of CHB patients, though the correlation is not ideal. The relationship between the ALT level and the liver histology may be better established by serial ALT levels rather than by a single ALT level measured at the time of liver biopsy.
The primary aim of the present study was to deter mine whether there was any relationship between HBV DNA levels and the degree of hepatic necroinflammation and fibrosis in both HBeAg- and anti-HBe-positive patients. If the relationship was affirmative, we would determine the HBV DNA level below which there was a significantly lesser degree of hepatic necroinflammation and fibrosis. The secondary aim was to examine the relationship between serial ALT levels and the degree of hepatic necroinflammation and fibrosis.
PATIENTS & METHODS
All CHB patients (n = 94) undergoing liver biopsy during the period of March 2001 and July 2003 in Queen Mary Hospital, The University of Hong Kong, Hong Kong, were recruited. None of the patients received any treatment for HBV prior to the biopsies. Patients must have at least one follow-up in the Hepatitis Clinic, Queen Mary Hospital, The University of Hong Kong, Hong Kong, before institution of therapy. All patients had informed consents for the liver biopsy procedure. The present study had been approved by the Ethics Committee, Institute Review Board, The University of Hong Kong and Queen Mary Hospital, Hong Kong. Patients were positive for hepatitis B surface antigen (HBsAg) for at least 6 months. Patients with concomitant liver diseases including chronic hepatitis C or D, alcoholic liver disease, autoimmune hepatitis, Wilson's disease, primary biliary cirrhosis, and drug-induced hepatitis were excluded. HBsAg, HBeAg, and anti-HBe were determined by the microparticle enzyme immunoassay, Abbott Laboratories, Chicago, IL. The liver biochemistry was measured on presentation, during subsequent follow-up, and within 2 wk prior to the liver biopsy. Serum taken 2 wk prior to the liver biopsy was also subjected to the measurement of HBV DNA levels by the Cobas Amplicor HBV Monitor test, Roche Diagnostics, Branchburg, NJ (lower limit of detection of 200 copies/ml).
The liver histology was assessed by a pathologist who was "blind" as to the results of the liver biochemistry and HBV DNA levels. The histology was graded by the histologic activity index (HAI) according to the criteria of Knodell et al. The total HAI score comprises two major components, namely necroinflammation (HAI-NI) and fibrosis (HAI-F). The HAI-NI was scored according to the following parameters: piecemeal necrosis, score 0-10; lobular necrosis and inflammation, score 0-4; and portal inflammation, score 0-4. The HAI-F was scored according to the different severity: absence of fibrosis, score 0; fibrous portal expansion, score 1; bridging fibrosis, score 3; and cirrhosis, score 4.
Core promoter and precore mutations of anti-HBe-positive patients were determined by a line probe assay, INNO-LiPA HBV Precore, Innogenetics NV, Gent, Belgium, as described previously.
RESULTS
Demographics
Of the 94 patients recruited, 43 (46.9%) patients were HBeAg positive and 51 (54.3%) patients were anti-HBe positive at the time of liver biopsy. Fourteen (14.9%) patients had normal ALT levels at that time. The demographics, liver biochemistry, and HBV DNA levels of all the patients are listed in Table 1. HBeAg-positive patients had a significantly younger median age and a higher median HBV DNA level compared to anti-HBe-positive patients (p= 0.001 and p< 0.001, respectively).
Relationship between HBV DNA Levels and HAI Scores
For HBeAg-positive patients, there were no correlations between the HBV DNA levels and the HAI-NI, HAI-F, and total HAI scores (p= 0.91, 0.88, and 0.93, respectively). There was no difference in the median HBV DNA level between patients (n = 27) with high-grade necroinflammation (defined as HAI-NI score >= 7) and patients (n = 16) with low-grade necroinflammation (defined as HAI-NI score < 7) [9.6 x 108 (range: 1.9 x 106-2 x 1013) vs 1.1 x 109 (range: 1.1 x 107-1.1 x 1013), respectively, p= 1.0]. Similarly, there was also no difference in the median HBV DNA level between patients (n = 20) with high-grade fibrosis (defined as HAI-F score >= 3) and patients (n = 23) with low-grade fibrosis (defined as HAI-F score <= 1) [8.2 x 108 (range: 1.9 x 106-9.2 x 1012) vs 1.1 x 109 (range: 1.1 x 107-2 x 1013) copies/ml, respectively, p= 0.39].
For anti-HBe-positive patients, HBV DNA levels correlated positively with the HAI-NI (r = 0.31, p= 0.014), HAI-F (r = 0.33, p= 0.017), and total HAI (r = 0.37, p= 0.008) scores. Patients (n = 25) with high-grade necroinflammation had a significantly higher median HBV DNA level compared to the patients (n = 26) with low-grade necroinflammation [1.8 x 107 (range: 2.9 x 105-8.5 x 108) vs 5.4 x 106 (range: <200-1.4 x 109), respectively, p= 0.009]. Patients (n = 29) with high-grade fibrosis also had a significantly higher median HBV DNA level compared to the patients (n = 22) with low-grade fibrosis [1.8 x 107 (range: 2,900-1.4 x 109) vs 6.1 x 106 (range: <200-6.4 x 107), respectively, p= 0.01].
The anti-HBe-positive patients were classified according to the HBV DNA levels of <105, <106, and <107 copies/ml. The majority of the anti-HBe-positive patients (85.7%) with HBV DNA levels <105 copies/ml had minimal necroinflammation (HAI-NI score < 3). Though 6 out of these 7 patients (85.7%) also had no fibrosis, one patient (14.3%) had cirrhosis (fibrosis score of 4).
Relationship between Core Promoter, Precore Mutations, and HAI Scores
Of the 51 anti-HBe-positive patients, 39 patients had the serum determined for the presence of core promoter and precore mutations. Comparing patients without core promoter mutations (n = 10), patients with core promoter mutations (n = 29) had a significantly higher median HAI-NI score [2.5 (range: 1-9) vs 7 (range: 0-11), respectively, p= 0.022], HAI-F score [0 (range: 0-3) vs 3 (range: 0-4), respectively, p= 0.028], and total HAI score [3 (range: 1-12) vs 10 (range: 0-14), respectively, p= 0.031]. However, there was no difference in the median HBV DNA level between the two groups [6.1 x 106 (range: 1.1 x 104-4.9 x 108) vs 1.3 x 107 (range: <200-1.4 x 109) copies/ml, respectively, p= 0.69].
Comparing patients with (n = 18) and without (n = 21) precore mutations, there were no differences in the median HAI-NI score [5 (range: 1-9) vs 5 (range: 0-11), respectively, p= 0.88], HAI-F score [1 (0-4) vs 3 (0-4), respectively, p= 0.43], and total HAI score [5.5 (1-12) vs 8 (range: 0-14), respectively, p= 0.99]. There was also no difference in the median HBV DNA level between the two groups [1.3 x 107 (range: <200-4.9 x 108) vs 8.3 x 106 (range: 11,700-1.4 x 109) copies/ml, respectively, p= 0.69].
Relationship between Liver Biochemistry at the Time of Liver Biopsy and HAI Scores
For HBeAg-positive patients, there were positive correlations between ALT levels and the HAI-NI (r = 0.43, p= 0.004) and total HAI (r = 0.40, p= 0.009) scores. Similarly, in anti-HBe patients, ALT levels also positively correlated with the HAI-NI (r = 0.35, p= 0.011) and total HAI (r = 0.31, p= 0.028) scores). However, there was no correlation between the ALT level and the HAI-F score in either HBeAg- or anti-HBe-positive patients.
The albumin level correlated inversely with the HAI-F score in both HBeAg-positive patients (r =-0.37, p= 0.016) and anti-HBe-positive patients (r =-0.49, p< 0.001). In the anti-HBe-positive patients, the bilirubin level was positively correlated with the HAI-F score (r = 0.44, p= 0.001).
Relationship between Serial ALT Levels and HAI Scores
Of the 94 patients, 56 patients had serial ALT level monitoring for at least three follow-ups (scheduled at 3-6 months interval) before the liver biopsies. The number of ALT readings during the follow-up are as follows: six patients had three serial ALT levels monitoring, 13 patients had 4-6, 18 patients had 7-9, and 19 patients had 10-12. These patients were classified into two groups according to the pattern of the ALT profiles. Group 1 (n = 33) represented patients with a pattern of persistently normal or elevated ALT levels and group 2 (n = 23) represented patients with a pattern of fluctuating ALT levels. The details of the classification according to the ALT levels and the respective median HAI-NI, HAI-F, and total HAI scores are listed in Table 3. Two patients in group 1 with persistently elevated ALT levels 5-10 x upper limit of normal (ULN) and one patient in group 2 with fluctuating ALT levels 5-10 x ULN were not listed in Table 3 because of the limited number of patients in these two subgroups. The only statistical difference detected was in the eight patients with ALT level persistently <1 x ULN who had significantly lower median HAI-NI and total HAI scores compared to all the other subgroups (Table 3). (The comparison with Group 2 patients with fluctuating ALT levels of 1-2 x ULN only showed a trend toward statistical significance which was probably due to the relatively small number of patients in these two groups.) However, there were no differences in the median HAI-F score when group 1 patients were compared to all other groups of patients. All eight patients with ALT level persistently <1 x ULN were anti-HBe positive. Three patients had HBV DNA levels >105 copies/ml. There were no differences in the histologic scores between all other subgroups indicating that there were no differences in the histology between patients with persistently elevated ALT levels and fluctuating pattern of ALT levels and between patients with elevated ALT levels of 1-2 x ULN and 2-5 x ULN.
AUTHOR DISCUSSION
In the present study, we primarily explored the relationship between HBV DNA levels and the degree of necroinflammation and fibrosis. The limitation of this study was that the HBV DNA levels were only measured at a single time point, at the time of the biopsies. However, at least for anti-HBe-positive patients, HBV DNA levels seldom fluctuate to any great extent.
We found that there was no relationship between HBV DNA levels and liver histology in term of the degree of necroinflammation and fibrosis in HBeAg-positive patients. This finding was in agreement with the studies conducted by Lindh et al. and Zavaglia et al. For patients in the immunotolerance phase, HBV DNA levels may show high viral replication but the immune-mediated attack has not yet set in. During the immunoclearance phase, there will also be a lack of correlation between the HBV DNA levels and the severity of liver histology because markedly enhanced immune-mediated response may lead to very low viremic levels even though ALT levels remain high at the time of the liver biopsies. This would account for the positive correlation between the ALT levels and the degree of necroinflammation in HBeAg-positive patients. The positive correlation between the ALT levels and the liver histology is in accord with other studies.
Though HBeAg seroconversion is commonly taken as a therapeutic endpoint in the past, increasing evidence shows that disease progression can continue in patients after HBeAg seroconversion, especially in the Asian and Mediterranean population. The problem is to identify the risk factors for anti-HBe-positive patients with the disease progression. In the National Institutes of Health Workshop on "Management of Hepatitis B," HBV DNA level of 105 copies/ml is arbitrarily chosen as a cut-off value to differentiate inactive carriers from patients with active hepatitis. The study conducted by Martinot-Peignoux et al. supports the use of this cut-off level since 98% of their patients with normal ALT had the HBV DNA levels <105 copies/ml. The histologic lesions were mild in all the cases. But all the patients recruited in that study had normal ALT levels. Two studies conducted by Manesis et al., however, suggest that a cut-off value of 3 x 104 copies/ml had a higher negative predictive value compared to 105 copies/ml. However this conclusion was based on ALT levels alone without supporting evidence from liver histology. Chu et al. recently show that using HBV DNA level <105 copies/ml as a cut-off level, 45% HBeAg-negative patients with active hepatitis (defined by the ALT levels) would be excluded if only the HBV DNA measurement on presentation is considered.
In the anti-HBe-positive patients of the present study, HBV DNA levels correlated positively with the degree of hepatic necroinflammation and fibrosis. This was in agreement with other previous studies. Though 85.7% anti-HBe-positive patients with HBV DNA level <105 copies/ml had minimal necroinflammation and fibrosis, there was still one patient (14.3%) with established fibrosis. The present study suggested that while HBV DNA level <105 copies/ml was associated with better histology, it was still not the ideal cut-off level for differentiating inactive patients with minimal liver damage from patients with advanced histology. Unfortunately, we were unable to examine the significance of HBV DNA level <104 copies/ml because only two patients had HBV DNA below this level.
Because of the preliminary finding of the association between core promoter mutations and poorer liver histology in the present study, we are currently conducting a study with the larger number of patients to confirm this finding. Several studies have shown that core promoter mutations are associated with the enhanced viral replication, though we were unable to show in the present study that patients with core promoter mutations had higher HBV DNA levels compared to the patients with core promoter wild-type. More studies are required to document whether core promoter mutations have any effects on the CHB disease.
There are very few data on the association between serial ALT levels and liver histology. We found that patients with persistently normal ALT levels had a low degree of necroinflammation (median HAI-NI score of 1.5) and of fibrosis (median HAI-F score of 0) when compared with patients with persistently abnormal ALT levels and patients with fluctuating levels of ALT. In addition, there were no differences in the liver histology between patients with persistently elevated ALT levels and with fluctuating ALT levels and between patients with ALT levels of 1-2 x ULN and of 2-5 x ULN. All these findings suggest that the only reliable marker of favorable liver histology is a persistently normal ALT level.
In conclusion, HBV DNA levels in HBeAg-positive patients have no prognostic significance on the liver histology. In anti-HBe-positive patients, HBV DNA levels positively correlated with the degree of necroinflammation and fibrosis. Though HBV DNA level of <105 copies/ml was associated with better histology, 14.3% patients already had established fibrosis. Further studies monitoring serial HBV DNA levels are necessary to define a better cut-off HBV DNA level for disease nonprogression.
|
|
| |
| |
|
|
|
