Notice: Wiley Online Library will be unavailable on Saturday 30th July 2016 from 08:00-11:00 BST / 03:00-06:00 EST / 15:00-18:00 SGT for essential maintenance. Apologies for the inconvenience.
Potential conflict of interest: Nothing to report.
Current guidelines for treatment of chronic hepatitis B include hepatitis B e antigen (HBeAg) status, levels of hepatitis B virus (HBV) DNA, and serum alanine aminotransferase (ALT) values in the setting of either chronic hepatitis or cirrhosis. Based on findings from a prospective study of hepatitis B surface antigen (HBsAg)-positive patients, we determined whether these guidelines included patients who developed hepatocellular carcinoma (HCC) and who died of non-HCC liver-related complications. The criteria for treatment from four published guidelines were matched to a cohort of 369 HBsAg-positive patients enrolled in the study. During a mean follow-up of 84 months, 30 patients developed HCC and 37 died of non-HCC liver-related deaths. Using criteria for antiviral therapy as stated by the four guidelines, only 20%-60% of the patients who developed HCC, and 27%-70% of patients who died of non-HCC liver-related deaths would have been identified for antiviral therapy according to current treatment recommendations. If baseline serum albumin levels of 3.5 mg/dL or less or platelet counts of 130,000 mm3 or less were added to criteria from the four treatment guidelines, then 89%-100% of patients who died of non-HCC liver-related complications, and 96%-100% of patients who developed HCC would have been identified for antiviral therapy. In addition, if basal core promoter T1762/A1764 mutants and precore A1896 mutants also were included, then 100% of patients who developed HCC would have been identified for treatment. Conclusion: This retrospective analysis showed that the current treatment guidelines for chronic hepatitis B excluded patients who developed serious liver-related complications. (HEPATOLOGY 2008.)
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
It is estimated that there are 1.25 million individuals who are chronically infected with the hepatitis B virus (HBV) in the United States.1 Most of these are of Asian ancestry and were infected by HBV at birth or early in childhood and before immigration to the United States. The morbidity and mortality rates from this chronic viral infection are high, and it is estimated that 15% to 40% will either succumb to chronic liver disease complications or develop hepatocellular carcinoma (HCC) during their lifetime.2 Therefore, to decrease the disease burden that may result from chronic hepatitis B infection, it is important to offer antiviral treatment before development of these liver-related complications.
At the present time, there are four published guidelines that specifically address the treatment of chronic hepatitis B patients. These include a consensus statement from the European Association for the Study of the Liver (EASL),3 a treatment algorithm by an independent panel of hepatologists in the United States (US panel),4 an Asian-Pacific consensus statement (Asian-Pacific panel),5 and the practice guidelines from the American Association for the Study of Liver Diseases (AASLD).6 These treatment guidelines are based mainly on data from published literature on the natural history of hepatitis B in Asian and European countries and on the personal experience of the authors of the guidelines. The criteria of the specific guidelines for treating chronic hepatitis B patients included baseline levels of serum alanine aminotransferase (ALT) and HBV DNA, the hepatitis B e antigen (HBeAg) status, and whether patients had chronic hepatitis or cirrhosis. Although these reports are detailed and complete, the recommendations may omit hepatitis B surface antigen (HBsAg)-positive patients who have the potential to develop HCC or die of liver-related complications. In addition, there are no retrospective or prospective studies to test the usefulness of these recommendations.
Recently, we reported a long-term follow-up study of HBsAg-positive patients from a clinic in Pasadena, California.7 At the time of recruitment into this prospective study, which began in 1989, the initial levels of serum ALT and albumin, platelet counts, and HBeAg status were recorded. In 2004, baseline sera that were stored at −70°C were measured for HBV DNA, basal core promoter T1762/A1764 mutants, and precore A1896 mutants in one research laboratory.8–10 During a mean follow-up of 84 months, we identified clinical and virological factors that were significantly associated with development of HCC and with non-HCC liver-related deaths.7, 11
Based on the recommendations from (1) the EASL consensus conference, (2) the US panel, (3) the Asian-Pacific panel, and (4) the AASLD guidelines, we matched the criteria for treatment from each of these guidelines with the baseline levels of ALT and HBV DNA and the HBeAg status of our HBsAg-positive patients who were enrolled in the prospective follow-up study. Our goal was to determine whether these four published guidelines on hepatitis B treatment included patients who developed HCC or died of non-HCC liver-related complications. Based on our findings, we offer additional criteria that may improve on the selection of HBsAg-positive patients who would benefit from antiviral therapy.
AASLD, American Association for the Study of Liver Diseases; ALT, alanine aminotransferase; EASL, European Association for the Study of the Liver; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HIV, human immunodeficiency virus.
Patients and Methods
From January 1989 to March 1998, 400 HBsAg-positive patients were enrolled in a prospective follow-up study.7 Patients who were co-infected with HCV or human immunodeficiency virus, or had a history of chronic alcoholism or other chronic liver diseases such as autoimmune hepatitis or hemochromatosis, were excluded. The mean age at recruitment was 48.4 years, 78.5% were of Asian ancestry, and 70.5% were male. The mean follow-up time was 83.6 ± 39.6 months. For HCC surveillance, abdominal ultrasound examinations and serum alpha-fetoprotein levels were performed every 6 months. If the serum alpha-fetoprotein was elevated or if a liver lesion was noted on ultrasound, further studies including magnetic resonance imaging, computed tomography scan, or biopsy of the lesion were used for HCC diagnosis. Of the 400 patients, 369 had baseline HBeAg, serum ALT, and albumin values, platelet counts, and HBV DNA levels and are the subjects of this analysis (Table 1). One hundred forty-four patients had chronic hepatitis, 129 patients had cirrhosis, and 96 were inactive carriers. All of the chronic hepatitis and cirrhosis patients had liver biopsy confirmations of their diagnoses. In the 369 patients, 186 (50.4%) were HBeAg positive, 62% had baseline HBV DNA levels of ≥105 copies/mL, and 49% had abnormal ALT levels (upper limit of normal was 45 U/L; Table 1). The mean baseline level of ALT was 57.8 ± 69.2 U/L, albumin 4.1 ± 0.6 mg/dL, and platelet counts were 193.6 ± 78.1 × 103 mm3.
Table 1. The Baseline Values of HBeAg, Serum Alanine Aminotransferase, and HBV DNA Levels in 369 HBsAg-Positive Patients
ALT, alanine aminotransferase; ULN, upper limits of normal.
1–2 × ULN
>2 × ULN
HBV DNA (copies/mL)
The guidelines from four publications are shown in Tables 2 and 3.3–6 The guidelines recommended that treatment be initiated in HBeAg-positive chronic hepatitis patients with HBV DNA levels 100,000 copies/mL or greater and serum ALT levels that are either greater than normal4 or greater than 2 times the upper limit of normal.3, 5, 6 For treatment of HBeAg-negative chronic hepatitis patients, the HBV DNA levels are required to be 10,000 copies/mL or greater4, 5 or 100,000 copes/mL or greater,3, 6 and serum ALT levels greater than normal4 or 2 times greater than normal.3, 5, 6 For patients with cirrhosis who are either HBeAg positive or negative, three guidelines included HBV DNA at levels of 104 copies/mL or greater with no ALT level specified,4–6 whereas one other guideline included HBV DNA levels of 105 copies/mL or greater and ALT levels above normal.6 We then matched the baseline HBeAg, ALT, and HBV DNA levels from our 369 chronic hepatitis B patients to criteria set forth in the four published treatment recommendations and determined whether these guidelines included patients who developed HCC or died of non-HCC liver-related complications during follow-up in our prospective study. In addition, to determine whether more of these latter patients would have been eligible for antiviral therapy, we included additional laboratory and virological tests to the criteria from the four treatment guidelines.
Table 2. Recommendations for Treatment in Chronic Hepatitis B Patients
The HBsAg-positive patients were enrolled into the current study before the time of approval of lamivudine for treatment of hepatitis B in 1998. Subsequently, two of the patients who developed HCC, three patients who died of non-HCC liver-related deaths, and 59 patients who remained alive received lamivudine for at least 1 year. The five patients who developed liver-related complications were followed for over 5 years before lamivudine treatment was initiated. Because the focus of this report is mainly on laboratory tests that were available during the first visit to our clinic, these five patients were included into the group who developed liver complications.
All statistical significance was assessed at the 0.05 level. Baseline data were descriptively summarized. Means and standard deviations were computed for all continuous data. Categorical data were summarized by using frequencies. The receiver operating characteristic curves were plotted for serum albumin values and platelet counts, and the test values that provided the best sensitivity and specificity for prediction of liver complications were obtained.
During a mean follow-up of 84 months, 37 of 369 (10%) patients died of non-HCC liver-related complications, including 26 with liver failure, seven with bleeding esophageal varices, and four with sepsis. All of these patients had cirrhosis. Also, during this time period, 30 of 369 (8.1%) progressed to HCC. At baseline, nine of the 30 had chronic hepatitis by liver biopsy, and the remaining 21 patients who developed HCC had cirrhosis on entry into the study. Thus, there were a total of 67 (18.2%) liver-related complications in this cohort of 369 HBsAg-positive patients.
The baseline HBeAg, levels of ALT and HBV DNA, and presence of either chronic hepatitis or cirrhosis in the 369 HBsAg-positive patients were established to determine who would have been eligible for antiviral treatment according to the four published guidelines (Table 4). Based on EASL, Asian Pacific panel, and AASLD, only 10, 10, and 9 chronic hepatitis patients, respectively, met criteria for treatment, whereas 68 chronic hepatitis patients were eligible for treatment according to the US panel criteria. This was because of the less stringent ALT level in the US panel recommendations. In patients with cirrhosis, 97 patients would have been treated according to the US panel, Asian-Pacific panel, and AASLD criteria, whereas only 47 patients would have received treatment according to the EASL recommendations. This was attributable to the requirement of higher HBV DNA levels and elevated ALT values in the EASL criteria compared with the other three recommendations, which required lower levels of HBV DNA with no ALT value specified.
Table 4. Number of Patients Eligible for Anti-Viral Therapy According to Current Guidelines
The number of HBsAg-positive patients who developed HCC or died of non-HCC liver-related complications who would have received treatment according to the four guidelines are shown in Table 5. For HCC development, EASL guidelines only identified six of 30 (20%) patients, the Asian-Pacific panel and AASLD each identified 14 (47%), and the US panel guidelines included 18 (60%) patients who progressed to HCC. As seen in Table 5, the EASL, Asian-Pacific Panel, and AASLD guidelines did not include any of the nine chronic hepatitis patients who progressed to HCC, whereas the US panel guidelines excluded five of these nine patients. Also, the EASL recommendations excluded 15 of 21 of the cirrhosis patients who developed HCC, whereas the US panel, Asian-Pacific Panel, and AASLD guidelines each excluded seven of 21 of these patients. Thus, the EASL guidelines excluded a total of 24 of 30 (80%) of the patients who developed HCC, the Asian-Pacific Panel and AASLD excluded 16 of 30 (53%), and the US panel excluded 12 of 30 (40%) of these patients. For non-HCC liver-related deaths, the EASL guidelines excluded 27 of 37 (73%) of patients whereas the US Panel, Asian-Pacific Panel, and AASLD guidelines each only excluded 11 of 37 (30%) of these patients (Table 5).
Table 5. Exclusion of Patients Who Developed HCC or Died of Non–HCC-Related Liver Deaths by the Current Guidelines for Treatment of Hepatitis B
The mean levels of serum albumin, ALT, and platelet counts in cirrhosis patients who (1) did not have complications (compensated cirrhosis; no. 71), (2) developed HCC (no. 30), and (3) died of non–HCC-related liver complications (no. 37) are shown in Table 6. These laboratory tests were obtained at baseline, midway during the follow-up period and before developing complications. During each of the three time periods, there was a significant difference in the mean serum albumin levels and platelet counts when the compensated cirrhosis patients were compared with those who died of non-HCC liver-related complications (ρ < 0.0001 for all observations). In addition, the mean serum albumin levels and platelet counts were consistently below normal values during the entire course of follow-up in patients who developed HCC as well in those who died of non-HCC liver complications (Fig. 1). Also, there was an increase in the mean ALT levels when patients with compensated cirrhosis were compared with those who died of non-HCC liver deaths during the three time periods (ρ = 0.06, 0.001, and 0.06, respectively). However, the mean ALT levels tended to fluctuate during follow-up and a trend could not be established to differentiate one category of patient from another (Fig. 1). Using receiver operating characteristic curves to identify the test values with the best sensitivity and specificity, we determined that if a baseline albumin level of ≤3.5 mg/dL or a platelet count of ≤130,000 mm3 were added to the four published criteria for antiviral therapy, then the number of patients who died of non–HCC-related liver complications who would have received treatment would increase to 89%, 100%, 100%, and 100% in the EASL, US Panel, Asian-Pacific Panel, and AASLD guidelines, respectively (Table 7). Similarly, the number of patients who developed HCC who would have been treated also would increase to 96% for EASL, Asian-Pacific Panel, and AASLD guidelines, respectively, and to 100% for the US Panel guidelines (Table 8).
Table 6. Serial Laboratory Tests in HBsAg-Positive Patients with Compensated Cirrhosis, Non-HCC Liver Deaths, and with HCC Development
Compensated Cirrhosis (71)
Non-HCC Liver Deaths (37)
HCC Development (30)
Compensated cirrhosis versus non-HCC liver deaths.
Time of tests. 1: baseline; 2: midway of follow-up period; 3: before non-HCC liver death or to HCC development.
Table 7. Number of Patients Who Died of Non-HCC Liver Complications Who Would Have Been Eligible for Antiviral Therapy with the Addition of Other Baseline Laboratory and Virologic Tests to the Four Guidelines
Non-HCC Liver-Related Deaths (Total 37)
No. recommended for treatment
No. excluded for treatment
No. treated if albumin ≤ 3.5 mg/dL or platelets ≤ 130 ×103 mm3 included
Total possible no. treated (%)
No. treated if albumin ≤ 3.5 mg/dL or platelets ≤ 130 × 103mm3, and basal core promoter mutants/precore mutant included
Total possible no. treated (%)
Table 8. Number of Patients Who Developed HCC Who Would Have Been Eligible for Antiviral Therapy with the Addition of Other Baseline Laboratory and Virologic Tests to the Four Guidelines
HCC (Total 30)
No. recommended for treatment
No. excluded for treatment
No. treated if albumin ≤ 3.5 mg/dL or platelets ≤ 130 × 103 mm3 included
Total possible no. treated (%)
No. treated if albumin ≤ 3.5 mg/dL or platelets ≤ 130 ×103 mm3, and basal core promoter mutant/precore mutant included
Total possible no. treated (%)
The virological markers in HBsAg-positive patients who developed HCC or died of non–HCC-related liver complications are shown in Table 9. Precore A1896 gene sequences were measurable in all 67 of these patients, whereas basal core promoter T1762/A1764 gene sequences were measurable in 56 of 67 of the patients. If basal core promoter mutants and precore mutants, and the albumin ≤3.5 mg/dL or platelets counts ≤130,000 mm3 were included into the published criteria for receiving antiviral therapy, the number of patients who developed HCC who would have been treated would increase to 100% in the EASL, US Panel, Asian-Pacific Panel, and AASLD guidelines (Table 8). Also, if these parameters were added to patients who died of non–HCC-related liver deaths, the numbers of these patients treated would increase to 92% in the EASL guidelines and again to 100% in the other three guidelines (Table 7).
Table 9. Virologic Markers in HBsAg-Positive Patients Who Developed HCC or Died of Non-HCC Liver-Related Complications
Non-HCC Liver-Related Deaths
In non-HCC liver-related deaths, precore sequences were detected in all 37 patients; BCP sequences were detectable in 31 of 37 patients. In HCC patients, precore sequences were detected in all 30 patients; BCP sequences were detectable in 25 of 30 patients.
Reports on the natural history of chronic hepatitis B infection have mainly originated from Asia and Europe. Our prospective study on HBsAg-positive patients was from a clinic in Pasadena, California, and is one of only a few reports on the natural history of chronic hepatitis B from the United States. As described herein, during a mean follow-up of 84 months, 37 of 369 (10%) patients died of liver complications, and 30 (8.1%) progressed to HCC. The annual rate of non–HCC-related liver deaths in this report was 3.9% and is similar to rates of 2.4% to 4% reported in other studies.12, 13 In addition, the annual rates for HCC development in our chronic hepatitis patients was 0.9%, and in cirrhosis patients was 2.3%, both of which were similar to HCC incidence rates of 1.5% to 3.8% in Europe14, 15 and rates of 0.7% to 2.2% in Asia.16, 17 Thus, deaths of liver disease complications and HCC development, the two major causes of mortality and morbidity from hepatitis B, appear to be similar among patients from different countries.
The long-term goals of antiviral treatment are to decrease or eliminate deaths of liver-related complications and to reduce the risk of HCC development. A recent randomized trial in chronic hepatitis B patients with advance fibrosis or cirrhosis showed that, compared with placebo, treatment with lamivudine significantly delayed liver disease progression and reduced the risk of hepatocellular carcinoma.18 Other reports have shown that use of antiviral therapy improved survival in HBsAg-positive patients with decompensated liver disease.19, 20 However, further treatments trials are required to determine whether antiviral therapy will have any impact on incidence of HCC, especially in patients with less active liver disease (in other words, low levels of viremia and normal or near normal ALT levels). The short-term goals of antiviral treatment include improvement in liver histology, normalization of serum ALT, suppression of HBV replication, and seroconversion to anti-HBe in HBeAg-positive patients. These objectives have been achieved in varying degrees in recent randomized trails of antiviral agents for treatment of chronic hepatitis B.21–24 However, whether these short-term goals will achieve significant reductions in HCC development or in liver-related deaths still remains to be proven.
The current guidelines for antiviral treatment include the baseline HBeAg status, level of HBV DNA, and serum ALT values and are dependent on whether patients have chronic hepatitis B or cirrhosis.3–6 Although these guidelines are detailed, there are no studies that test the validity of this recommendation or whether they actually include patients who will die of liver-related complications or develop HCC. In the study reported herein, the four published guidelines excluded 30% to 73% of our patients who died of non-HCC liver-related complications and 53% to 80% who developed HCC. The least accurate of the guidelines were the EASL recommendations, which was an earlier paper published in 2003. The reasons for the poor performance of the guidelines were because at baseline, up to 61% of our patients who progressed to liver failure or developed HCC were anti-HBe positive, 30% had HBV DNA levels of 104 copies/mL or less, and 40% had normal ALT values. Thus, many of these patients would not have been selected for antiviral therapy.
Criteria for patient selection for anti-viral therapy and treatment endpoints are still not agreed on by all investigators. In a recent commentary, the treatment guidelines from the AASLD, EASL, and Asian Pacific Associations for the Study of the Liver were questioned.25 These authors reasoned that most Asian patients contracted hepatitis B at birth or during early childhood as compared with Caucasian patients, who usually contracted hepatitis during adolescence or adulthood via sexual contact or intravenous drug usage. They indicated that HBeAg seroconversion was not a significant end point to stop therapy because more than 70% of patients who develop HCC or complications from cirrhosis are anti-HBe positive. Also, these authors indicated that a significant portion of Asian patients continue to be at risk for liver complications even if their HBV DNA levels decrease to less than 105 copies/mL or to less than 104 copies/mL. Another study showed that Asian patients with ALT levels between 0.5 and 2 times upper limits of normal still had risks for liver disease complications.26 Therefore, the treatment endpoints of HBeAg seroconversion to anti-HBe, ALT normalization, and decrease of HBV DNA levels to less than 105 copies/mL may not apply to Asian patients with chronic hepatitis B.
The serum albumin level has been a traditional test for measurement of liver synthetic function, and albumin synthesis has a significant correlation with Child-Pugh classification.27 In our report herein, cirrhosis patients without complications had consistently normal albumin levels compared with those who developed complications (Table 6, Fig. 1). Also, platelets counts have been reported to be associated with the stage of hepatic fibrosis.28 In this report, our cirrhosis patients without complications had normal platelet counts during follow-up compared with those who developed complications (Table 6, Fig. 1), and in these latter patients, both the serum albumin levels and platelets counts continued to decrease during the follow-up period. Thus, these tests are useful for identifying patients who progress to liver failure or develop HCC at baseline as well as during the course of their disease. The mean levels of serum albumin and platelet counts were within the normal range during follow-up in patients with chronic hepatitis who developed HCC (Fig. 1). In these latter patients, the mean serum ALT levels were elevated above normal at all time points (Fig. 1). Abnormal ALT values also were observed for patients with compensated cirrhosis as well as for those who developed HCC or progressed to non-HCC liver deaths. In patients with liver disease, elevated ALT levels may be related to sex (higher in men), body mass, metabolic syndrome, nonalcoholic steatohepatitis, chronic alcohol consumption, or co-existing hepatitis C infection and may be associated with increase mortality risk.26, 29 In our patients herein, we excluded HBsAg-positive patients who had co-existing liver disease such as alcoholic liver disease, autoimmune hepatitis, hemochromatosis, nonalcoholic steatohepatitis, and chronic HCV infection. However, most patients with chronic hepatitis B will experience viral and biochemical flares that result in increases in the serum ALT and aspartate aminotransferase values. Although indicative of liver inflammation, elevated serum transaminase levels will not predict HCC development nor identify patients with cirrhosis who are progressing to liver decompensation. As shown in our report herein, the platelet counts and serum albumin are more reliable indicators of liver disease progression.
Recently, we showed that basal core promoter T1762/A1764 mutants and precore A1896 mutants were independent risk factors for HCC development.11, 30 Other studies have shown that basal core promoters T1762/A1764 mutants were more frequently detected in HCC patients and in patients with decompensated liver disease.31–,34 The role of precore A1896 mutants is less clear. In our study, the presence of precore A1896 mutants was an independent factor for HCC development and was detected more frequently in HCC patients than in chronic carriers.11, 27 The precore A1896 mutant also has been described in HBeAg-negative chronic hepatitis.35 In medical facilities where the measurement of basal core promoter sequences and precore sequences are available, detection of basal core promoter T1762/A1764 mutants and precore A1896 mutants would imply that the patient has the potential to develop serious liver complications. In these instances, initiation of antiviral therapy may be warranted, but further studies are needed to determine the feasibility of this approach.
In an attempt to identify more patients who have the potential to die of serious liver complications or develop HCC, we propose the addition of baseline serum albumin and platelet counts to the current treatment guidelines. We showed that if albumin levels of 3.5 mg/dL or less or platelets counts of 130,000 mm3 or less were included, then 89% to 100% of patients who died of non-HCC liver complication and 96% to 100% who developed HCC would have been recommended for antiviral therapy. In addition, a decrease in albumin levels or platelets counts during follow-up clinic visits also would identify patients who have the potential of developing serious complications (Table 6, Fig. 1). Finally, addition of basal core promoter T1762/A1764 and precore A1896 mutants also would further increase identification of patients who developed HCC to 100% as well as 92% to 100% of the patients who died of non-HCC liver-related complications. However, measurement of serum albumin and platelets would already have included most of these patients for antiviral therapy.
Although this is a retrospective study, this type of analysis is important to determine the usefulness of the treatment guidelines and to gather further information so that more patients, who are at high risk of developing HCC or dying of non-HCC liver-related complications, will be included in the treatment recommendations for chronic hepatitis B. Further studies to validate the use of serum albumin levels of 3.5 mg/dL or less, platelet counts of 130,000 mm3 or less, and HBV DNA mutational analysis for inclusion in treatment guidelines are warranted.