Despite significant progress in its prevention and treatment in the past few decades, hepatitis B virus (HBV) infection remains a challenging global public health problem. Approximately 5% of the world population or 350-400 million individuals are chronically infected with HBV. Long-term consequences of chronic HBV infection include cirrhosis, hepatic decompensation, and hepatocellular carcinoma (HCC), to which more than 1 million people succumb annually.1 Although the number of new cases of acute HBV infection has been decreasing in some of the endemic countries as well as the United States,2, 3 the prevalence and disease burden among individuals already chronically infected with HBV remain substantial. In fact, widespread immigration from HBV-endemic regions has led to an increased number of Americans chronically infected with HBV as well as rising rates of HBV-related hospitalizations, HCC, and deaths.4
One of the challenges in individual patient management, as well as public health intervention, in chronic HBV infection is the large degree of intra- and interindividual variability in the natural history of chronic HBV infection. A given patient with chronic HBV infection may be in one of several phases of infection, such as immune tolerance, immune active, and inactive carrier phases.1 It is generally accepted that progressive liver damage occurs mostly in patients in the immune active phase, characterized by abnormal alanine aminotransferase (ALT) and significant levels of circulating virus. Thus, at the population level, disease burden, at present and in the future, would depend on the proportion of subjects at risk of progressive liver damage.
In the current issue of Hepatology, Szpakowski and Tucker studied the natural history of HBV infection in a U.S. population in the era preceding widespread use of antiviral agents5. Using a large, prospectively collected database from a managed healthcare plan in Northern California between March 1996 and December 2005, they analyzed crude mortality rates, causes of death, and demographic risk factors for cause-specific mortality in individuals with serologic evidence of chronic HBV infection. The study cohort included a total of 6,689 patients with chronic HBV. Nearly 70% of them were of Asian-Pacific Islander (API) descent and only 12% were non-Hispanic whites. One of the findings unexpected by the researchers was that nearly half of the deaths in individuals chronically infected with HBV were related to HBV infection. The death rate from HCC was twice that of decompensated cirrhosis. HCC deaths represented 70% of cancer deaths in males and 37% in females. On multivariable analysis incorporating available variables, age was the only significant predictor of HBV mortality. Somewhat surprisingly, HBV mortality was not higher among subjects of API descent, compared to non-Hispanic whites or other/unknown origin.
This work is unique in several aspects. First, the size of the population represented in the database is quite large, particularly for a U.S. study, which allows analyses of relatively infrequent events (namely, HBV-related deaths) with sufficient degree of precision. Second, the study period was chosen such that there was minimal contamination with patients treated with antiviral therapy. More recent data would have been enriched with patients receiving treatment that is expected to have a significant effect on the natural history. Third, although the researchers' claim of this study being truly population based may be disputed, it is likely that enrollment into the health insurance coverage was random with respect to HBV infection and that the study sample was representative of the HBV-infected individuals in the part of the United States in which the study was done.
Long term follow-up studies in HBV-infected patients have reported variable mortality outcome data, depending upon the study setting. Studies in patients with incidentally detected HBV infection, mostly conducted in blood donors in Western countries, tend to portray a benign course. For example, studies based on individuals with incidentally discovered HBV infection reported that their incidence of complications of chronic liver disease (CLD), HCC, or liver-related mortality is not significantly higher than that in hepatitis B surface antigen–negative healthy controls.6-8 On the other hand, studies conducted in patients diagnosed with chronic hepatitis B, including those from endemic areas, report much more serious consequences in HBV-infected individuals.9-12 For example, the lifetime risk of a Chinese patient to die of an HBV-related cause has been estimated to be up to 50% in men and 15% in women.13
Subsequent data have elucidated individual characteristics that influence disease progression in patients with HBV infection. In hepatitis B e antigen (HBeAg)-positive patients, prolonged periods of hepatitis activity before HBeAg seroconversion portends poor prognosis.14 For example, patients who have persistently and often mildly elevated ALT sometimes well beyond 40 years of age may have a smoldering, yet progressive, liver damage because of inability to mount an effective immune response to achieve seroconversion. Other patients may experience repeated HBeAg seroconversion and reversion and be subjected to bouts of necroinflammatory activities.15 In patients who are HBeAg negative, the best characterized determinant of disease progression is the HBV DNA titer. In large cohort studies, DNA levels correlated with future occurrence of cirrhosis and HCC.16 In addition, viral characteristics, such as precore or basal core promoter variants as well as initial genotype, may add to the prognostic equation. Regardless of HBeAg status, male gender, alcohol use, and concomitant HCV, hepatitis D virus, and human immunodeficiency virus infection worsen the prognosis.
Given the heterogeneity of individuals with HBV infection included in various studies, it is likely that disparate outcome results among different studies are, at least in part, attributable to the differences in the study population. For example, in the study by Szpakowski and Tucker, the large majority of patients represented were of API ancestry, making it likely that the biological characteristics of the study subjects were likely to be more similar to those of Asian studies than to otherwise healthy subjects in Western studies. One important limitation of the present study is the lack of detailed systematic laboratory follow-up data. If they were available, it is likely that well-established variables, such as HBV DNA level, may have been included in their multivariable survival model.
Even in the era of effective potent antiviral agents, accurate understanding of the natural history is important in tailoring treatment. Although skeptics have pointed to the absence of studies demonstrating improvement in hard endpoints, such as survival in treated, compared to untreated, patients, there is a broad consensus among experts that potent suppression of viral replication in patients at risk of disease progression fundamentally alters the natural history of the patient. The strongest evidence for this comes from the well-publicized data by Liaw et al., which was a placebo-controlled trial evaluating lamivudine in viremic patients with bridging fibrosis or compensated cirrhosis.17 Patients who received antiviral therapy were markedly less likely to experience hepatic decompensation, which led to termination of the study earlier than scheduled. In addition, the incidence of HCC was lowered by lamivudine. More recently, dramatic demonstration of regression of fibrosis and cirrhosis in paired biopsy studies in patients receiving antiviral agents adds to the confidence that antiviral agents make a meaningful difference in long-term outcome.18,19 These patient-level data also translate to population-wide epidemiological statistics. Liver transplant wait-list data in the United States suggest that coincident to widespread use of antiviral agents, the need for liver transplantation for hepatic decompensation from chronic HBV cirrhosis has decreased.20 Finally, a recent article from Taiwan showed that antiviral therapy subsequent to HCC resection was associated with significant reduction in risk of recurrence.21 These data help us to derive the most benefit from antiviral agents, by identifying treatment candidates and timing of treatment such that their natural history of HBV infection is meaningfully altered.
In summary, the work by Szpakowski and Tucker is significant as a rare piece of U.S. data that documents the real-life effect of HBV infection in the absence of antiviral therapy. It also identifies demographic subgroups that are at higher risk of adverse outcome over time. For clinicians, it highlights the importance of proper screening for, and diagnosis of, chronic HBV infection to identify patients at risk of future morbidity and mortality as well as those that will benefit from therapy. For health policy makers, these data will be a useful baseline data to which future data may be compared to quantify the effect of antiviral therapy at the population level and to formulate means to optimize its application. In an era of effective vaccines and potent antiviral agents, HBV should be eliminated as a cause of CLD and end-stage liver disease. Finally, continued work is needed to prevent and control HCC at the population level, because antiviral therapy has only limited effect on the carcinogenic effects of HBV.