| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Editorial |
Veterans Affairs Medical Center, Washington, DC, and, George Washington University Medical Center, Washington, DC, E-mail d.robert.dufour{at}med.va.gov
Hepatitis B virus (HBV) is the most common significant chronic viral infection world-wide. Approximately 350–400 million people have active HBV infections, and as many as one third of the world’s population has evidence of exposure to HBV. Spread of the virus occurs primarily through contact with infected serum, sexual contact with infected individuals, and vertical transmission from mother to infant. Since shortly after its discovery as the "Australia antigen" (1), hepatitis B surface antigen (HBsAg) has been the principal target for laboratory testing to identify active infection by HBV. Accumulating evidence, including the article by Chen and Kaplan (2) in this issue of Clinical Chemistry, calls into question the degree to which HBsAg assays can be used as the "gold standard" for detecting HBV infection. The increasing availability of automated assays for viral markers on instruments traditionally found in immunochemistry laboratories makes it imperative for those working in such laboratories to be familiar with issues in HBsAg testing.
natural history of hbv infection
Exposure to HBV has a variable outcome that is largely determined by the age and immune status of an infected individual. In healthy adolescents and adults, the most common population exposed to HBV in northern Europe and most of North America, exposure to HBV typically leads to acute hepatitis, followed by loss of detectable HBsAg and loss of circulating HBV DNA (as measured by most assays). Chronic infection, identified by persistence of HBsAg, may follow infections in infants, persons with immunosupression, and a small number of otherwise healthy individuals. Such chronic infection may be associated with active viral replication (detected by circulating HBV DNA), or with "nonreplicating" infection, in which HBsAg continues to be produced but HBV DNA is undetectable (by most assays) in the circulation. Replicating forms of infection can convert to nonreplicating forms, either spontaneously (5%–10%/year) or after treatment with either interferon or nucleot(s)ide analogs. Rarely (1%–2%/year), HBsAg will also disappear from the circulation, followed by appearance of antibody to HBsAg (anti-HBs). This antibody is thought to be protective, and is the antibody produced with successful responses to the HBV vaccine.
falsely positive hbsag results and neutralization assays
HBsAg is typically detected by immunoassays that use anti-HBs to capture antigen in the sample. As with all infectious disease assays, nonspecific binding can occur. Manufacturers determine a "cutoff value" (C) that balances the necessary high sensitivity for detecting the antigen (or anti-HBs antibody) with the need to avoid false-positive results. Samples with signal values (S) above the "cutoff value" are called positive (S/C ratio >1). In the case of tests for antibodies to viral antigens (e.g., anti-HCV and anti-HIV), weakly positive results are often falsely positive and require confirmation with tests using isolated viral antigens (recombinant immunoblot for HCV, Western blot for HIV). In the case of tests for HBsAg, manufacturers provide a neutralization test that can be used to confirm true positivity of the results. In this test, samples with S/C ratio >1 are incubated with anti-HBs; if HBsAg is truly present, the anti-HBs in the neutralization step blocks binding to the reagent antibodies, reducing or eliminating the signal from HBsAg (positive neutralization test result).
The article by Chen and Kaplan in this issue of Clinical Chemistry (2) found a high frequency of weakly positive HBsAg results with 2 lots of reagent sets from one manufacturer; most of these weakly positive results were falsely positive (negative neutralization test results). In a large retrospective study of positive HBsAg results of 2 different immunoassays, O’Brien also found that weakly positive results were more likely to fail neutralization than were strongly positive results, and suggested that this confirmatory test be used only in samples with weakly positive results. (3) Similar recommendations have been made by the Centers for Disease Control for weakly positive anti-HCV results, for example (4). Weakly positive results are also more likely to be falsely positive with other infectious disease assays, including anti-HIV (5)(6) and anti-HBc (7)(8). Interestingly, in Chen and Kaplan’s study, a third lot from the same manufacturer had much lower rates of weakly positive and falsely positive results. We have shown similar lot-to-lot variability in anti-HCV reagents (9), which suggests that assignment of optimal cutoff values is an area that could be improved.
false-negative hbsag results
Chen and Kaplan (2) also measured HBV DNA in randomly selected samples that were initially reactive in HBsAg testing but failed neutralization; their interpretation should, therefore, have been HBsAg-negative. Of 89 samples tested, 6 were HBV DNA–positive, all at low viral loads (<9000 copies/mL). Why might falsely negative HBsAg results occur? It has been recognized for many years that there is a delay between HBV infection and appearance of HBsAg; during this "window" period, the individual may still be infectious. This is thought to account for the higher risk of transmission of HBV by blood products (1:600 000 units, compared with 
1:2 000 000 for hepatitis C and HIV), and the pending addition of HBV DNA testing of donor units is meant to address this issue. Diagnostic ("clinical") sensitivity may be related to the analytical performance of the assay used. For example, Scheiblauer et al. evaluated 17 different diagnostic reagent sets available in Europe, and found that detection limits varied by 5- to 10-fold between the most sensitive and least sensitive assays, depending on the serotype of HBsAg tested (10).
Over the last 10 years, increasing attention has been paid to 2 other forms of HBsAg-negative HBV infection: "occult" infection and presence of HBsAg mutants that escape detection by HBsAg assays and produce disease despite "protective" titers of anti-HBs following immunization.
occult hbv infection
Until more sensitive HBV DNA assays became available in the late 1990’s, it was thought that individuals with negative HBsAg were always HBV DNA–negative. While most adolescents and adults with acute HBV infection lose HBsAg and develop both anti-HBc and anti-HBs, HBV DNA remains present in small amounts, both in the circulation and within the liver (11); viral DNA also remains present after resolution of chronic HBV (12). The infecting strains of HBV in such cases often have mutants that impair HBV replication, leading to low concentrations of HBV DNA in the circulation (13). In most cases, the infection remains quiescent, but when the individual’s immune system becomes suppressed (e.g., with chemotherapy or with use of immunosuppressant drugs), viral replication may increase. When the immune status of the person recovers, recurrence of hepatic injury (often termed "reactivation") may develop and, in some cases, will progress to acute liver failure (14). At present, it is not clear whether assays with better analytical sensitivity would be able to detect such "occult" infections, nor what the clinical significance would be to reporting the presence of HBsAg in such settings (15).
hbsag "escape" mutants
Of perhaps even greater concern is the presence of HBsAg mutants that are not detected by HBsAg assays and that can cause infection despite the presence of anti-HBs. Genotypes of HBsAg have a mixture of serologic determinants, all of which include a common "a" determinant. The "a" determinant is the major region detected both by HBsAg assays and by naturally occurring anti-HBs, whether following exposure to the virus or to HBsAg vaccine. Individuals exposed to strains with mutations in critical positions in the HBsAg "a" determinant may become infected despite having what are thought to be protective titers of anti-HBs; additionally, such mutants may escape detection by HBsAg assays (16). Concern over mutants led one of the manufacturers of hepatitis serologic assays to sponsor a meeting of an expert panel, which developed a consensus statement on the significance of such mutants (17). The panel concluded that the prevalence of such undetectable mutants is unknown; while such mutants are generally felt to be rare, the panel felt that prevalence is likely higher than generally appreciated. It is not clear whether such mutant strains could also affect the ability of HBV DNA assays to amplify the virus (18). Recently, an entire supplement to the Journal of Medical Virology was devoted to the issue of such HBsAg mutants (Volume 78, Supplement 1, 2006). Because mutants can occur at many positions in the "a" determinant, the expert panel consensus statement suggested that assays will need to include more than one monoclonal antibody to detect all mutant strains (17).
what should laboratories and manufacturers do to improve hbv detection?
The article by Chen and Kaplan (2) indicates that laboratories need to be aware of the performance of their HBsAg assays. At the least, laboratories should know the analytical performance of their assays near the cutoff concentration, and use neutralization assays in samples with weakly positive HBsAg results. Samples with discordant results (e.g., those positive for both HBsAg and anti-HBs, or samples with positive HBeAg or HBV DNA but negative HBsAg) should be evaluated for the presence of mutant strains; this implies that panels of tests should be performed whenever HBsAg results are suspect. Laboratories should communicate with clinicians on the status of patients with questionable HBsAg results; important parts of the clinical history include likely source of exposure (neonates may be more likely to harbor infection by vaccine-escape mutant strains), immune status of the individual, and results of any previous tests that may have been performed using another assay. Manufacturers should provide information about the region(s) of the HBsAg molecule that can be detected by the reagent antibody or antibodies. Users also need data from manufacturers on the analytical performance of their methods, especially regarding imprecision at concentrations near that of the cutoff. Similarly, the detection limit reflects both analytical sensitivity (the change of the signal for a unit change of analyte) and imprecision. Manufacturers may also need to improve the lot-to-lot reproducibility of infectious disease test cutoff values, particularly those used to identify weakly positive samples that require additional testing (such as neutralization assays for HBsAg). Additional research is needed to address several questions relating to the prevalence and significance of mutants that are not detected by HBsAg assays, and to the need for increased sensitivity of HBsAg assays to reduce the window period and the occurrence of occult HBV infections.
Disclosure: In the past year, the author has received honoraria from Abbott Diagnostics and Ortho Clinical Diagnostics.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
