• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of jcmPermissionsJournals.ASM.orgJournalJCM ArticleJournal InfoAuthorsReviewers
J Clin Microbiol. Aug 1999; 37(8): 2639–2647.
PMCID: PMC85302

Improved Detection of Hepatitis B Virus Surface Antigen by a New Rapid Automated Assay

Abstract

The performance of hepatitis B virus (HBV) surface antigen (HBsAg) screening assays is continuously improved in order to reduce the residual risk of transfusion-associated hepatitis B. In a multicenter study, a new automated rapid screening assay, Elecsys HBsAg (Roche Diagnostics), was compared to well-established tests (Auszyme Monoclonal [overnight incubation] version B and IMx HBsAg [Abbott]). Included in the evaluation were 23 seroconversion panels; sera from the acute and chronic phases of infection; dilution series of various HBsAg standards, HBV subtypes, and S gene mutants; and isolated anti-HBV core antigen-positive samples. To challenge the specificity of the new assay, sera from HBsAg-negative blood donors, pregnant women, and dialysis and hospitalized patients and potentially cross-reactive samples were investigated. Elecsys HBsAg showed a higher sensitivity for HBsAg subtypes ad, ay, adw2, adw4, ayw1, ayw2, ayw4, and adr detection in dilution series of different standards or sera than Auszyme Monoclonal version B and/or IMx HBsAg. Acute hepatitis B was detected in 11 to 16 of 23 seroconversion panels between 2 and 16 days earlier with Elecsys HBsAg than with the alternative assays. Elecsys HBsAg and Auszyme Monoclonal version B detected HBsAg surface mutants with equal sensitivity. The sensitivity and specificity of Elecsys HBsAg were 100%. Auszyme Monoclonal version B had a 99.9% specificity, and its sensitivity was 96.6%. IMx HBsAg showed a poorer sensitivity and specificity than the other assays. In conclusion, Elecsys HBsAg permits earlier detection of acute hepatitis B and different HBV subtypes than the alternative assays. By using highly sensitive HBsAg screening assays, low-level HBsAg carriers among isolated anti-HBV core antigen-positive individuals can be detected.

The envelope protein of hepatitis B virus (HBV), hepatitis B surface antigen (HBsAg), is a glycosylated lipoprotein usually shed in large amounts in the serum of infected individuals, where it is found as spherical particles with diameters of 22 nm or filaments of similar diameter (39). The a determinant of HBsAg, a predicted double-loop structure projecting from the surface of the HBV particle (36), is the major neutralizing epitope. Antibodies to the a determinant confer protection in adults against all the common subtypes of HBV. Subtype determinants d or y and w or r are also located within the predicted loop regions. Nine serotypes have been described (10). These have been related to six genomic groups, A through F, based on sequencing of the S gene of isolates from different geographical regions (26, 27).

HBsAg is one of the first serum markers to appear during the course of HBV infection and can be detected 2 to 8 weeks before biochemical evidence of liver dysfunction and the onset of jaundice. HBsAg is cleared within a few months in self-limiting illness. If HBsAg persists for more than 6 months, spontaneous clearance is very improbable and the infected individual is considered a chronic HBV carrier.

Among the many commercially licensed HBsAg assays available, enzyme-linked immunosorbent assays are currently the most frequently used. These assays use either monoclonal or polyclonal anti-HBs bound to a solid phase and a second labelled anti-HBs to detect the captured antigen. Despite the high-level performance of screening assays, transfusion-associated HBV infection is still reported (17, 21). There are three possible explanations of false-negative results in commercial assays. In chronic HBV carriers, the HBsAg level may be below the detection limit, i.e., a high proportion of individuals with antibodies against HBV core antigen (anti-HBc) as the only serological marker of infection are low-level chronic carriers of the virus (14, 20). Another explanation is that virus variants yield sequences that are not recognized by the antibodies employed in the assays. In different geographic locations, vaccine-escape mutants are emerging under the selective pressure of active immunization, and there is a danger that they will become dominant strains as vaccination becomes universal (6, 19). Breakthrough infections due to point mutations of the a determinant have been identified in Europe, Africa, and Asia (8, 13, 28, 40). Vaccine-escape mutants within the a determinant of the S gene are not recognized as effectively by conventional diagnostic tests as the wild-type particle (9, 18). A third possible explanation is that there are variants in other parts of the genome that downregulate the production of HBsAg (7).

To reduce the residual risk of transfusion-associated hepatitis B, the sensitivity of HBsAg screening assays is continuously improved. Elecsys HBsAg (Roche Diagnostics, Penzberg, Germany) is a new, fully automated, rapid assay which permits the qualitative detection of HBsAg directly from the patient blood collection tube in a total incubation time of 18 min. In a multicenter study, it was compared with two alternative well-established serological assays.

MATERIALS AND METHODS

Elecsys HBsAg.

Elecsys HBsAg is a two-step sandwich assay for the qualitative detection of HBsAg in human serum or plasma, with a total incubation time of 18 min. All the reaction steps of Elecsys HBsAg are performed automatically by the Elecsys 2010 system. The assay can be performed directly from the primary patient blood collection tube, and random access is possible.

In the first incubation step, biotinylated and ruthenylated monoclonal antibodies directed against HBsAg are added to 50 μl of sample, and the mixture is incubated for 9 min at 37°C. If HBsAg is present in the sample, it binds to the biotinylated and ruthenium [Tris (2,2′-bipyridyl)ruthenium (II)-complex] [RU(bpy)32+]-labelled antibodies to form a sandwich complex. During the second incubation period (9 min at 37°C), immune complexes bind to streptavidin-coated magnetic microparticles which are added to the mixture. In the measuring cell of the Elecsys 2010 system, the microparticles are magnetically captured on the surface of the electrode. Unbound substances are removed with ProCell. Application of voltage to the electrode induces chemiluminescence, which is measured with a photomultiplier.

Results are calculated with the Elecsys software by comparing the chemiluminescence signal obtained from the sample with the cutoff value previously obtained by HBsAg calibration. Samples with a signal/cutoff ratio (s/co) of ≥1 are considered positive. If the s/co is <1, the sample is considered negative.

Elecsys HBsAg confirmatory test.

Repeatedly HBsAg-reactive samples are preincubated during 30 min at room temperature or overnight at 2 to 8°C in a ratio depending on the index value with anti-HBs-positive plasma in order to confirm the presence of HBsAg by specific antibody neutralization. In parallel, the same sample volume is incubated with control reagent (anti-HBs-negative serum). If HBsAg is present in the specimen, it will form an immune complex with the blocking anti-HBs antibodies, thereby reducing the number of binding sites available for the biotinylated and ruthenylated antibodies of Elecsys HBsAg. A specimen is considered positive if the reduction in absorbance of the neutralized specimen is 50% or greater compared to that of the nonneutralized control.

Comparative assays.

Elecsys HBsAg was compared to two antigen detection assays, Auszyme Monoclonal and IMx HBsAg (Abbott, Delkenheim, Germany). For the Auszyme Monoclonal assay, overnight incubation was performed (version B) in order to achieve a high level of sensitivity, and the assay will be referred to hereafter as Auszyme Monoclonal B.

Anti-HBs, anti-HBc immunoglobulin M (IgM), hepatitis B early antigen (HBeAg), and anti-HBe determinations were performed using IMx assays (IMx AUSAB, IMx CORE, IMx CORE-M, IMx HBeAg, and IMx anti-HBe; Abbott) and AxSYM tests (AxSYM AUSAB, AxSYM CORE, AxSYM CORE-M, AxSYM HBeAg, and AxSYM anti-HBe; Abbott).

Quantitative determination of anti-HBc-IgM was performed with VIDAS HBc IgM (Biomérieux, Marcy-l’Etoile, France). VIDAS HBc IgM is an automated enzyme-linked immunofluorescence assay based on μ-capture technology with a detection limit of 10 Paul Ehrlich Institute units (PEIU)/ml.

All the tests were performed and interpreted in accordance with the manufacturers’ recommendations.

HBV DNA PCR.

HBV DNA PCR was performed as described by Berger et al. (1). Briefly, DNA was extracted from serum samples (25 μl) by using the QIAmp blood kit (Qiagen, Hilden, Germany) according to the manufacturer’s recommendations. For DNA amplification, a nested PCR protocol was performed. For the first PCR, primers HBV1 (5′-TCg TgT TAC Agg Cgg ggT TT) and HBV2 (5′-CgA ACC ACT gAA CAA Atg gC) flanking a 513-bp fragment of the S genome region were used. With the inner primers HBV3 (5′-gCC AAA ATT CgC AgT CCC CA) and HBV4 (5′-AgA TgA ggC ATA gCA gCA gg), a 129-bp sequence was amplified. The reaction mixture (50 μl) consisted of 5 μl of extracted DNA, 5 μl of 10× PCR buffer (Boehringer Mannheim, Mannheim, Germany), 5 μl of deoxynucleoside triphosphate (dNTP) mix (2 mM), 1 μl of each primer, and 0.4 μl of Taq polymerase (5 U/μl) in 32.6 μl of H2O. Amplification was performed with a Gene Amp 9600 device (Perkin-Elmer, Norwalk, Conn.) and involved 30 cycles in which primers were denatured at 95°C (30 sec), annealed at 55°C (30 sec), and extended at 72°C (45 sec). Amplificates were visualized by 4% agarose gel electrophoresis and ethidium bromide staining.

Specimens.

The intra- and interassay precision was assessed by testing replicates of calibrators, controls, and positive and negative serum samples (Tables (Tables11 and and2).2). The technical performance of Elecsys HBsAg was assessed by testing HBsAg-positive and HBsAg-negative samples spiked with hemoglobin, bilirubin, triglyceride, biotin, or rheumatoid factor and sera with increasing concentrations of IgG, IgA, and IgM.

TABLE 1
Intra-assay precision of testing with 20 replicates of calibrators, controls, and human sera
TABLE 2
Interassay reproducibility as determined by measuring calibrators, controls, and serum samples in nine different runs

The following specimens were tested to evaluate the sensitivity of the assays. (i) Serial dilutions of the Paul Ehrlich Institute (PEI) (Langen, Germany) standard for HBsAg subtypes ad (1,000 PEIU/ml [2,360 IU/ml]) and ay (1,000 PEIU/ml [3,210 IU/ml]), HBsAg reference material WHO (HBsAg subtype ad, first international standard established in 1985; code 80/549; 100 IU/ml), and two highly HBsAg-positive serum samples (subtypes ad and ay) in an HBsAg- and anti-HBs-negative serum pool.

(ii) Three sensitivity panels, including the HBsAg sensitivity panel PHA 805 (Boston Biomedica Inc. [BBI], West Bridgewater, Mass.), the Abbott sensitivity panel (lot B, 03564M400), and the Laboratoire National de Santé [LNS] panel (Etablissement National de Transfusion Sanguine, Nord-Pas-de-Calais, France). HBsAg sensitivity panel PHA 805 includes 10 subtype ad and 10 subtype ay sera with decreasing HBsAg concentrations (2.5 to 0.1 ng/ml). The Abbott sensitivity panel consists of 16 subtype ad and ay serum samples with HBsAg concentrations ranging from 0.06 to 3.99 ng/ml. Included in the LNS panel are HBsAg-positive serum samples of different subtypes (adw2, adw4, adr, ayw1, ayw2, ayw3, ayw4, and ayr) with concentrations ranging from 0.61 to 0.80 ng/ml.

(iii) Two low-titer panels (PHA 102 and 103 [BBI]) composed of 11 or 15 serum or plasma samples, respectively, with low HBsAg concentrations (0.4 to 2.3 ng/ml).

(iv) Two mixed-titer panels (PHA 202 and 203 [BBI]) including 25 undiluted plasma and 25 serum samples from asymptomatic blood donors with HBsAg concentrations ranging from 0.2 to 2.6 ng/ml.

(v) Serial dilutions of four HBsAg-positive samples with concentrations greater than 50,000 IU/ml and of a serum sample which was spiked with purified HBsAg up to a concentration of 2,000,000 IU/ml.

(vi) Twenty-three commercially available seroconversion panels (BBI) consisting of follow-up samples which were collected at weekly or monthly intervals from patients suffering from acute hepatitis B. All the panels were characterized for HBV-specific serological markers (anti-HBs, anti-HBc, anti-HBc-IgM, HBeAg, and anti-HBe).

(vii) HBsAg-positive sera from patients at different stages of HBV infection (n = 228). All these serum samples were preselected on the basis of clinical diagnoses. Included were samples from patients with acute hepatitis B (n = 20), chronic hepatitis B (n = 180), and chronic active hepatitis B (n = 4), and samples from asymptomatic HBsAg carriers (n = 24).

(viii) Serum samples preselected on the basis of the seroconstellation of HBV-specific markers (n = 233), including samples with the serological profile of acute hepatitis (HBsAg, HBeAg, and anti-HBc-IgM positive; n = 46), and 179 HBsAg-positive samples preselected irrespective of the results of other HBV-specific serological markers.

(ix) Isolated anti-HBc-positive samples (n = 147). All the samples had been tested with IMx and AxSYM CORE assays and showed inhibition values of >90%. Human immunodeficiency virus (HIV) and hepatitis C virus (HCV) coinfection were present in 19.5 and 35.2% of isolated anti-HBc-positive individuals, respectively. Of the isolated anti-HBc positives, 35.6% were coinfected with both HIV and HCV. Five samples were HBV DNA positive. Two samples were anti-HBc-IgM antibody positive with IMx CORE-M. With the more sensitive VIDAS HBc IgM, specific IgM antibody was detected in 10 samples (6.8%) at concentrations ranging from 10 to >200 PEIU/ml.

(x) Dilution series of different HBsAg-positive serum samples of different HBV subtypes and surface antigen mutants in a negative serum pool were tested in order to assess the influence of genetic variability on HBsAg assays.

The Centre National de la Transfusion Sanguine (CNTS) Panel (CNTS Foundation, Paris, France) consists of nine HBsAg-positive specimens of different subtypes (ayw1, ayw2, ayw3, ayw4, ayr, adw2, adw4, adrq and adrq+) with an analyte concentration of 10 ng/ml. The HBsAg subtype of each specimen was serologically determined.

HBsAg-positive specimens (n = 7) were obtained from different geographical regions. The S genome region of HBV DNA was amplified by PCR. Sequencing and sequence analysis were performed by MediGene (Martinsried, Germany). DNA sequences and deduced amino acid sequences were classified according to their sequence homology with HBV strains of defined genotypes and subtypes. The genotypes and subtypes of the HBsAg-positive specimens were as follows: A and adw2, B and adw2, C and adr, D and ayw2, D and ayw3, and E and ayw4 (n = 2).

Three surface antigen mutants with amino acid substitutions in positions 133 (M→L), 103 (G→T), and 126 (I→N) and one with an amino acid insertion behind position 113 (T) were tested.

For the evaluation of specificity, the following selected specimens were comparatively tested with different HBsAg assays: samples from HBsAg-negative blood donors from South Africa (n = 490) and a high number of potentially interfering samples, including rheumatoid factor-positive sera (n = 40); sera from anti-HBV-vaccinated individuals (n = 20); sera from patients suffering from alcohol-related liver disease (n = 28), acute or chronic viral and bacterial infections (n = 119), and autoimmune diseases (n = 35); sera from dialysis patients (n = 182), hemophiliacs (n = 20), intravenous drug abusers (IVDAs) (n = 116), homosexuals (n = 97), and pregnant women (n = 384); and specimens from hospitalized patients who were treated for cancer, nephropathy, and gastroenterologic symptoms (n = 449).

Table Table33 gives an overview of the serum samples that were analyzed with the different assays.

TABLE 3
Specimens tested and results obtained with three HBsAg screening assays

Statistical analysis.

The performance of Elecsys HBsAg was compared to those of Auszyme Monoclonal B and IMx HBsAg for the seroconversion panels. The mean number of days by which HBsAg was detected with Elecsys HBsAg in comparison to Auszyme Monoclonal B and IMx HBsAg was determined for the 23 seroconversion panels tested. The statistical significance of the results for each test was determined by using the Wilcoxon test for matched pairs (5). Anti-HBc, anti-HBc-IgM, anti-HBs, anti-HBe, and HBeAg determinations were performed for the resolution of discrepant results between HBsAg assays. Discordant samples were considered for the calculation of sensitivity and specificity only if results from at least three comparative assays and/or a neutralization assay were available. A sample was considered a true positive if it was repeatedly reactive in at least two assays and confirmed by neutralization assay (if enough sample material was available) and if it was anti-HBc positive and anti-HBs negative. For seroconversion panels and sensitivity panels, repeated reactivity in a single assay was considered as a true positive result. Conversely, a test result was interpreted as a true negative if it was negative with Elecsys HBsAg and in at least one comparative assay and confirmatory testing (if enough sample material was available) and if it showed one of the following serological constellations: anti-HBc and anti-HBs negative (HBV negative), anti-HBc and anti-HBs positive (resolved infection), or anti-HBc negative and anti-HBs positive (HBV vaccination).

RESULTS

Technical performance.

The intra-assay precision was validated by testing 20 replicates of two calibrators, positive and negative controls, and one HBsAg-negative and two HBsAg-positive serum samples (Table (Table1).1). The intra-assay coefficients of variation (CV) ranged from 1.2 to 6.9%. The interassay reproducibility of Elecsys HBsAg was evaluated by testing in nine different runs the positive and negative calibrators, one positive and one negative control, one borderline-negative, one low-positive, and two negative serum samples (Table (Table2).2). The interassay CV ranged between 3.1 and 11.4%. No interference was observed with HBsAg-positive and -negative serum samples spiked with hemoglobin (up to 1.4 g/dl), triglycerides (up to 2,884 mg/dl), bilirubin (up to 30.9 mg/dl), and rheumatoid factor (up to 3,500 IU/ml) and samples with high concentrations of IgG (up to 6,900 mg/dl), IgM (up to 3,400 mg/dl), and IgA (up to 3,600 mg/dl). With biotin concentrations higher than 60 ng/ml, a weakly HBsAg-positive sample (0.06 IU/ml) was determined to be a false negative.

Sensitivity.

Overall, Elecsys HBsAg showed a higher sensitivity than Auszyme Monoclonal B and IMx HBsAg for detecting HBsAg subtypes ad and ay in dilution series of HBsAg PEI and WHO standards, BBI HBsAg sensitivity panel PHA 805, and the Abbott sensitivity panel (Table (Table4).4). The calculation of the detection limit by interpolation yielded 0.014 PEIU/ml (0.033 IU/ml) and 0.017 PEIU/ml (0.055 IU/ml) for PEI standards ad and ay, respectively. No difference in sensitivity between Elecsys HBsAg and Auszyme Monoclonal B was observed for HBsAg-positive specimens 4/ay and 10/ad and BBI HBsAg low-titer panels 102 and 103 (Table (Table4).4). IMx HBsAg showed a higher detection limit in HBsAg low-titer panel 102 than in the alternative assays (Table (Table4).4). Two specimens of low-titer panel 103 tested negative with IMx HBsAg but were positive with Elecsys HBsAg and Auszyme Monoclonal B.

TABLE 4
Results of titration of HBsAg standards and HBsAg-positive serum samples and testing of sensitivity panels with different HBsAg assays

The specimens with the lowest HBsAg concentrations of the mixed-titer panels (0.6 to 0.7 ng/ml) were all detected with Elecsys HBsAg, Auszyme Monoclonal B, and IMx HBsAg (Table (Table44).

All the different subtypes (adw2, adw4, adr, ayw1, ayw2, ayw3, ayw4, and ayr) of the LNS panel were detected with Elecsys HBsAg and Auszyme Monoclonal B. For PEI standards ad and ay, Elecsys HBsAg showed a linear dilution behavior in the concentration ranges from 0 to 236 IU/ml and 0 to 321 IU/ml, respectively. A saturation effect was present at 2.36 IU/ml for Auszyme Monoclonal B (data not shown). A prozone effect was observed for Elecsys HBsAg at concentrations higher than 500 IU/ml. However, even at concentrations greater than 2,000,000 IU/ml, no false-negative results due to a high-dose hook effect were obtained.

HBsAg was detected in 11 of 23 seroconversion panels between 2 and 16 days earlier with Elecsys HBsAg than with Auszyme Monoclonal B (Table (Table5).5). Elecsys HBsAg showed a higher sensitivity for acute HBV infections due to subtype ad, since HBsAg was detected earlier in 9 of 16 subtype ad seroconversion panels. One of four subtype ay infections was detected earlier with Elecsys HBsAg than with Auszyme Monoclonal B. In comparison to IMx HBsAg, Elecsys HBsAg showed a higher sensitivity for detection of acute hepatitis B, independent of HBV subtype. In 16 of 23 seroconversion panels, HBsAg was detected between 2 and 16 days earlier with Elecsys HBsAg (Table (Table5).5). Overall, in the 23 seroconversion panels tested, Elecsys HBsAg detected acute hepatitis B an average of 3.6 and 5.3 days earlier than Auszyme Monoclonal B and IMx HBsAg, respectively. The performance of Elecsys HBsAg was significantly better (P < 0.05) than that of Auszyme Monoclonal B and IMx HBsAg.

TABLE 5
Comparison of the performance of different HBsAg assays for detection of acute hepatitis B in seroconversion panels

All the samples from patients suffering from acute hepatitis B tested positive with Elecsys HBsAg and Auszyme Monoclonal B (Table (Table3).3). A total of 169 of 170 chronic hepatitis B patients gave congruently positive results with Elecsys HBsAg and Auszyme Monoclonal B (Table (Table3).3). One sample from a patient with primary hepatocarcinoma was repeatedly reactive with Elecsys HBsAg but was negative according to Auszyme Monoclonal B. The Auszyme radioimmunoassay (Abbott, North Chicago, Ill.) confirmed the positive Elecsys HBsAg result. The pattern of serological markers of this sample was HBeAg negative, anti-HBe positive, and weakly postiive for anti-HBc-IgM. Elecsys HBsAg and IMx HBsAg showed concordant results with the 46 samples with serological patterns of acute hepatitis B (Table (Table3).3). All of the 179 preselected HBsAg-positive sera were positive in Auszyme Monoclonal B and Elecsys HBsAg (Table (Table33).

Of the 147 isolated anti-HBc-positive specimens, four (2.7%) were repeatedly reactive with Elecsys HBsAg, and this result was confirmed by the neutralization assay. HBsAg was not detected with IMx HBsAg in these four serum samples. One sample was HBV DNA positive and anti-HBc-IgM negative (<10 PEIU/ml). The remaining samples were HBV DNA and anti-HBc-IgM negative.

HBsAg subtypes and mutants.

Elecsys HBsAg showed a two- to eightfold-higher sensitivity than Auszyme Monoclonal B for subtypes ayw1, ayw2, ayw4, adw2, adw4, adrq+, and adrq of the CNTS panel (Table (Table6).6). Of the seven subtyped sera of various geographical origins, Elecsys HBsAg detected subtypes adw2 and ayw4 and the mixed subtype adr/ayr with a sensitivity better than that of Auszyme Monoclonal B (Table (Table6).6). No difference in sensitivity was observed between both tests for HBsAg surface mutants (Table (Table6).6). The insertion mutant was detected only at a low dilution of 1/5 with both assays.

TABLE 6
Sensitivity of Elecsys HBsAg and Auszyme Monoclonal B for detection of HBsAg subtypes and mutants

Specificity.

One sample of 490 preselected HBsAg-negative specimens from South African blood donors was repeatedly reactive and confirmed by the Elecsys confirmatory assay (Table (Table7).7). IMx HBsAg gave a negative result, while Auszyme Monoclonal B showed a very weak reactivity. The serological profile (anti-HBc positive, anti-HBc-IgM negative, anti-HBe positive, HBeAg negative, and anti-HBs negative) was compatible with an HBsAg-positive HBV carrier (see Table Table9).9). Further, five samples were repeatedly weakly reactive with IMx HBsAg but negative in the neutralization assay and Auszyme Monoclonal B. All five serum samples showed the serological profile of resolved HBV infection (anti-HBs positive, anti-HBc positive, anti-HBe positive, and anti-HBc-IgM and HBeAg negative) (see Table Table9).9).

TABLE 7
Specificity of Elecsys HBsAg and IMx HBsAg assays
TABLE 9
Specificity of three HBsAg assays and resolution of discrepant resultsa

The results obtained by testing potentially cross-reactive serum samples are summarized in Table Table8.8. Among the rheumatoid factor-positive specimens, 3 of 40 were found to be reactive with Elecsys HBsAg and were confirmed positive with the Elecsys HBsAg confirmatory test and Auszyme Monoclonal B. One anti-HCV-positive serum sample was weakly reactive with Elecsys HBsAg but negative with Auszyme Monoclonal B (Table (Table8).8). There was no more sample material for further investigations; therefore, that sample was not taken into consideration for the calculation of specificity (Table (Table9).9). Two samples from dialysis patients were repeatedly reactive with Elecsys HBsAg and confirmed positive with the Elecsys HBsAg confirmatory test and Auszyme Monoclonal B (Table (Table8).8).

TABLE 8
Determination of the specificity of Elecsys HBsAg and Auszyme Monoclonal B assays

Three serum samples from pregnant women were repeatedly reactive with Elecsys HBsAg and IMx HBsAg (Table (Table7).7). In contrast, 11 samples were repeatedly reactive with IMx HBsAg but negative with Elecsys HBsAg. Only one IMx HBsAg-reactive sample was confirmed positive after retesting with Auszyme Monoclonal B. The result for this serum sample, which was reactive in both Abbott assays, could not be confirmed by IMx HBsAg neutralization assay. The sample was probably from a vaccinated individual, since anti-HBs was isolated positive (Table (Table9).9). Consequently, the Elecsys HBsAg result was considered a true negative. The other 10 Elecsys HBsAg-negative and IMx HBsAg-positive samples all tested negative with Auszyme Monoclonal B (Table (Table9).9). The HBV serological constellations of these serum samples (anti-HBs and anti-HBc negative, anti-HBc and anti-HBs positive, and anti-HBc negative and anti-HBs positive) were compatible with true negative results of Elecsys HBsAg and Auszyme Monoclonal B (Table (Table99).

Overall, 17 of the 449 specimens from hospitalized patients were reactive with Elecsys HBsAg and IMx HBsAg (Table (Table7).7). Discrepant results were observed for 27 serum samples. One Elecsys HBsAg-positive and IMx HBsAg-negative sample was confirmed HBsAg positive with the Elecsys HBsAg confirmatory test and Auszyme Monoclonal B. The serological profile of this patient (anti-HBc and anti-HBe positive, anti-HBc-IgM and anti-HBs negative) is compatible with that of an HBV carrier (Table (Table9).9). However, isolated anti-HBc reactivity or resolved infection with loss of anti-HBs cannot be definitively ruled out. There was not enough sample material from the 26 IMx-reactive specimens for confirmatory testing. All the samples tested negative with Auszyme Monoclonal B. The serological profiles of these samples (all markers negative; anti-HBs, anti-HBe, and anti-HBc positive; and anti-HBe and anti-HBs positive) were compatible with the true-negative results obtained with Elecsys HBsAg and Auszyme Monoclonal B (Table (Table99).

Of the 116 specimens from IVDAs, 13 samples were congruently repeatedly reactive with Elecsys and IMx HBsAg, and 10 samples gave discordant results with both assays (Table (Table7).7). One Elecsys HBsAg-positive and IMx HBsAg-negative sample (anti-HBc and anti-HBe positive, anti-HBc-IgM and anti-HBs negative) was confirmed to be HBsAg positive by retesting with Auszyme Monoclonal B and the Elecsys HBsAg confirmatory test (Table (Table9).9). The HBV seroconstellation was compatible with that of a true positive Elecsys HBsAg and Auszyme Monoclonal B result (Table (Table9).9). Nine remaining discordant samples (Elecsys HBsAg negative and IMx HBsAg positive) were confirmed as true negatives by Auszyme Monoclonal B. HBV serology was in favor of the Elecsys HBsAg and Auszyme negative results in eight of nine cases (anti-HBc positive and anti-HBs positive, anti-HBc negative and anti-HBs negative, anti-HBc negative and anti-HBs positive; Table Table9).9). One sample was considered indeterminate since the seroconstellation was compatible with that of an HBsAg carrier but two of the three screening assays were negative. Furthermore, isolated anti-HBc reactivity or resolved infection could not be definitively excluded.

Seven of 90 samples from homosexuals were congruently reactive with Elecsys HBsAg and IMx HBsAg (Table (Table7).7). Two samples gave discordant results (IMx HBsAg positive but Elecsys HBsAg negative). Further testing with Auszyme Monoclonal B and IMx HBsAg neutralization assay (inhibition, <50%) and anti-HBc and anti-HBs testing confirmed one Elecsys HBsAg result as a true negative (Table (Table9).9). The second sample was considered indeterminate since two of the three HBsAg assays were negative and the serological profile was anti-HBc positive and anti-HBs negative.

The sensitivity and specificity of Elecsys HBsAg were 100% (Table (Table10).10). Auszyme Monoclonal B had 99.9% specificity, and its sensitivity was 96.6%. A high number of false-positive results was observed with the IMx HBsAg; its specificity was 97.6%. IMx HBsAg also showed a poorer sensitivity than the other assays (Table (Table10).10).

TABLE 10
Sensitivities, specificities, and predictive values of three HBsAg assays

DISCUSSION

The results of the present study show that Elecsys has a sensitivity better than those of well-established HBsAg screening assays. Furthermore, Elecsys HBsAg combines a high sensitivity (100%) and specificity (100%), whereas usually these criteria are mutually exclusive. By testing samples from a total of 4,120 unselected blood donors in different blood banks (supplied by A. Mühlbacher and U. Michl, Blutzentrale der Landeskrankenanstalten Salzburg, Salzburg, Austria; R. Camacho, Hospital Egas Moniz, Laboratorio di Imunohemoterapia, Lisbon, Portugal; and J. H. Hernandez, Banco de Sangre de ICS, Hospital Valle de Hebron, Barcelona, Spain), a specificity of 100% was obtained (data not shown).

The benefit of high sensitivity was particularly evident in seroconversion panels (Table (Table5).5). Elecsys HBsAg detected acute hepatitis B significantly earlier than the comparative assays. Elecsys was also more sensitive than the Monolisa Ag HBs second generation assay (Sanofi Pasteur, Marnes-la-Coquette, France). Among 15 commercially available HBsAg tests, including Auszyme Monoclonal B and IMx HBsAg, Elecsys HBsAg achieved the highest sensitivity in BBI seroconversion panels (29). Comparing the results obtained from 21 BBI seroconversion panels as reported by Palmer et al. (29) with our results, Elecsys HBsAg detected acute HBV infection earlier than the Monolisa Ag HBs second generation assay in 8 of 21 seroconversion panels. Elecsys HBsAg also showed a better performance in BBI seroconversion panels than an alternative new automated microparticle chemiluminescence immunoassay (16).

A high rate (2.4%) of false-positive results was obtained with IMx HBsAg for specimens from blood donors, hospitalized patients, pregnant women, and IVDAs (Tables (Tables3,3, ,7,7, and and9).9). In contrast, Auszyme Monoclonal B showed 99.9% specificity. Previous reports (31, 35) showed a false-positivity rate of 0.2 to 0.3% with Auszyme Monoclonal B. Nonspecific reactivity may be reduced by overnight storage of samples at 2 to 8°C (31).

For the resolution of discrepant results of HBsAg screening assays, it is important to consider additional HBV markers. HBsAg in combination with anti-HBc determination shows a positive and negative predictive value of 100%. However, during acute hepatitis B, anti-HBc may not be detected in the first days to weeks in up to 8% of the cases (24). Another explanation for isolated HBsAg-positive results may be the absence of anti-HBc antibody response as a consequence of iatrogenic or acquired immunodeficiency (25). Occasionally, anti-HBc reactivity is absent in patients with selective immunodeficiency or immunocompromised HBsAg carriers (11, 23, 24, 30).

Depending on the prevalence of HBV infection and the patient group investigated, 1 to 32% of positive anti-HBc results are isolated positive findings (15). Isolated anti-HBc reactivity is frequently observed in IVDAs, HIV-infected individuals (32, 33), HBV and HCV coinfected patients (20), and pregnant women (14). Several reports have shown that 10 to 40% of individuals positive solely for anti-HBc are chronic carriers (2, 3, 14, 20, 34, 41) and that they can transmit HBV via blood transfusion or from mother to child (38). Anti-HBc-positive organ donors are a potential source of transmission of HBV to their recipients (12). The infectivity of these individuals can be confirmed by the detection of HBV DNA by hybridization (22) or PCR (14, 32, 37, 38, 41). A collection of isolated anti-HBc-positive individuals was included in our study in order to test the hypothesis that a highly sensitive HBsAg assay may detect low-level HBsAg in HBV carriers. With Elecsys HBsAg, 2.7% of isolated anti-HBc-reactive samples that were negative with IMx HBsAg tested positive (Table (Table1);1); only one of these samples was PCR positive. Of 2,000 antenatal clinical attendees in Papua New Guinea, 5% of HBsAg-positive subjects were negative by a widely used monoclonal assay but positive by PCR. The monoclonal assay had a sensitivity of 0.5 to 1 ng/ml. These samples were reactive in another assay with a sensitivity of 0.1 ng/ml. Over 50% of these discordant samples had rare or unique variants of the major hydrophilic region of HBsAg (7). These data also demonstrate that the sensitivity of current screening HBsAg assays should be increased, arguing for implementation of detection of HBV DNA in blood donors. In the clinical diagnostic laboratory, potentially infectious low-level HBV carriers may also be detected by using highly sensitive anti-HBc-IgM assays (4). Another reason for the possible failure of commercial assays in detecting surface antigen in isolated anti-HBc reactive HBV carriers is that the a determinant is more variable and shows a larger variability of the whole protein than HBsAg-positive controls (43). Polyclonal-antibody-based assays do not guarantee full sensitivity (42). Modification of commercial assays is necessary to increase the sensitivity of detection of S gene variants. By using monoclonal antibodies directed against different S gene mutants in a prototype assay, it was possible to achieve a higher sensitivity in dilution series of HBsAg variants than with serological assays (unpublished data).

Elecsys HBsAg and Auszyme Monoclonal B showed a high sensitivity for three surface antigen mutants with amino acid substitutions in positions 133 (M→L), 103 (G→T), and 126 (I→N). However, one mutant with an amino acid insertion behind position 113 (T) was detected only at a low dilution of 1/5 (Table (Table6).6). These data illustrate that the sensitivity of commercial assays is variable for the different surface mutants. It is conceivable that false-negative results may be obtained with serological assays, especially for low-level HBV mutant carriers.

Elecsys HBsAg detected subtypes adw2 and ayw1 (genotype A), adw4 (genotype F), ayw2 (genotype D), ayw4 (genotype E), and adr (genotype C) with a sensitivity better than that of Auszyme Monoclonal B. According to these data, Elecsys HBsAg seems to be more accurate for HBsAg screening in different geographical locations.

In conclusion, Elecsys HBsAg is highly sensitive and specific and represents a major improvement over the alternative assays for the detection of HBsAg in blood donors and in routine laboratory diagnostics. Future developments of serological assays should include monoclonal antibodies that recognize epidemiologically relevant surface antigen mutants and further optimization of sensitivity. An alternative would be the screening of blood donors with nucleic acid amplification techniques. However, technology for nucleic acid testing in blood screening laboratories has not yet been developed. The limitations of these procedures are mainly linked to the difficulties related to the automation of sample processing and to the possibilities of cross-contamination of samples due to the high sensitivity of amplification methods.

REFERENCES

1. Berger, A., H. W. Doerr, H. F. Rabenau, M. Teebrücke, and B. Weber. Epidemiological analysis of the significance of positive test results for antibody to hepatitis B core antigen (anti-HBc). Submitted for publication.
2. Bréchot C, Degos F, Lugassy C, Thiers V, Zafrani S, Franco D, Bismuth H, Trépo C, Benhamou J, Wands J, Isselbacher K, Tiollais P, Berthelot P. Hepatitis B virus DNA in patients with chronic liver disease and negative tests for hepatitis B surface antigen. N Engl J Med. 1985;312:270–276. [PubMed]
3. Bréchot C, Kremsdorf D, Paterlini P, Thiers V. Hepatitis B virus DNA in HBsAg-negative patients: molecular characterization and clinical implications. J Hepatol. 1991;4:S49–S55. [PubMed]
4. Brunetto M R, Cerenzia M T, Oliveri F, Piantino P, Randone A, Calvo P L, Rocca G, Galli C, Bonino F. Monitoring the natural course and response to therapy of chronic hepatitis B with a semi-quantitative assay for IgM anti-HBc. J Hepatol. 1993;19:431–436. [PubMed]
5. Bünung H, Trenkler G. Nicht-parametrierbare Methoden. Berlin, Germany: Springer-Verlag; 1978.
6. Carman W F. The clinical significance of surface antigen variants of hepatitis B virus. J Viral Hepat. 1997;4(Suppl. 1):S11–S20. [PubMed]
7. Carman W F, Mimms L T. Pre-S/S gene variants of hepatitis B virus. In: Rizetto M, Purcell R H, Gerin J L, Verne G, editors. Viral hepatitis and liver disease. Torino, Italy: Edizioni Minerva Medica; 1997. pp. 108–115.
8. Carman W F, Zanetti A R, Karayiannis P, Waters J, Manzillo G, Tanzi E, Zuckerman A J, Thomas H C. Vaccine-induced escape mutant of hepatitis B virus. Lancet. 1990;336:325–329. [PubMed]
9. Carman W F, Van Deursen F J, Mimms L T, Hardie D, Coppola R, Decker R, Sanders R. The prevalence of surface antigen variants of hepatitis B virus in Papua New Guinea, South Africa, and Sardinia. Hepatology. 1997;26:1658–1666. [PubMed]
10. Courouce-Pauty A M, Plancon A, Soulier J P. Distribution of HBsAg subtypes in the world. Vox Sang. 1983;35:197–211. [PubMed]
11. Coursaget P, Yvonnet B, Bourdil C, Mevelec M N, Adamowicz P, Barres J L, Chotard J, N’Doye R, Diop Mar I, Chiron J P. HBsAg positive reactivity in a man not due to hepatitis B virus. Lancet. 1987;ii:1354–1358. [PubMed]
12. Dodson S F, Issa S, Araya V, Gayowski T, Pinna A, Eghtesad B, Iwatsuki S, Montalvo E, Rakela J, Fung J. Infectivity of hepatic allografts with antibodies to hepatitis B virus. Transplantation. 1997;64:1582–1584. [PubMed]
13. Fortuin M, Karthigesu V, Allison L, Howard C, Hoare S, Mendy M, Witthle H C. Breakthrough infections and identification of a viral variant in Gambian children immunized with hepatitis B vaccine. J Infect Dis. 1994;169:1374–1376. [PubMed]
14. Gross A, Joller-Jemelka H I, Wicki A N, Grob P J. Der Hepatitis-serologische Befund “Anti-HBc allein”, zirkulierende virale DNS und Befund-Interpretation. Schweiz Med Wochenschr. 1993;123:1193–1202. [PubMed]
15. Hadler S C, Murphy B L, Schable C A, Heyward W L, Francis D P, Kane M A. Epidemiological analysis of the significance of low-positive test results for antibody to hepatitis B surface and core antigens. J Clin Microbiol. 1984;19:521–525. [PMC free article] [PubMed]
16. Hall J, Jilg W, Hottenträger B, Bonnar P, Fang C, Baker L. Performance of a chemiluminescent immunoassay for HBsAg on the new high-throughput and fully automated ACS: CentaurTM system. Clin Lab. 1998;44:349–354.
17. Hoofnagle J. Post-transfusion hepatitis B. Transfusion. 1990;30:384–386. [PubMed]
18. Howard C R. The structure of hepatitis B envelope and molecular variants of hepatitis B virus. J Viral Hepat. 1995;2:165–170. [PubMed]
19. Howard C R, Allison L M C. Hepatitis B surface antigen variation and protective immunity. Intervirology. 1995;38:35–40. [PubMed]
20. Jilg W, Sieger E, Zachoval R, Schätzl H. Individuals with antibodies against hepatitis B core antigen as the only serological marker for hepatitis B infection: high percentage of carriers of hepatitis B and C virus. J Hepatol. 1995;23:14–20. [PubMed]
21. Kojima M, Shinizu M, Tsuchimochi T, Koyasu M, Tanaka S, Iizuka H, Tanaka T, Okamoto H, Tsuda F, Miyakowa Y. Post-transfusion fulminant hepatitis B associated with pre-core defective HBV mutants. Vox Sang. 1991;60:34–39. [PubMed]
22. Lai K N, Lai F M M, Leung N W Y, Lo S, Tam J S. Hepatitis with isolated serum antibody to hepatitis B core antigen. Am J Clin Pathol. 1990;93:79–83. [PubMed]
23. Lazizi Y, Dubreuil P, Pillot J. Excess HBcAg in HBc antibody-negative chronic hepatitis B virus carriers. Virology. 1993;17:966–970. [PubMed]
24. Lee H S, Rajagopalan M S, Chien D, Cordell R, Perkins H A, Vyas G N. Specificity of enzyme immunoassays for hepatitis B core antibody used in screening blood donors. Transfusion. 1987;27:103–106. [PubMed]
25. Lee J H, Paglieroni T G, Holland P V, Zeldis J B. Chronic hepatitis B virus infection in an HBc antibody-nonreactive blood donor: variant virus or defective immune response? Hepatology. 1992;16:24–30. [PubMed]
26. Norder H, Hammas B, Lofdahl S, Courouce A M, Magnius L O. Comparison of the amino acid sequences of nine different serotypes of hepatitis B surface antigen and the genomic classification of the corresponding hepatitis B virus strains. J Gen Virol. 1992;73:1201–1208. [PubMed]
27. Norder H, Hammas B, Lee S D, Bile K, Courouce A M, Mushahwar I K, Magnius L O. Genetic relatedness of hepatitis B viral strains of diverse geographical origin and natural variations in the primary structure of the surface antigen. J Gen Virol. 1993;74:1341–1348. [PubMed]
28. Oon C J, Lim G K, Ye Z, Goh K T, Tan K I, Yo S L, Hopes E. Molecular epidemiology of hepatitis B virus vaccine variants in Singapore. Vaccine. 1995;13:699–702. [PubMed]
29. Palmer D R, Perry K R, Mortimer P P, Parry J V, Garden P, Dhell J. An evaluation of fifteen HBsAg screening assays. London, England: Medical Devices Agency; 1995.
30. Ramesh A B, Ulrich P P, Vyas G N. Molecular characterization of a new variant of hepatitis B virus in a persistently infected homosexual man. Hepatology. 1990;11:271–276. [PubMed]
31. Ratnam S, Stead F, Head C B. False-positive results with third-generation monoclonal hepatitis B surface antigen enzyme immunoassay. J Clin Microbiol. 1989;27:2102–2104. [PMC free article] [PubMed]
32. Sanchez-Quijano A, Jauregui J I, Leal M, Pineda J A, Castilla A, Abad M A, Civeira M P, Garcia de Pesquera F, Prieto J, Lissen E. Hepatitis B virus occult infection in subjects with persistent isolated anti-HBc reactivity. J Hepatol. 1993;17:288–293. [PubMed]
33. Scheitlin T H, Grob P J. Infektionen mit den Hepatitisviren HAV, HBV und HCV sowie mit dem AIDS Virus HIV bei Drogenabhängigen der Cassenszene Zürichs: eine Prävalenzstudie. Schweiz Med Wochenschr. 1992;120:621–629. [PubMed]
34. Shih L, Sheu J, Wang J, Huang G, Yang P, Lee H, Sung J, Wang T, Chen D. Serum hepatitis B virus DNA in healthy HBsAg-negative Chinese adults evaluated by polymerase chain reaction. J Med Virol. 1990;32:257–260. [PubMed]
35. Skurrie I J, Garland S M. False positivity with third-generation (monoclonal) assay for hepatitis B surface antigen. Lancet. 1988;i:299–300. [PubMed]
36. Stirk H J, Thornton J M, Howard C R. A topological model for hepatitis B surface antigen. Intervirology. 1992;33:148–158. [PubMed]
37. Sumazaki R, Motz M, Wolf H, Heinig J, Jilg W, Deinhardt F. Detection of hepatitis B virus in serum using amplification of DNA by means of polymerase chain reaction. J Med Virol. 1989;27:304–308. [PubMed]
38. Thiers V, Nakajima E, Kremsdorf D, Mack D, Schellekens H, Driss F, Goudeau A, Wands J, Sninsky J, Tiollais P, Brechot C. Transmission of hepatitis B from hepatitis B-seronegative subjects. Lancet. 1988;ii:1273–1276. [PubMed]
39. Vyas G N, Cohen S N, Schmid R, editors. Viral hepatitis. Philadelphia, Pa: Franklin Institute Press; 1978.
40. Wallace L A, Echevarria J E, Echevarria J M, Carman W F. Molecular characterization of envelope antigenic variants of hepatitis B virus from Spain. J Infect Dis. 1994;170:1300–1303. [PubMed]
41. Wang J T, Wang T H, Sheu J C, Shih L N, Lin J T, Chen D S. Detection of hepatitis B virus DNA by polymerase chain reaction in plasma of volunteer blood donors negative for hepatitis B surface antigen. J Infect Dis. 1991;163:397–399. [PubMed]
42. Waters J, Kennedy M, Voet P, Hauser P, Petre J, Carman W, Thomas H C. Loss of the common “A” determinant of hepatitis B surface antigen by a vaccine-induced escape mutant. J Clin Investig. 1992;90:2543–2547. [PMC free article] [PubMed]
43. Weinberger K M, Kreuzpainter E A, Hottenträger B, Neifer S, Jilg W. Mutations in the s-gene of hepatitis B virus isolates from chronic carriers with anti-HBc as the only serological marker of HBV infection. In: Rizetto M, Purcell R H, Gerin J L, Verne G, editors. Viral hepatitis and liver disease. Torino, Italy: Edizioni Minerva Medica; 1997. pp. 138–143.

Articles from Journal of Clinical Microbiology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...