Laboratory Evaluation of a Fully Automated Chemiluminescence ...

4 downloads 93519 Views 85KB Size Report
Jul 18, 2003 - Corresponding author. Mailing address: Clinical Microbiology Di- ... cence system (Ortho/ECi) for qualitative detection of HBsAg, antibodies to ...
JOURNAL OF CLINICAL MICROBIOLOGY, Feb. 2004, p. 610–617 0095-1137/04/$08.00⫹0 DOI: 10.1128/JCM.42.2.610–617.2004 Copyright © 2004, American Society for Microbiology. All Rights Reserved.

Vol. 42, No. 2

Laboratory Evaluation of a Fully Automated Chemiluminescence Immunoassay for Rapid Detection of HBsAg, Antibodies to HBsAg, and Antibodies to Hepatitis C Virus Nahed Ismail,* Geoffrey E. Fish, and Michael B. Smith Department of Pathology, Clinical Microbiology Division, University of Texas Medical Branch, Galveston, Texas 77555 Received 18 July 2003/Returned for modification 31 August 2003/Accepted 4 November 2003

The performance of a fully automated, random access, enhanced chemiluminescence immunoassay (Ortho/ ECi) for the detection of antibody to hepatitis C virus (HCV) (anti-HCV), HBsAg, and antibody to HBsAg (anti-HBsAg), in human serum was compared to a Abbott second-generation enzyme immunoassay (EIA 2.0). The Ortho/ECi assays employ an immunometric technique with enhanced chemiluminescence for optimal assay performance. With regard to the study of clinical laboratory performance, six groups of sera prescreened with Abbott EIAs were assayed: anti-HCV-negative samples (n ⴝ 318), anti-HCV-positive samples (n ⴝ 177), anti-HBsAg-negative samples (n ⴝ 241), anti-HBsAg-positive samples (n ⴝ 239), HBsAg-positive samples (n ⴝ 158), and HBsAg-negative samples (n ⴝ 312). Sera with discrepant results in the two serological assays were resolved by confirmatory tests. Sera with indeterminate results by one or more confirmatory tests were evaluated by reviewing medical records. The overall concordance between the Ortho/ECi assay and the Abbott EIA were 97.78, 93.54, and 97.66% for anti-HCV antibodies, anti-HBsAg antibodies, and HBsAg, respectively. After resolving the discrepancies, the specificities of the new assay for anti-HCV and anti-HBsAg antibodies and HBsAg were 98.1, 92.8, and 100%, respectively. The sensitivities of the new assay for anti-HCV, anti-HBsAg, and HBsAg were 100, 98.8, and 97.4%, respectively. In conclusion, The Ortho/ECi assays for diagnosis of HCV and hepatitis B virus (HBV) infections are highly specific and sensitive assays. The rapid turnaround time, random access, full automation, and high throughput make it an effective assay system for clinical laboratory diagnosis of HCV and HBV infections. ation assay, introduced in 1991, incorporated recombinant antigens from nonstructural regions (NS3 and NS4) of the putative HCV genome (3, 38, 45). These antigens are shown to be detectable in the late phase of acute infection and become undetectable a few months to a few years after recovery in subjects not progressing to chronicity. The first- and the second-generation anti-HCV enzyme immunoassays, thus, had important limitations, notably, a high rate of false-positive and -negative results (7, 18). To increase both sensitivity and specificity, a greater number of HCV-encoded antigens are now included in the third-generation enzyme immunoassay (EIA), allowing an increase in the specificity (8, 19). The addition of core and NS5 region-encoded antigens on the solid phase of serological assays also resulted in earlier detection of antiHCV during acute infection, a marked increase in the sensitivity, and a dramatic reduction in the incidence of posttransfusion hepatitis in blood banks (25). Detection of antibodies to the NS5 region-encoded antigens in the immunoassay, together with a different format and antigen concentration on the solid phase contributes to increased sensitivity of the screening test (25). A strip immunoassay developed by the Chiron Corporation (Emeryville, Calif.) is being used to help differentiate true-positive from false-positive EIA results. The Food and Drug Administration approved the second-generation recombinant immunoblot assay (RIBA) in 1993 followed by approval of the third-generation RIBA in 1999. The strip immunoassays include the EIA antigens and human superoxide dismutase (hSOD). The RIBA is considered positive if there are reactions with at least two antigens with intensities

The four immunoassays described here are associated with the diagnosis of hepatitis C and B virus (HCV and HBV, respectively) infections. HCV, an enveloped positive-stranded RNA virus of the Flaviviridae family, has been demonstrated to be the etiologic agent of 90% of chronic non-A, non-B hepatitis (2, 5, 10, 15). HCV infection is often asymptomatic; however, the vast majority of HCV-infected individuals (⬎85%) develop persistent chronic infection and chronic hepatitis (9, 32, 45, 46). Diagnosis of HCV infection has serious implications, especially for the high-risk patients such as those undergoing hemodialysis (20, 27, 35). The presence of anti-HCV antibodies indicates that an individual may have been infected with HCV and/or may be capable of transmitting HCV infection while active infection is marked by the presence of HCV RNA detected by reverse transcriptase PCR (6, 17, 35, 42). Detection of viremia is often required in certain cases of acute infection and/or immunodeficiency, where individuals may fail to produce antibodies specific for HCV (32, 39). The HCV genome consists of seven functional regions: the core, the envelope, including the E1 and E2 regions, and the nonstructural region, including NS2, NS3, NS4, and NS5 (31). The first commercially available HCV test was an enzymelinked immunosorbent assay, which was introduced in 1989. This assay incorporated recombinant C100-3 antigen derived from the nonstructural region of the virus. The second-gener* Corresponding author. Mailing address: Clinical Microbiology Division, Department of Pathology, University of Texas Medical Branch, 301 University Blvd., Galveston, TX 77555. Phone: (409) 772-3111. Fax: (409) 772-3568. E-mail: [email protected]. 610

VOL. 42, 2004

SEROLOGICAL DIAGNOSIS OF HCV AND HBV INFECTIONS

greater than or equal to that for the weak immunoglobulin G (IgG) control and no reactivity with hSOD. Indeterminate RIBA are those in which there are reactions with only one antigen or with the hSOD plus one or more HCV antigens. Replacement of the C100 and C22 recombinant proteins with synthetic peptides in the version 3.0 RIBA has significantly reduced the number of indeterminate RIBA results (16, 36, 40, 41). Although diagnostic tests that employ EIA and RIBA are used for serological screening and confirmation, respectively, nucleic acid testing and HCV antigen detection methods are required to demonstrate active infection (11, 17). As HCV antigen tests are still under development, qualitative nucleic acid testing is currently the method of choice to confirm active infection (11, 17, 42). In addition, the measurement of HCV RNA levels in plasma by a branched DNA assay has become an important tool for monitoring individuals during antiviral therapy (9). Viral hepatitis due to HBV is a major public health problem, with an estimated 300 million persistent carriers of HBV worldwide (54, 55). HBV infection produces an array of unique antigens and antibody responses that follow distinct serological patterns (55, 56). The envelope protein of HBV, 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 a glycosylated lipoprotein that is usually shed in large amounts in the serum of infected individuals. HBsAg is cleared within a few months in self-limited illness. If HBsAg persists for more than 6 months, spontaneous clearance is unlikely and the infected individual is considered a chronic HBV carrier (56). The determinant of HBsAg is the major neutralizing epitope. Serological detection of HBsAg involves the use of either monoclonal or polyclonal anti-HBs bound to a solid-phase or second labeled anti-HBs to detect the captured antigen. Despite the high performance of third-generation enzyme-linked immunosorbent assay in the detection of HBsAg, a high incidence of false-negative results has been reported (22–24, 43). Antibodies to HBsAg indicate recovery from infection and serve as an indicator of protection from infection (53). Individuals who have resolved their HBV infection usually demonstrate both anti-HBsAg and antibody to HBcAg (anti-HBc) in their serum. Anti-HBsAg testing is useful for identifying HBV-susceptible individuals in pre- and postvaccination screening programs, where absence of these antibodies is indicative of susceptibility to HBV infection (26, 28, 53). This could be an indication for vaccination against HBV. Serological immunoassays for qualitative determination of antibodies to HBsAg in serum involves the reaction of anti-HBsAg in the sample with HBsAg coated onto the wells. While the diagnosis of HBV infection is based on serologic detection of the HBsAg and anti-HBsAg antibodies as described above, HBV replication status in these patients is usually monitored through assessment of HBV DNA levels. Monitoring of serum HBV DNA levels is a consistent method for the assessment of potency of antiviral therapy (1, 37). The aim of this study was to assess the performance of a new, fully automated rapid immunodiagnostic chemiluminescence system (Ortho/ECi) for qualitative detection of HBsAg, antibodies to HBsAg, and antibodies to HCV in terms of specificity, sensitivity, and suitability for use in the diagnosis of

611

viral hepatitis compared to commercially available and commonly used screening Abbott EIA, version 2.0. Supplemental assays based on either neutralization or RIBA that possess high specificity were applied to resolve discrepant results between the two immunoassays. MATERIALS AND METHODS Serological assays. The HBsAg, HBsAg confirmatory tests, anti-HBsAg, and anti-HCV assays were performed by using the Vitros reagent pack and immunodiagnostic calibrator on the Vitros ECi immunodiagnostic system. Ortho/ECi anti-HCV assay. The Ortho/ECi anti-HCV assay is a two-step sandwich chemiluminescence assay for the qualitative detection of human antibodies in serum or plasma to various proteins of HCV, with a total incubation time of 45 min. Recombinant antigens from the HCV core, NS3, NS4, and NS5 regions of the HCV genome are immobilized on the solid phase (coated wells). After incubation of controls and specimens and subsequent washing, horseradish peroxidase (HRP)-labeled mouse anti-human IgG conjugate is added. Unbound materials are removed by washing, and a reagent containing luminogenic substrates and an electron transfer agent are added to the wells. The HRP in the bound conjugates catalyzes the oxidation of the luminol derivative, producing light. The electron transfer agent increases the level and duration of the light produced. The light signals are read by the Ortho/ECi system. Results are calculated as a normalized signal relative to the cutoff value (signal/cutoff [S/C] ratio). During the calibration process, a lot-specific parameter, encoded on the lot calibration card, is used to determine a valid stored cutoff value for the Ortho/ECi system. Therefore, the results are calculated by dividing the signal for test sample by the cutoff value. Patient samples with single S/C ratio of ⱖ1.00 are considered to test positive. If the S/C ratio is ⬍0.90, the sample is considered negative. Samples with an S/C ratio of ⱖ0.90 and ⬍1.00 were retested in duplicate based on the manufacturer’s recommendation. The Ortho/ECi system uses a small sample volume (20 ␮l) for each determination compared to 100 ␮l of sample used for the anti-HCV Abbott EIA. Discrepant results between Ortho/ ECi and Abbott EIA were resolved by Chiron HCV strip RIBA, version 3.0. Ortho/ECi HBsAg. Detection of HBsAg by Ortho/ECi involves the simultaneous reaction of HBsAg in the sample with mouse monoclonal anti-HBs antibodies coated onto the wells and an HRP-labeled mouse monoclonal anti-HBs antibody in the conjugate. The total incubation time is 29 min. Results are calculated similar to that described above. Samples with an S/C ratio of ⱖ5.00 are considered positive. If the S/C ratio is ⬍0.90, the sample is considered negative. Samples with an S/C ratio between ⱖ0.90 and ⬍5.00 are retested in duplicate. The Ortho/ECi HBsAg assay uses 80 ␮l of sample for each determination compared to 200 ␮l of sample used for the Abbott EIA for HBsAg. Indeterminate or discrepant results were resolved by the Ortho HBsAg confirmatory kit. As part of the evaluation process, the performance of the Vitros HBsAg confirmatory neutralization kit was also evaluated and the results were compared with those obtained by the same test performed by another laboratory. The test utilizes the principal of specific antibody neutralization to confirm the presence of HBsAg. The sample is tested twice: one aliquot is incubated with a neutralizing reagent containing high-titer human anti-HBs (the confirmatory antibody), and the second aliquot is incubated with a nonneutralizing control reagent (the sample diluent). The confirmatory antibody binds to HBsAg in the sample, inhibiting its reaction in the Vitros HBsAg assay. This leads to a reduced S/C ratio compared to that for the nonneutralized control samples. Ortho/ECi anti-HBsAg antibody assay. Detection of anti-HBsAg antibodies by Ortho/ECi involves the simultaneous reaction of antibodies in serum with purified native HBsAg coated onto the wells and an HRP-labeled mouse monoclonal anti-HBs antibody in the conjugate. The results are calculated in a way similar to that for the anti-HCV assay. The incubation time is 45 min. Samples with an S/C ratio of ⱖ1.20 are considered positive. This cutoff value was found to be consistent with a level of anti-HBs antibodies of ⱖ10 mIU/ml determined by quantitative anti-HBsAg assay. The accepted criterion for immunity to HBV is ⱖ10 mIU of anti-HBs/ml, with milli-international units defined by the World Health Organization reference preparation (14). If the S/C ratio is ⬍0.50, the sample is considered negative. Samples with an S/C ratio of ⱖ0.50 and ⬍1.20 are reported as indeterminate and retested in duplicate. The Ortho/ECi anti-HBsAg assay uses 80 ␮l of sample for each determination compared to 200 ␮l of sample used for Abbott EIA anti-HBsAg. Indeterminate or discrepant results were resolved by quantitative anti-HBs assay to determine the immunological status of the patient. Additionally, further chart reviews, described below, were undertaken to determine the clinical and immu-

612

ISMAIL ET AL.

J. CLIN. MICROBIOL.

TABLE 1. Intra-assay precision of testing with 10 replicates of controls and positive and negative human sera Sample

Meanb

SDc

CV (%)a

Anti-HCV Negative control Positive control Negative serum Positive serum

0.56 6.38 0.49 5.43

0.04 0.05 0.03 0.04

6.4 0.7 1.1 0.8

HBsAg Negative control Positive control Negative serum Positive serum

0.62 2.81 0.28 3.70

0.05 0.07 0.03 0.03

5.3 2.5 6.9 1.7

Anti-HBsAg Negative control Positive control Negative serum Positive serum

0.50 2.28 0.51 2.23

0.04 0.03 0.03 0.02

8.0 1.3 5.3 1.0

a

CV, coefficient of variation. The results expressed as the S/C ratio were assessed in terms of central trend index (mean). c Standard deviation of S/C ratio of 10 triplicate wells. b

nological status of the individual as well as the presence or absence of risk factors and/or interfering substances. Design of evaluation and procedures. Intra-assay precision was determined as follows: negative and positive controls of the test kit as well as known positive or negative patient samples were measured repeatedly within one run. Evaluation of laboratory performance. Six different sample panels were selected: sera negative (n ⫽ 318) or positive (n ⫽ 177) for anti-HCV prescreened with Abbott EIA, version 2.0; sera negative (n ⫽ 214) or positive (n ⫽ 239) for anti-HBsAg prescreened with Abbott EIA, version 2.0; and sera negative (n ⫽ 312) or positive (n ⫽ 158) for HBsAg prescreened with Abbott EIA. Serum specimens were stored both refrigerated and frozen prior to testing. Some of the serum samples were preselected on the basis of serological profile as determined by Abbott EIA, version 2.0. These samples were frozen and tested subsequently with the Ortho/ECi assay. Other samples were refrigerated and tested simultaneously within 72 h of specimen collection with Ortho/ECi and Abbott EIA. Comparative and supplemental assays. The Ortho/ECi chemiluminescence assay (Ortho Clinical Diagnostic) was compared to the Abbott EIA, version 2.0 (Abbott Laboratories). The Ortho/ECi test results obtained in each assay format were independently evaluated against the Abbott EIA. Confirmatory tests for each assay and marker were performed on all specimens, with discrepant results between Ortho/ECi and Abbott EIA to determine false-positive and -negative test results. For example, a third-generation RIBA (Chiron) was performed on discrepant anti-HCV samples according to the manufacture’s instructions. Discrepant HBsAg samples were additionally tested with the Ortho/ECi HBsAg confirmatory kit. Discrepant anti-HBsAg samples were further evaluated by Abbott EIA AUSAB anti-HBsAg quantitation panel according to the manufacturer’s instructions. Discrepant samples with indeterminate results by confirmatory and supplemental tests were further evaluated by chart review. Chart review and clinical evaluation. Patient samples used in this study were sent to our laboratory under suspicion for HCV or HBV infection due to either elevated liver enzyme values (alanine aminotransferase, ⬎45 U/liter), clinical signs of hepatitis (jaundice and upper abdominal pain), or risk factors for parenterally transmitted diseases, such as chronic hemodialysis, blood transfusion, or intravenous drug use. Samples that were tested by Ortho/ECi, Abbott EIA, and a supplemental assay and had indeterminate results were investigated further through chart review. This chart review involved examination of clinical signs of hepatitis described above, elevated liver enzymes, and the presence or absence of past and present clinical and other laboratory evidence of infection with hepatitis A virus, HBV, and HCV. For example, we looked for the presence of anti-hepatitis A virus IgM antibodies, anti-HBcAg IgM, and IgG. In addition, these samples were investigated for the presence of clinical conditions and substances that could interfere with the performance and specificity of Ortho or Abbott assays, such as acute and chronic viral and bacterial infections, autoimmune diseases, dialysis, pregnancy, and rheumatoid factor. Evaluation of results. Evaluation of sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the Ortho/ECi assay and Abbott EIA were based on agreement between Ortho/ECi and Abbott EIA as

TABLE 2. Results of clinical evaluation of Ortho/ECi on a panel of predefined anti-HCV, HBsAg, and anti-HBsAg positive and negative sera, comparing performance to Abbott EIA No. of specimens with result

Sample source

Positive

Negative

Agreement (%)

177 158 239

318 312 214

99 98 94

Anti-HCV HBsAg Anti-HBsAg

Sensitivity (%)a

Specificity (%)a

PPV (%)a

NPV (%)a

98.9 93.7 97.5

97.2 99.7 89.6

95.1 99.3 90.3

99.4 96.9 97.3

a Sensitivity, Specificity, PPV, and NPV of Ortho assays were calculated on the basis of comparison with the Abbott EIA only.

well as confirmation of discrepant results by supplemental tests or clinical chart review, when indicated. A sample was considered a true positive for anti-HCV, HBsAg, or anti-HBsAg if it was reactive on both Ortho and Abbott assays or if it was only reactive on one assay but confirmed as a true positive by a supplemental assay. Serum samples with insufficient volume due to the consumption of a large volume during the comparative analysis and repeating tests were further investigated based on clinical data and chart review described above. Conversely, a test result was interpreted as a true negative if it was negative by the Ortho/ECi assay and the Abbott EIA or if it was only negative on one assay but confirmed as a true negative by supplemental tests or chart review and other serological markers.

RESULTS Technical performance and precision evaluation. The intraassay precision was validated by testing 10 replicates of positive and negative controls and one positive and one negative sample of HBsAg, anti-HBsAg, and anti-HCV (Table 1). The interassay reproducibility of the Ortho EIA was not evaluated in this study, since this was evaluated as part of the manufacturer’s validation of the assay. Laboratory evaluation on routine samples. The clinical performance of the Ortho/ECi anti-HCV, HBsAg, and antiHBsAg test versus that of the Abbott EIA is shown in Table 2. Overall, approximately 500 specimens were tested for a panel of predefined positive and negative sera. Detection of anti-HCV antibodies. With regard to the antiHCV antibody test, the correlation of all positive and negative test results between Abbott EIA and Ortho/ECi was 98.9% (484 of 495) (Table 2). Of 177 predefined positive serums, two were negative initially by Ortho and remained negative on repetition. The RIBA HCV assay confirmed the negative Ortho/ECi anti-HCV result for two samples. Of 318 predefined negative serums, seven were positive and two were indeterminate initially and these remained positive and indeterminate, TABLE 3. Results of confirmatory anti-HCV RIBA on discrepant sera Sample no.

Abbott result

Ortho result

RIBA result

Ortho interpretation

1 2 3 4 5 6 7 8 9 10 11

Negative Negative Negative Negative Negative Negative Negative Negative Negative Positive Positive

Indeterminate Positive Positive Positive positive positive Indeterminate Positive Positive Negative Negative

Negative Indeterminate Indeterminate Negative Negative Indeterminate Negative Negative Negative Negative Negative

Ortho indeterminate RIBA indeterminate RIBA indeterminate False positive False positive RIBA indeterminate Ortho indeterminate False positive False positive True negative True negative

VOL. 42, 2004

SEROLOGICAL DIAGNOSIS OF HCV AND HBV INFECTIONS

613

TABLE 4. Results of screening and confirmatory anti-HCV tests on sera with discrepant results between Ortho/ECi and Abbott EIAa HCV sample no.

Abbott result

Ortho result

1 2 3 4 5 6 7 8 9 10 11

Negative Negative Negative Negative Negative Negative Negative Negative Negative Positive Positive

Indeterminate Positive Positive Positive Positive Positive Indeterminate Positive Positive Negative Negative

RIBA 3.0 result for:b C100-3

C33

C22-3

NS5

RIBA-3 result interpretation

Ortho interpretation

⫾ ⫾ ⫾ ⫺ ⫺ ⫾ ⫾ ⫺ ⫺ ⫺ ⫾

⫾ 2⫹ 1⫹ ⫾ ⫺ 2⫹ ⫺ ⫾ ⫾ ⫾ ⫺

⫾ ⫺ ⫺ ⫾ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺

⫾ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺

Negative Indeterminate Indeterminate Negative Negative Indeterminate Negative Negative Negative Negative Negative

Ortho indeterminate RIBA indeterminate RIBA indeterminate False positive False positive RIBA indeterminate Ortho indeterminate False positive False positive True negative True negative

a Due to insufficient sample volume, samples with indeterminate RIBA results were excluded from the calculation of the performance of the Ortho/ECi and Abbott EIA. b ⫾, indeterminate; 2⫹, strong reactive; 1⫹, weakly reactive; ⫺, nonreactive.

respectively, on repetition. The two Ortho/ECi anti-HCV indeterminate and Abbott EIA anti-HCV negative samples were confirmed as anti-HCV negative with the RIBA confirmatory test (indeterminate Ortho results are shown in Tables 3 and 4). Among the 7 Ortho/ECi anti-HCV-positive and Abbott EIA anti-HCV-negative samples, RIBA was negative for four samples and indeterminate for three samples (indeterminate RIBA results are shown in Tables 3 and 4). Thus, discrepant Ortho/ ECi results only agreed with 18% (2 of 11) of interpretable RIBA results. On the other hand, Abbott EIA agreed with 55% (6 of 11) of interpretable RIBA results. Interestingly, all seven samples that were positive by Ortho but did not confirm by RIBA had an S/C ratio of ⬎1 but ⬍8. Due to insufficient sample volume, samples with indeterminate results by RIBA were not further tested by other supplemental assays such as nucleic acid testing. However, those samples had significantly high alanine aminotransferase levels in the absence of other markers for HBV (data not shown), suggesting a possible early stage of HCV infection. Based on the current algorithm used to define true positive and negative serostatus for HCV, the sensitivity and specificity of the Ortho/ECi assay for anti-HCV based on agreement with RIBA were 98.9 and 97.2%, respectively (Table 2). The sensitivity of the Ortho/ECi anti-HCV assay assessed on the basis of RIBA, excluding samples with indeterminate RIBA results, was 100%, compared to 98.9% for the Abbott EIA, and the specificity of the Ortho/ECi assay was 98.1%, compared to 100% for the Abbott EIA (Tables 5 and 6). Detection of HBsAg. With regard to the HBsAg test, the correlation of test results between the Abbott EIA and Ortho/ ECi was 98% (459 of 470) (Table 2). Of 157 predetermined positive HBsAg tests by Abbott EIA, 9 samples were initially negative by Ortho/ECi and remained negative on repetition. Of these nine Ortho HBsAg-negative and Abbott HBsAg-positive samples, four samples were confirmed positive by Ortho HBsAg neutralization assay (i.e., Ortho false negative). Two samples were confirmed negative by HBsAg confirmation assay (true negative). Since we did not have a sufficient amount to test the remaining three samples by HBsAg confirmatory test, we examined the serological profiles of these samples with Ortho/ECi negative HBsAg results. The HBV serological profiles of these last three 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 Ortho/ECi HBsAg (Table 7). Similarly, of 312 predefined negative sera, 1 was positive by Ortho/ECi initially, remained positive on repetition, and was confirmed as a true positive by the HBsAg neutralization test. The serological profile of this patient (anti-HBc, anti-HBc IgM, and anti-HBsAg negative) is also compatible with that of an HBV carrier (Table 7). Based on the algorithm used to define true positive and negative serostatus for HBsAg, the sensitivity and specificity of the Ortho/ECi assay for HBsAg based on comparison with Abbott EIA were 93.7 and 99.7%, respectively (Table 2). The sensitivity of the Ortho/ECi HBsAg assay assessed on the basis of HBsAg confirmation test results and serological profile was 97.4%, compared to 96.8% for the Abbott EIA, and the specificity of the Ortho/ECi assay was 100%, compared to 99.7% for the Abbott EIA (Tables 5 and 6). Detection of anti-HBsAg antibodies. With regard to the anti-HBsAg test, the correlation of test results between the Abbott EIA and Ortho/ECi was 94% (449 of 480) (Table 2). Discrepant results were confirmed by quantitative measurement of anti-HBsAg antibodies, with a cutoff value of more than 10 IU/ml indicating immunity and corresponding to a positive qualitative HBsAg. Of 239 predetermined positive

TABLE 5. Results of clinical evaluation of diagnostic Ortho/ECi assay and Abbott EIA following resolution of discrepant samplesa Ortho/ECi result

Abbott/EIA result

Sample type

Positive

Negative

Positive

Negative

Anti-HCVb Positive Negative

175 6

0 311

175 0

2 315

HBsAgb Positive Negative

149 0

4 316

152 1

5 311

Anti-HBsAg Positive Negative

241 17

3 219

236 8

3 233

a Concordant positive or negative samples between the two assays and discrepant samples that were confirmed as true positive or negative based on supplemental tests and chart review were employed to determine false-positive or -negative Ortho/ECi and Abbott EIA results. b Concordant samples plus samples confirmed as positive or negative by supplemental tests and/or chart review.

614

ISMAIL ET AL.

J. CLIN. MICROBIOL. TABLE 8. Results of medical chart review and quantitative anti-HBsAg supplemental assays on initially discordant Ortho and Abbott anti-HBsAg samples (n ⫽ 31)

TABLE 6. Sensitivities, specificities, and predictive values of Ortho and Abbott assays on samples confirmed positive or negative by confirmatory tests or chart review Ortho hepatitis serology result (%)

Sample type

Anti-HCV HBsAg Anti-HBsAg

Abbott hepatitis serology result (%)

Test

Sensitivity

Specificity

PPV

NPV

Sensitivity

Specificity

PPV

NPV

100 97.4 98.8

98.1 100 92.8

96.7 100 93.5

100 98.8 98.7

98.9 96.8 98.7

100 99.7 96.7

100 99.3 96.7

99.4 98.4 98.7

anti-HBsAg tests by Abbott EIA, 6 samples were initially negative by Ortho/ECi and remained negative on repetition. Among the six discrepant negative samples detected by Ortho/ECi, three of them were negative by quantitative anti-HBsAg (i.e., true negative). Similarly, of 241 predefined negative serums, 28 samples were positive initially. Of these 28 samples, 3 samples were negative, 8 samples were positive, and 17 samples were indeterminate on repetition. Among the 8 discordant positive samples by Ortho/ECi, 3 samples were positive (38%) by quantitative anti-HBsAg and 5 samples were negative (62%). Of 17 indeterminate samples, 12 (71%) were negative and 5 (29%) were positive by quantitative anti-HBsAg (Table 8). Based on the algorithm used to define true positive and negative serostatus for anti-HBsAg, the sensitivity and specificity of the Ortho/ECi assay for anti-HBsAg were 97.5 and 89.6%, respectively (Table 2). The sensitivity of the Ortho/ ECi anti-HBsAg assay assessed on the basis of anti-HBsAg quantitative assay results and serological profile was 98.8%, similar to the sensitivity of Abbott EIA anti-HBsAg, and the specificity of the Ortho/ECi assay was 92.8%, compared to 96.7% for the Abbott EIA (Tables 5 and 6). Effect of potentially cross-reacting or interfering subgroups on the performance of Ortho immunoassays. The specificity of Ortho/ECi and Abbott EIA HBsAg, anti-HBsAg, and antiHCV antibody immunoassays were evaluated by testing 125 samples from potentially cross-reacting subgroups. The Ortho/ ECi HBsAg assay was not affected by either cross-reacting or interfering substance in 25 samples (data not shown). Ortho/ ECi false-positive anti-HCV samples were associated with the presence of red blood cells, hyperlipidemia, and cross-reaction with HBsAg (data not shown). Thirteen samples from specific groups that contained interfering substances had indeterminate Ortho anti-HBsAg assay, negative Abbott EIA, and negative confirmatory test results. Three samples with Ortho/ECi false-negative anti-HBsAg results were derived from a dialysis

TABLE 7. Results of medical chart review and confirmatory HBsAg assays on initially discordant Ortho and Abbott HBsAg samples (n ⫽ 10) Test

Ortho

Initial result

Positive Negative Abbott Positive Negative

No. of samples

1 9 9 1

No. of samples with result Confirmed positive

Confirmed negative

Clinically negative

Clinically positive

1 4 4 1

0 2 2 0

0 3 3 0

1 0 0 1

Ortho Abbott

Initial result

Positive Indeterminate Negative Positive Indeterminate Negative

No. of samples

No. of samples with confirmed result Positive

Negative

3 5 3 3 0 8

5 12 3 3 0 17

8 17 6 6 0 25

patient and from sera with a high bilirubin or lipid level (Table 9). Analysis of labor and cost requirements of the two immunodiagnostic assays for serological diagnosis of hepatitis C and B viruses. As part of our evaluation of the new Ortho/ECi method for the diagnosis of HCV and HBV, we further performed a time and cost analysis to compare the cost of reagents, sample volume, labor involved in testing and instrument maintenance, time to result, and hands-on time of the Ortho/ECi anti-HCV, HBsAg, and anti-HBsAg assays versus the Abbott EIA, version 2.0, equivalent hepatitis assays, our present standard protocol. The Ortho assays are very easy to perform, have random access, require a small sample volume (Table 10), and are fully automated compared to the partially automated Abbott EIA. The time to results and hands-on time per sample were measured once in each individual assay setup (Table 10). Both are substantially better for Ortho/ECi analysis, even when a large number of specimens is tested. Consequently, labor costs are lower for the Ortho/ECi assay. In addition, the availability of random access and the short assay time of the Ortho/ECi tests allow the repetition of positive samples and performance of supplemental tests, if needed, in the same day. The general TABLE 9. Potentially cross-reactive or interfering subgroups on the performance of Ortho/ECi anti-HBsAg assays No. of samples with result

Sample type, source, or condition

No. of samples tested

False positive or indeterminate

False negative

Autoimmune disease Rheumatoid factor positive Anti-HCV positive Rubella virus IgG positive EBV VCAa IgG positive HIVb Dialysis patients CMVc IgG Pregnancy Transplant Renal dialysis Lipemic Bilirubin Blood

4 6 20 7 3 25 4 2 10 3 4 8 8 6

1 0 2 2 0 0 0 0 3 1 1 0 1 2

0 0 0 0 0 0 1 0 0 0 0 1 1 0

110

13

3

Total a b c

EBV VCA, Epstein-Barr virus virus capsid antigen. HIV, human immunodeficiency virus. CMV, cytomegalovirus.

VOL. 42, 2004

SEROLOGICAL DIAGNOSIS OF HCV AND HBV INFECTIONS

TABLE 10. Detailed analysis of hands-on and hands-off time for anti-HCV, HBsAg, and anti-HBsAg diagnosis by the Abbott EIA, version 2.0, and the Ortho/ECi assaya Assay

Hands-on time

Hands-off time

Time to first result

Ortho Anti-HCV HBsAg Anti-HBsAg

2 h, 30 min 2 h, 30 min 2h

45 min 29 min 45 min

55 min 37 min 55 min

Abbott Anti-HCV HBsAg Anti-HBsAg

3 h, 30 min 3 h, 30 min 3 h, 30 min

2 h, 30 min 3 h, 30 min 4 h, 30 min

6h 6 h, 30 min 7 h, 30 min

Sample vol (␮l)

20 80 80 10 200 200

a Hands-on time refers to the time that the technologist was actively manipulating the specimen or reagents; hands-off time refers to incubation time periods when the technologist was not actively involved in the assay procedures.

cost of disposables is also lower for the Ortho/ECi assay. Comparing list prices for both assays, the reagent cost of the Abbott EIA is significantly lower than that of the Ortho/ECi assay, although the manufacturers’ volume discounts certainly influence final pricing. DISCUSSION The newly developed Ortho/ECi anti-HCV, HBsAg, and anti-HBsAg tests showed good reproducibility in the precision assay as judged by the intra-assay coefficients of variation. With regard to the evaluation of laboratory performance, since no recognized “gold standard” exists for establishing the presence or absence of anti-HCV, HBsAg, or anti-HBsAg, the sensitivity and specificity of Ortho/ECi and Abbott EIA were calculated in relation to the sum of concordant results between the two assays and agreement with the confirmatory test and clinical data. Our data show that the Ortho/ECi immunoassays are highly specific and sensitive automated tests for detection of HBsAg and antibodies to HBsAg and HCV. The overall concordances of the Ortho/ECi assay with Abbott EIA for detection of anti-HCV antibodies, HBsAg, and anti-HBs antibodies were 98, 98, and 94%, respectively. The Ortho/ECi test gives a number of positive results with negative or indeterminate confirmatory Chiron strip RIBA HCV, version 3.0, and Abbott EIA, respectively. The inability of RIBA to resolve the discrepancy in one-third of discrepant HCV samples greatly limits its usefulness for confirmation of EIA results, which is consistent with previous reports (46). Confirmation by RIBA has not been very useful for resolving weakly positive samples, and a molecular HCV RNA detection is then recommended (45, 46). In our laboratory, we perform approximately 85,000 antiHCV tests per year. Of all anti-HCV reactive samples, approximately 10 to 15% are weakly positive (S/C ratio of ⬍8, as determined by Ortho/ECi anti-HCV test, or ⬍3.8, as determined by Abbott EIA), and the other 85 to 90% are strongly positive, with an S/C ratio of ⱖ8, as determined by Ortho/ECi, or ⱖ3.8, as determined by Abbott EIA (6). Possible explanations for low positive anti-HCV results include false-positive serological tests, early testing during acute infection (i.e., before complete seroconversion), decreased immune responses due to immunosuppression, and waning antibody due to re-

615

solved infection (44, 47, 48, 49). Under these conditions, nucleic acid testing can determine if HCV RNA is present in these samples. Similarly, in patients with strongly positive antiHCV serology, nucleic acid testing is required to confirm active HCV infection (50). Confirmation of active infection is essential when assessing individuals who undergo treatment (30, 42). Therefore, if the nucleic acid testing result is negative for persons with a positive screening test result, the HCV antibody or infection status cannot be determined. Among persons with these results, additional testing with RIBA is necessary to verify the anti-HCV result and determine the need for counseling and medical evaluation (6). If the anti-HCV screening test results are judged falsely positive (i.e., RIBA negative), no further evaluation of the person is needed, whereas if the anti-HCV screening test results are verified as positive by RIBA, the person should undergo medical evaluation, including serial determinations of HCV RNA and serum transaminase activity (6). Nevertheless, certain situations exist in which the HCV RNA result can be negative for persons with active HCV infection. For example, HCV RNA is not detectable in certain persons during the acute phase of hepatitis C infection (9). In addition, intermittent positive HCV RNA has been detected among individuals with chronic HCV infection (9). Therefore, in the absence of additional clinical information, the significance of a single negative HCV RNA result is unknown, and the need for further medical evaluation is determined by verifying anti-HCV status. A negative HCV RNA result also can indicate resolved infection (44, 47). To determine if HCV infection has resolved, a negative HCV RNA result should be demonstrated on multiple occasions. In addition, several studies on hepatitis C patients treated with interferon showed that HCV RNA can be undetectable even if the virus is still present, as demonstrated by the recurrence of HCV RNA after the end of the therapeutic course. For these reasons, HCV RNA is able to confirm whether an infection is established but its undetectability cannot preclude the true presence of anti-HCV (25, 31, 32, 57). For samples with indeterminate anti-HCV RIBA results in this study, nucleic acid testing was not performed due to insufficient specimen volume. Since the true status of these specimens could not be determined, they (n ⫽ 3) were excluded from our analysis of specificity. The results of the present study, including a group of individuals with high anti-HCV prevalence and another group with low anti-HCV prevalence (pregnant women and outpatient clinic patients), show that the Ortho/ECi assay combines high sensitivity and specificity. On the population tested, the Ortho/ECi system showed a higher sensitivity (100%) than the Abbott EIA (98.9%). However, the specificity of the Ortho/ECi anti-HCV assay was lower than that of the Abbott EIA (98.1 versus 100%). With regard to detection of HBsAg, the results of the present study show that the Ortho/ECi assay has 100% specificity and 97.4% sensitivity for detection of HBsAg, which is higher than that of Abbott EIA HBsAg assay (99.7% specificity and 96.8% sensitivity). Nevertheless, both assays have a high rate of false-negative HBsAg results (2.6% for Ortho/ECi and 3.2% for Abbott EIA). There are three possible explanations for the high false-negative results. First, in low-level chronic HBV carriers, the HBsAg level may be below the detection limit.

616

ISMAIL ET AL.

Antibodies against HBcAg (anti-HBc) could be the only serological marker of infection in the low-level chronic carriers (25, 27). Second, very low levels of HBsAg can be observed in fulminant hepatitis and in neonatal infections (29). Third, virus variants can yield sequences that are not recognized by the antibodies employed in the assay. In different locations, vaccine escape mutants are emerging under the selective pressure of active immunization (13, 24, 28). Vaccine escape mutants within the determinant of the S gene are not recognized as effectively by conventional diagnostic tests as the wild-type determinant (12, 52). Similarly, mutation in the a determinant may also affect the performance of the HBsAg immunoassay (51). Samples that had discordant results upon repetition were resolved by confirmatory HBsAg neutralization assay. For those samples that were not resolved by confirmatory assay due to insufficient sample volume to perform the test, it is important to consider additional HBV markers. HBsAg in combination with anti-HBc determination shows a very high PPV and NPV 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 (33). Alternatively, isolated HBsAg positive results could be encountered due to absence of anti-HBc antibody response as a consequence of immunodeficiency or an immunocompromised patient (34). With regard to detection of anti-HBsAg antibodies, a higher percentage of false-positive results (4.7%) was obtained with the Ortho/ECi anti-HBsAg test compared to 3.3% false positives with the Abbott EIA. One possible reason for falsepositive anti-HBsAg results could be due to presence of potentially cross-reactive substance in the patient samples, as indicated in Table 10. The majority of discrepant results that accounted for high false-positive results in our data were indeterminate by the Ortho/ECi test (Table 8). The higher frequency of indeterminate results by the Ortho/ECi assay than by the Abbott EIA could be due to the fact that the S/C ratio of positive Ortho test was established at a relatively high cutoff that corresponds to immunity to HBV infection (⬎10 mIU/ ml). According to the manufacturer’s instructions, the Abbott EIA reports the specimen to be anti-HBsAg positive if only two of three replicates are positive, whereas the Ortho/ECi assay considers the specimen to be positive only if all three replicates are ⱖ1.20 (i.e., reactive). Therefore, indeterminate results will be reported if only one of three replicates is between 0.50 (i.e., nonreactive) and 1.20 (reactive). Nevertheless, the infectivity of these individuals is usually confirmed by detection of HBV DNA by hybridization (21) or PCR (1, 21). However, due to insufficient samples in our study, we could not further test these samples. In addition to precision and reliability, technical complexity and result turnaround time have important bearings on the usefulness of an assay in the clinical management of patients. In this regard, the Ortho/ECi analyzer, which permits a highspeed throughput, random access, primary tube sampling, and full automation, offers the great advantage of being technically simpler and faster in yielding results than the Abbott EIA, particularly for high-volume laboratories. In conclusion, Ortho/ECi (chemiluminescence) assays for serological diagnosis of HCV and HBV infection demonstrated excellent concordance with Abbott EIA and other

J. CLIN. MICROBIOL.

markers of liver disease. Our findings demonstrate that the Ortho anti-HCV and HBsAg assays show a better sensitivity than the corresponding Abbott EIA (100 versus 98.9% for the anti-HCV assay, respectively, and 97.4 versus 96.8% for the HBsAg assay, respectively). The increased sensitivity of the antiHCV assay performed in conjunction with the molecular detection of HCV RNA may be increasingly important in HCV diagnostic testing as screening and new therapeutic strategies evolve. In contrast, the anti-HCV Abbott EIA show a better specificity than the Ortho/ECi assay (100 versus 98.1%, respectively). Therefore, confirmation of positive HCV ECi results, particularly those of weakly reactive samples with low S/C ratios, by supplemental tests is recommended. Finally, because larger numbers of samples can be processed in a shorter time than for the Abbott EIA, the Ortho/ECi assay may be the assay of choice in a laboratory which requires short turnaround times. In addition, the Ortho/ECi assays are an excellent choice for laboratories with large numbers of qualitative HCV and HBV immunodiagnostic tests. REFERENCES 1. Abe, A., K. Inoue, T. Tanaka, J. Kato, N. Kajiyama, R. Kawaguchi, S. Tanaka, M. Yoshiba, and M. Kohara. 1999. Quantitation of hepatitis B virus genomic DNA by real-time detection PCR. J. Clin. Microbiol. 37:2899–2903. 2. Alter, H. 1999. Discovery of non-A, non-B hepatitis and identification of its etiology. Am. J. Med. 107(Suppl. 6B):16S–20S. 3. Alter, H. J. 1992. New kit on the block: evaluation of second-generation assays for detection of antibody to the hepatitis C virus. Hepatology 15:350– 353. 4. Alter, H. J., and L. B. Seeff. 2000. Recovery, persistence, and sequelae in hepatitis C virus infection: a perspective on long-term outcome. Semin. Liver Dis. 20:17–35. 5. Alter, M. J., H. S. Margolis, and K. Krawczynski. 1992. Natural history of community-acquired hepatitis C in the United States. N. Engl. J. Med. 327: 899–905. 6. Alter, M. J., W. L. Kuhnert, and L. Finelli. 2003. Guidelines for laboratory testing and result reporting of antibody to hepatitis C virus. Morb. Mortal. Wkly. Rep. 52(RR-3):1–16. 7. Aoyagi, K., K. Iida, C. Ohue, Y. Matsunaga, E. Tanaka, K. Kiyosawa, and S. Yagi. 2001. Performance of a conventional enzyme immunoassay for hepatitis C virus core antigen in the early phases of hepatitis C infection. Clin. Lab. 47:119–127. 8. Barrera, J. M., B. Francis, G. Ercilla, M. Nelles, D. Achord, J. Darner, and S. R. Lee. 1995. Improved detection of anti-HCV in post-transfusion hepatitis by a third-generation ELISA. Vox Sang. 68:15–18. 9. Beld, M., R. Sentjens, S. Rebers, C. Weegink, J. Weel, C. Sol, R. Boom. 2002. Performance of the new Bayer VERSANT HCV RNA 3.0 assay for quantitation of hepatitis C virus RNA in plasma and serum: conversion to international units and comparison with the Roche COBAS Amplicor HCV Monitor, version 2.0, assay. J. Clin. Microbiol. 40:788–793. 10. Bendinelli, M., M. Pistello, F. Maggi, and M. Vatteroni. 2000. Blood-borne hepatitis viruses: hepatitis B, C, D, and G viruses and TT virus, p. 306–337. In S. Specter, R. L. Hodinka, and S. A. Young (ed.), Clinical virology manual, 3rd ed. ASM Press, Washington, D.C. 11. Busch, M. P., S. H. Kleinman, B. Jackson, S. L. Stramer, I. Hewlett, and S. Preston. 2000. Committee report. Nucleic acid amplification testing of blood donors for transfusion-transmitted infectious diseases: report of the Interorganizational Task Force on Nucleic Acid Amplification of Blood Donors. Transfusion 40:143–159. 12. Carman, W. F., A. R. Zanetti, P. Karayiannis, J. Waters, G. Manzillo, E. Tanzi, A. J. Zuckerman, and H. C. Thomas. 1990. Vaccine-induced escape mutant of hepatitis B virus. Lancet 336:325–329. 13. Carman, W. F., J. Korula, L. Wallace, R. MacPhee, L. Mimms, and R. Decker. 1995. Fulminant reactivation of hepatitis B due to envelope protein mutant that escaped detection by monoclonal HBsAg ELISA. Lancet 345: 1406–1407. 14. Centers for Disease Control and Prevention. 1991. Viral hepatitis: a comprehensive strategy for eliminating transmission in the United States through universal childhood vaccination. Recommendations of the Immunization Practices Advisory Committee (ACIP). Morb. Mortal. Wkly. Rep. 40(RR13):1–25. 15. Choo, Q.-L., G. Kuo, A. J. Weiner, L. R. Overby, D. W. Bradley, and M. Houghton. 1989. Isolation of a cDNA clone derived from a blood-borne non-A, non-B viral hepatitis genome. Science 244:359–362. 16. Colin, C., D. Lanoir, S. Touzet, L. Meyaud-Kraemer, F. Bailly, and C. Trepo.

VOL. 42, 2004

17.

18. 19.

20. 21.

22. 23. 24. 25.

26. 27.

28. 29.

30. 31.

32. 33. 34. 35. 36. 37. 38.

SEROLOGICAL DIAGNOSIS OF HCV AND HBV INFECTIONS

2001. Sensitivity and specificity of third-generation hepatitis C virus antibody detection assays: an analysis of the literature. J. Viral Hepat. 8:87–95. Gretch, D. R., C. de la Rosa, R. L. Carithers, R. A. Willson, B. Williams, and L. Corey. 1995. Assessment of hepatitis C viremia using molecular amplification technologies: correlations and clinical implications. Ann. Intern. Med. 123:321–329. Gretch, D. R. 1997. Diagnostic tests for hepatitis C. Hepatology 26(Suppl. 1):43S–47S. Hennig, H., P. Schlenke, H. Kirchner, I. Bauer, B. Schulte-Kellinghaus, and H. Bludau. 2000. Evaluation of newly developed microparticle enzyme immunoassays for the detection of HCV antibodies. J. Virol. Methods 84:181– 190. Herrine, S. K., B. Michael, W. L. Ma, S. Rossi, S. R. Dunn, T. Hyslop. 2002. Development of an HCV infection risk stratification algorithm for patients on chronic hemodialysis. Am. J. Gastroenterol. 97:2619–2622. Ho, S. K., W. C. Yam, E. T. Leung, L. P. Wong, J. K. Leung, K. N. Lai, and T. M. Chan. 2003. Rapid quantification of hepatitis B virus DNA by realtime PCR using fluorescent hybridization probes. J. Med. Microbiol. 52:397– 402. Hoofnagle, J. H. 1990. Posttransfusion hepatitis B. Transfusion 30:384–386. Hoofnagle, J. H., and A. M. Di Bisceglie. 1991. Serologic diagnosis of acute and chronic viral hepatitis. Semin. Liver Dis. 11:73–83. Howard, C. R. 1995. The structure of hepatitis B envelope and molecular variants of hepatitis B virus. J. Viral Hepat. 2:165–170. Jilg, W., E. Sieger, R. Zachoval, and H. Schatzl. 1995. 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. 23:14–20. Jilg, W., M. Schmidt, F. Deinhardt, and R. Zachoval. 1984. Hepatitis B vaccination: how long does protection last? Lancet ii:458–462. Kao, J.-H., M.-Y. Lai, Y.-T. Hwang, P.-M. Yang, P.-J. Chen, J.-C. Sheu, T.-H. Wang, H.-C. Hsu, and D.-S. Chen. 1996. Chronic hepatitis C without antihepatitis C antibodies by second-generation assay: a clinicopathologic study and demonstration of the usefulness of a third-generation assay. Dig. Dis. Sci. 41:161–165. Koff, R. S. 2002. Immunogenicity of hepatitis B vaccines: implications of immune memory. Vaccine 20:3695–3701. Kojima, M., M. Shimizu, T. Tsuchimochi, M. Koyasu, S. Tanaka, H. Iizuka, T. Tanaka, H. Okamoto, F. Tsuda, and Y. Miyakawa. 1991. Posttransfusion fulminant hepatitis B associated with precore-defective HBV mutants. Vox Sang. 60:34–39. Krajden, M. 2000. Hepatitis C virus diagnosis and testing. Can. J. Public Health 91(Suppl. 1):S34–S39. Kuo, G., Q.-L. Choo, H. J. Alter, G. L. Gitnick, A. G. Redeker, R. H. Purcell, T. Miyamura, J. L. Dienstag, et al. 1989. An assay for circulating antibodies to a major etiologic virus of human non-A, non-B hepatitis. Science 244: 362–364. Lauer, G. M., and B. D. Walker. 2001. Hepatitis C virus infection. N. Engl. J. Med. 345:41–52. Lee, H. S., M. S. Rajagopalan, D. Chien, R. Cordell, H. A. Perkins, and G. N. Vyas. 1987. Specificity of enzyme immunoassay for hepatitis B core antibody used in screening blood donors. Transfusion 27:103–106. Lee, J. H., T. G. Paglieroni, P. V. Holland, and J. B. Zeldis. 1992. Chronic hepatitis B virus infection in an anti-HBc-nonreactive blood donor: variant virus or defective immune response? Hepatology 16:24–30. Le Pogam, S., D. Le Chapois, R. Christen, F. Dubois, F. Barin, and A. Goudeau. 1998. Hepatitis C in a hemodialysis unit: molecular evidence for nosocomial transmission. J. Clin. Microbiol. 36:3040–3043. Lin, H. J., J. Y. Lau, I. J. Lauder, N. Shi, C. L. Lai, and F. B. Hollinger. 1993. The hepatitis C virus genome: a guide to its conserved sequences and candidate epitopes. Virus Res. 30:27–41. Lopez, V. A., E. J. Bourne, M. W. Lutz, and L. D. Condreay. 2002. Assessment of the COBAS Amplicor HBV Monitor test for quantitation of serum hepatitis B virus DNA levels. J. Clin. Microbiol. 40:1972–1976. Okamoto, H., F. Tsuda, A. Machida, E. Munekata, Y. Akahane, Y. Sugai, K.

39. 40.

41.

42.

43. 44. 45. 46. 47.

48. 49.

50.

51.

52.

53. 54. 55. 56. 57.

617

Mashiko, T. Mitsui, T. Tanaka, and Y. Miyakawal. 1992. Antibodies against synthetic oligopeptides deduced from the putative core gene for the diagnosis of hepatitis virus infection. Hepatology 15:180–186. O’Meara, D., Z. Yun, A. Sonnerborg, and J. Lundeberg. 1998. Cooperative oligonucleotides mediating direct capture of hepatitis C virus RNA from serum. J. Clin. Microbiol. 36:2454–2459. Pawlotsky, J.-M., A. Bastie, C. Pellet, J. Remire, F. Darthuy, L. Wolfe, C. Sayada, J. Duval, and D. Dhumeaux. 1996. Significance of indeterminate third-generation hepatitis C virus recombinant immunoblot assay. J. Clin. Microbiol. 34:80–83. Pawlotsky, J.-M., A. Fleury, V. Choukroun, L. Deforges, F. Roudot-Thoraval, P. Aumont, J. Duval, and D. Dhumeaux. 1994. Significance of highly positive c22–3 “indeterminate” second-generation hepatitis C virus (HCV) recombinant immunoblot assay (RIBA) and resolution by third-generation HCV RIBA. J. Clin. Microbiol. 32:1357–1359. Polyak, S. J., and D. R. Gretch. 1997. Molecular diagnostic testing for viral hepatitis: methods and applications, p. 1–33. In R. A Willson (ed.), Viral hepatitis: diagnosis, treatment, and prevention, vol. 4. Marcel Dekker, Inc., New York, N.Y. Ratnam, S., F. Stead, and C. B. Head. 1989. False-positive results with third-generation monoclonal hepatitis B surface antigen enzyme immunoassay. J. Clin. Microbiol. 27:2102–2104. Read, S. J. 2001. Recovery efficiencies of nucleic acid extraction kits as measured by quantitative LightCycler™ PCR. J. Clin. Pathol. 54:86–90. Richter, S. S. 2002. Laboratory assays for diagnosis and management of hepatitis C virus infection. J. Clin. Microbiol. 40:4407–4412. Richter, S. S. 2003. Laboratory assays for diagnosis and management of hepatitis C virus infection. J. Clin. Microbiol. 41:530–534. Sarrazin, C., G. Teuber, R. Kokka, H. Rabenau, and S. Zeuzem. 2000. Detection of residual hepatitis C virus RNA by transcription-mediated amplification in patients with complete virologic response according to polymerase chain reaction-based assays. Hepatology 32:818–823. Schafer, D. F., and M. F. Sorrell. 2000. Conquering hepatitis C, step by step. N. Engl. J. Med. 343:1723–1724. Takaki, A., M. Wiese, G. Maertens, E. Depla, U. Seifert, A. Liebetrau, J. L. Miller, M. P. Manns, and B. Rehermann. 2000. Cellular immune responses persist and humoral responses decrease two decades after recovery from a single-source outbreak of hepatitis C. Nat. Med. 6:578–582. Tanaka, E., C. Ohue, K. Aoyagi, K. Yamaguchi, S. Yagi, K. Kiyosawa, and H. J. Alter. 2000. Evaluation of a new enzyme immunoassay for hepatitis C virus (HCV) core antigen with clinical sensitivity approximating that of genomic amplification of HCV RNA. Hepatology 32:388–393. Waters, J. A., M. Kennedy, P. Voet, P. Hauser, J. Petre, W. Carman, and H. C. Thomas. 1992. Loss of the common “A” determinant of hepatitis B surface antigen by a vaccine-induced escape mutant. J. Clin. Investig. 90: 2543–2547. Weinberger, K. M., E. A. Kreuzainter, B. Hottentrager, S. Neifer, and W. Jilg. 1997. Mutation in the s-gene of hepatitis B virus isolates from chronic carriers with anti-HBc as the only serological marker of HBV infection, p. 138–143. In M. Rizetto, R. H. Purcell, J. L. Gerin, and G. Verne (ed.), Viral hepatitis and liver disease. Edizioni Minerva Medica, Turin, Italy. West, D. J., and G. B. Calandra. 1996. Vaccine induced immunologic memory for hepatitis B surface antigen: implications for policy on booster vaccination. Vaccine 14:1019–1027. World Health Organization. 1994. Epidemiology of infection, p. 54–59. IARC monographs on the evaluation of carcinogenic risks to humans, hepatitis viruses, vol 59. World Health Organization, Geneva, Switzerland. Wright, T. L., and J. Y. Lau. 1993. Clinical aspects of hepatitis B virus infection. Lancet 342:1340–1344. Vyas, G. N., S. N. Cohen, and S. Schmid (ed.). 1978. Viral hepatitis. Franklin Institute Press, Philadelphia, Pa. Young, K. K., J. J. Archer, O. Yokosuka, M. Omata, and R. M. Resnick. 1995. Detection of hepatitis C virus RNA by a combined reverse transcription PCR assay: comparison with nested amplification and antibody testing. J. Clin. Microbiol. 33:654–657.