- Email: [email protected]
Evaluation of the performance of the EIAgen HCV test for detection of hepatitis C virus infection
Journal of Virological Methods 162 (2009) 203–207
Contents lists available at ScienceDirect
Journal of Virological Methods journal homepage: www.elsevier.com/locate/jviromet
Evaluation of the performance of the EIAgen HCV test for detection of hepatitis C virus infection Hui-Ying Rao a,1 , Fu-Rong Ren b,1 , Wen-Li Guan a , Michel Houde c , Shao-Cai Du a , Chang-Li Liu b , Xiao-Yan Gong b , Lai Wei a,∗ a b c
Peking University People’s Hospital, Peking University Hepatology Institute, Beijing 100044, China Beijing Red Cross Blood Center, Beijing, China Adaltis Development Inc., Laval, Quebec, Canada
a b s t r a c t Article history: Received 30 March 2009 Received in revised form 11 August 2009 Accepted 17 August 2009 Available online 22 August 2009 Keywords: Hepatitis C Anti-HCV Performance evaluation Sensitivity Specificity
The EIAgen HCV test (Adaltis Inc., Montreal, Canada) is an enzyme immunoassay (EIA) for the detection of anti-hepatitis C virus (HCV) antibodies. This study compared the performance of this test side-byside with the current Ortho HCV 3.0 Anti-HCV assay (Ortho-Clinical Diagnostics Inc., Johnson & Johnson Company, Raritan, NY, USA). Among 2559 specimens examined, 178 were true positives, 2376 were true negatives and 5 were indeterminate. The sensitivity of the EIAgen HCV test was 100%, versus 98.3% for the Ortho HCV test, while their respective specificities were 98.1% and 98.2%. The EIAgen HCV test gave a positive predictive value of 79.8% and a negative predictive value of 100%. Overall, the concordance of this test with the Ortho HCV test was 98.2%. Specimens from potentially interfering substances, such as sera from pregnant women, sera from patients with acute non-C hepatitis, autoimmune diseases, lipidemia, or from patients undergoing hemolysis, showed no interference with either EIA. An EIAgen HCV test signal-to-cut-off ratio of >5.9 would be highly predictive of a true-positive finding in these specimens. The EIAgen HCV test is well suited for screening blood and blood products in antibodies to HCV. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Hepatitis C virus (HCV) infection is a major global healthcare problem. The World Health Organization (WHO) estimates that up to 3% of the world’s population has been infected with HCV, equating to more than 170 million carriers of HCV worldwide. The infection is often asymptomatic, but the infection becomes persistent in more than 80% of infected people and can be associated with chronic hepatitis, cirrhosis or hepatocellular carcinomas (Di Bisceglie, 2000; Hoofnagle, 2002; Wei et al., 2002; Afdhal, 2004). Transmission is associated mainly with infected blood products or intravenous drug abuse, although other less common routes such as perinatal or sexual transmission have been described (Vandelli et al., 2004). HCV was first identified in 1989 using a recombinant DNA approach (Choo et al., 1989). Currently, routine diagnosis of HCV is based on detecting anti-HCV IgG antibodies in serum or plasma by enzyme immunoassay (EIA) (Centers for
Abbreviations: HCV, hepatitis C virus; WHO, World Health Organization; EIA, enzyme immunoassay; RIBA, recombinant immunoblot assay; NAT, nucleic acid amplification test; S/CO, signal-to-cut-off ratio. ∗ Corresponding author. Tel.: +86 10 88325566; fax: +86 10 68322662. E-mail addresses: [email protected], [email protected] (L. Wei). 1 These authors contributed equally. 0166-0934/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jviromet.2009.08.008
Disease Control and Prevention, 1991). Screening of blood and blood products for anti-HCV antibodies is an important element part of prevention. Infection is defined as repeated reactivity by a screening assay and confirmed reactivity of ≥1+ for at least two HCV bands by a recombinant immunoblot assay (RIBA), with or without the presence of detectable viral RNA (Allain, 2005). To increase the safety of blood products, many countries have now implemented donor testing of pooled or single blood samples for viral RNA using a nucleic acid amplification test (NAT) (Centers for Disease Control and Prevention, 1998). In addition to NAT screening, serological testing for HCV antibodies has improved in both sensitivity and specificity. Implementation of blood donation screening for anti-HCV antibodies by EIA has led to a marked decline in the risk of transfusion-transmitted hepatitis (Donahue et al., 1992). The EIAgen HCV test (Adaltis Inc., Montreal, Canada) is a diagnostic tool using EIA for qualitative detection of anti-HCV IgG and IgM antibodies in serum and plasma. The Ortho HCV 3.0 Anti-HCV (Ortho HCV) assay (Ortho-Clinical Diagnostics Inc., Johnson & Johnson Company, Raritan, NY, USA) is also a microplate EIA used to detect qualitatively anti-HCV antibodies. This test has been widely used, and its clinical performance is well documented (Tobler et al., 2003). This study was undertaken to compare the performance of the two tests with regards to sensitivity and specificity. The supplemental tests applied were the Chiron RIBA HCV 3.0 strip
204
H.-Y. Rao et al. / Journal of Virological Methods 162 (2009) 203–207
Fig. 1. Procedure for testing and reporting assay results.
immunoblot assay (RIBA 3.0; Chiron Corporation, Emeryville, CA, USA) and a commercial NAT (Beijing Hepatitis Reagent Research & Production Center, Beijing, China).
2. Materials and methods 2.1. Materials EIAgen HCV tests were obtained from Adaltis Inc. (Montreal, Canada). These use a two-step EIA for the detection of HCV infections in serum and plasma. Microplate wells are coated with HCV-specific antigens which capture anti-HCV antibodies in the sample. Upon completion of the assay, the development of color indicates the presence of HCV. The Ortho HCV tests were purchased from Ortho-Clinical Diagnostics Inc. (Johnson & Johnson Company, Raritan, NY, USA). Confirmatory assays were the RIBA 3.0 and a NAT as detailed above. The specificity of the NAT reverse transcription polymerase chain reaction (RT-PCR) assay is 90.9%, and the sensitivity is 98.8% when compared with the Amplicor HCV test (Roche Molecular Diagnostics, Basel, Switzerland). The high sensitivity of the NAT allows the detection of samples containing at least 21 viral copies of HCV per milliliter (Du et al., 2007).
2.2. Specimens Serum or plasma samples were collected from 2559 individuals: 2082 blood donors (Beijing Red Cross Blood Center) and 477 patients (Peking University Hepatology Institute). These specimens were stored frozen at −80 ◦ C before testing for anti-HCV antibodies and for HCV RNA. Samples included various HCV genotypes as well as samples from patients with non-C hepatitis (hepatitis A, B and E) and autoimmune diseases. They also included samples from pregnant women, hemolyzed sera and lipidemia sera. The ethics boards of the hospital and institute approved the study.
2.3. Assay procedures Both EIAs yield their final results as ratios of the specimen signal (in relative light units) to the cut-off value (signal-to-cut-off ratio, S/CO). S/CO ratios ≥1.0 were considered reactive for anti-HCV antibodies while those <1.0 were considered nonreactive. Specimen preparation and testing were carried out according to the manufacturers’ instructions. For the EIAgen HCV assay, 200 L of sample diluent followed by 10 L aliquots of specimens or controls and 50 L of assay diluent were added to each of the microplate wells. Wells were incubated at 37 ◦ C for 45 min then washed five times. Aliquots of 100 L of conjugate were added to all wells, which were then incubated for 45 min at 37 ◦ C. The plate was washed five times, then 100 L of substrate was added to each of the wells. The microplate was incubated for 15 min at room temperature (18–24 ◦ C). Finally, the stop solution was added, and the plates were read at 450/630 nm. The EIAgen HCV tests and Ortho HCV tests were always carried out side-by-side. Concordant nonreactive results of the two assays were considered as true negatives. All concordant reactive samples were then forwarded for supplemental testing. To reduce the costs of supplemental testing, an HCV NAT was performed first. Only HCV NAT-negative samples were tested by RIBA. Confirmed HCV infection was defined by reactive results by either NAT or RIBA tests. Discordant samples were retested with both tests in duplicate, and final result was determined following the recommendation of the manufacturers. If the final results were nonreactive concordantly, they were classified as true negative. If the results were still discordant, specimens were forwarded for confirmation testing by the HCV NAT and RIBA. RIBA 3.0-positive and/or NAT-positive samples were classified as true positive while RIBA 3.0-negative and NATnegative results were classified as true negative. Samples which were indeterminate by the RIBA but negative with the HCV NAT were classified as indeterminate. Those that were RIBA negative but HCV NAT positive also were classified as indeterminate because these samples might have been in a “window phase”. Indetermi-
H.-Y. Rao et al. / Journal of Virological Methods 162 (2009) 203–207 Table 3 Sensitivity of the EIAgen HCV test and Ortho HCV test.
Table 1 Results of the EIAgen HCV test and Ortho HCV test. Ortho HCV test
205
EIAgen HCV test
Total
+(S/CO ≥ 1)
−(S/CO < 1)
+(S/CO ≥ 1) −(S/CO < 1)
216 10
4 2329
220 2339
Total
226
2333
2559
EIAgen HCV test
Ortho HCV test
True positives 178 False negatives 0 Total 178 Sensitivity 100 (178/178 × 100)% 95% confidence intervals 100%
175 3 178 98.3 (175/178 × 100)% 97.4–99.3%
S/CO, signal-to-cut-off ratio.
nate RIBA and NAT results were not included in the calculation of sensitivity or specificity (Fig. 1). Precision was evaluated using a proficiency panel consisting of the anti-HCV calibrator and positive and negative controls.
value = true negative/(true negative + false negative) × 100%. Concordance rate = (true positive + true negative)/(true positive + true negative + false positive + false negative) × 100%.
3. Results
2.4. HCV NAT HCV RNA was sought using a NAT (Beijing Hepatitis Reagent Research & Production Center, Beijing, China). Testing and interpretation of results followed the manufacturer’s instructions. 2.5. RIBA RIBA was used as the “gold standard” for anti-HCV testing. Testing followed the manufacturer’s instructions. The minimum requirement for a positive result is two HCV-specific bands with at least 1+ reactivity. An indeterminate result was defined as: (1) one HCV-specific band with reactivity equal to or greater than 1+, or (2) reactivity to human superoxide dismutase and one or more HCV-specific bands with at least 1+ reactivity. A negative result was defined as no HCV-specific band with reactivity.
3.1. Overall results The results of 2559 sera tested by the EIAgen HCV test and the Ortho HCV Test are summarized in Table 1. Fourteen specimens (0.6%) had discordant results and were tested further by supplemental assays, as shown in Table 2. Five samples were RIBA 3.0 indeterminate, 3 were positive and 6 were negative. Overall, testing by the EIAgen HCV test yielded 178 true positives, no false negatives, 2331 true negatives and 45 false positives. Testing with the Ortho HCV test yielded 175 true positives, 3 false negatives, 2333 true negatives and 43 false positives. Five specimens were RIBA 3.0 indeterminate. For the EIAgen HCV test, positive and negative predictive values were 79.8% and 100%, respectively, and the overall concordance rate was 98.2%. For the Ortho HCV test, positive and negative predictive values were 80.3% and 99.9%, respectively, and the overall concordance rate was 98.2%.
2.6. Calculations Sensitivity was defined as the correct identification of anti-HCVpositive specimens, using the following formula: sensitivity = true positive/(true positive + false negative) × 100%. Specificity was defined as the correct identification of anti-HCV-negative specimens, using the following formula: specificity = true negative/(true negative + false positive) × 100%. True positives and true negatives were defined as the numbers of anti-HCV-positive and -negative specimens identified correctly by each test. False positives and false negatives were defined as the numbers of anti-HCV-negative or -positive specimens identified incorrectly as reactive repeatedly by each method. Other outcome measures were as follows. Positive predictive value = true positive/(true positive + false positive) × 100%. Negative predictive
3.2. Sensitivity The ability of the EIAgen HCV test and Ortho HCV test to detect anti-HCV antibodies were evaluated with 178 true-positive serum samples (Table 3). The sensitivity of the EIAgen HCV test was 100% and that of the Ortho HCV test was 98.3% (95% confidence intervals, CI 97.4–99.3%), with three specimens being classified as false negatives. One of these false negatives was RIBA positive while the two others were NAT positive and RIBA indeterminate. A series of 27 specimens with various HCV genotypes (10 of genotype 1b, 11 of genotype 2a, 2 of genotype 1a, 2 of genotype 3 and 2 of genotype 6) was part of that panel. These specimens were all shown to be true positives by both tests.
Table 2 Supplemental test results of 14 discordant specimens (RT-PCR cut-off 0.132). No.
1 2 3 4 5 6 7 8 9 10 11 12 13 14
EIAgen
Ortho
RIBA
S/CO
S/CO
NS4
NS3
Core
NS5
RT-PCR
5.2 1.3 2.6 2.4 6.4 1.4 1.1 0.8 0.6 4.3 1.0 4.1 0.02 0.27
0.42 0.05 0.64 0.10 0.11 0.01 0.32 2.0 4.0 0.03 0.08 0.05 1.3 1.1
− −/+ −/+ −/+ − − − − − − − − − −
2+ − −/+ −/+ − − − − − 1+ − 1+ − 2+
1+ 1+ 1+ −/+ −/+ − −/+ − − − 2+ − − −
− −/+ − − − − − − −/+ − − − − −
EIAgen result
OD + Indeterminate Indeterminate − − − − − − Indeterminate Indeterminate Indeterminate − Indeterminate
0.057 1.361 1.353 0.095 0.069 0.063 0.076 0.060 0.058 0.067 0.075 0.056 1.138 0.060
− + + − − − − − − − − − + −
True positive True positive True positive False positive False positive False positive False positive True negative True negative Indeterminate Indeterminate Indeterminate Indeterminate Indeterminate
206
H.-Y. Rao et al. / Journal of Virological Methods 162 (2009) 203–207
Table 4 Specificity of the EIAgen HCV test and Ortho HCV test.
True negatives False positives Total Specificity 95% confidence intervals
4. Discussion
EIAgen HCV test
Ortho HCV test
2331 45 2376 98.1 (2331/2376 × 100)% 97.8–98.4%
2333 43 2376 98.2 (2333/2376 × 100)% 97.9–98.5%
3.3. Specificity A total of 2376 specimens were found to be true negatives based on their results with EIAs and confirmatory assays (nonreactive with both RIBA and NAT tests). For the EIAgen HCV test, 45 specimens were found to be false positives, resulting in a final specificity of 98.1% (95% CI 97.8–98.4%). For the Ortho HCV, 43 specimens tested as falsely positive, resulting in a final specificity of 98.2% (95% CI 97.9–98.5%) (Table 4). Specimens from individuals with medical conditions unrelated to HCV and containing potentially interfering substances were tested with both methods. These specimens consisted of 20 samples from patients with hepatitis A infection, 70 with hepatitis B infection, 10 with hepatitis E infection, 18 from patients with systemic lupus erythematosus, 19 from patients with primary biliary cirrhosis and 13 sera from pregnant women. Both the EIAgen HCV test and Ortho HCV test gave an apparent specificity of 100%. 3.4. Hemolysis or sera with lipidemia Specimens from patients with hemolysis or lipidemia were also tested by both tests. Hemolyzed HCV-negative specimens had free hemoglobin concentrations ranging from 2.5 to 10.0 g/L, while the 20 lipidemia specimens had chyle indexes ranging from 16 to 32. As shown in Table 5, all specimens were nonreactive by both assays, giving a specificity of 100%. 3.5. Signal-to-cut-off ratio (S/CO) When the subgroup of 178 true-positive samples was considered, S/CO ratios of the EIAgen HCV test ranged from 1.1 to 25.0 while those of the Ortho HCV test ranged from 1.1 to 5.0. Among specimens with an Ortho HCV test S/CO result of ≥3.8, 88.8% were also ≥5.9 with the EIAgen HCV test. Of the EIAgen HCV test S/CO results, 90.5% (161/178) of ratios were above 5.9, compared with 81.5% (145/178) above 3.8 for the Ortho HCV test (Table 6).
The EIAgen HCV test is a diagnostic EIA intended for screening patients with HCV infection using qualitative detection of anti-HCV IgG and IgM. The Ortho HCV test is a commercial assay, also using EIA, for qualitative detection of anti-HCV antibodies. The performance of the EIAgen HCV test with the current Ortho HCV test was evaluated using a series of specimens collected in clinical settings. The main criteria for an HCV screening assay are to attain the highest sensitivity possible in combination with excellent specificity. The challenge of any screening assay is to enhance sensitivity while not compromising assay specificity. In countries where HCV NAT testing for blood donor screening is not mandatory, an antiHCV assay serves as the only test in detecting of blood with HCV. The importance of detecting anti-HCV antibodies is stressed further by the fact that a significant number of specimens were demonstrated to be anti-HCV-positive but NAT negative when tested using a minipool NAT (Janatpour et al., 2002; Busch et al., 2003). It should be recognized that the level of virus declines and fluctuates widely in patients following seroconversion, or when they are in remission and treated with antiviral drugs (Kuramoto et al., 2002). Specimens from these patients may be negative transiently for viral nucleic acid and/or viral antigens but might still be infectious despite the presence of HCV antibodies. Therefore, the accuracy of an HCV screening assay is important. A loss of sensitivity is not acceptable, because this affects blood donor safety and delays the treatment of patients infected with HCV. In our hands, the sensitivity of the EIAgen HCV test was 100%, which was higher than the Ortho HCV test (98.3%). There are various explanations. First, the EIAgen HCV test can detect IgM in addition to IgG antibodies, whereas the Ortho HCV test can detect only IgG antibodies. Second, formulations used to prepare the EIA (the microplate wells and conjugate) might confer a better integrity to the antigens used in the EIAgen HCV test. Third, the EIAgen HCV test might use antigens with higher sequence homology to the most prevalent HCV strains found in China. This latter hypothesis seems improbable since all the HCV genotypes examined in this study (genotypes 1b, 2a, 1a, 3 and 6) were detected by both assays. False-positive results also cause distress and confusion, because the clinical significance of these results is not known or, if known, may not be understood by the donor. During the 1990s, several reports raised concerns over this problem (Sakugawa et al., 1995; Bar-Shany et al., 1996; Ownby et al., 1997; Sharma et al., 2003). In certain clinical settings, false-positive anti-HCV results are rare because the majority of persons being tested have evidence of liver disease and the sensitivity and specificity of the screening assays
Table 5 EIAgen HCV test results with hemolyzed specimens. Degree of hemolysis (free Hb g/L)
Control (0.0)
Mild (2.5)
Medium (5.0)
Severe (10.0)
True positives True negatives False positives False negatives
0 19 0 0
0 19 0 0
0 19 0 0
0 19 0 0
Total Specificity (%)
19 100
19 100
19 100
19 100
Table 6 Proportion of S/CO ratios for the EIAgen HCV test and Ortho HCV test. Ortho HCV test
EIAgen HCV test S/CO ≥ 5.9
Total S/CO < 5.9
S/CO ≥ 3.8 S/CO < 3.8
143 (88.8%) 18 (11.2%)
2 (12%) 15 (88%)
145 (81.5%) 33 (18.5%)
Total
161 (90.5%)
17 (10%)
178
H.-Y. Rao et al. / Journal of Virological Methods 162 (2009) 203–207
are high. However, among populations with a low (<10%) prevalence of HCV infection, false-positive results do occur (Hyams et al., 2001). Among immunocompetent populations with an antiHCV prevalence lower than 10%, the proportion of false-positive results with the Ortho HCV test average about 35% (range: 15–60%) (Alter et al., 2003), although the specificity is ≥99%. With regard to the populations tested, the specificity of the EIAgen HCV test was 98.1%, compared with 99.2% for the Ortho HCV test. This level is considered excellent, based on the standard requirement of 99.5% for regulatory approval of HCV screening tests throughout the world. However, specimens with low S/CO ratios (from 1.0 to 3.0) as determined by the EIAgen assay should be verified using an independent supplemental test with high specificity. None of the sera from patients with hepatitis A, B, E, systemic lupus erythematosus, primary biliary cirrhosis or pregnant women) interfered with the assays. The potential interfering effect of hemolysis and lipidemia was also examined. All specimens with various concentrations of hemoglobin and lipids were found to be true negative, which resulted in complete specificity of the two EIAs. Analysis of early versions of anti-HCV EIA results tested in volunteer blood donors indicated that the average repeatedly reactive S/CO ratios could be used to predict positive results from supplemental testing (Tobler et al., 2000). The USA Center for Disease Control determined which specific S/CO ratio would predict a true antibody positive result >95% of the time, regardless of the prevalence of anti-HCV antibodies or the characteristics of the population tested. In the case of the Ortho HCV test, an average S/CO ratio >3.8 appeared to be highly predictive of the true anti-HCV status (Alter et al., 2003). These results were confirmed in Chinese blood donors (Ren et al., 2005). In the current study, this same cut-off value of 3.8 yielded a proportion of 81.5% for the Ortho HCV test, confirming these previous observations. This proportion increased to 90.5% for the EIAgen HCV test using a cut-off level of 5.9. Within that group of specimens with an EIAgen HCV test S/CO ratio ≥5.9, 88.8% had an Ortho HCV test S/CO ratio ≥3.8. Reciprocally, among the truepositive specimens with an EIAgen HCV test S/CO ratio <5.9, 88.2% had an Ortho HCV test S/CO ratio <3.8. Therefore, specimens with average S/CO ratios >5.9 in the EIAgen HCV test would be highly predictive of true HCV-positive status. The EIAgen HCV test is well suited for screening of blood and blood products. Acknowledgments This work was partially supported by grants from the Chinese Basic Research Foundation 973 (2005CB522902 and 2007CB512900), 863 program (2006AA02A410), NSFC (30571639), the National Science and Technology Key Project during the 11th Five-Year Plan Period (2008ZX10002-012 and 2008ZX10002-013) and Key Clinical Research Program of Ministry of Health. References Afdhal, N.H., 2004. The natural history of hepatitis C. Semin. Liver Dis. 24, S3–S8.
207
Allain, J.P., 2005. Hepatitis C virus in blood donation. Lancet 365, 276– 278. Alter, M.J., Kuhnert, W.L., Finelli, L., 2003. Guidelines for laboratory testing and result reporting of antibody to hepatitis C virus. M. M. W. R. 52, 1–15. Bar-Shany, S., Green, M.S., Shinar, E., 1996. False positive tests for anti-hepatitis C antibodies and the problem of notifying blood donors. Int. J. Epidemiol. 25, 674–678. Busch, M.P., Tobler, L.H., Gerlich, W.H., Schaefer, S., Giachetti, C., Smith, R., 2003. Very low level viremia in HCV infectious unit missed by NAT. Transfusion 43, 1173–1174. Centers for Disease Control and Prevention, 1991. Public health service inter-agency guidelines for screening donors of blood, plasma, organ, tissues and semen for evidence of hepatitis B and hepatitis C. M. M. W. R. 40, 1–17. Centers for Disease Control and Prevention, 1998. Recommendations for prevention and control of hepatitis C virus (HCV) infection and HCV-related chronic disease. M. M. W. R. 47, 1-39. Choo, Q.L., Kuo, G., Weiner, A.J., Overby, L.R., Bradley, D.W., Houghton, M., 1989. Isolation of a cDNA clone derived from a blood-borne non-A, non-B viral hepatitis genome. Science 244, 359–362. Di Bisceglie, A.M., 2000. Natural history of hepatitis C: its impact on clinical management. Hepatology 31, 1014–1018. Donahue, J.G., Munoz, A., Ness, P.M., Brown Jr., D.E., Yawn, D.H., McAllister Jr., H.A., Reitz, B.A., Nelson, K.E., 1992. The declining risk of post-transfusion hepatitis C virus infection. N. Engl. J. Med. 327, 369–373. Du, S.C., Zhang, R., Li, J.Q., Wei, L., 2007. Utilization of uracil-DNA glycosylase for combining reverse transcription and anti-contamination with polymerase chain reaction in hepatitis C virus. J. Peking Univ. [Health Sci.] 39, 426–428. Hoofnagle, J.H., 2002. Course and outcome of hepatitis C. Hepatology 36, S21– S29. Hyams, K.C., Riddle, J., Rubertone, M., Trump, D., Alter, M.J., Cruess, D.F., Han, X., Nainam, O.V., Seeff, L.B., Mazzuchi, J.F., Bailey, S., 2001. Prevalence and incidence of hepatitis C virus infection in the US military: a seroepidemiologic survey of 21,000 troops. Am. J. Epidemiol. 153, 764–770. Janatpour, K.A., Kemper, M., Aceituno, S., Holland, P.V., 2002. Incidence of minipool HCV NAT negative, single unit NAT positive units among serological positive blood donations. Transfusion 42, 80S. Kuramoto, I.K., Moriya, T., Schoening, V., Holland, P.V., 2002. Fluctuation of serum HCV-RNA levels in untreated blood donors with chronic hepatitis C virus infection. J. Virol. Hepatitis 9, 36–42. Ownby, H.E., Korelitz, J.J., Busch, M.P., Williams, A.E., Kleinman, S.H., Gilcher, R.O., Nourjah, P., Retrovirus Epidemiology Donor Study, 1997. Loss of volunteer blood donors because of unconfirmed enzyme immunoassay screening results. Transfusion 37, 199–205. Ren, F.R., Lv, Q.S., Zhuang, H., Li, J.J., Gong, X.Y., Gao, G.J., Liu, C.L., Wang, J.X., Yao, F.Z., Zhen, Y.R., Zhu, F.M., Tiemuer, M.H., Bai, X.H., Shan, H., 2005. Significance of the signal-to-cutoff ratios of anti-hepatitis C virus enzyme immunoassays in screening of Chinese blood donors. Transfusion 45, 1816–1822. Sakugawa, H., Nakasone, H., Nakayoshi, T., Kinjo, F., Saito, A., Yakabi, S., Zukeran, H., Miyagi, Y., Taira, R., Koja, K., 1995. High proportion of false positive reactions among donors with anti-HCV antibodies in a low prevalence area. J. Med. Virol. 46, 334–338. Sharma, U.K., Stramer, S.L., Wright, D.J., Glynn, S.A., Hermansen, S., Schreiber, G.B., Kleinman, S.H., Busch, M.P., Retrovirus Epidemiology Donor Study, 2003. Impact of changes in viral marker screening assays. Transfusion 43, 202–214. Tobler, L.H., Tegtmeier, G., Stramer, S.L., Quan, S., Dockter, J., Giachetti, C., Busch, M.P., 2000. Lookback on donors who are repeatedly reactive on first-generation hepatitis C virus assays: justification and rational implementation. Transfusion 40, 15–24. Tobler, L.H., Stramer, S.L., Lee, S.R., Masecar, B.L., Peterson, J.E., Davis, E.A., Andrews, W.E., Brodsky, J.P., Kleinman, S.H., Phelps, B.H., Busch, M.P., 2003. Impact of HCV 3.0 EIA relative to HCV 2.0 EIA on blood-donor screening. Transfusion 43, 1452–1459. Vandelli, C., Renzo, F., Romano, L., Tisminetzky, S., De Palma, M., Stroffolini, T., Ventura, E., Zanetti, A., 2004. Lack of evidence of sexual transmission of hepatitis C among monogamous couples: results of a 10-year prospective follow-up study. Am. J. Gastroenterol. 99, 855–859. Wei, L., Wang, Q.X., Xu, X.Y., Wan, H., Gao, Y., Tian, X.L., Yu, M., Sun, D.G., Fan, C.L., Jin, J., Fan, W.M., Yin, L.M., Zhu, W.F., Chen, H.S., Zhuang, H., Wang, Y., 2002. 12–25year follow-up of hepatitis C virus infection in a rural area of Hebei province. China J. Peking Univ. [Health Sci.] 34, 574–578.
Contents lists available at ScienceDirect
Journal of Virological Methods journal homepage: www.elsevier.com/locate/jviromet
Evaluation of the performance of the EIAgen HCV test for detection of hepatitis C virus infection Hui-Ying Rao a,1 , Fu-Rong Ren b,1 , Wen-Li Guan a , Michel Houde c , Shao-Cai Du a , Chang-Li Liu b , Xiao-Yan Gong b , Lai Wei a,∗ a b c
Peking University People’s Hospital, Peking University Hepatology Institute, Beijing 100044, China Beijing Red Cross Blood Center, Beijing, China Adaltis Development Inc., Laval, Quebec, Canada
a b s t r a c t Article history: Received 30 March 2009 Received in revised form 11 August 2009 Accepted 17 August 2009 Available online 22 August 2009 Keywords: Hepatitis C Anti-HCV Performance evaluation Sensitivity Specificity
The EIAgen HCV test (Adaltis Inc., Montreal, Canada) is an enzyme immunoassay (EIA) for the detection of anti-hepatitis C virus (HCV) antibodies. This study compared the performance of this test side-byside with the current Ortho HCV 3.0 Anti-HCV assay (Ortho-Clinical Diagnostics Inc., Johnson & Johnson Company, Raritan, NY, USA). Among 2559 specimens examined, 178 were true positives, 2376 were true negatives and 5 were indeterminate. The sensitivity of the EIAgen HCV test was 100%, versus 98.3% for the Ortho HCV test, while their respective specificities were 98.1% and 98.2%. The EIAgen HCV test gave a positive predictive value of 79.8% and a negative predictive value of 100%. Overall, the concordance of this test with the Ortho HCV test was 98.2%. Specimens from potentially interfering substances, such as sera from pregnant women, sera from patients with acute non-C hepatitis, autoimmune diseases, lipidemia, or from patients undergoing hemolysis, showed no interference with either EIA. An EIAgen HCV test signal-to-cut-off ratio of >5.9 would be highly predictive of a true-positive finding in these specimens. The EIAgen HCV test is well suited for screening blood and blood products in antibodies to HCV. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Hepatitis C virus (HCV) infection is a major global healthcare problem. The World Health Organization (WHO) estimates that up to 3% of the world’s population has been infected with HCV, equating to more than 170 million carriers of HCV worldwide. The infection is often asymptomatic, but the infection becomes persistent in more than 80% of infected people and can be associated with chronic hepatitis, cirrhosis or hepatocellular carcinomas (Di Bisceglie, 2000; Hoofnagle, 2002; Wei et al., 2002; Afdhal, 2004). Transmission is associated mainly with infected blood products or intravenous drug abuse, although other less common routes such as perinatal or sexual transmission have been described (Vandelli et al., 2004). HCV was first identified in 1989 using a recombinant DNA approach (Choo et al., 1989). Currently, routine diagnosis of HCV is based on detecting anti-HCV IgG antibodies in serum or plasma by enzyme immunoassay (EIA) (Centers for
Abbreviations: HCV, hepatitis C virus; WHO, World Health Organization; EIA, enzyme immunoassay; RIBA, recombinant immunoblot assay; NAT, nucleic acid amplification test; S/CO, signal-to-cut-off ratio. ∗ Corresponding author. Tel.: +86 10 88325566; fax: +86 10 68322662. E-mail addresses: [email protected], [email protected] (L. Wei). 1 These authors contributed equally. 0166-0934/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jviromet.2009.08.008
Disease Control and Prevention, 1991). Screening of blood and blood products for anti-HCV antibodies is an important element part of prevention. Infection is defined as repeated reactivity by a screening assay and confirmed reactivity of ≥1+ for at least two HCV bands by a recombinant immunoblot assay (RIBA), with or without the presence of detectable viral RNA (Allain, 2005). To increase the safety of blood products, many countries have now implemented donor testing of pooled or single blood samples for viral RNA using a nucleic acid amplification test (NAT) (Centers for Disease Control and Prevention, 1998). In addition to NAT screening, serological testing for HCV antibodies has improved in both sensitivity and specificity. Implementation of blood donation screening for anti-HCV antibodies by EIA has led to a marked decline in the risk of transfusion-transmitted hepatitis (Donahue et al., 1992). The EIAgen HCV test (Adaltis Inc., Montreal, Canada) is a diagnostic tool using EIA for qualitative detection of anti-HCV IgG and IgM antibodies in serum and plasma. The Ortho HCV 3.0 Anti-HCV (Ortho HCV) assay (Ortho-Clinical Diagnostics Inc., Johnson & Johnson Company, Raritan, NY, USA) is also a microplate EIA used to detect qualitatively anti-HCV antibodies. This test has been widely used, and its clinical performance is well documented (Tobler et al., 2003). This study was undertaken to compare the performance of the two tests with regards to sensitivity and specificity. The supplemental tests applied were the Chiron RIBA HCV 3.0 strip
204
H.-Y. Rao et al. / Journal of Virological Methods 162 (2009) 203–207
Fig. 1. Procedure for testing and reporting assay results.
immunoblot assay (RIBA 3.0; Chiron Corporation, Emeryville, CA, USA) and a commercial NAT (Beijing Hepatitis Reagent Research & Production Center, Beijing, China).
2. Materials and methods 2.1. Materials EIAgen HCV tests were obtained from Adaltis Inc. (Montreal, Canada). These use a two-step EIA for the detection of HCV infections in serum and plasma. Microplate wells are coated with HCV-specific antigens which capture anti-HCV antibodies in the sample. Upon completion of the assay, the development of color indicates the presence of HCV. The Ortho HCV tests were purchased from Ortho-Clinical Diagnostics Inc. (Johnson & Johnson Company, Raritan, NY, USA). Confirmatory assays were the RIBA 3.0 and a NAT as detailed above. The specificity of the NAT reverse transcription polymerase chain reaction (RT-PCR) assay is 90.9%, and the sensitivity is 98.8% when compared with the Amplicor HCV test (Roche Molecular Diagnostics, Basel, Switzerland). The high sensitivity of the NAT allows the detection of samples containing at least 21 viral copies of HCV per milliliter (Du et al., 2007).
2.2. Specimens Serum or plasma samples were collected from 2559 individuals: 2082 blood donors (Beijing Red Cross Blood Center) and 477 patients (Peking University Hepatology Institute). These specimens were stored frozen at −80 ◦ C before testing for anti-HCV antibodies and for HCV RNA. Samples included various HCV genotypes as well as samples from patients with non-C hepatitis (hepatitis A, B and E) and autoimmune diseases. They also included samples from pregnant women, hemolyzed sera and lipidemia sera. The ethics boards of the hospital and institute approved the study.
2.3. Assay procedures Both EIAs yield their final results as ratios of the specimen signal (in relative light units) to the cut-off value (signal-to-cut-off ratio, S/CO). S/CO ratios ≥1.0 were considered reactive for anti-HCV antibodies while those <1.0 were considered nonreactive. Specimen preparation and testing were carried out according to the manufacturers’ instructions. For the EIAgen HCV assay, 200 L of sample diluent followed by 10 L aliquots of specimens or controls and 50 L of assay diluent were added to each of the microplate wells. Wells were incubated at 37 ◦ C for 45 min then washed five times. Aliquots of 100 L of conjugate were added to all wells, which were then incubated for 45 min at 37 ◦ C. The plate was washed five times, then 100 L of substrate was added to each of the wells. The microplate was incubated for 15 min at room temperature (18–24 ◦ C). Finally, the stop solution was added, and the plates were read at 450/630 nm. The EIAgen HCV tests and Ortho HCV tests were always carried out side-by-side. Concordant nonreactive results of the two assays were considered as true negatives. All concordant reactive samples were then forwarded for supplemental testing. To reduce the costs of supplemental testing, an HCV NAT was performed first. Only HCV NAT-negative samples were tested by RIBA. Confirmed HCV infection was defined by reactive results by either NAT or RIBA tests. Discordant samples were retested with both tests in duplicate, and final result was determined following the recommendation of the manufacturers. If the final results were nonreactive concordantly, they were classified as true negative. If the results were still discordant, specimens were forwarded for confirmation testing by the HCV NAT and RIBA. RIBA 3.0-positive and/or NAT-positive samples were classified as true positive while RIBA 3.0-negative and NATnegative results were classified as true negative. Samples which were indeterminate by the RIBA but negative with the HCV NAT were classified as indeterminate. Those that were RIBA negative but HCV NAT positive also were classified as indeterminate because these samples might have been in a “window phase”. Indetermi-
H.-Y. Rao et al. / Journal of Virological Methods 162 (2009) 203–207 Table 3 Sensitivity of the EIAgen HCV test and Ortho HCV test.
Table 1 Results of the EIAgen HCV test and Ortho HCV test. Ortho HCV test
205
EIAgen HCV test
Total
+(S/CO ≥ 1)
−(S/CO < 1)
+(S/CO ≥ 1) −(S/CO < 1)
216 10
4 2329
220 2339
Total
226
2333
2559
EIAgen HCV test
Ortho HCV test
True positives 178 False negatives 0 Total 178 Sensitivity 100 (178/178 × 100)% 95% confidence intervals 100%
175 3 178 98.3 (175/178 × 100)% 97.4–99.3%
S/CO, signal-to-cut-off ratio.
nate RIBA and NAT results were not included in the calculation of sensitivity or specificity (Fig. 1). Precision was evaluated using a proficiency panel consisting of the anti-HCV calibrator and positive and negative controls.
value = true negative/(true negative + false negative) × 100%. Concordance rate = (true positive + true negative)/(true positive + true negative + false positive + false negative) × 100%.
3. Results
2.4. HCV NAT HCV RNA was sought using a NAT (Beijing Hepatitis Reagent Research & Production Center, Beijing, China). Testing and interpretation of results followed the manufacturer’s instructions. 2.5. RIBA RIBA was used as the “gold standard” for anti-HCV testing. Testing followed the manufacturer’s instructions. The minimum requirement for a positive result is two HCV-specific bands with at least 1+ reactivity. An indeterminate result was defined as: (1) one HCV-specific band with reactivity equal to or greater than 1+, or (2) reactivity to human superoxide dismutase and one or more HCV-specific bands with at least 1+ reactivity. A negative result was defined as no HCV-specific band with reactivity.
3.1. Overall results The results of 2559 sera tested by the EIAgen HCV test and the Ortho HCV Test are summarized in Table 1. Fourteen specimens (0.6%) had discordant results and were tested further by supplemental assays, as shown in Table 2. Five samples were RIBA 3.0 indeterminate, 3 were positive and 6 were negative. Overall, testing by the EIAgen HCV test yielded 178 true positives, no false negatives, 2331 true negatives and 45 false positives. Testing with the Ortho HCV test yielded 175 true positives, 3 false negatives, 2333 true negatives and 43 false positives. Five specimens were RIBA 3.0 indeterminate. For the EIAgen HCV test, positive and negative predictive values were 79.8% and 100%, respectively, and the overall concordance rate was 98.2%. For the Ortho HCV test, positive and negative predictive values were 80.3% and 99.9%, respectively, and the overall concordance rate was 98.2%.
2.6. Calculations Sensitivity was defined as the correct identification of anti-HCVpositive specimens, using the following formula: sensitivity = true positive/(true positive + false negative) × 100%. Specificity was defined as the correct identification of anti-HCV-negative specimens, using the following formula: specificity = true negative/(true negative + false positive) × 100%. True positives and true negatives were defined as the numbers of anti-HCV-positive and -negative specimens identified correctly by each test. False positives and false negatives were defined as the numbers of anti-HCV-negative or -positive specimens identified incorrectly as reactive repeatedly by each method. Other outcome measures were as follows. Positive predictive value = true positive/(true positive + false positive) × 100%. Negative predictive
3.2. Sensitivity The ability of the EIAgen HCV test and Ortho HCV test to detect anti-HCV antibodies were evaluated with 178 true-positive serum samples (Table 3). The sensitivity of the EIAgen HCV test was 100% and that of the Ortho HCV test was 98.3% (95% confidence intervals, CI 97.4–99.3%), with three specimens being classified as false negatives. One of these false negatives was RIBA positive while the two others were NAT positive and RIBA indeterminate. A series of 27 specimens with various HCV genotypes (10 of genotype 1b, 11 of genotype 2a, 2 of genotype 1a, 2 of genotype 3 and 2 of genotype 6) was part of that panel. These specimens were all shown to be true positives by both tests.
Table 2 Supplemental test results of 14 discordant specimens (RT-PCR cut-off 0.132). No.
1 2 3 4 5 6 7 8 9 10 11 12 13 14
EIAgen
Ortho
RIBA
S/CO
S/CO
NS4
NS3
Core
NS5
RT-PCR
5.2 1.3 2.6 2.4 6.4 1.4 1.1 0.8 0.6 4.3 1.0 4.1 0.02 0.27
0.42 0.05 0.64 0.10 0.11 0.01 0.32 2.0 4.0 0.03 0.08 0.05 1.3 1.1
− −/+ −/+ −/+ − − − − − − − − − −
2+ − −/+ −/+ − − − − − 1+ − 1+ − 2+
1+ 1+ 1+ −/+ −/+ − −/+ − − − 2+ − − −
− −/+ − − − − − − −/+ − − − − −
EIAgen result
OD + Indeterminate Indeterminate − − − − − − Indeterminate Indeterminate Indeterminate − Indeterminate
0.057 1.361 1.353 0.095 0.069 0.063 0.076 0.060 0.058 0.067 0.075 0.056 1.138 0.060
− + + − − − − − − − − − + −
True positive True positive True positive False positive False positive False positive False positive True negative True negative Indeterminate Indeterminate Indeterminate Indeterminate Indeterminate
206
H.-Y. Rao et al. / Journal of Virological Methods 162 (2009) 203–207
Table 4 Specificity of the EIAgen HCV test and Ortho HCV test.
True negatives False positives Total Specificity 95% confidence intervals
4. Discussion
EIAgen HCV test
Ortho HCV test
2331 45 2376 98.1 (2331/2376 × 100)% 97.8–98.4%
2333 43 2376 98.2 (2333/2376 × 100)% 97.9–98.5%
3.3. Specificity A total of 2376 specimens were found to be true negatives based on their results with EIAs and confirmatory assays (nonreactive with both RIBA and NAT tests). For the EIAgen HCV test, 45 specimens were found to be false positives, resulting in a final specificity of 98.1% (95% CI 97.8–98.4%). For the Ortho HCV, 43 specimens tested as falsely positive, resulting in a final specificity of 98.2% (95% CI 97.9–98.5%) (Table 4). Specimens from individuals with medical conditions unrelated to HCV and containing potentially interfering substances were tested with both methods. These specimens consisted of 20 samples from patients with hepatitis A infection, 70 with hepatitis B infection, 10 with hepatitis E infection, 18 from patients with systemic lupus erythematosus, 19 from patients with primary biliary cirrhosis and 13 sera from pregnant women. Both the EIAgen HCV test and Ortho HCV test gave an apparent specificity of 100%. 3.4. Hemolysis or sera with lipidemia Specimens from patients with hemolysis or lipidemia were also tested by both tests. Hemolyzed HCV-negative specimens had free hemoglobin concentrations ranging from 2.5 to 10.0 g/L, while the 20 lipidemia specimens had chyle indexes ranging from 16 to 32. As shown in Table 5, all specimens were nonreactive by both assays, giving a specificity of 100%. 3.5. Signal-to-cut-off ratio (S/CO) When the subgroup of 178 true-positive samples was considered, S/CO ratios of the EIAgen HCV test ranged from 1.1 to 25.0 while those of the Ortho HCV test ranged from 1.1 to 5.0. Among specimens with an Ortho HCV test S/CO result of ≥3.8, 88.8% were also ≥5.9 with the EIAgen HCV test. Of the EIAgen HCV test S/CO results, 90.5% (161/178) of ratios were above 5.9, compared with 81.5% (145/178) above 3.8 for the Ortho HCV test (Table 6).
The EIAgen HCV test is a diagnostic EIA intended for screening patients with HCV infection using qualitative detection of anti-HCV IgG and IgM. The Ortho HCV test is a commercial assay, also using EIA, for qualitative detection of anti-HCV antibodies. The performance of the EIAgen HCV test with the current Ortho HCV test was evaluated using a series of specimens collected in clinical settings. The main criteria for an HCV screening assay are to attain the highest sensitivity possible in combination with excellent specificity. The challenge of any screening assay is to enhance sensitivity while not compromising assay specificity. In countries where HCV NAT testing for blood donor screening is not mandatory, an antiHCV assay serves as the only test in detecting of blood with HCV. The importance of detecting anti-HCV antibodies is stressed further by the fact that a significant number of specimens were demonstrated to be anti-HCV-positive but NAT negative when tested using a minipool NAT (Janatpour et al., 2002; Busch et al., 2003). It should be recognized that the level of virus declines and fluctuates widely in patients following seroconversion, or when they are in remission and treated with antiviral drugs (Kuramoto et al., 2002). Specimens from these patients may be negative transiently for viral nucleic acid and/or viral antigens but might still be infectious despite the presence of HCV antibodies. Therefore, the accuracy of an HCV screening assay is important. A loss of sensitivity is not acceptable, because this affects blood donor safety and delays the treatment of patients infected with HCV. In our hands, the sensitivity of the EIAgen HCV test was 100%, which was higher than the Ortho HCV test (98.3%). There are various explanations. First, the EIAgen HCV test can detect IgM in addition to IgG antibodies, whereas the Ortho HCV test can detect only IgG antibodies. Second, formulations used to prepare the EIA (the microplate wells and conjugate) might confer a better integrity to the antigens used in the EIAgen HCV test. Third, the EIAgen HCV test might use antigens with higher sequence homology to the most prevalent HCV strains found in China. This latter hypothesis seems improbable since all the HCV genotypes examined in this study (genotypes 1b, 2a, 1a, 3 and 6) were detected by both assays. False-positive results also cause distress and confusion, because the clinical significance of these results is not known or, if known, may not be understood by the donor. During the 1990s, several reports raised concerns over this problem (Sakugawa et al., 1995; Bar-Shany et al., 1996; Ownby et al., 1997; Sharma et al., 2003). In certain clinical settings, false-positive anti-HCV results are rare because the majority of persons being tested have evidence of liver disease and the sensitivity and specificity of the screening assays
Table 5 EIAgen HCV test results with hemolyzed specimens. Degree of hemolysis (free Hb g/L)
Control (0.0)
Mild (2.5)
Medium (5.0)
Severe (10.0)
True positives True negatives False positives False negatives
0 19 0 0
0 19 0 0
0 19 0 0
0 19 0 0
Total Specificity (%)
19 100
19 100
19 100
19 100
Table 6 Proportion of S/CO ratios for the EIAgen HCV test and Ortho HCV test. Ortho HCV test
EIAgen HCV test S/CO ≥ 5.9
Total S/CO < 5.9
S/CO ≥ 3.8 S/CO < 3.8
143 (88.8%) 18 (11.2%)
2 (12%) 15 (88%)
145 (81.5%) 33 (18.5%)
Total
161 (90.5%)
17 (10%)
178
H.-Y. Rao et al. / Journal of Virological Methods 162 (2009) 203–207
are high. However, among populations with a low (<10%) prevalence of HCV infection, false-positive results do occur (Hyams et al., 2001). Among immunocompetent populations with an antiHCV prevalence lower than 10%, the proportion of false-positive results with the Ortho HCV test average about 35% (range: 15–60%) (Alter et al., 2003), although the specificity is ≥99%. With regard to the populations tested, the specificity of the EIAgen HCV test was 98.1%, compared with 99.2% for the Ortho HCV test. This level is considered excellent, based on the standard requirement of 99.5% for regulatory approval of HCV screening tests throughout the world. However, specimens with low S/CO ratios (from 1.0 to 3.0) as determined by the EIAgen assay should be verified using an independent supplemental test with high specificity. None of the sera from patients with hepatitis A, B, E, systemic lupus erythematosus, primary biliary cirrhosis or pregnant women) interfered with the assays. The potential interfering effect of hemolysis and lipidemia was also examined. All specimens with various concentrations of hemoglobin and lipids were found to be true negative, which resulted in complete specificity of the two EIAs. Analysis of early versions of anti-HCV EIA results tested in volunteer blood donors indicated that the average repeatedly reactive S/CO ratios could be used to predict positive results from supplemental testing (Tobler et al., 2000). The USA Center for Disease Control determined which specific S/CO ratio would predict a true antibody positive result >95% of the time, regardless of the prevalence of anti-HCV antibodies or the characteristics of the population tested. In the case of the Ortho HCV test, an average S/CO ratio >3.8 appeared to be highly predictive of the true anti-HCV status (Alter et al., 2003). These results were confirmed in Chinese blood donors (Ren et al., 2005). In the current study, this same cut-off value of 3.8 yielded a proportion of 81.5% for the Ortho HCV test, confirming these previous observations. This proportion increased to 90.5% for the EIAgen HCV test using a cut-off level of 5.9. Within that group of specimens with an EIAgen HCV test S/CO ratio ≥5.9, 88.8% had an Ortho HCV test S/CO ratio ≥3.8. Reciprocally, among the truepositive specimens with an EIAgen HCV test S/CO ratio <5.9, 88.2% had an Ortho HCV test S/CO ratio <3.8. Therefore, specimens with average S/CO ratios >5.9 in the EIAgen HCV test would be highly predictive of true HCV-positive status. The EIAgen HCV test is well suited for screening of blood and blood products. Acknowledgments This work was partially supported by grants from the Chinese Basic Research Foundation 973 (2005CB522902 and 2007CB512900), 863 program (2006AA02A410), NSFC (30571639), the National Science and Technology Key Project during the 11th Five-Year Plan Period (2008ZX10002-012 and 2008ZX10002-013) and Key Clinical Research Program of Ministry of Health. References Afdhal, N.H., 2004. The natural history of hepatitis C. Semin. Liver Dis. 24, S3–S8.
207
Allain, J.P., 2005. Hepatitis C virus in blood donation. Lancet 365, 276– 278. Alter, M.J., Kuhnert, W.L., Finelli, L., 2003. Guidelines for laboratory testing and result reporting of antibody to hepatitis C virus. M. M. W. R. 52, 1–15. Bar-Shany, S., Green, M.S., Shinar, E., 1996. False positive tests for anti-hepatitis C antibodies and the problem of notifying blood donors. Int. J. Epidemiol. 25, 674–678. Busch, M.P., Tobler, L.H., Gerlich, W.H., Schaefer, S., Giachetti, C., Smith, R., 2003. Very low level viremia in HCV infectious unit missed by NAT. Transfusion 43, 1173–1174. Centers for Disease Control and Prevention, 1991. Public health service inter-agency guidelines for screening donors of blood, plasma, organ, tissues and semen for evidence of hepatitis B and hepatitis C. M. M. W. R. 40, 1–17. Centers for Disease Control and Prevention, 1998. Recommendations for prevention and control of hepatitis C virus (HCV) infection and HCV-related chronic disease. M. M. W. R. 47, 1-39. Choo, Q.L., Kuo, G., Weiner, A.J., Overby, L.R., Bradley, D.W., Houghton, M., 1989. Isolation of a cDNA clone derived from a blood-borne non-A, non-B viral hepatitis genome. Science 244, 359–362. Di Bisceglie, A.M., 2000. Natural history of hepatitis C: its impact on clinical management. Hepatology 31, 1014–1018. Donahue, J.G., Munoz, A., Ness, P.M., Brown Jr., D.E., Yawn, D.H., McAllister Jr., H.A., Reitz, B.A., Nelson, K.E., 1992. The declining risk of post-transfusion hepatitis C virus infection. N. Engl. J. Med. 327, 369–373. Du, S.C., Zhang, R., Li, J.Q., Wei, L., 2007. Utilization of uracil-DNA glycosylase for combining reverse transcription and anti-contamination with polymerase chain reaction in hepatitis C virus. J. Peking Univ. [Health Sci.] 39, 426–428. Hoofnagle, J.H., 2002. Course and outcome of hepatitis C. Hepatology 36, S21– S29. Hyams, K.C., Riddle, J., Rubertone, M., Trump, D., Alter, M.J., Cruess, D.F., Han, X., Nainam, O.V., Seeff, L.B., Mazzuchi, J.F., Bailey, S., 2001. Prevalence and incidence of hepatitis C virus infection in the US military: a seroepidemiologic survey of 21,000 troops. Am. J. Epidemiol. 153, 764–770. Janatpour, K.A., Kemper, M., Aceituno, S., Holland, P.V., 2002. Incidence of minipool HCV NAT negative, single unit NAT positive units among serological positive blood donations. Transfusion 42, 80S. Kuramoto, I.K., Moriya, T., Schoening, V., Holland, P.V., 2002. Fluctuation of serum HCV-RNA levels in untreated blood donors with chronic hepatitis C virus infection. J. Virol. Hepatitis 9, 36–42. Ownby, H.E., Korelitz, J.J., Busch, M.P., Williams, A.E., Kleinman, S.H., Gilcher, R.O., Nourjah, P., Retrovirus Epidemiology Donor Study, 1997. Loss of volunteer blood donors because of unconfirmed enzyme immunoassay screening results. Transfusion 37, 199–205. Ren, F.R., Lv, Q.S., Zhuang, H., Li, J.J., Gong, X.Y., Gao, G.J., Liu, C.L., Wang, J.X., Yao, F.Z., Zhen, Y.R., Zhu, F.M., Tiemuer, M.H., Bai, X.H., Shan, H., 2005. Significance of the signal-to-cutoff ratios of anti-hepatitis C virus enzyme immunoassays in screening of Chinese blood donors. Transfusion 45, 1816–1822. Sakugawa, H., Nakasone, H., Nakayoshi, T., Kinjo, F., Saito, A., Yakabi, S., Zukeran, H., Miyagi, Y., Taira, R., Koja, K., 1995. High proportion of false positive reactions among donors with anti-HCV antibodies in a low prevalence area. J. Med. Virol. 46, 334–338. Sharma, U.K., Stramer, S.L., Wright, D.J., Glynn, S.A., Hermansen, S., Schreiber, G.B., Kleinman, S.H., Busch, M.P., Retrovirus Epidemiology Donor Study, 2003. Impact of changes in viral marker screening assays. Transfusion 43, 202–214. Tobler, L.H., Tegtmeier, G., Stramer, S.L., Quan, S., Dockter, J., Giachetti, C., Busch, M.P., 2000. Lookback on donors who are repeatedly reactive on first-generation hepatitis C virus assays: justification and rational implementation. Transfusion 40, 15–24. Tobler, L.H., Stramer, S.L., Lee, S.R., Masecar, B.L., Peterson, J.E., Davis, E.A., Andrews, W.E., Brodsky, J.P., Kleinman, S.H., Phelps, B.H., Busch, M.P., 2003. Impact of HCV 3.0 EIA relative to HCV 2.0 EIA on blood-donor screening. Transfusion 43, 1452–1459. Vandelli, C., Renzo, F., Romano, L., Tisminetzky, S., De Palma, M., Stroffolini, T., Ventura, E., Zanetti, A., 2004. Lack of evidence of sexual transmission of hepatitis C among monogamous couples: results of a 10-year prospective follow-up study. Am. J. Gastroenterol. 99, 855–859. Wei, L., Wang, Q.X., Xu, X.Y., Wan, H., Gao, Y., Tian, X.L., Yu, M., Sun, D.G., Fan, C.L., Jin, J., Fan, W.M., Yin, L.M., Zhu, W.F., Chen, H.S., Zhuang, H., Wang, Y., 2002. 12–25year follow-up of hepatitis C virus infection in a rural area of Hebei province. China J. Peking Univ. [Health Sci.] 34, 574–578.