Sensitivity of second-generation enzyme immunoassay for detection of hepatitis C virus infection among oncology patients

Results A total of 264 samples were included in the study. We identified 92 HCV infections, 76 of which were Abbott EIA 2 positive. Abbott EIA 2 sensitivity was 83% (76/92), lower than that reported among healthy adults (90%) ( p = 0.01) and poor sensitivity was associated with receipt of chemotherapy during the outbreak period ( p = 0.02). Only 1 (6%) of the 16 EIA 2-negative cases had elevated ALT. Conclusions In this study, EIA 2 sensitivity among oncology patients was lower than that previously reported among immunocompetent persons. Impaired antibody production related to cancer and/or chemotherapy might explain the reduced sensitivity. These findings indicate that, when assessing HCV status in oncology patients, a NAT test should be routinely considered in addition to EIA. Abbreviations ALT alanine aminotransferase CIA chemiluminescence immunoassay EIA enzyme immunoassay HCV hepatitis C virus NAT nucleic acid testing PCR polymerase chain reaction RIBA recombinant immunoblot assay TMA transcription-mediated amplification Keywords Hepatitis C virus Sensitivity Immunosuppression Oncology Chemotherapy 1 Introduction Enzyme immunoassays (EIAs), both second- and third-generations, are antibody-detection tests recommended for diagnosing hepatitis C virus (HCV) infection in the United States ( Alter et al., 2003 ). Since these tests depend on antibody levels at or above test detection limits, sensitivity among persons with low antibody levels because of recent infection or immunosupression may be decreased. In addition, antibody tests do not distinguish active infection, characterized by the presence of HCV ribonucleic acid (RNA) detected by nucleic acid testing (NAT), from resolved infection ( Erensoy, 2001 ). Because of the possibility of EIA–false-positive results, more specific tests (e.g., recombinant immunoblot assay [RIBA] and NAT) are recommended to confirm positive results. EIA-negative results are considered final and no additional tests are recommended unless recent infection is suspected or in immunosuppressed persons when other evidence of HCV infection (e.g., elevated alanine aminotransferase [ALT] level) exists ( Alter et al., 2003 ). Previous studies have shown poor EIA sensitivity among persons undergoing hemodialysis and transplant recipients, because of the immunosuppression associated with these conditions ( Feucht et al., 1995; Schneeberger et al., 1998; Pawlotsky, 1999 ). For these patients, HCV infection diagnosis can be made using a combination of NAT and EIA. NAT includes qualitative and quantitative reverse transcriptase polymerase chain reaction (PCR) and the newly developed transcription-mediated amplification (TMA) assay ( Giachetti et al., 2002; Gorrin et al., 2003 ). Despite some reports of HCV EIA poor performance in oncology patients, its sensitivity in this group of persons remains unknown ( Paydas et al., 2003 ). In this study, we evaluated the sensitivity of a second-generation EIA (EIA 2) to detect HCV infection in oncology patients. 2 Methods 2.1 Subject selection We invited oncology patients from a hematology/oncology clinic in eastern Nebraska to take part in this study. During March 2000–June 2001, an HCV outbreak occurred at this clinic due to cross-contamination of shared saline bags ( Centers for Disease Control and Prevention, 2003; Macedo de Oliveira et al., 2005 ). Variables potentially associated with antibody-detection tests performance such as sex, age, type of cancer and receipt of chemotherapy during the outbreak period were collected from clinic medical records. 2.2 Sample collection Blood was collected in October 2002, 16 months after the end of the outbreak, which is beyond the seronegative window period for HCV EIA. After blood collection in 7 mL SST Vacutainers ® , serum was separated by centrifugation within 4 h. We aliquoted the serum in 2 mL tubes and stored at −20 °C until processing. All specimen handling and testing were performed in accordance with manufacturer's guidelines, as described below. 2.3 Serological testing HCV serological testing was done using EIA 2 (Abbott HCV EIA 2.0, Abbott Laboratories, Abbott Park, Illinois). This assay consists of a microtiter plate with viral antigens embedded in the wells. Antibodies present in patients’ sera adhere to the well and anti-immunoglobulins containing a colorimetric marker are added to allow antibody detection ( Carithers et al., 2000 ). In samples with two Abbott EIA 2-positive repeats, HCV infection was confirmed using qualitative PCR (Cobas Amplicor™ HCV Test, v2.0; Roche Molecular Diagnostic Systems, Branchburg, New Jersey) or, in case of qualitative PCR-negative samples, RIBA (Chiron RIBA ® HCV 3.0 SIA, Chiron Corporation, Emeryville, California). 2.4 TMA testing We tested all samples using TMA (Gen-Probe Incorporated, San Diego, California). This assay involves three steps within a single tube: sample preparation with target capture, HCV RNA target amplification by TMA and detection of the amplification products by hybridization. For verification purposes, an internal control is added to each reaction ( Giachetti et al., 2002 ). 2.5 Abbott EIA 2-negative/TMA-positive samples We further tested the Abbott EIA 2-negative/TMA-positive specimens using enhanced chemiluminescence immunoassay (CIA) (Vitros ® Anti-HCV Assay, Ortho-Clinical Diagnostics, Raritan, New Jersey). This test is one of the third-generation HCV immunoassays, which have reported sensitivity of at least 97% ( Colin et al., 2001 ). For these samples, we confirmed the TMA results using quantitative PCR (Cobas Amplicor™ HCV Monitor Test, v2.0; Roche Molecular Diagnostic Systems, Branchburg, New Jersey) and determined serum ALT levels (Roche Diagnostics Corporation, Indianapolis, Indiana). 2.6 Case definition HCV infection was defined as a positive RIBA or TMA test in an oncology patient. 2.7 Statistical analysis We used the z -test to compare Abbott EIA 2 sensitivity with previous reports in the literature. Abbott EIA 2 sensitivity was also stratified by sex, age, type of cancer and use of chemotherapy during the outbreak period and was evaluated using the Fisher exact and the chi square tests, as appropriate. 3 Results 3.1 Subject demographics Of the 472 patients seen at the clinic during the outbreak period, 269 (57%) oncology patients agreed to be tested. Five of those were excluded from analysis: three with previous HCV treatment, one with insufficient material for complete testing and one with indeterminate laboratory results. Among the 264 patients included in the analysis, the mean age was 67 years and 183 (69%) were females. The most common cancer diagnosis was breast cancer (114 [43%] patients), followed by colon cancer (55 [21%] patients) and leukemia/lymphoma (37 [14%] patients). One hundred and two (39%) patients received chemotherapy during the outbreak period. 3.2 HCV test results We identified 92 (35%) cases among the 264 valid samples. Seventy-four cases were positive by Abbott EIA 2, qualitative PCR and TMA, and 16 cases were Abbott EIA 2-negative and TMA-positive ( Fig. 1 ). Additionally, two of the 92 cases were EIA 2-positive, TMA-negative and qualitative PCR-negative. In these two cases, we confirmed the EIA 2-positive result by RIBA. 3.3 Abbott EIA 2-negative/TMA-positive samples To exclude the possibility of TMA false positivity, the 16 Abbott EIA 2-negative/TMA-positive samples were retested with quantitative PCR and all yielded positive results. Four (25%) of these samples were CIA-positive. We determined ALT levels in these 16 specimens, and 15 (94%) were within the normal range (<31 IU/mL). 3.4 Abbott EIA 2 sensitivity In this study, Abbott EIA 2 sensitivity was 83% (76/92), lower than that in previous studies (90%), which were based on immunocompetent persons ( p = 0.01) ( Whyte and Beal, 1995 ). All 16 Abbott EIA 2–false-negative results were seen among patients who received chemotherapy during the outbreak period. The difference in Abbott EIA 2 sensitivity between patients receiving and not receiving chemotherapy during the outbreak period (78% versus 100%, respectively) was statistically significant ( p = 0.02) ( Table 1 ). Abbott EIA 2 sensitivity did not statistically differ when stratified by sex, age, or type of cancer. 4 Discussion Abbott EIA 2 sensitivity was estimated at 83% among oncology patients from a Nebraska clinic where an HCV outbreak occurred ( Centers for Disease Control and Prevention, 2003; Macedo de Oliveira et al., 2005 ). The 16 patients with Abbott EIA 2–false-negative results had HCV infection demonstrated by NAT. Previous studies have shown lower sensitivity of HCV EIA in persons immunosuppression due to hemodialysis and organ transplantation, cases of acute HCV infection and infants born to HCV-infected mothers ( Bukh et al., 1993; Soffredini et al., 2004 ). However, few studies have assessed the sensitivity of HCV immunoassays among patients with cancer or receiving chemotherapy. The present study indicates that EIA 2 sensitivity in oncology patients is lower than that reported among immunocompetent persons ( Whyte and Beal, 1995; Abdel-Hamid et al., 2002 ). A more sensitive immunoassay, CIA, was positive in 4 (25%) of the 16 EIA 2–false-negative cases. EIA 2-positive samples ( n = 76) were not tested by CIA, but assuming that all these cases were also CIA-positive, we can estimate a maximum CIA sensitivity of 87% (80/92). This performance is below the sensitivity of third-generation HCV immunoassays among immunocompetent persons and HIV-infected patients (>99%) ( Abdel-Hamid et al., 2002; Thio et al., 2000 ). One limitation inherent in the analysis of HCV test sensitivity is the lack of a consensus regarding the gold standard for HCV infection ( Pawlotsky, 1999 ). In this investigation, we used a combination of serological (RIBA) and NAT (TMA) assays as the gold standards, a strategy adopted in previous studies ( Janot et al., 1994; Colin et al., 2001 ). Decreased EIA 2 sensitivity was associated with receipt of chemotherapy. We were unable to distinguish the relative contributions of cancer and chemotherapy to decreased EIA 2 sensitivity because all false-negative results were seen among patients undergoing chemotherapy and few patients without chemotherapy exposure ( n = 20) were available for comparison. Future studies documenting HCV-antibody response among oncology patients are needed. Sex, age and type of cancer were not significantly associated with EIA 2 performance; however, our study might have lacked the power to find such association if it existed. Cancer and its treatment may cause cellular and humoral immune deficiencies, but the impact on HCV EIA performance is unknown. Previous reports have shown reduced EIA sensitivity for diagnosing adenovirus infection among bone marrow transplant recipients ( Raboni et al., 2003 ). Additionally, other studies have shown decreased vaccine-induced antibody levels after chemotherapy and recommended patient revaccination ( Reinhardt et al., 2003 ). It follows that among oncology patients, cancer- and/or chemotherapy-induced immunosupression might have contributed to low levels of HCV antibodies and the reduced EIA sensitivity seen in our study. Three generations of HCV immunoassays have been developed to detect circulating HCV antibodies. The first-generation HCV EIA was developed in 1989 but had poor sensitivity and specificity ( Baath et al., 1992 ). The second- and third-generation assays improved sensitivity to 90% and 97%, respectively, and are equally recommended as screening tests in the United States ( Whyte and Beal, 1995; Abdel-Hamid et al., 2002; Colin et al., 2001 ). According to current guidelines, EIA-negative results are considered final, except when recent infection is suspected or in immunosuppressed persons with a high index of suspicion for HCV infection, such as abnormal ALT levels ( Alter et al., 2003 ). In this study, 15 (94%) of the 16 Abbott EIA 2–false-negative cases had ALT levels within the normal range and would have gone undiagnosed if NAT had not been performed. We recognize that, based on previously published reports, clinicians already employ a combination of EIA and NAT testing to assess HCV status in oncology patients ( Paydas et al., 2003 ). However, because of the limited number of study subjects, these reports could not offer definitive conclusions about EIA sensitivity and appropriate testing algorithms within this patient population. In summary, our study suggests that EIA sensitivity among oncology patients is low, which limits the role of this assay as a screening test. The ultimate importance of HCV testing is to identify patients with active infection, i.e. circulating HCV RNA, and provide treatment. However, exclusive use of NAT is also suboptimal, because it will not identify patients with resolved HCV infection (EIA-positive and NAT-negative patients), who should be identified, regularly monitored and treated if their disease recrudesces ( Lauer et al., 2001 ). Therefore, both EIA and NAT should be routinely performed in all oncology patients when assessing HCV status. Acknowledgements We are indebted to Alice Fournell, Erica Hamilton and Andrew Stuart for logistic coordination and data entry. For special assistance, we would like to thank Thomas McKnight, M.D.; Jean Schafersman, R.N.; the testing clinic staff and most of all the clinic patients. References Abdel-Hamid et al., 2002 M. Abdel-Hamid M. El Daly S. El Kafrawy N. Mikhail G.T. Strickland A.D. Fix Comparison of second- and third-generation enzyme immunoassays for detecting antibodies to hepatitis C virus J Clin Microbiol 40 5 2002 1656 1659 Alter et al., 2003 M.J. Alter W.L. Kuhnert L. Finelli Guidelines for laboratory testing and result reporting of antibody to hepatitis C virus centers for disease control and prevention MMWR Recomm Rep 52 RR-3 2003 1 13 Baath et al., 1992 L. Baath A. Widell E. Nordenfelt A comparison between one first generation and three second generation anti-HCV ELISAs: an investigation in high- and low-risk subjects in correlation with recombinant immunoblot assay and polymerase chain reaction J Virol Methods 40 3 1992 287 296 Bukh et al., 1993 J. Bukh P. Wantzin K. Krogsgaard F. Knudsen R.H. Purcell R.H. Miller Copenhagen Dialysis HCV Study Group High prevalence of hepatitis C virus (HCV) RNA in dialysis patients: failure of commercially available antibody tests to identify a significant number of patients with HCV infection J Infect Dis 168 6 1993 1343 1348 Carithers et al., 2000 R.L. Carithers Jr. A. Marquardt D.R. Gretch Diagnostic testing for hepatitis C Semin Liver Dis 20 2 2000 159 171 Centers for Disease Control and Prevention, 2003 Centers for Disease Control and Prevention. Transmission of hepatitis B and C viruses in outpatient settings—New York Oklahoma, and Nebraska, 2000–2002. MMWR Morb Mortal Wkly Rep 2003;52(38):901–906. Colin et al., 2001 C. Colin D. Lanoir S. Touzet L. Meyaud-Kraemer F. Bailly C. Trepo Sensitivity and specificity of third-generation hepatitis C virus antibody detection assays: an analysis of the literature J Viral Hepat 8 2 2001 87 95 Erensoy, 2001 S. Erensoy Diagnosis of hepatitis C virus (HCV) infection and laboratory monitoring of its therapy J Clin Virol 21 3 2001 271 281 Feucht et al., 1995 H.H. Feucht B. Zollner S. Polywka R. Laufs Study on reliability of commercially available hepatitis C virus antibody tests J Clin Microbiol 33 3 1995 620 624 Giachetti et al., 2002 C. Giachetti J.M. Linnen D.P. Kolk J. Dockter K. Gillotte-Taylor M. Park Highly sensitive multiplex assay for detection of human immunodeficiency virus type 1 and hepatitis C virus RNA J Clin Microbiol 40 7 2002 2408 2419 Gorrin et al., 2003 G. Gorrin M. Friesenhahn P. Lin M. Sanders R. Pollner B. Eguchi Performance evaluation of the VERSANT HCV RNA qualitative assay by using transcription-mediated amplification J Clin Microbiol 41 1 2003 310 317 Janot et al., 1994 C. Janot A.M. Courouce F. Barin F. Lunel-Fabiani C. Trepo C. Botte Screening tests of anti-HVC antibodies used in France Analysis of sensitivity Transfus Clin Biol 1 4 1994 295 301 Lauer et al., 2001 G.M. Lauer B.D. Walker Hepatitis C virus infection N Engl J Med 345 1 2001 41 52 Macedo de Oliveira et al., 2005 A. Macedo de Oliveira K.L. White D.P. Leschinsky B.D. Beecham T.M. Vogt R.L. Moolenaar An outbreak of Hepatitis C virus infections among outpatients at a hematology/oncology clinic Ann Int Med 142 2005 898 902 Pawlotsky, 1999 J.M. Pawlotsky Diagnostic tests for hepatitis C J Hepatol 31 Suppl. 1 1999 71 79 Paydas et al., 2003 S. Paydas B. Kilic S. Yavuz U. Disel K. Tanriverdi B. Sahin Anti-HCV and HCV-RNA prevalence and clinical correlations in cases with non-Hodgkin's lymphoma Am J Hematol 74 2 2003 89 93 Raboni et al., 2003 S.M. Raboni M.M. Siqueira S.R. Portes R. Pasquini Comparison of PCR, enzyme immunoassay and conventional culture for adenovirus detection in bone marrow transplant patients with hemorrhagic cystitis J Clin Virol 27 3 2003 270 275 Reinhardt et al., 2003 D. Reinhardt K. Houliara A. Pekrun M. Lakomek B. Krone Impact of conventional chemotherapy on levels of antibodies against vaccine-preventable diseases in children treated for cancer Scand J Infect Dis 35 11–12 2003 851 857 Schneeberger et al., 1998 P.M. Schneeberger I. Keur W. van der Vliet K. van Hoek H. Boswijk A.M. van Loon Hepatitis C virus infections in dialysis centers in The Netherlands: a national survey by serological and molecular methods J Clin Microbiol 36 6 1998 1711 1715 Soffredini et al., 2004 R. Soffredini M.G. Rumi M.L. Parravicini G. Ronchi E. Del Ninno A. Russo Serum levels of hepatitis C virus core antigen as a marker of infection and response to therapy Am J Gastroenterol 99 9 2004 1738 1743 Thio et al., 2000 C.L. Thio K.R. Nolt J. Astemborski D. Vlahov K.E. Nelson D.L. Thomas Screening for hepatitis C virus in human immunodeficiency virus-infected individuals J Clin Microbiol 38 2000 575 577 Whyte and Beal, 1995 G. Whyte R. Beal Screening—sensitivity, specificity and hepatitis C Med J Aust 163 2 1995 63 64