- Email: [email protected]
Determination of EBV serostatus prior to kidney transplantation: Comparison of VIDAS®, LIAISON® and immunofluorescence assays
Journal of Virological Methods 203 (2014) 107–111
Contents lists available at ScienceDirect
Journal of Virological Methods journal homepage: www.elsevier.com/locate/jviromet
Determination of EBV serostatus prior to kidney transplantation: Comparison of VIDAS® , LIAISON® and immunofluorescence assays I. Johannessen a,∗ , M. Noel a , A. Galloway a , S. Black a , M. Shearman a , C. Graham b a b
Department of Laboratory Medicine (Virology), Royal Infirmary of Edinburgh, Little France, Edinburgh EH16 4SA, UK Epidemiology and Statistics Core, Wellcome Trust Clinical Research Facility, Western General Hospital, Edinburgh EH4 2XU, UK
a b s t r a c t Article history: Received 16 June 2013 Received in revised form 18 March 2014 Accepted 24 March 2014 Available online 2 April 2014 Keywords: Epstein–Barr Virus Serostatus Kidney transplantation Post-transplant lymphoproliferative disease Enzyme immunosorbent assay Immunofluorescent assay
Immunosuppression following solid organ transplantation reduces T cell-mediated immune control of Epstein–Barr Virus (EBV), which may then drive development of post-transplant lymphoproliferative disease. Serology plays a key role in determination of risk of outgrowth of such lesions following transplantation. The study compared the VIDAS® (bioMérieux) and LIAISON® (DiaSorin) enzyme immunoassays (EIAs) and immunofluorescence assays (IFA; MBL-Bion) in the kidney transplantation setting. Sera from 100 live kidney donors [51 males; age range 20–82 years (mean 51.2 years)] and 100 cadaveric kidney recipients [70 males; age range 17–77 years (mean 51.0 years)] were tested. Overall proportional agreement ranged from 96% to 100% for VIDAS® and LIAISON® . Sensitivity ranged from 91% to 100% and 92% to 100% for VIDAS® /IFA and LIAISON® /IFA, respectively. The VIDAS® and LIAISON® approaches gave similar results. Such automated random access EIAs are well suited to busy clinical virology laboratories and rapid determination of donor and recipient EBV serostatus prior to transplantation. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Epstein–Barr Virus (EBV) infects approximately 90% of adults worldwide and persists for life in host B cells. Primary infection usually occurs in childhood and is generally asymptomatic at that time, but later acquisition of EBV may lead to clinical manifestations of infectious mononucleosis (Macsween et al., 2010). T cell-mediated immunity controls persistent EBV infection and immunosuppression tips the balance in favour of the virus, which expresses growth-promoting oncogenes (Long et al., 2011). In solid organ transplant recipients, this may lead to uncontrolled proliferation of virus-infected B cells and, in up to 10% of organ graft recipients, culminate in post-transplant lymphoproliferative disease (Crawford et al., 1980; Bollard et al., 2012). The two main risk factors for such disease are (1) EBV seronegativity prior to solid organ transplant surgery and (2) level and duration of immunosuppression (Thomas et al., 1991). Delineation of donor and recipient EBV serostatus prior to solid organ transplantation forms part of UK and European guidelines for transplant assessments. Although EBV carriage is very common
∗ Corresponding author. Tel.: +44 131 242 6003. E-mail address: [email protected] (I. Johannessen). http://dx.doi.org/10.1016/j.jviromet.2014.03.022 0166-0934/© 2014 Elsevier B.V. All rights reserved.
in adults, serological results may highlight susceptibility to EBV infection following transplantation surgery at a time when the recipient is ill equipped to control infection – a high risk situation for development of EBV-associated lymphoproliferative disease (Höcker et al., 2013; Morton et al., 2013). Historically, EBV immunofluorescence assay (IFA) has been used as gold standard for serological diagnostics. However, the assay is labour intensive and includes an element of operator subjectivity. Equally, it does not lend itself to automated approaches. Thus, the aim of this study was to compare VIDAS® (bioMérieux, France) and LIAISON® (DiaSorin, USA) enzyme immunosorbent assays (EIA) as well as IFA (MBL-Bion, USA) for determination of EBV serostatus prior to kidney transplantation. 2. Materials and methods 2.1. Study patients The study comprised of a panel of 200 serum samples obtained pre-kidney transplantation from unrelated live renal donors (n = 100) and cadaveric renal recipients (n = 100). One hundred serum samples from live kidney transplant donors [51 male; age range 20–82 years (mean (sd) 51.2 (12.2) years)] were tested for VCA IgG and EBNA IgG in all three assay systems (Table 1). One
108
I. Johannessen et al. / Journal of Virological Methods 203 (2014) 107–111
Table 1 Sensitivity and proportional agreement data when assessing kidney donors. The following tables show the number of all positive and negative results with LIAISON® vs. VIDAS® [first column of tables], LIAISON® vs. IFA [middle column of tables] and VIDAS® vs. IFA [last column of tables]; in each table, the sensitivity and the proportional agreement are presented along with a 95% CI. VIDAS®
VCAG
N
VCAG P
®
LIAISON N 1 0 1 0 99 99 P 1 99 100 All Sensitivity (LIAISON vs. VIDAS) 100% (Lower 95%CI 97%) Proportional agreement 100% (Lower 95%CI 97%) EBNA
VIDAS® N
P
IFA
All
P
IFA
All
®
N
P
All
®
LIAISON N 1 0 1 P 1 98 99 All 2 98 100 Sensitivity (LIAISON vs. IFA) 100% (Lower 95%CI 97%) Proportional agreement 99% (95% CI (95, 100))
VIDAS N 1 0 1 P 1 98 99 All 2 98 100 Sensitivity (VIDAS vs. IFA) 100% (Lower 95% CI 96) Proportional agreement 99% (95% CI (95, 100))
EBNA
EBNA
All
LIAISON® N 3 2 5 0 95 95 P 3 97 100 All Sensitivity (LIAISON vs. VIDAS) 98% (95% CI (93, 100)) Proportional agreement 98% (95% CI (93, 100))
VCAG
N
IFA N
P
All
LIAISON® N 1 4 5 P 2 93 95 All 3 97 100 Sensitivity (LIAISON vs. IFA) 96% (95% CI (90, 99)) Proportional agreement 94% (95% CI (87, 98))
IFA N
P
All
VIDAS® N 1 2 3 P 2 95 97 All 3 97 100 Sensitivity (VIDAS vs. IFA) 98% (95% CI (93, 100)) Proportional agreement 96% (95% CI (90, 99))
P: positive; N: negative; IFA: immunofluorescence assay; VCAG: VCA IgG; EBNA: EBNA IgG.
hundred cadaveric kidney transplant recipients [70 male; age range 17–77 years (mean (sd) 51.0 (14.5) years)] were tested for both parameters in all three assay systems (Table 2). Blood samples tested were retrospective (archived left over) serum samples that had been stored at −20 ± 6 ◦ C for <2 years. 2.2. Ethical considerations Consent was obtained for EBV testing at time of sampling, and the study (classified as service development by local ethics authority) re-analysed anonymised residual blood tube specimens left over from that time. 2.3. Samples Serum samples [stored at −20 ◦ C (±6 ◦ C) for <2 years] were homogenised and tested singly in each assay. Insufficient samples were excluded from analysis; equivocal results were assigned a positive (‘reactive’) status. Due to volume restriction, samples were not re-tested. The study did not ascertain whether or not
additional samples were available from the same individual(s) for further/extended tests. 2.4. Experimental plan The following qualitative assays were used in accordance with the manufacturers’ instructions: (1) VIDAS® Automated System (bioMérieux, Lyon, France): ‘EBV VCA/EA IgG’ (ref. 30236) and ‘EBV EBNA IgG’ (ref. 30235) for detection of IgG antibodies (Ab) directed against EBV VCA (p18 synthetic peptide) and EA (p54 synthetic peptide), and IgG Ab directed against EBNA-1 (p72 synthetic peptide), respectively; (2) LIAISON® Automated System (DiaSorin, Saluggia Vercelli, Italy): ‘EBV VCA IgG’ (ref. 310510) and ‘EBV EBNA IgG’ (ref. 310520) for detection of IgG Ab directed against EBV VCA (p18 synthetic peptide), and IgG Ab directed against EBNA-1 (p72 synthetic peptide), respectively; (3) MBL-Bion Manual IFA System (Des Plaines, IL, USA): ‘EBV VCA Antigen Substrate Slide’ (P3HR1 EBV strain-infected lymphocytes; ref. EB-5012) and ‘EBV EBNA Antigen Substrate Slide’ (Raji EBV strain-infected lymphocytes; ref. NA-5112) were used in the context of indirect IF (IIF) employing anti-human conjugate detection Ab enabling detection of human
Table 2 Sensitivity and proportional agreement data when assessing kidney recipients. The following tables show the number of all positive and negative results with LIAISON® vs. VIDAS® [first column of tables], LIAISON® vs. IFA [middle column of tables] and VIDAS® vs. IFA [last column of tables]; in each table, the sensitivity and the proportional agreement are presented along with a 95% CI. VCAG
VIDAS® N
VCAG P
All
LIAISON® N 6 2 8 0 92 92 P 6 94 100 All Sensitivity (LIAISON vs. VIDAS) 98% (95% CI (93, 100)) Proportional agreement 98% (95% CI (93, 100)) EBNA
VIDAS® N
P
All
LIAISON® N 10 2 12 2 86 88 P 12 88 100 All Sensitivity (LIAISON vs. VIDAS) 98% (95% CI (92, 100)) Proportional agreement 96% (95% CI (90, 99))
IFA N
VCAG P
All
IFA N
P
All
LIAISON® N 2 6 8 P 0 92 92 All 2 98 100 Sensitivity (LIAISON vs. IFA) 94% (95% CI (87, 98))0.938776 Proportional agreement 94% (95% CI (87, 98))
VIDAS® N 2 4 6 P 0 94 94 All 2 98 100 Sensitivity (VIDAS vs. IFA) 96% (95% CI (90, 99)) Proportional agreement 96% (95% CI (90, 99))
EBNA
EBNA
IFA N
P
All
LIAISON® N 5 7 12 P 6 82 88 All 11 89 100 Sensitivity (LIAISON vs. IFA) 92% (95% CI (84, 97)) Proportional agreement 87% (95% CI (79, 93))
P: positive; N: negative; IFA: immunofluorescence assay; VCAG: VCA IgG; EBNA: EBNA IgG.
IFA N
P
All
VIDAS® N 4 8 12 P 7 81 88 All 11 89 100 Sensitivity (VIDAS vs. IFA) 91% (95% CI (83, 96)) Proportional agreement 85% (95% CI (76, 91))
I. Johannessen et al. / Journal of Virological Methods 203 (2014) 107–111
109
Table 3 Sensitivity and proportional agreement when equivocal data are excluded. LIAISON® vs. VIDAS®
LIAISON® vs. IFA
VIDAS® vs. IFA
Sensitivity Proportional agreement Sensitivity Proportional agreement
100% (lower 95%CI 96%) 100% (lower 95%CI 97%) 98% (95% CI (92, 100)) 98% (95% CI (93, 100))
100% (lower 95%CI 96%) 99% (95% CI (95, 100)) 97% (95% CI (91, 99)) 95% (95% CI (88, 98))
100% (lower 95% CI 96) 99% (95% CI (95, 100)) 98% (95% CI (92, 100)) 96% (95% CI (90, 99))
Sensitivity Proportional agreement Sensitivity Proportional agreement
98% (95% CI (93, 100)) 98% (95% CI (93, 100)) 100% (lower 95% CI 96%) 100% (lower 95% CI 97%)
95% (95% CI (88, 98)) 95% (95% CI (88, 98)) 91% (95% CI (83, 96)) 90% (95% CI (81, 95))
97% (95% CI (91, 99)) 97% (95% CI (91, 99)) 90% (95% CI (82, 96)) 85% (95% CI (76, 91))
Assay Kidney donors VCA IgG EBNA IgG Kidney recipients VCA IgG EBNA IgG
Table 4 Overview of discrepant results. LIAISON® vs. IFA
VIDAS® vs. IFA
Kideny donors 0 VCA IgG EBNA IgG 2b Total 2
1a 6b , c 7
1a 4c 5
Kidney recipients 2d VCA IgG EBNA IgG 4f , h Total 6
6d , e 13f , g 19
4e 15g , h 19
Assay
LIAISON® vs. VIDAS®
a
Same sample. Two shared samples. c Four shared samples. d Two shared samples. e Four shared samples. f One shared sample. g Twelve shared samples. h Three shared samples. IFA: immunofluorescence assay. b
IgG Ab directed against VCA (IIF), and IgG Ab directed against EBNA (anti-compliment IF; ACIF), respectively; ‘IF IgG TEST’ (ref. CXCG9972) detection system was used in the context of the substrate slides for detection of human IgG. 2.5. Statistical analysis Cross-tabulations of results were generated and proportional agreement and sensitivity were calculated as appropriate along with 95% confidence intervals (CIs). Due to the limited study sample numbers, it was not always possible to calculate both the upper and lower bounds of the 95% CI and where this is the case, only the bound calculated has been presented. Similarly, due to limited study sample numbers, it has not been possible to calculate specificity in a meaningful manner (although such calculations can be made from the raw data presented in Tables 1 and 2). McNemar’s exact test was used to assess proportional agreement between two (matched pairs; VCA and EBNA) IgG tests in Table 5. 3. Results For assessment of pre-kidney transplant EBV serostatus, VIDAS® EA/VCA IgG and EBNA IgG results were compared with LIAISON® and IFA VCA IgG and EBNA IgG results, respectively. The kidney graft groups were studied separately since they comprised of live donors and cadaveric organ recipients. Summary cross-tabulations of all results are shown in Tables 1 and 2. Importantly, exclusion of equivocal results (Table 3) did not affect study outcome. Overview of discrepant results is shown in Table 4. Sensitivity and proportional agreement between the VCA and EBNA IgG assays is shown in Table 5.
Looking across donor and recipient results in Tables 1 and 2, proportional agreement ranged from 96% to 100% for VIDAS® and LIAISON® . Sensitivity ranged from 91% to 100% and 92% to 100% for VIDAS® /IFA and LIAISON® /IFA, respectively.
3.1. Kidney donors (Tables 1 and 3–5) 3.1.1. VCA IgG Proportional agreement between VIDAS® and LIAISON® was 100% (lower bound of 95% CI: 97). The sensitivity of the VIDAS® assay when compared to IFA was 100% (lower bound of 95% CI: 96). Sensitivity of the LIAISON® assay when compared to IFA was 100% (lower bound of 95% CI: 97). VCA IgG equivocal results were obtained for 0, 1 and 0 samples when tested by LIAISON® , VIDAS® and IFA, respectively.
3.1.2. EBNA IgG Proportional agreement between VIDAS® and LIAISON® was 98% (95% CI: 93,100). Two samples showed discrepant results and both were found to be positive on VIDAS® but negative on LIAISON® . The sensitivity of the VIDAS® assay when compared to IFA was 98% (95% CI: 93,100). Four samples showed discrepant results; two were positive on VIDAS® but negative by IFA whereas two samples were negative on VIDAS® but positive upon IFA testing. Sensitivity of the LIAISON® assay when compared to IFA was 96% (95% CI: 90,99). Six samples showed discrepant results; four were negative on LIAISON® but positive on IFA whereas two were positive on LIAISON® but negative on IFA. EBNA IgG equivocal results were obtained for 5, 1 and 3 samples when tested by LIAISON® , VIDAS® and IFA tests, respectively – only one of which gave equivocal results in more than one assay (LIAISON® and VIDAS® ). Although specificity can be calculated using the raw data presented in Table 1, such calculations have not been carried out due to limited study sample numbers that preclude meaningful assessment of specificity. An overview of sensitivity and proportional agreement when equivocal data are excluded is shown in Table 3. An overview of discrepant VCA and EBNA IgG results is given in Table 4, which shows that (overall) 2 and 7 samples gave discrepant results in LIAISON® vs. VIDAS® and LIAISON® vs. IFA, respectively, and 5 samples gave discrepant results in VIDAS® vs. IFA. A comparison of proportional agreement between VCA and EBNA IgG assays is shown in Table 5A. Overall, for LIAISON® and VIDAS® , sensitivity ranged from 96% to 98% and proportional agreement from 96% to 98%. Statistical assessment of distribution of discrepant results obtained for donors across the VCA and EBNA IgG tests for both LIAISON® and VIDAS® could not be carried out as a result of the donor data containing zero values.
110
I. Johannessen et al. / Journal of Virological Methods 203 (2014) 107–111
Table 5 Sensitivity and proportional agreement between VCA and EBNA IgG assays. The following tables show the number of positive and negative VCA and EBNA IgG results with VIDAS® [first column of tables] and LIAISON® [second column of tables]; in each table, the sensitivity of EBNA compared to VCA IgG and proportional agreement is presented along with a 95% CI and McNemar’s assessment of proportional agreement between the two IgG tests (for recipients only since the donor data includes zero values that precludes McNemar’s testing). 5A – Donors VIDAS®
N EBNA N P All Sensitivity Proportional agreement
LIAISON®
VCAG
1 0 1 96% 96%
P 4 90 94 95% CI (89, 99) 95% CI (90, 99)
All
VCAG N
EBNA N P All Sensitivity Proportional agreement
5 90 95
1 0 1 98% 98%
P
All
2 95 97 95% CI (93, 100) 95% CI (93, 100)
3 95 98
P
All
8 78 86 95% CI (82, 96) 95% CI (80, 94)
12 81 93
5B – Recipients VIDAS®
N EBNA N P All Sensitivity Proportional agreement McNemar’s test p = 0.070 (exact test)
LIAISON®
VCAG
5 1 6 92% 92%
P 7 84 91 95% CI (85, 97) 95% CI (84, 96)
All
VCAG N
EBNA N P All Sensitivity Proportional agreement McNemar’s test p = 0.227 (exact test)
12 85 97
4 3 7 91% 88%
P: positive; N: negative; IFA: immunofluorescence assay; VCAG: VCA IgG; EBNA: EBNA IgG.
3.2. Kidney recipients (Tables 2–5) 3.2.1. VCA IgG Proportional agreement between VIDAS® and LIAISON® was 98% (95% CI: 93,100). Two samples showed discrepant results; both were negative on LIAISON® but positive on VIDAS® . The sensitivity of the VIDAS® assay when compared to IFA was 96% (95% CI: 90,99). Four samples showed discrepant results; all were negative on VIDAS® but positive on IFA. Sensitivity of the LIAISON® assay when compared to IFA was 94% (95% CI: 87,98). Six samples showed discrepant results; all were negative on LIAISON® but positive on IFA. VCA IgG equivocal results were obtained for 0, 0 and 3 samples when tested by LIAISON® , VIDAS® and IFA, respectively. 3.2.2. EBNA IgG Proportional agreement between VIDAS® and LIAISON® was 96% (95% CI: 90,99). Four samples showed discrepant results; two were negative on LIAISON® but positive on VIDAS® whereas two were positive on LIAISON® and negative on VIDAS® . The sensitivity of the VIDAS® assay when compared to IFA was 91% (95% CI: 83,96). Fifteen samples showed discrepant results; eight were negative on VIDAS® but positive on IFA whereas seven were positive on VIDAS® but negative on IFA. Sensitivity of the LIAISON® assay when compared to IFA was 92% (95% CI: 84,97). Thirteen samples showed discrepant results; seven were negative on LIAISON® but positive on IFA whereas six were positive on LIAISON® but negative on IFA. EBNA IgG equivocal results were obtained for 7, 3 and 5 samples when tested by LIAISON® , VIDAS® and IFA tests, respectively – only one of which gave equivocal results in more than one assay (LIAISON® and VIDAS® ). Although specificity can be calculated using the raw data presented in Table 2, such calculations have not been carried out due to limited study sample numbers that preclude meaningful assessment of specificity. An overview of sensitivity and proportional agreement when equivocal data are excluded is shown in Table 3. An overview of discrepant VCA and EBNA IgG results is given in Table 4, which shows that (overall) 6 and 19 samples gave discrepant results in LIAISON® vs. VIDAS® and LIAISON® vs. IFA,
respectively, and 19 samples gave discrepant results in VIDAS® vs. IFA. A comparison of proportional agreement between VCA and EBNA IgG assays is shown in Table 5B. Overall, for LIAISON® and VIDAS® , sensitivity ranged from 91% to 92% and proportional agreement from 88% to 92%. The distribution of discrepant results obtained for recipients was evenly distributed across the VCA and EBNA IgG tests for both LIAISON® and VIDAS® (p = 0.227 and p = 0.070, respectively).
4. Discussion The aim of EBV serology in the transplant setting is to delineate organ transplant donor and recipient serostatus with reference to past infection (seropositive status) or susceptibility to infection (seronegative status). To this end, serological assays are carried out. The current study assessed delineation of live kidney adult donor (‘D’) and cadaveric kidney adult recipient (‘R’) EBV serostatus by 3 methods. In this context, the aim of EBV testing prior to transplantation is to define EBV D and R serostatus with a view to assess risk of EBV-associated lymphoproliferative disease. The high risk D+/R− setting prompts close surveillance of recipients following solid organ transplant surgery to ensure early detection of any signs or symptoms of EBV-associated lesions thereby facilitating early therapeutic intervention [see British Committee for Standards in Haematology (BCSH) and British Transplantation Society (BTS) guidance; full reference below]. To this end, UK transplant centres carry out EBV VCA (or EA/VCA) IgG and/or EBNA IgG tests. There is considerable evidence base for agreement of serological EBV profiles across various immunoassays when compared with gold standard reference tests such as IFA (Rea et al., 2002; Gärtner et al., 2003; Binnicker et al., 2008; Klutts et al., 2009). However, to our knowledge, this is the first study that compares the VIDAS® , LIAISON® and IFA systems for determination of EBV serostatus in the context of kidney transplantation. Other studies have carried out such assessments in the context of primary EBV infection (for a recent review of EBV diagnostic tools, see Luzuriaga and Sullivan, 2010; Odumade et al., 2011).
I. Johannessen et al. / Journal of Virological Methods 203 (2014) 107–111
The results in Tables 1 and 2 show high overall proportional agreement and sensitivity between the assays examined, which is in line with prior reports (for example, Koidl et al., 2011; Lupo et al., 2012). Exclusion of equivocals does not change this assessment (see Table 3). This is not wholly unexpected since both VIDAS® and LIAISON® use the same VCA (p18) and EBNA-1 (p72) target antigens in their assays. The VIDAS® VCA/EA also contains the EA (p54) target antigen, which may explain discrepant recipient results. An overall summary of discrepant results is given in Table 4. The table highlights similar results when using the LIAISON® and VIDAS® assays. However, IFA assessments generate discrepant results when compared with the other two assays. In particular, differences were observed when using the EBNA IgG test, which is likely to reflect challenges to the operator when scoring the EBNA immunostaining IFA results since the EBNA nuclear staining pattern can be particularly difficult to determine despite considerable operator experience. Thus, for EBNA IgG testing, EIAs should be preferred to IFA. Most UK laboratories carry out only one EBV IgG test prior to transplantation surgery, and each discrepant result may result in inaccurate designation. In the UK, the Department of Health’s Advisory Committee on the Safety of Blood, Tissues and Organs (SaBTO) issued new guidance on virology testing in the solid organ transplant context in 2011 that requires assessment of donor EBV serostatus but does not specify whether this should be based on VCA and/or EBNA IgG testing (www.gov.uk – see full URL web reference below). To this end, parallels can be drawn with UK Standard for Microbiology Investigation V26, ‘Epstein–Barr Virus Serology’, that suggests the use of EBNA IgG for assessment of past EBV infection (www.hpa.org.uk – see full URL below). Interestingly, a comparison of proportional agreement between the VCA and EBNA IgG test results for each of the VIDAS® and LIAISON® assays (see Table 5) in the kidney transplant setting showed high proportional agreement between both tests. Equally, discrepant results between the VCA and EBNA IgG tests were evenly distributed when using both the VIDAS® (p = 0.070) and LIAISON® assays (p = 0.227) suggesting that either IgG test could be used at outset. However, in light of the small sample size, the current study does not permit a statistically valid conclusion that the use of VCA IgG and EBNA IgG should be the routine approach to determination of EBV serostatus pre-transplantation. Interestingly, the data does suggest that there may be a clinical advantage to such an approach since the study has shown an increase in detection of seropositive samples with the addition of the EBNA IgG test. However, such an approach would necessitate clear guidance on the interpretation of discrepant results. Historically, the IFA approach to EBV serological diagnosis has been considered to be ‘gold standard’ for such assessments. However, the approach contains an element of observer subjectivity and is susceptible to non-specific IF reactivity, which will explain the observed discrepancies between IFA and the automated assays. As a result of such observations, IFA has been largely replaced in UK clinical virology laboratories with more automated (random access) EIA approaches that do not require as much assayspecific expertise for interpretation as well as lending themselves to high through-put, out-of-hours use and short turn-around times necessary. In the current study, the LIAISON® and VIDAS® EIAs gave comparable results and could, therefore, be considered for assessments of EBV serostatus prior to kidney transplantation.
111
Conflict of interest The study was funded by bioMérieux, Lyon, France. However, study design and execution of work as well as analysis and interpretation of results were performed independently by the authors. Acknowledgments We are grateful to Dr. K. Cuschieri and Dr. K. Eastick of the Scottish Human Papillomavirus Reference Laboratory and Scottish Bacterial Sexually Transmitted Infections Reference Laboratory, respectively, for their valuable input and advice on the study. References Advisory Committee on the Safety of Blood, Tissues and Organs 2014 (SaBTO), https://www.gov.uk/government/publications/guidance-on-the-microbiological-safety-of-human-organs-tissues-and-cells-used-in-transplantation.(accessed 05.02.14). Binnicker, M.J., Jespersen, D.J., Harring, J.A., Rollins, L.O., Beito, E.M., 2008. Evaluation of a multiplex flow immunoassay for detection of Epstein–Barr virus-specific antibodies. Clin. Vaccine Immunol. 15, 1410–1413. Bollard, C.M., Rooney, C.M., Heslop, H.E., 2012. T-cell therapy in the treatment of post-transplant lymphoproliferative disease. Nat. Rev. Clin. Oncol. 9, 510–519. Crawford, D.H., Thomas, J.A., Janossy, G., Sweny, P., Fernando, O.N., Moorhead, J.F., Thompson, J.H., 1980. Epstein–Barr virus nuclear antigen positive lymphoma after cyclosporin A treatment in a patient with renal allograft. Lancet 1, 1355–1356. Gärtner, B.C., Hess, R.D., Bandt, D., Kruse, A., Rethwilm, A., Roemer, K., MuellerLantzsch, N., 2003. Evaluation of four commercially available Epstein–Barr virus enzyme immunoassays with an immunofluorescence assay as the reference method. Clin. Diagn. Lab. Immunol. 10, 78–82. Haemato-oncology subgroup of the British Committee for Standards in Haematology (BCSH), the British Transplantation Society (BTS), 2010. Managementof post-transplant lymphoproliferative disorder in adult solid organ trans-plant recipients – BCSH and BTS Guidelines. Brit. J. Haematol. 149, 693–705. Höcker, B., Fickenscher, H., Delecluse, H.-J., Böhm, S., Küsters, U., Schnitzler, P., Pohl, M., John, U., Kemper, M.J., Fehrenbach, H., Wigger, M., Holder, M., Schröder, M., Billing, H., Fichtner, A., Feneberg, R., Sander, A., Köpf-Shakib, S., Süsal, C., Tönshoff, B., 2013. Epidemiology and morbidity of Epstein–Barr virus infection in pediatric renal transplant recipients: a multicenter, prospective study. Clin. Infect. Dis. 56, 84–92. Klutts, J.S., Ford, B.A., Perez, N.R., Gronowski, A.M., 2009. Evidence-based approach for interpretation of Epstein–Barr virus serological patterns. J. Clin. Microbiol. 47, 3204–3210. Koidl, C., Riedl, R., Schweighofer, B., Fett, S., Bozic, M., Marth, E., 2011. Performance of new enzyme-linked fluorescent assays for detection of Epstein–Barr virus specific antibodies in routine diagnostics. Wien. Klin. Wochenschr. 123, 230–234. Long, H.M., Taylor, G.S., Rickinson, A.B., 2011. Immune defence against EBV and EBVassociated disease. Curr. Opin. Immunol. 23, 258–264. Lupo, J., Germi, R., Semenova, T., Buisson, M., Seigneurin, J.M., Morand, P., 2012. Performance of two commercially available automated immunoassays for the determination of Epstein–Barr virus serological status. Clin. Vaccine Immunol. 19, 929. Luzuriaga, K., Sullivan, J.L., 2010. Infectious mononucleosis. N. Engl. J. Med. 362, 1993–2000. Macsween, K.F., Higgins, C.D., McAulay, K.A., Williams, H., Harrison, N., Swerdlow, A.J., Crawford, D.H., 2010. Infectious mononucleosis in university students in the United Kingdom: evaluation of the clinical features and consequences of the disease. Clin. Infect. Dis. 50, 699–706. Morton, M., Coupes, B., Roberts, S.A., Klapper, P.E., Byers, R.J., Vallely, P.J., Ryan, K., Picton, M.L., 2013. Epidemiology of posttransplantation lymphoproliferative disorder in adult renal transplant recipients. Transplantation 95, 470–478. Odumade, O.A., Hogquist, K.A., Balfour Jr., H.H., 2011. Progress and problems in understanding and managing primary Epstein–Barr virus infections. Clin. Microbiol. Rev. 24, 193–209. Rea, T.D., Ashley, R.L., Russo, J.E., Buchwald, D.S., 2002. A systematic study of Epstein–Barr virus serologic assays following acute infection. Am. J. Clin. Pathol. 117, 156–161. Thomas, J.A., Allday, M., Crawford, D.H., 1991. Epstein–Barr virus-associated lymphoproliferative disorders in immunocompromised individuals. Adv. Cancer Res. 57, 329–380. UK Standards for Microbiology Investigations (UK SMI), 2014. Epstein–BarrVirus (SMI V26i4). http://www.hpa.org.uk/webc/HPAwebFile/HPAweb C/1317131313375.(accessed 05.02.14).
Contents lists available at ScienceDirect
Journal of Virological Methods journal homepage: www.elsevier.com/locate/jviromet
Determination of EBV serostatus prior to kidney transplantation: Comparison of VIDAS® , LIAISON® and immunofluorescence assays I. Johannessen a,∗ , M. Noel a , A. Galloway a , S. Black a , M. Shearman a , C. Graham b a b
Department of Laboratory Medicine (Virology), Royal Infirmary of Edinburgh, Little France, Edinburgh EH16 4SA, UK Epidemiology and Statistics Core, Wellcome Trust Clinical Research Facility, Western General Hospital, Edinburgh EH4 2XU, UK
a b s t r a c t Article history: Received 16 June 2013 Received in revised form 18 March 2014 Accepted 24 March 2014 Available online 2 April 2014 Keywords: Epstein–Barr Virus Serostatus Kidney transplantation Post-transplant lymphoproliferative disease Enzyme immunosorbent assay Immunofluorescent assay
Immunosuppression following solid organ transplantation reduces T cell-mediated immune control of Epstein–Barr Virus (EBV), which may then drive development of post-transplant lymphoproliferative disease. Serology plays a key role in determination of risk of outgrowth of such lesions following transplantation. The study compared the VIDAS® (bioMérieux) and LIAISON® (DiaSorin) enzyme immunoassays (EIAs) and immunofluorescence assays (IFA; MBL-Bion) in the kidney transplantation setting. Sera from 100 live kidney donors [51 males; age range 20–82 years (mean 51.2 years)] and 100 cadaveric kidney recipients [70 males; age range 17–77 years (mean 51.0 years)] were tested. Overall proportional agreement ranged from 96% to 100% for VIDAS® and LIAISON® . Sensitivity ranged from 91% to 100% and 92% to 100% for VIDAS® /IFA and LIAISON® /IFA, respectively. The VIDAS® and LIAISON® approaches gave similar results. Such automated random access EIAs are well suited to busy clinical virology laboratories and rapid determination of donor and recipient EBV serostatus prior to transplantation. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Epstein–Barr Virus (EBV) infects approximately 90% of adults worldwide and persists for life in host B cells. Primary infection usually occurs in childhood and is generally asymptomatic at that time, but later acquisition of EBV may lead to clinical manifestations of infectious mononucleosis (Macsween et al., 2010). T cell-mediated immunity controls persistent EBV infection and immunosuppression tips the balance in favour of the virus, which expresses growth-promoting oncogenes (Long et al., 2011). In solid organ transplant recipients, this may lead to uncontrolled proliferation of virus-infected B cells and, in up to 10% of organ graft recipients, culminate in post-transplant lymphoproliferative disease (Crawford et al., 1980; Bollard et al., 2012). The two main risk factors for such disease are (1) EBV seronegativity prior to solid organ transplant surgery and (2) level and duration of immunosuppression (Thomas et al., 1991). Delineation of donor and recipient EBV serostatus prior to solid organ transplantation forms part of UK and European guidelines for transplant assessments. Although EBV carriage is very common
∗ Corresponding author. Tel.: +44 131 242 6003. E-mail address: [email protected] (I. Johannessen). http://dx.doi.org/10.1016/j.jviromet.2014.03.022 0166-0934/© 2014 Elsevier B.V. All rights reserved.
in adults, serological results may highlight susceptibility to EBV infection following transplantation surgery at a time when the recipient is ill equipped to control infection – a high risk situation for development of EBV-associated lymphoproliferative disease (Höcker et al., 2013; Morton et al., 2013). Historically, EBV immunofluorescence assay (IFA) has been used as gold standard for serological diagnostics. However, the assay is labour intensive and includes an element of operator subjectivity. Equally, it does not lend itself to automated approaches. Thus, the aim of this study was to compare VIDAS® (bioMérieux, France) and LIAISON® (DiaSorin, USA) enzyme immunosorbent assays (EIA) as well as IFA (MBL-Bion, USA) for determination of EBV serostatus prior to kidney transplantation. 2. Materials and methods 2.1. Study patients The study comprised of a panel of 200 serum samples obtained pre-kidney transplantation from unrelated live renal donors (n = 100) and cadaveric renal recipients (n = 100). One hundred serum samples from live kidney transplant donors [51 male; age range 20–82 years (mean (sd) 51.2 (12.2) years)] were tested for VCA IgG and EBNA IgG in all three assay systems (Table 1). One
108
I. Johannessen et al. / Journal of Virological Methods 203 (2014) 107–111
Table 1 Sensitivity and proportional agreement data when assessing kidney donors. The following tables show the number of all positive and negative results with LIAISON® vs. VIDAS® [first column of tables], LIAISON® vs. IFA [middle column of tables] and VIDAS® vs. IFA [last column of tables]; in each table, the sensitivity and the proportional agreement are presented along with a 95% CI. VIDAS®
VCAG
N
VCAG P
®
LIAISON N 1 0 1 0 99 99 P 1 99 100 All Sensitivity (LIAISON vs. VIDAS) 100% (Lower 95%CI 97%) Proportional agreement 100% (Lower 95%CI 97%) EBNA
VIDAS® N
P
IFA
All
P
IFA
All
®
N
P
All
®
LIAISON N 1 0 1 P 1 98 99 All 2 98 100 Sensitivity (LIAISON vs. IFA) 100% (Lower 95%CI 97%) Proportional agreement 99% (95% CI (95, 100))
VIDAS N 1 0 1 P 1 98 99 All 2 98 100 Sensitivity (VIDAS vs. IFA) 100% (Lower 95% CI 96) Proportional agreement 99% (95% CI (95, 100))
EBNA
EBNA
All
LIAISON® N 3 2 5 0 95 95 P 3 97 100 All Sensitivity (LIAISON vs. VIDAS) 98% (95% CI (93, 100)) Proportional agreement 98% (95% CI (93, 100))
VCAG
N
IFA N
P
All
LIAISON® N 1 4 5 P 2 93 95 All 3 97 100 Sensitivity (LIAISON vs. IFA) 96% (95% CI (90, 99)) Proportional agreement 94% (95% CI (87, 98))
IFA N
P
All
VIDAS® N 1 2 3 P 2 95 97 All 3 97 100 Sensitivity (VIDAS vs. IFA) 98% (95% CI (93, 100)) Proportional agreement 96% (95% CI (90, 99))
P: positive; N: negative; IFA: immunofluorescence assay; VCAG: VCA IgG; EBNA: EBNA IgG.
hundred cadaveric kidney transplant recipients [70 male; age range 17–77 years (mean (sd) 51.0 (14.5) years)] were tested for both parameters in all three assay systems (Table 2). Blood samples tested were retrospective (archived left over) serum samples that had been stored at −20 ± 6 ◦ C for <2 years. 2.2. Ethical considerations Consent was obtained for EBV testing at time of sampling, and the study (classified as service development by local ethics authority) re-analysed anonymised residual blood tube specimens left over from that time. 2.3. Samples Serum samples [stored at −20 ◦ C (±6 ◦ C) for <2 years] were homogenised and tested singly in each assay. Insufficient samples were excluded from analysis; equivocal results were assigned a positive (‘reactive’) status. Due to volume restriction, samples were not re-tested. The study did not ascertain whether or not
additional samples were available from the same individual(s) for further/extended tests. 2.4. Experimental plan The following qualitative assays were used in accordance with the manufacturers’ instructions: (1) VIDAS® Automated System (bioMérieux, Lyon, France): ‘EBV VCA/EA IgG’ (ref. 30236) and ‘EBV EBNA IgG’ (ref. 30235) for detection of IgG antibodies (Ab) directed against EBV VCA (p18 synthetic peptide) and EA (p54 synthetic peptide), and IgG Ab directed against EBNA-1 (p72 synthetic peptide), respectively; (2) LIAISON® Automated System (DiaSorin, Saluggia Vercelli, Italy): ‘EBV VCA IgG’ (ref. 310510) and ‘EBV EBNA IgG’ (ref. 310520) for detection of IgG Ab directed against EBV VCA (p18 synthetic peptide), and IgG Ab directed against EBNA-1 (p72 synthetic peptide), respectively; (3) MBL-Bion Manual IFA System (Des Plaines, IL, USA): ‘EBV VCA Antigen Substrate Slide’ (P3HR1 EBV strain-infected lymphocytes; ref. EB-5012) and ‘EBV EBNA Antigen Substrate Slide’ (Raji EBV strain-infected lymphocytes; ref. NA-5112) were used in the context of indirect IF (IIF) employing anti-human conjugate detection Ab enabling detection of human
Table 2 Sensitivity and proportional agreement data when assessing kidney recipients. The following tables show the number of all positive and negative results with LIAISON® vs. VIDAS® [first column of tables], LIAISON® vs. IFA [middle column of tables] and VIDAS® vs. IFA [last column of tables]; in each table, the sensitivity and the proportional agreement are presented along with a 95% CI. VCAG
VIDAS® N
VCAG P
All
LIAISON® N 6 2 8 0 92 92 P 6 94 100 All Sensitivity (LIAISON vs. VIDAS) 98% (95% CI (93, 100)) Proportional agreement 98% (95% CI (93, 100)) EBNA
VIDAS® N
P
All
LIAISON® N 10 2 12 2 86 88 P 12 88 100 All Sensitivity (LIAISON vs. VIDAS) 98% (95% CI (92, 100)) Proportional agreement 96% (95% CI (90, 99))
IFA N
VCAG P
All
IFA N
P
All
LIAISON® N 2 6 8 P 0 92 92 All 2 98 100 Sensitivity (LIAISON vs. IFA) 94% (95% CI (87, 98))0.938776 Proportional agreement 94% (95% CI (87, 98))
VIDAS® N 2 4 6 P 0 94 94 All 2 98 100 Sensitivity (VIDAS vs. IFA) 96% (95% CI (90, 99)) Proportional agreement 96% (95% CI (90, 99))
EBNA
EBNA
IFA N
P
All
LIAISON® N 5 7 12 P 6 82 88 All 11 89 100 Sensitivity (LIAISON vs. IFA) 92% (95% CI (84, 97)) Proportional agreement 87% (95% CI (79, 93))
P: positive; N: negative; IFA: immunofluorescence assay; VCAG: VCA IgG; EBNA: EBNA IgG.
IFA N
P
All
VIDAS® N 4 8 12 P 7 81 88 All 11 89 100 Sensitivity (VIDAS vs. IFA) 91% (95% CI (83, 96)) Proportional agreement 85% (95% CI (76, 91))
I. Johannessen et al. / Journal of Virological Methods 203 (2014) 107–111
109
Table 3 Sensitivity and proportional agreement when equivocal data are excluded. LIAISON® vs. VIDAS®
LIAISON® vs. IFA
VIDAS® vs. IFA
Sensitivity Proportional agreement Sensitivity Proportional agreement
100% (lower 95%CI 96%) 100% (lower 95%CI 97%) 98% (95% CI (92, 100)) 98% (95% CI (93, 100))
100% (lower 95%CI 96%) 99% (95% CI (95, 100)) 97% (95% CI (91, 99)) 95% (95% CI (88, 98))
100% (lower 95% CI 96) 99% (95% CI (95, 100)) 98% (95% CI (92, 100)) 96% (95% CI (90, 99))
Sensitivity Proportional agreement Sensitivity Proportional agreement
98% (95% CI (93, 100)) 98% (95% CI (93, 100)) 100% (lower 95% CI 96%) 100% (lower 95% CI 97%)
95% (95% CI (88, 98)) 95% (95% CI (88, 98)) 91% (95% CI (83, 96)) 90% (95% CI (81, 95))
97% (95% CI (91, 99)) 97% (95% CI (91, 99)) 90% (95% CI (82, 96)) 85% (95% CI (76, 91))
Assay Kidney donors VCA IgG EBNA IgG Kidney recipients VCA IgG EBNA IgG
Table 4 Overview of discrepant results. LIAISON® vs. IFA
VIDAS® vs. IFA
Kideny donors 0 VCA IgG EBNA IgG 2b Total 2
1a 6b , c 7
1a 4c 5
Kidney recipients 2d VCA IgG EBNA IgG 4f , h Total 6
6d , e 13f , g 19
4e 15g , h 19
Assay
LIAISON® vs. VIDAS®
a
Same sample. Two shared samples. c Four shared samples. d Two shared samples. e Four shared samples. f One shared sample. g Twelve shared samples. h Three shared samples. IFA: immunofluorescence assay. b
IgG Ab directed against VCA (IIF), and IgG Ab directed against EBNA (anti-compliment IF; ACIF), respectively; ‘IF IgG TEST’ (ref. CXCG9972) detection system was used in the context of the substrate slides for detection of human IgG. 2.5. Statistical analysis Cross-tabulations of results were generated and proportional agreement and sensitivity were calculated as appropriate along with 95% confidence intervals (CIs). Due to the limited study sample numbers, it was not always possible to calculate both the upper and lower bounds of the 95% CI and where this is the case, only the bound calculated has been presented. Similarly, due to limited study sample numbers, it has not been possible to calculate specificity in a meaningful manner (although such calculations can be made from the raw data presented in Tables 1 and 2). McNemar’s exact test was used to assess proportional agreement between two (matched pairs; VCA and EBNA) IgG tests in Table 5. 3. Results For assessment of pre-kidney transplant EBV serostatus, VIDAS® EA/VCA IgG and EBNA IgG results were compared with LIAISON® and IFA VCA IgG and EBNA IgG results, respectively. The kidney graft groups were studied separately since they comprised of live donors and cadaveric organ recipients. Summary cross-tabulations of all results are shown in Tables 1 and 2. Importantly, exclusion of equivocal results (Table 3) did not affect study outcome. Overview of discrepant results is shown in Table 4. Sensitivity and proportional agreement between the VCA and EBNA IgG assays is shown in Table 5.
Looking across donor and recipient results in Tables 1 and 2, proportional agreement ranged from 96% to 100% for VIDAS® and LIAISON® . Sensitivity ranged from 91% to 100% and 92% to 100% for VIDAS® /IFA and LIAISON® /IFA, respectively.
3.1. Kidney donors (Tables 1 and 3–5) 3.1.1. VCA IgG Proportional agreement between VIDAS® and LIAISON® was 100% (lower bound of 95% CI: 97). The sensitivity of the VIDAS® assay when compared to IFA was 100% (lower bound of 95% CI: 96). Sensitivity of the LIAISON® assay when compared to IFA was 100% (lower bound of 95% CI: 97). VCA IgG equivocal results were obtained for 0, 1 and 0 samples when tested by LIAISON® , VIDAS® and IFA, respectively.
3.1.2. EBNA IgG Proportional agreement between VIDAS® and LIAISON® was 98% (95% CI: 93,100). Two samples showed discrepant results and both were found to be positive on VIDAS® but negative on LIAISON® . The sensitivity of the VIDAS® assay when compared to IFA was 98% (95% CI: 93,100). Four samples showed discrepant results; two were positive on VIDAS® but negative by IFA whereas two samples were negative on VIDAS® but positive upon IFA testing. Sensitivity of the LIAISON® assay when compared to IFA was 96% (95% CI: 90,99). Six samples showed discrepant results; four were negative on LIAISON® but positive on IFA whereas two were positive on LIAISON® but negative on IFA. EBNA IgG equivocal results were obtained for 5, 1 and 3 samples when tested by LIAISON® , VIDAS® and IFA tests, respectively – only one of which gave equivocal results in more than one assay (LIAISON® and VIDAS® ). Although specificity can be calculated using the raw data presented in Table 1, such calculations have not been carried out due to limited study sample numbers that preclude meaningful assessment of specificity. An overview of sensitivity and proportional agreement when equivocal data are excluded is shown in Table 3. An overview of discrepant VCA and EBNA IgG results is given in Table 4, which shows that (overall) 2 and 7 samples gave discrepant results in LIAISON® vs. VIDAS® and LIAISON® vs. IFA, respectively, and 5 samples gave discrepant results in VIDAS® vs. IFA. A comparison of proportional agreement between VCA and EBNA IgG assays is shown in Table 5A. Overall, for LIAISON® and VIDAS® , sensitivity ranged from 96% to 98% and proportional agreement from 96% to 98%. Statistical assessment of distribution of discrepant results obtained for donors across the VCA and EBNA IgG tests for both LIAISON® and VIDAS® could not be carried out as a result of the donor data containing zero values.
110
I. Johannessen et al. / Journal of Virological Methods 203 (2014) 107–111
Table 5 Sensitivity and proportional agreement between VCA and EBNA IgG assays. The following tables show the number of positive and negative VCA and EBNA IgG results with VIDAS® [first column of tables] and LIAISON® [second column of tables]; in each table, the sensitivity of EBNA compared to VCA IgG and proportional agreement is presented along with a 95% CI and McNemar’s assessment of proportional agreement between the two IgG tests (for recipients only since the donor data includes zero values that precludes McNemar’s testing). 5A – Donors VIDAS®
N EBNA N P All Sensitivity Proportional agreement
LIAISON®
VCAG
1 0 1 96% 96%
P 4 90 94 95% CI (89, 99) 95% CI (90, 99)
All
VCAG N
EBNA N P All Sensitivity Proportional agreement
5 90 95
1 0 1 98% 98%
P
All
2 95 97 95% CI (93, 100) 95% CI (93, 100)
3 95 98
P
All
8 78 86 95% CI (82, 96) 95% CI (80, 94)
12 81 93
5B – Recipients VIDAS®
N EBNA N P All Sensitivity Proportional agreement McNemar’s test p = 0.070 (exact test)
LIAISON®
VCAG
5 1 6 92% 92%
P 7 84 91 95% CI (85, 97) 95% CI (84, 96)
All
VCAG N
EBNA N P All Sensitivity Proportional agreement McNemar’s test p = 0.227 (exact test)
12 85 97
4 3 7 91% 88%
P: positive; N: negative; IFA: immunofluorescence assay; VCAG: VCA IgG; EBNA: EBNA IgG.
3.2. Kidney recipients (Tables 2–5) 3.2.1. VCA IgG Proportional agreement between VIDAS® and LIAISON® was 98% (95% CI: 93,100). Two samples showed discrepant results; both were negative on LIAISON® but positive on VIDAS® . The sensitivity of the VIDAS® assay when compared to IFA was 96% (95% CI: 90,99). Four samples showed discrepant results; all were negative on VIDAS® but positive on IFA. Sensitivity of the LIAISON® assay when compared to IFA was 94% (95% CI: 87,98). Six samples showed discrepant results; all were negative on LIAISON® but positive on IFA. VCA IgG equivocal results were obtained for 0, 0 and 3 samples when tested by LIAISON® , VIDAS® and IFA, respectively. 3.2.2. EBNA IgG Proportional agreement between VIDAS® and LIAISON® was 96% (95% CI: 90,99). Four samples showed discrepant results; two were negative on LIAISON® but positive on VIDAS® whereas two were positive on LIAISON® and negative on VIDAS® . The sensitivity of the VIDAS® assay when compared to IFA was 91% (95% CI: 83,96). Fifteen samples showed discrepant results; eight were negative on VIDAS® but positive on IFA whereas seven were positive on VIDAS® but negative on IFA. Sensitivity of the LIAISON® assay when compared to IFA was 92% (95% CI: 84,97). Thirteen samples showed discrepant results; seven were negative on LIAISON® but positive on IFA whereas six were positive on LIAISON® but negative on IFA. EBNA IgG equivocal results were obtained for 7, 3 and 5 samples when tested by LIAISON® , VIDAS® and IFA tests, respectively – only one of which gave equivocal results in more than one assay (LIAISON® and VIDAS® ). Although specificity can be calculated using the raw data presented in Table 2, such calculations have not been carried out due to limited study sample numbers that preclude meaningful assessment of specificity. An overview of sensitivity and proportional agreement when equivocal data are excluded is shown in Table 3. An overview of discrepant VCA and EBNA IgG results is given in Table 4, which shows that (overall) 6 and 19 samples gave discrepant results in LIAISON® vs. VIDAS® and LIAISON® vs. IFA,
respectively, and 19 samples gave discrepant results in VIDAS® vs. IFA. A comparison of proportional agreement between VCA and EBNA IgG assays is shown in Table 5B. Overall, for LIAISON® and VIDAS® , sensitivity ranged from 91% to 92% and proportional agreement from 88% to 92%. The distribution of discrepant results obtained for recipients was evenly distributed across the VCA and EBNA IgG tests for both LIAISON® and VIDAS® (p = 0.227 and p = 0.070, respectively).
4. Discussion The aim of EBV serology in the transplant setting is to delineate organ transplant donor and recipient serostatus with reference to past infection (seropositive status) or susceptibility to infection (seronegative status). To this end, serological assays are carried out. The current study assessed delineation of live kidney adult donor (‘D’) and cadaveric kidney adult recipient (‘R’) EBV serostatus by 3 methods. In this context, the aim of EBV testing prior to transplantation is to define EBV D and R serostatus with a view to assess risk of EBV-associated lymphoproliferative disease. The high risk D+/R− setting prompts close surveillance of recipients following solid organ transplant surgery to ensure early detection of any signs or symptoms of EBV-associated lesions thereby facilitating early therapeutic intervention [see British Committee for Standards in Haematology (BCSH) and British Transplantation Society (BTS) guidance; full reference below]. To this end, UK transplant centres carry out EBV VCA (or EA/VCA) IgG and/or EBNA IgG tests. There is considerable evidence base for agreement of serological EBV profiles across various immunoassays when compared with gold standard reference tests such as IFA (Rea et al., 2002; Gärtner et al., 2003; Binnicker et al., 2008; Klutts et al., 2009). However, to our knowledge, this is the first study that compares the VIDAS® , LIAISON® and IFA systems for determination of EBV serostatus in the context of kidney transplantation. Other studies have carried out such assessments in the context of primary EBV infection (for a recent review of EBV diagnostic tools, see Luzuriaga and Sullivan, 2010; Odumade et al., 2011).
I. Johannessen et al. / Journal of Virological Methods 203 (2014) 107–111
The results in Tables 1 and 2 show high overall proportional agreement and sensitivity between the assays examined, which is in line with prior reports (for example, Koidl et al., 2011; Lupo et al., 2012). Exclusion of equivocals does not change this assessment (see Table 3). This is not wholly unexpected since both VIDAS® and LIAISON® use the same VCA (p18) and EBNA-1 (p72) target antigens in their assays. The VIDAS® VCA/EA also contains the EA (p54) target antigen, which may explain discrepant recipient results. An overall summary of discrepant results is given in Table 4. The table highlights similar results when using the LIAISON® and VIDAS® assays. However, IFA assessments generate discrepant results when compared with the other two assays. In particular, differences were observed when using the EBNA IgG test, which is likely to reflect challenges to the operator when scoring the EBNA immunostaining IFA results since the EBNA nuclear staining pattern can be particularly difficult to determine despite considerable operator experience. Thus, for EBNA IgG testing, EIAs should be preferred to IFA. Most UK laboratories carry out only one EBV IgG test prior to transplantation surgery, and each discrepant result may result in inaccurate designation. In the UK, the Department of Health’s Advisory Committee on the Safety of Blood, Tissues and Organs (SaBTO) issued new guidance on virology testing in the solid organ transplant context in 2011 that requires assessment of donor EBV serostatus but does not specify whether this should be based on VCA and/or EBNA IgG testing (www.gov.uk – see full URL web reference below). To this end, parallels can be drawn with UK Standard for Microbiology Investigation V26, ‘Epstein–Barr Virus Serology’, that suggests the use of EBNA IgG for assessment of past EBV infection (www.hpa.org.uk – see full URL below). Interestingly, a comparison of proportional agreement between the VCA and EBNA IgG test results for each of the VIDAS® and LIAISON® assays (see Table 5) in the kidney transplant setting showed high proportional agreement between both tests. Equally, discrepant results between the VCA and EBNA IgG tests were evenly distributed when using both the VIDAS® (p = 0.070) and LIAISON® assays (p = 0.227) suggesting that either IgG test could be used at outset. However, in light of the small sample size, the current study does not permit a statistically valid conclusion that the use of VCA IgG and EBNA IgG should be the routine approach to determination of EBV serostatus pre-transplantation. Interestingly, the data does suggest that there may be a clinical advantage to such an approach since the study has shown an increase in detection of seropositive samples with the addition of the EBNA IgG test. However, such an approach would necessitate clear guidance on the interpretation of discrepant results. Historically, the IFA approach to EBV serological diagnosis has been considered to be ‘gold standard’ for such assessments. However, the approach contains an element of observer subjectivity and is susceptible to non-specific IF reactivity, which will explain the observed discrepancies between IFA and the automated assays. As a result of such observations, IFA has been largely replaced in UK clinical virology laboratories with more automated (random access) EIA approaches that do not require as much assayspecific expertise for interpretation as well as lending themselves to high through-put, out-of-hours use and short turn-around times necessary. In the current study, the LIAISON® and VIDAS® EIAs gave comparable results and could, therefore, be considered for assessments of EBV serostatus prior to kidney transplantation.
111
Conflict of interest The study was funded by bioMérieux, Lyon, France. However, study design and execution of work as well as analysis and interpretation of results were performed independently by the authors. Acknowledgments We are grateful to Dr. K. Cuschieri and Dr. K. Eastick of the Scottish Human Papillomavirus Reference Laboratory and Scottish Bacterial Sexually Transmitted Infections Reference Laboratory, respectively, for their valuable input and advice on the study. References Advisory Committee on the Safety of Blood, Tissues and Organs 2014 (SaBTO), https://www.gov.uk/government/publications/guidance-on-the-microbiological-safety-of-human-organs-tissues-and-cells-used-in-transplantation.(accessed 05.02.14). Binnicker, M.J., Jespersen, D.J., Harring, J.A., Rollins, L.O., Beito, E.M., 2008. Evaluation of a multiplex flow immunoassay for detection of Epstein–Barr virus-specific antibodies. Clin. Vaccine Immunol. 15, 1410–1413. Bollard, C.M., Rooney, C.M., Heslop, H.E., 2012. T-cell therapy in the treatment of post-transplant lymphoproliferative disease. Nat. Rev. Clin. Oncol. 9, 510–519. Crawford, D.H., Thomas, J.A., Janossy, G., Sweny, P., Fernando, O.N., Moorhead, J.F., Thompson, J.H., 1980. Epstein–Barr virus nuclear antigen positive lymphoma after cyclosporin A treatment in a patient with renal allograft. Lancet 1, 1355–1356. Gärtner, B.C., Hess, R.D., Bandt, D., Kruse, A., Rethwilm, A., Roemer, K., MuellerLantzsch, N., 2003. Evaluation of four commercially available Epstein–Barr virus enzyme immunoassays with an immunofluorescence assay as the reference method. Clin. Diagn. Lab. Immunol. 10, 78–82. Haemato-oncology subgroup of the British Committee for Standards in Haematology (BCSH), the British Transplantation Society (BTS), 2010. Managementof post-transplant lymphoproliferative disorder in adult solid organ trans-plant recipients – BCSH and BTS Guidelines. Brit. J. Haematol. 149, 693–705. Höcker, B., Fickenscher, H., Delecluse, H.-J., Böhm, S., Küsters, U., Schnitzler, P., Pohl, M., John, U., Kemper, M.J., Fehrenbach, H., Wigger, M., Holder, M., Schröder, M., Billing, H., Fichtner, A., Feneberg, R., Sander, A., Köpf-Shakib, S., Süsal, C., Tönshoff, B., 2013. Epidemiology and morbidity of Epstein–Barr virus infection in pediatric renal transplant recipients: a multicenter, prospective study. Clin. Infect. Dis. 56, 84–92. Klutts, J.S., Ford, B.A., Perez, N.R., Gronowski, A.M., 2009. Evidence-based approach for interpretation of Epstein–Barr virus serological patterns. J. Clin. Microbiol. 47, 3204–3210. Koidl, C., Riedl, R., Schweighofer, B., Fett, S., Bozic, M., Marth, E., 2011. Performance of new enzyme-linked fluorescent assays for detection of Epstein–Barr virus specific antibodies in routine diagnostics. Wien. Klin. Wochenschr. 123, 230–234. Long, H.M., Taylor, G.S., Rickinson, A.B., 2011. Immune defence against EBV and EBVassociated disease. Curr. Opin. Immunol. 23, 258–264. Lupo, J., Germi, R., Semenova, T., Buisson, M., Seigneurin, J.M., Morand, P., 2012. Performance of two commercially available automated immunoassays for the determination of Epstein–Barr virus serological status. Clin. Vaccine Immunol. 19, 929. Luzuriaga, K., Sullivan, J.L., 2010. Infectious mononucleosis. N. Engl. J. Med. 362, 1993–2000. Macsween, K.F., Higgins, C.D., McAulay, K.A., Williams, H., Harrison, N., Swerdlow, A.J., Crawford, D.H., 2010. Infectious mononucleosis in university students in the United Kingdom: evaluation of the clinical features and consequences of the disease. Clin. Infect. Dis. 50, 699–706. Morton, M., Coupes, B., Roberts, S.A., Klapper, P.E., Byers, R.J., Vallely, P.J., Ryan, K., Picton, M.L., 2013. Epidemiology of posttransplantation lymphoproliferative disorder in adult renal transplant recipients. Transplantation 95, 470–478. Odumade, O.A., Hogquist, K.A., Balfour Jr., H.H., 2011. Progress and problems in understanding and managing primary Epstein–Barr virus infections. Clin. Microbiol. Rev. 24, 193–209. Rea, T.D., Ashley, R.L., Russo, J.E., Buchwald, D.S., 2002. A systematic study of Epstein–Barr virus serologic assays following acute infection. Am. J. Clin. Pathol. 117, 156–161. Thomas, J.A., Allday, M., Crawford, D.H., 1991. Epstein–Barr virus-associated lymphoproliferative disorders in immunocompromised individuals. Adv. Cancer Res. 57, 329–380. UK Standards for Microbiology Investigations (UK SMI), 2014. Epstein–BarrVirus (SMI V26i4). http://www.hpa.org.uk/webc/HPAwebFile/HPAweb C/1317131313375.(accessed 05.02.14).