© 2015 APMIS. Published by John Wiley & Sons Ltd. DOI 10.1111/apm.12413
APMIS 123: 648–651
Comparison of chemiluminescent immunoassay and ELISA for measles IgG and IgM 1 and JUAN C. SANZ2,3 FERNANDO DE ORY,1,2 TEODORA MINGUITO,1 PILAR BALFAGON 1
Laboratorio de Serologıa, Centro Nacional de Microbiologıa, Instituto de Salud Carlos III, Majadahonda, Madrid; 2CIBER de Epidemiologıa y Salud P ublica (CIBERESP), Barcelona; and 3Laboratorio de Salud P ublica de la Comunidad de Madrid, Madrid, Spain
de Ory F, Minguito T, Balfag on P, Sanz JC. Comparison of chemiluminescent immunoassay and ELISA for measles IgG and IgM. APMIS 2015 123: 648–651. In the context of measles elimination, the identification of recent infections is important for clinical laboratories. Serological diagnosis is achieved by detecting specific IgG and IgM. Recently an automated chemiluminescent immunoassay (CLIA) (Liaison, DiaSorin, Italy) has been used to quantify the measles antibody. The aim of this study was to compare this assay with Enzygnost ELISA (Siemens, Germany), with final classification of discrepancies by indirect immunofluorescence (Euroimmun, Germany). For measles IgM, 204 sera were analyzed: 50 IgM-positive, 104 IgM-negative/ IgG-positive, and 50 from other viral infections (B19V, rubella, mumps, CMV, and EBV). For the measles IgG assay, 162 samples were tested: 106 were positive and 56 were negative. For measles IgM, the sensitivity and specificity of CLIA against ELISA were 94% (95% CI: 83.2–98.6) and 100% (95% CI: 97.1–100), respectively; the corrected figures after the final classification of discrepancies were 100% (95% CI: 91.0–100) and 99.4% (95% CI: 96.1–100), respectively. In relation to IgG, the sensitivity and specificity of CLIA against ELISA were, respectively, 97.2% (95% CI: 91.7–99.4) and 92.9% (95% CI: 82.5–97.7), and 95.5% (95% CI: 89.5–98.3) and 100% (95% CI: 91.8–100) after the final classification. CLIA showed excellent sensitivity and specificity in detecting measles IgG and IgM antibodies, eliminating the need to aliquot specimens before carrying out the assay. Key words: Chemiluminescent immunoassay; measles; enzyme immunoassay. Fernando de Ory, Laboratorio de Serologıa, Centro Nacional de Microbiologıa, Instituto de Salud Carlos III, Majadahonda, 28220 Madrid, Spain. e-mail: [email protected]
Measles virus (family Paramyxoviridae, genus Morbillivirus) (MV) causes a highly infectious but vaccine-preventable disease. In 2001, a plan for the elimination of measles was established by the Spanish Public Health authorities in accordance with the recommendations of the World Health Organization (WHO) (1). With the goal of eliminating measles, the identification of recent measles infection is important for clinical laboratories. Laboratory diagnosis of measles includes the reverse-transcription polymerase chain reaction (RT-PCR), done during the first 3 days after the exanthema, and serology by means of detection of measles-specific IgM (from the third day after onset) (2, 3) or seroconversion of IgG if the acute serum available is taken in the first 5 days, being negative for measles-specific IgM. RT-PCR in pharyngeal swabs and IgM detection have been the diagnostic Received 10 December 2014. Accepted 19 March 2015
648
methods of choice in a low-incidence epidemiological setting (4). Annual measles incidence in Spain markedly declined after the introduction of vaccination; between 1999 and 2007 it was ≤1 per 100 000 (5). However, the Spanish situation does not currently meet the elimination target of fewer than one case of measles per million in the population during the previous 12 months (6). Therefore, the use of reliable diagnostic methods is still recommended in Spain. There are various assays for detecting measles IgG and IgM that are mainly based on indirect immunofluorescence (IIF) and enzyme immunoassay (ELISA) in indirect or capture formats (7–10). Recently, fully automated chemiluminescent immunoassays (CLIAs) (Liaison, DiaSorin, Italy) have been used to detect measles IgM and IgG antibodies. The aim of this study was to compare this new assay with a commercially available ELISA (Enzygnost measles IgM and IgG, Siemens, Germany).
COMPARISON OF ASSAYS FOR MEASLES IgG AND IgM
MATERIAL AND METHODS Samples were selected on the basis of the results obtained by indirect ELISA, (Enzygnost) and were stored at 20 °C until use. The samples were received in our laboratory for diagnosing the corresponding agents or for determining the measles immune status. No information on vaccination was available from these samples. For measles IgM detection, 204 serum samples were analyzed. Of these, 50 were IgM ELISA-positive, obtained from individuals known to have had a recent measles infection; 104 were ELISA IgM-negative, showing specific IgG, from healthy people; 50 samples came from individuals with a recent infection by other viruses (Human Parvovirus B19 [B19V], family Parvoviridae, genus Erythroparvovirus; rubella virus, family Togaviridae, genus Rubivirus; mumps, family Paramyxoviridae, genus Rubulavirus; cytomegalovirus [CMV], family Herpesviridae, genus Cytomegalovirus; and Epstein–Barr virus [EBV], family Herpesviridae, genus Lymphocryptovirus; 10 of each). For the measles IgG assay, 162 samples were tested: 106 and 56 were positive and negative, respectively, in ELISA for measles IgG, from healthy people. All the samples were subjected to CLIA and ELISA for qualitative IgM and qualitative and quantitative IgG measles detection. Quantitative Enzygnost IgG results were calculated by the alpha method (Siemens) and expressed as mIU/mL. In the case of discrepancies, results were retested by both ELISA and CLIA, and assayed by IIF (Euroimmun, Germany); the final result was that obtained from this assay. Rubella, mumps and CMV IgM were assessed by indirect ELISA (Enzygnost), B19V IgM was examined by capture ELISA (Biotrin, Ireland), and EBV IgM was quantified by IIF (Meridian, USA). The IgG and IgM CLIA assays both use a baculovirusexpressed recombinant antigen. The assay for IgM is based on capture methodology. Magnetic particles are coated with anti-human IgM, which captures the IgM from the sample. Measles-specific IgM is identified by the addition of the antigen linked to an isoluminol derivative. The assay for IgG is based on indirect methodology. Magnetic particles are coated with the antigen. If present, measles-specific IgG binds to the solid phase, and is identified by an antiIgG linked to an isoluminol derivative. In both cases, the starter reagents are added, and a flash chemiluminescence reaction is generated. The light signal is measured in a photomultiplier, indicating the presence of IgM or IgG measles-specific antibodies. The determinations were done with the Liaison Platform (DiaSorin). In the IgG assay, the analyzer automatically calculates the antibody concentration, which is expressed in arbitrary units (AU/mL). All the serological assays were performed according to the manufacturer’s instructions. For calculations of sensitivity and specificity of CLIA against ELISA, indeterminate CLIA results were considered the most adverse; i.e., the indeterminate samples were classified as positive if negative results were obtained from ELISA, and classified as negative if ELISA results were positive. The 95% confidence interval (95% CI) of the percentages of sensitivity and specificity were calculated using GraphPad (La Jolla, CA, USA). The coefficient of determination (R2) of the pairs of quantitative IgG values from CLIA (range, ≥5 to ≤300 AU/mL) and Enzygnost (range, ≥150 to ≤3500 IU/
© 2015 APMIS. Published by John Wiley & Sons Ltd
mL) was calculated, and a scatterplot produced in Microsoft Office Excel 2003 (Redmond, WA, USA).
RESULTS Overall results are shown in Table 1. For measles IgM, when samples showing IgM positivity in ELISA were tested by CLIA, results were positive in 47, indeterminate in one, and negative in two cases. Thus, the relative sensitivity of CLIA against ELISA was 94% (95% CI: 83.2–98.6). All three discordant samples were negative with IIF, thereby giving a final sensitivity of 100% (95% CI: 91.0–100). On the other hand, the relative specificity of CLIA against ELISA was 100% (95% CI: 97.1–100) (154/154), with a sensitivity of 99.4% (95% CI: 96.1–100) after final classification of discrepancies (156/157). The results obtained when IgG ELISA-positive samples were tested by CLIA are shown in Table 1. The relative sensitivity and relative specificity of CLIA against ELISA were respectively 97.2% (95% CI: 91.7–99.4) and 92.9% (95% CI: 82.5–97.7). After IIF testing of discrepant samples, 110 were finally classified as positive and 52 as negative. Of the positive samples, 105 were positive, three were indeterminate and two were negative in CLIA, giving a relative sensitivity was 95.5% (95% CI: 89.5– 98.3). On the other hand, all 52 negative samples gave a negative result (specificity 100%; 95% CI: 91.8–100). The very high correlation between CLIA and Enzygnost (R2: 0.9156) is illustrated in Fig. 1.
DISCUSSION The WHO recognizes specific IgM detection as a method for confirming suspected measles cases (1). Table 1. Results for measles IgG and IgM obtained by chemiluminescent immunoassay (CLIA) in comparison to those of ELISA (Enzygnost, Siemens) and final classification of samples Assay and result ELISA Final classification1 Positive Negative Positive Negative CLIA-IgM Positive 47 0 47 0 Indeterminate 1 0 0 1 Negative 2 154 0 156 CLIA-IgG Positive 103 2 105 0 Indeterminate 1 2 3 0 Negative 2 52 2 52 1 Samples with discrepant CLIA and ELISA results were tested by IIF (Euroimmun). The final result was that obtained by this assay.
649
DE ORY et al.
There are a variety of immunoassays for this purpose, including indirect and capture ELISA and IIF, all of which are commercially available. The performance of some of these assays has been previously established (7–10). Recently, DiaSorin have developed a fully automated CLIA system, which has been applied to detect specific responses for the diagnosis of viral infections, including those to herpes simplex virus-1 and -2 IgG (11), hepatitis B surface antibodies (12), and to define seroprofiles to EBV (13, 14). In all the cited studies, appropriate sensitivity and specificity estimates relative to reference assays were obtained. This system has recently been applied to measles IgG and IgM. The aim of our study was to compare the system with a commercially available ELISA (Enzygnost), which has proved to be the best serological approach so far for diagnosing measles infections (9, 15), using an IIF assay for final classification of discrepant samples. The sensitivity of CLIA IgM compared with ELISA was 94%, rising to 100% after the final classification of samples by IIF. An important aspect of measles IgM detection, especially in an elimination scenario, is the differential diagnosis with other exanthematic diseases, such as rubella and B19V, as well as EBV and CMV. No false-positive Liaison measles IgM results were obtained from healthy individuals or those known to have been recently infected by other organisms (data not shown), thus ensuring the specificity of the assay. The main application of measles IgG assays is serosurveillance as part of the strategy for eliminating y = 0.0852x + 1.3106 R² = 0.9156
350
300
CLIA AU/mL
250
200
the disease. For years, the hemagglutination inhibition (HAI) test and the plaque reduction neutralization test (PRNT) (16) have been considered as reference assays. However, their standardization is hampered by the difficulties of obtaining adequate reagents, which makes them impractical for most clinical laboratories, especially for testing the large number of samples that are required for seroepidemiological studies. Automated approaches avoid the difficulties of HAI and PRNT. Enzygnost antimeasles IgG ELISA is the method of choice for this purpose, as it has excellent sensitivity and specificity compared with a functional assay such PRNT (17). When compared with Enzygnost ELISA, CLIA showed excellent relative sensitivity and specificity (97.2% and 92.9%, respectively). The relative specificity increased to 100% after the classification of discrepancies by means of IIF, with a minor drop in sensitivity (to 95.5%). Moreover, the R2 of >0.9 detected in this study indicates not only a good qualitative concordance of IgG but also a good correlation of quantitative results between the CLIA and ELISA methods. Additional potential advantages of CLIA are the automatic dilution of the samples from primary tubes, the possibility of running up to 15 assays simultaneously and the ability to process up to 18 samples per hour. One limitation of this study is the use of a criterion for the final classification of discrepant results based on an IIF technique. This is not a true gold standard for measles IgG (as PRNT or HAI) or IgM and was used as a tiebreaker in calculating the sensitivity and specificity of CLIA. As the samples were not characterized using a true gold standard, it is more accurate to say that the values of sensitivity and specificity obtained are a measure of the percentage agreement between the techniques. In conclusion, the fully automated CLIA from DiaSorin is a good alternative to a well-established commercial ELISA when used for measles IgG and IgM. This easy and rapid assay could be a very useful technique for diagnosing measles infections that eliminates the need to aliquot specimens prior to the assay.
150
This work was supported in part by a contract from DiaSorin Iberia SA and the Institute of Health Carlos III (MVP1235/12).
100
50
REFERENCES 0
0
500
1000
1500
2000
2500
3000
3500
4000
Enzygnost mIU/mL
Fig. 1. Scatterplot of semi-quantitative results from the IgG-measles CLIA and ELISA.
650
1. Centro Nacional de Epidemiologıa. ISCIII. Plan de Eliminaci on del Sarampi on en Espa~ na. 2000. http:// www.isciii.es/ISCIII/es/contenidos/fd-servicios-cientifi co-tecnicos/fd-vigilancias-alertas/fd-enfermedades/fd-en © 2015 APMIS. Published by John Wiley & Sons Ltd
COMPARISON OF ASSAYS FOR MEASLES IgG AND IgM
2.
3.
4.
5.
6. 7.
8.
9.
fermedades-prevenibles-vacunacion/PLANSARAMPION.pdf Mosquera MM, de Ory F, Gallardo V, Cuenca L, Morales M, S anchez Yebra W, et al. Evaluation of diagnostic markers for measles infection in the context of an outbreak in Spain. J Clin Microbiol 2005;43: 5117–21. Woo GK, Wong AH, Lee WY, Lau CS, Cheng PK, Leung PC, et al. Comparison of laboratory diagnostic methods for measles infection and identification of measles virus genotypes in Hong Kong. J Med Virol 2010;2010:1773–81. Jayamaha J, Binns PL, Fennell M, Ferson MJ, Newton P, Tran T, et al. Laboratory diagnosis, molecular characteristics, epidemiological and clinical features of an outbreak of measles in a low incidence population in Australia. J Clin Virol 2012;54:168–73. Pe~ na-Rey I, de Martınez Arag on V, Mosquera M, de Ory F, Echevarrıa JE. Measles Elimination Plan Working Group in Spain. Measles risk groups in Spain: implications for the European measles-elimination target. Vaccine 2009;27:3927–34. ECDC Surveillance Report. Measles and rubella monitoring. 2013. http://ecdc.europa.eu/en/publications/ Publications/measles-rubella-monitoring-may-2013.pdf Kleiman MB, Bockhold CA, Zimmerman SE, Griffin C, French ML, Barrett C. Adaptation of a commercially available indirect fluorescent antibody slide test for measuring measles-specific immunoglobulins. Diagn Microbiol Infect Dis 1986;4:285–90. Arista S, Ferraro D, Cascio A, Vizzi E, di Stefano R. Detection of IgM antibodies specific for measles virus by capture and indirect enzyme immunoassays. Res Virol 1995;146:225–32. Ratnam S, Tipples G, Head C, Fauvel M, Fearon M, Ward BJ. Performance of indirect immunoglobulin M (IgM) serology tests and IgM capture assays for laboratory diagnosis of measles. J Clin Microbiol 2000;38:99–104.
© 2015 APMIS. Published by John Wiley & Sons Ltd
10. De Ory F. Evaluaci on de un ensayo de enzimoinmunoan alisis (ELISA) de captura para el diagn ostico de la infecci on por el virus del sarampi on. Enferm Infecc Microbiol Clin 2001;19:409–11. 11. Maters AW, Wright CV, Lee MT, Schwichtenberg G, Detrick B. Detection of type-specific antibodies to HSV-1 and HSV-2: comparative analysis of a chemiluminescence immunoassay with a conventional ELISA. Diagn Microbiol Infect Dis 2012;73: 273–4. 12. Kinn S, Akhavan S, Agut H, Thibault V. Performance of the DiaSorin LIAISON(Ò) anti-HBs II for the detection of hepatitis B surface antibodies: comparison with the Abbott Architect anti-HBs assay. J Clin Virol 2011;50:297–302. 13. Berth M, Bosmans E. Comparison of three automated immunoassay methods for the determination of Epstein-Barr virus-specific immunoglobulin M. Clin Vaccine Immunol 2010;17:559–63. 14. de Ory F, Guisasola ME, Sanz JC, Garcıa-Bermejo I. Evaluation of four commercial systems for the diagnosis of Epstein-Barr virus primary infections. Clin Vaccine Immunol 2011;18:444–8. 15. Sanz JC, Mosquera M, Ramos B, Ramırez R, de Ory F, Echevarria JE. Assessment of RNA amplification by multiplex RT-PCR and IgM detection by indirect and capture ELISAs for the diagnosis of measles and rubella. APMIS 2010;118:203–9. 16. Cohen BJ, Doblas D, Andrews N. Comparison of plaque reduction neutralisation test (PRNT) and measles virus-specific IgG ELISA for assessing immunogenicity of measles vaccination. Vaccine 2008;26: 6392–7. 17. Tischer A, Andrews N, Kafatos G, Nardone A, Berbers G, Davidkin I, et al. Standardization of measles, mumps and rubella assays to enable comparisons of seroprevalence data across 21 European countries and Australia. Epidemiol Infect 2007;135: 787–97.
651
APMIS 123: 648–651
Comparison of chemiluminescent immunoassay and ELISA for measles IgG and IgM 1 and JUAN C. SANZ2,3 FERNANDO DE ORY,1,2 TEODORA MINGUITO,1 PILAR BALFAGON 1
Laboratorio de Serologıa, Centro Nacional de Microbiologıa, Instituto de Salud Carlos III, Majadahonda, Madrid; 2CIBER de Epidemiologıa y Salud P ublica (CIBERESP), Barcelona; and 3Laboratorio de Salud P ublica de la Comunidad de Madrid, Madrid, Spain
de Ory F, Minguito T, Balfag on P, Sanz JC. Comparison of chemiluminescent immunoassay and ELISA for measles IgG and IgM. APMIS 2015 123: 648–651. In the context of measles elimination, the identification of recent infections is important for clinical laboratories. Serological diagnosis is achieved by detecting specific IgG and IgM. Recently an automated chemiluminescent immunoassay (CLIA) (Liaison, DiaSorin, Italy) has been used to quantify the measles antibody. The aim of this study was to compare this assay with Enzygnost ELISA (Siemens, Germany), with final classification of discrepancies by indirect immunofluorescence (Euroimmun, Germany). For measles IgM, 204 sera were analyzed: 50 IgM-positive, 104 IgM-negative/ IgG-positive, and 50 from other viral infections (B19V, rubella, mumps, CMV, and EBV). For the measles IgG assay, 162 samples were tested: 106 were positive and 56 were negative. For measles IgM, the sensitivity and specificity of CLIA against ELISA were 94% (95% CI: 83.2–98.6) and 100% (95% CI: 97.1–100), respectively; the corrected figures after the final classification of discrepancies were 100% (95% CI: 91.0–100) and 99.4% (95% CI: 96.1–100), respectively. In relation to IgG, the sensitivity and specificity of CLIA against ELISA were, respectively, 97.2% (95% CI: 91.7–99.4) and 92.9% (95% CI: 82.5–97.7), and 95.5% (95% CI: 89.5–98.3) and 100% (95% CI: 91.8–100) after the final classification. CLIA showed excellent sensitivity and specificity in detecting measles IgG and IgM antibodies, eliminating the need to aliquot specimens before carrying out the assay. Key words: Chemiluminescent immunoassay; measles; enzyme immunoassay. Fernando de Ory, Laboratorio de Serologıa, Centro Nacional de Microbiologıa, Instituto de Salud Carlos III, Majadahonda, 28220 Madrid, Spain. e-mail: [email protected]
Measles virus (family Paramyxoviridae, genus Morbillivirus) (MV) causes a highly infectious but vaccine-preventable disease. In 2001, a plan for the elimination of measles was established by the Spanish Public Health authorities in accordance with the recommendations of the World Health Organization (WHO) (1). With the goal of eliminating measles, the identification of recent measles infection is important for clinical laboratories. Laboratory diagnosis of measles includes the reverse-transcription polymerase chain reaction (RT-PCR), done during the first 3 days after the exanthema, and serology by means of detection of measles-specific IgM (from the third day after onset) (2, 3) or seroconversion of IgG if the acute serum available is taken in the first 5 days, being negative for measles-specific IgM. RT-PCR in pharyngeal swabs and IgM detection have been the diagnostic Received 10 December 2014. Accepted 19 March 2015
648
methods of choice in a low-incidence epidemiological setting (4). Annual measles incidence in Spain markedly declined after the introduction of vaccination; between 1999 and 2007 it was ≤1 per 100 000 (5). However, the Spanish situation does not currently meet the elimination target of fewer than one case of measles per million in the population during the previous 12 months (6). Therefore, the use of reliable diagnostic methods is still recommended in Spain. There are various assays for detecting measles IgG and IgM that are mainly based on indirect immunofluorescence (IIF) and enzyme immunoassay (ELISA) in indirect or capture formats (7–10). Recently, fully automated chemiluminescent immunoassays (CLIAs) (Liaison, DiaSorin, Italy) have been used to detect measles IgM and IgG antibodies. The aim of this study was to compare this new assay with a commercially available ELISA (Enzygnost measles IgM and IgG, Siemens, Germany).
COMPARISON OF ASSAYS FOR MEASLES IgG AND IgM
MATERIAL AND METHODS Samples were selected on the basis of the results obtained by indirect ELISA, (Enzygnost) and were stored at 20 °C until use. The samples were received in our laboratory for diagnosing the corresponding agents or for determining the measles immune status. No information on vaccination was available from these samples. For measles IgM detection, 204 serum samples were analyzed. Of these, 50 were IgM ELISA-positive, obtained from individuals known to have had a recent measles infection; 104 were ELISA IgM-negative, showing specific IgG, from healthy people; 50 samples came from individuals with a recent infection by other viruses (Human Parvovirus B19 [B19V], family Parvoviridae, genus Erythroparvovirus; rubella virus, family Togaviridae, genus Rubivirus; mumps, family Paramyxoviridae, genus Rubulavirus; cytomegalovirus [CMV], family Herpesviridae, genus Cytomegalovirus; and Epstein–Barr virus [EBV], family Herpesviridae, genus Lymphocryptovirus; 10 of each). For the measles IgG assay, 162 samples were tested: 106 and 56 were positive and negative, respectively, in ELISA for measles IgG, from healthy people. All the samples were subjected to CLIA and ELISA for qualitative IgM and qualitative and quantitative IgG measles detection. Quantitative Enzygnost IgG results were calculated by the alpha method (Siemens) and expressed as mIU/mL. In the case of discrepancies, results were retested by both ELISA and CLIA, and assayed by IIF (Euroimmun, Germany); the final result was that obtained from this assay. Rubella, mumps and CMV IgM were assessed by indirect ELISA (Enzygnost), B19V IgM was examined by capture ELISA (Biotrin, Ireland), and EBV IgM was quantified by IIF (Meridian, USA). The IgG and IgM CLIA assays both use a baculovirusexpressed recombinant antigen. The assay for IgM is based on capture methodology. Magnetic particles are coated with anti-human IgM, which captures the IgM from the sample. Measles-specific IgM is identified by the addition of the antigen linked to an isoluminol derivative. The assay for IgG is based on indirect methodology. Magnetic particles are coated with the antigen. If present, measles-specific IgG binds to the solid phase, and is identified by an antiIgG linked to an isoluminol derivative. In both cases, the starter reagents are added, and a flash chemiluminescence reaction is generated. The light signal is measured in a photomultiplier, indicating the presence of IgM or IgG measles-specific antibodies. The determinations were done with the Liaison Platform (DiaSorin). In the IgG assay, the analyzer automatically calculates the antibody concentration, which is expressed in arbitrary units (AU/mL). All the serological assays were performed according to the manufacturer’s instructions. For calculations of sensitivity and specificity of CLIA against ELISA, indeterminate CLIA results were considered the most adverse; i.e., the indeterminate samples were classified as positive if negative results were obtained from ELISA, and classified as negative if ELISA results were positive. The 95% confidence interval (95% CI) of the percentages of sensitivity and specificity were calculated using GraphPad (La Jolla, CA, USA). The coefficient of determination (R2) of the pairs of quantitative IgG values from CLIA (range, ≥5 to ≤300 AU/mL) and Enzygnost (range, ≥150 to ≤3500 IU/
© 2015 APMIS. Published by John Wiley & Sons Ltd
mL) was calculated, and a scatterplot produced in Microsoft Office Excel 2003 (Redmond, WA, USA).
RESULTS Overall results are shown in Table 1. For measles IgM, when samples showing IgM positivity in ELISA were tested by CLIA, results were positive in 47, indeterminate in one, and negative in two cases. Thus, the relative sensitivity of CLIA against ELISA was 94% (95% CI: 83.2–98.6). All three discordant samples were negative with IIF, thereby giving a final sensitivity of 100% (95% CI: 91.0–100). On the other hand, the relative specificity of CLIA against ELISA was 100% (95% CI: 97.1–100) (154/154), with a sensitivity of 99.4% (95% CI: 96.1–100) after final classification of discrepancies (156/157). The results obtained when IgG ELISA-positive samples were tested by CLIA are shown in Table 1. The relative sensitivity and relative specificity of CLIA against ELISA were respectively 97.2% (95% CI: 91.7–99.4) and 92.9% (95% CI: 82.5–97.7). After IIF testing of discrepant samples, 110 were finally classified as positive and 52 as negative. Of the positive samples, 105 were positive, three were indeterminate and two were negative in CLIA, giving a relative sensitivity was 95.5% (95% CI: 89.5– 98.3). On the other hand, all 52 negative samples gave a negative result (specificity 100%; 95% CI: 91.8–100). The very high correlation between CLIA and Enzygnost (R2: 0.9156) is illustrated in Fig. 1.
DISCUSSION The WHO recognizes specific IgM detection as a method for confirming suspected measles cases (1). Table 1. Results for measles IgG and IgM obtained by chemiluminescent immunoassay (CLIA) in comparison to those of ELISA (Enzygnost, Siemens) and final classification of samples Assay and result ELISA Final classification1 Positive Negative Positive Negative CLIA-IgM Positive 47 0 47 0 Indeterminate 1 0 0 1 Negative 2 154 0 156 CLIA-IgG Positive 103 2 105 0 Indeterminate 1 2 3 0 Negative 2 52 2 52 1 Samples with discrepant CLIA and ELISA results were tested by IIF (Euroimmun). The final result was that obtained by this assay.
649
DE ORY et al.
There are a variety of immunoassays for this purpose, including indirect and capture ELISA and IIF, all of which are commercially available. The performance of some of these assays has been previously established (7–10). Recently, DiaSorin have developed a fully automated CLIA system, which has been applied to detect specific responses for the diagnosis of viral infections, including those to herpes simplex virus-1 and -2 IgG (11), hepatitis B surface antibodies (12), and to define seroprofiles to EBV (13, 14). In all the cited studies, appropriate sensitivity and specificity estimates relative to reference assays were obtained. This system has recently been applied to measles IgG and IgM. The aim of our study was to compare the system with a commercially available ELISA (Enzygnost), which has proved to be the best serological approach so far for diagnosing measles infections (9, 15), using an IIF assay for final classification of discrepant samples. The sensitivity of CLIA IgM compared with ELISA was 94%, rising to 100% after the final classification of samples by IIF. An important aspect of measles IgM detection, especially in an elimination scenario, is the differential diagnosis with other exanthematic diseases, such as rubella and B19V, as well as EBV and CMV. No false-positive Liaison measles IgM results were obtained from healthy individuals or those known to have been recently infected by other organisms (data not shown), thus ensuring the specificity of the assay. The main application of measles IgG assays is serosurveillance as part of the strategy for eliminating y = 0.0852x + 1.3106 R² = 0.9156
350
300
CLIA AU/mL
250
200
the disease. For years, the hemagglutination inhibition (HAI) test and the plaque reduction neutralization test (PRNT) (16) have been considered as reference assays. However, their standardization is hampered by the difficulties of obtaining adequate reagents, which makes them impractical for most clinical laboratories, especially for testing the large number of samples that are required for seroepidemiological studies. Automated approaches avoid the difficulties of HAI and PRNT. Enzygnost antimeasles IgG ELISA is the method of choice for this purpose, as it has excellent sensitivity and specificity compared with a functional assay such PRNT (17). When compared with Enzygnost ELISA, CLIA showed excellent relative sensitivity and specificity (97.2% and 92.9%, respectively). The relative specificity increased to 100% after the classification of discrepancies by means of IIF, with a minor drop in sensitivity (to 95.5%). Moreover, the R2 of >0.9 detected in this study indicates not only a good qualitative concordance of IgG but also a good correlation of quantitative results between the CLIA and ELISA methods. Additional potential advantages of CLIA are the automatic dilution of the samples from primary tubes, the possibility of running up to 15 assays simultaneously and the ability to process up to 18 samples per hour. One limitation of this study is the use of a criterion for the final classification of discrepant results based on an IIF technique. This is not a true gold standard for measles IgG (as PRNT or HAI) or IgM and was used as a tiebreaker in calculating the sensitivity and specificity of CLIA. As the samples were not characterized using a true gold standard, it is more accurate to say that the values of sensitivity and specificity obtained are a measure of the percentage agreement between the techniques. In conclusion, the fully automated CLIA from DiaSorin is a good alternative to a well-established commercial ELISA when used for measles IgG and IgM. This easy and rapid assay could be a very useful technique for diagnosing measles infections that eliminates the need to aliquot specimens prior to the assay.
150
This work was supported in part by a contract from DiaSorin Iberia SA and the Institute of Health Carlos III (MVP1235/12).
100
50
REFERENCES 0
0
500
1000
1500
2000
2500
3000
3500
4000
Enzygnost mIU/mL
Fig. 1. Scatterplot of semi-quantitative results from the IgG-measles CLIA and ELISA.
650
1. Centro Nacional de Epidemiologıa. ISCIII. Plan de Eliminaci on del Sarampi on en Espa~ na. 2000. http:// www.isciii.es/ISCIII/es/contenidos/fd-servicios-cientifi co-tecnicos/fd-vigilancias-alertas/fd-enfermedades/fd-en © 2015 APMIS. Published by John Wiley & Sons Ltd
COMPARISON OF ASSAYS FOR MEASLES IgG AND IgM
2.
3.
4.
5.
6. 7.
8.
9.
fermedades-prevenibles-vacunacion/PLANSARAMPION.pdf Mosquera MM, de Ory F, Gallardo V, Cuenca L, Morales M, S anchez Yebra W, et al. Evaluation of diagnostic markers for measles infection in the context of an outbreak in Spain. J Clin Microbiol 2005;43: 5117–21. Woo GK, Wong AH, Lee WY, Lau CS, Cheng PK, Leung PC, et al. Comparison of laboratory diagnostic methods for measles infection and identification of measles virus genotypes in Hong Kong. J Med Virol 2010;2010:1773–81. Jayamaha J, Binns PL, Fennell M, Ferson MJ, Newton P, Tran T, et al. Laboratory diagnosis, molecular characteristics, epidemiological and clinical features of an outbreak of measles in a low incidence population in Australia. J Clin Virol 2012;54:168–73. Pe~ na-Rey I, de Martınez Arag on V, Mosquera M, de Ory F, Echevarrıa JE. Measles Elimination Plan Working Group in Spain. Measles risk groups in Spain: implications for the European measles-elimination target. Vaccine 2009;27:3927–34. ECDC Surveillance Report. Measles and rubella monitoring. 2013. http://ecdc.europa.eu/en/publications/ Publications/measles-rubella-monitoring-may-2013.pdf Kleiman MB, Bockhold CA, Zimmerman SE, Griffin C, French ML, Barrett C. Adaptation of a commercially available indirect fluorescent antibody slide test for measuring measles-specific immunoglobulins. Diagn Microbiol Infect Dis 1986;4:285–90. Arista S, Ferraro D, Cascio A, Vizzi E, di Stefano R. Detection of IgM antibodies specific for measles virus by capture and indirect enzyme immunoassays. Res Virol 1995;146:225–32. Ratnam S, Tipples G, Head C, Fauvel M, Fearon M, Ward BJ. Performance of indirect immunoglobulin M (IgM) serology tests and IgM capture assays for laboratory diagnosis of measles. J Clin Microbiol 2000;38:99–104.
© 2015 APMIS. Published by John Wiley & Sons Ltd
10. De Ory F. Evaluaci on de un ensayo de enzimoinmunoan alisis (ELISA) de captura para el diagn ostico de la infecci on por el virus del sarampi on. Enferm Infecc Microbiol Clin 2001;19:409–11. 11. Maters AW, Wright CV, Lee MT, Schwichtenberg G, Detrick B. Detection of type-specific antibodies to HSV-1 and HSV-2: comparative analysis of a chemiluminescence immunoassay with a conventional ELISA. Diagn Microbiol Infect Dis 2012;73: 273–4. 12. Kinn S, Akhavan S, Agut H, Thibault V. Performance of the DiaSorin LIAISON(Ò) anti-HBs II for the detection of hepatitis B surface antibodies: comparison with the Abbott Architect anti-HBs assay. J Clin Virol 2011;50:297–302. 13. Berth M, Bosmans E. Comparison of three automated immunoassay methods for the determination of Epstein-Barr virus-specific immunoglobulin M. Clin Vaccine Immunol 2010;17:559–63. 14. de Ory F, Guisasola ME, Sanz JC, Garcıa-Bermejo I. Evaluation of four commercial systems for the diagnosis of Epstein-Barr virus primary infections. Clin Vaccine Immunol 2011;18:444–8. 15. Sanz JC, Mosquera M, Ramos B, Ramırez R, de Ory F, Echevarria JE. Assessment of RNA amplification by multiplex RT-PCR and IgM detection by indirect and capture ELISAs for the diagnosis of measles and rubella. APMIS 2010;118:203–9. 16. Cohen BJ, Doblas D, Andrews N. Comparison of plaque reduction neutralisation test (PRNT) and measles virus-specific IgG ELISA for assessing immunogenicity of measles vaccination. Vaccine 2008;26: 6392–7. 17. Tischer A, Andrews N, Kafatos G, Nardone A, Berbers G, Davidkin I, et al. Standardization of measles, mumps and rubella assays to enable comparisons of seroprevalence data across 21 European countries and Australia. Epidemiol Infect 2007;135: 787–97.
651
