IgG - Journal of Clinical Microbiology - American Society for

JOURNAL OF CLINICAL MICROBIOLOGY, May 2008, p. 1847–1849 0095-1137/08/$08.00⫹0 doi:10.1128/JCM.00468-08 Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Vol. 46, No. 5

Evaluation of Rubella Virus Immunoglobulin G (IgG) and IgM Assays with the New Vidia Instrument䌤 Maria Cristina Medici,* Monica Martinelli, Valeria Albonetti, Carlo Chezzi, and Giuseppe Dettori Section of Microbiology, Unit of Virology, Department of Pathology and Laboratory Medicine, Medical School, University of Parma, Parma, Italy Received 10 March 2008/Accepted 12 March 2008

For rubella virus immunoglobulin G (IgG) and IgM detection, Vidia assays were compared to Vidas, AxSYM, and Liaison assays with 419 serum samples. Only Vidia produced a sensitivity of 100% for IgG and IgM. Vidia specificities were 98.4% for IgG and 99.8% for IgM, versus Vidas specificities of 100 and 99.3%. Vidia IgG and IgM assays performed equally well. ufacturers’ instructions. A positive or negative status was assigned to each sample analyzed when at least three concordant results were available. Samples for which an equivocal or unresolved status was assigned, 21 serum samples (5.0%) for IgG (Table 1) and 7 serum samples (1.7%) for IgM (Table 2), were excluded from the analysis. A discrepancy analysis was performed by repeating the test with the discordant result and, for IgM, also by IgG avidity testing with the BEIA rubella IgG avidity assay (Bouty, Milan, Italy). For RV IgG, the sensitivity of the Vidia assay in both studies was 100% (Table 1). The overall sensitivities of Vidas, AxSYM, and Liaison assays were 99.4, 99.1, and 98.2%, respectively. The overall specificity of the Vidia RV IgG assay has been established to be 98.4%, whereas Vidas and Liaison assays gave specificities of 100% and the AxSYM assay gave a specificity of 76.6%. When the samples with discordant results were reanalyzed (Table 3), among the 334 samples ultimately determined to be positive for RV IgG, Vidia gave equivocal results for 2 samples, AxSYM gave equivocal results for 3 samples, and Liaison gave equivocal results for 4 samples. Liaison gave two false-negative results. No false-negative or equivocal result was resolved by repeating the assay with the discordant result, except one equivocal result from AxSYM, which increased the sensitivity of this assay to 99.4%. For the 64 samples ultimately determined to be negative for RV IgM, Vidia gave 1 equivocal result whereas AxSYM gave 5 equivocal and 10 false-positive results. By repeating the assays with the discordant results, one equivocal result from the AxSYM assay was resolved, which increased the specificity of this assay to 78.1%. For RV IgM, in the retrospective study the sensitivity was 100% for all methods except AxSYM (25.0%) (Table 2). In the prospective study, the sensitivity was not determinable as no positive serum was obtained. The overall specificity was 99.8% for the Vidia RV IgM assay, compared to 99.3% for Vidas, 98.0% for AxSYM, and 97.8% for Liaison. When the samples with discordant results were reanalyzed (Table 3), for the four samples ultimately determined to be positive for RV IgM, AxSYM gave one equivocal and two false-negative results. For the 408 samples determined to be negative for RV IgM, Vidia gave 1 false-positive result, Vidas gave 3 equivocal results, AxSYM gave 2 equivocal and 6 false-positive results, and Li-

Infections caused by rubella virus (RV) are usually mild and often subclinical (3, 9). However, serious sequelae for the newborn result from maternal infection. Although rubella vaccination has reduced the incidence of RV infection substantially, it is still important to determine the immune status of women of childbearing age, pregnant women, and health care workers. RV-specific immunoglobulin M (RV IgM) is detectable after the incubation period and usually disappears after approximately 8 weeks. RV IgG levels rise more slowly and persist for life (9). As the RV IgG response matures, the avidity of the antibody reaction to RV increases (5, 6). Accurate RV IgM and IgG assays are critical to diagnose RV infection and to clearly understand the serological status of the subject. The purpose of this study was to evaluate the analytical performances of the new automated Vidia Rub IgG and IgM assays (bioMe´rieux, Marcy l’Etoile, France) in comparison with those of Vidas Rub IgG and IgM assays (bioMe´rieux), AxSYM rubella IgG and IgM assays (Abbott Laboratories, Abbott Park, IL), and Liaison rubella IgG and IgM assays (DiaSorin, Saluggia, Italy). A total of 419 serum samples, 209 frozen (retrospective study) and 210 fresh (prospective study), obtained for diagnostic RV IgG and IgM testing at the Virology Unit of the University Hospital in Parma, Italy, were examined. Before testing, frozen samples were thawed and clarified by centrifugation at 2,000 ⫻ g for 10 min. Both thawed and fresh samples were kept at 4°C until use (no more than 5 days). The 419 samples corresponded to 405 subjects (195 for the retrospective study and 210 for the prospective study; median age, 32 years; age range, 3 months to 87 years): 232 pregnant or prepregnancy women, 69 immunocompromised patients, 35 children or infants (3 of which were immunocompromised), 54 healthy adults, and 15 subjects for which no clinical data were available. Each sample was tested in parallel or within 24 h and in singlicate by the four automated systems according to the man-

* Corresponding author. Mailing address: Section of Microbiology, Department of Pathology and Laboratory Medicine, University of Parma, Viale Antonio Gramsci, 14, 43100 Parma, Italy. Phone: 39-0521-988885. Fax: 39-0521-993620. E-mail: mariacristina.medici @unipr.it. 䌤 Published ahead of print on 26 March 2008. 1847

1848

NOTES

J. CLIN. MICROBIOL. TABLE 1. RV IgG assay performancea

Study type(s) and assay Retrospective study Vidia Vidas AxSYM Liaison

No. of samples for study(ies) 209

No. No. No. with final No. with No. of No. of with with equivocal or initially false false negative positive unresolved equivocal positives negatives b result result status result

210

419

% Agreement

42 42 28 42

154 153 154 152

42

154

21 22 21 22

180 179 177 176

22

180

63 64 49 64

334 332 331 328

64

334

0 0 10 0

0 0 0 1

0 1 4 1

100.0 (97.5–100.0) 100.0 (91.3–100.0) 100.0 (98.0–100.0) 99.4 (97.5–100.0) 100.0 (91.3–100.0) 99.5 (97.1–99.9) 100.0 (97.5–100.0) 66.7 (57.7–85.2) 92.9 (88.3–95.7) 98.7 (96.3–99.9) 100.0 (91.3–100.0) 99.0 (96.3–99.7)

0 0 0 0

0 0 0 1

1 1 4 3

100.0 (97.8–100.0) 95.5 (84.0–100.0) 99.4 (97.8–100.0) 100.0 (84.6–100.0) 98.3 (97.8–100.0) 95.5 (84.0–100.0) 97.8 (96.8–99.9) 100.0 (84.6–100.0)

99.5 (97.2–99.9) 99.5 (97.2–99.9) 98.0 (94.9–99.2) 98.0 (94.9–99.2)

0 0 10 0

0 0 0 2

1 2 8 4

100.0 (98.8–100.0) 98.4 (94.0–100.0) 99.4 (98.8–100.0) 100.0 (94.1–100.0) 99.1 (98.8–100.0) 76.6 (71.3–90.6) 98.2 (97.8–99.8) 100.0 (94.1–100.0)

99.7 (98.6–100.0) 99.5 (98.1–99.9) 95.5 (92.9–97.1) 98.5 (96.7–99.3)

8

Total for study type All study types Vidia Vidas AxSYM Liaison

% Specificity

13

Total for study type Prospective study Vidia Vidas AxSYM Liaison

% Sensitivity

21

Total for all study types a

Values in parentheses are 95% confidence intervals. Samples for which the ultimate status was equivocal or unresolved were excluded from the analysis and are not included in the number of samples with an initially equivocal result. b

aison gave 1 equivocal and 8 false-positive results. After the repetition of the assays with discordant results, the specificities of the AxSYM and Liaison RV IgM assays increased to 98.3 and 98.5%, respectively. In the past, RV IgG assays have been rarely evaluated. In

comparative studies with the AxSYM and Vidas assays, the sensitivities were 99.8 and 99.6%, respectively, and the specificities ranged from 81.5% for AxSYM to an undeterminable value for Vidas due to the high vaccination rate (1, 8). No previous data are available for Vidia and Liaison RV assays. In

TABLE 2. RV IgM assay performancea Study type(s) and assay Retrospective study Vidia Vidas AxSYM Liaison

No. of samples for study(ies) 209

No. No. No. with final No. with No. of No. of with with equivocal or initially false false negative positive unresolved equivocal positives negatives b result result status result

210

Total for all study types a

419

% Agreement

199 198 194 192

4 4 1 4

99.5 (97.1–99.9) 99.0 (98.0–100.0) 97.0 (94.2–98.9) 96.0 (92.8–98.3)

99.5 (97.2–99.9) 99.0 (96.4–99.7) 95.6 (91.7–97.7) 96.1 (92.4–98.0)

200

4

208 207 206 207

0 0 0 0

208

0

407 405 400 399

4 4 1 4

408

4

1 0 5 7

0 0 2 0

0 2 2 1

100.0 (50.0–100.0) 100.0 (50.0–100.0) 25.0 (6.0–79.7) 100.0 (50.0–100.0)

0 0 1 1

0 0 0 0

0 1 1 0

ND ND ND ND

1 0 6 8

0 0 2 0

0 3 3 1

100.0 (50.0–100.0) 100.0 (50.0–100.0) 25.0 (6.0–79.7) 100.0 (50.0–100.0)

2

Total for study type All study types Vidia Vidas AxSYM Liaison

% Specificity

5

Total for study type Prospective study Vidia Vidas AxSYM Liaison

% Sensitivity

100.0 (98.1–100.0) 100.0 (98.1–100.0) 99.5 (98.1–100.0) 99.5 (97.3–99.9) 99.0 (97.2–99.9) 99.0 (96.5–99.7) 99.5 (97.3–99.9) 99.5 (97.3–99.9)

7 99.8 (98.6–100.0) 99.3 (99.0–100.0) 98.0 (96.8–99.3) 97.8 (96.1–99.0)

99.8 (98.6–100.0) 99.3 (97.8–99.8) 97.3 (95.2–98.5) 97.8 (95.9–98.9)

Values in parentheses are 95% confidence intervals. ND, not determined due to low prevalence in population. Samples for which the ultimate status was equivocal or unresolved were excluded from the analysis and are not included in the number of samples with an initially equivocal result. b

VOL. 46, 2008

NOTES

1849

TABLE 3. Discrepancy analysis of four RV assays testing 419 samplesa

Antibody tested and assay

No. of samples determined to be positive

IgG Vidia Vidas AxSYM Liaison

334

IgM Vidia Vidas AxSYM Liaison

4

No. of samples with the following result among samples determined to be positive: Equivocal

False negative

% Sensitivity after analysis

Before

After

Before

After

0 2 3 4

0 2 2 4

0 0 0 2

0 0 0 2

100 99.4 99.4 98.2

0 0 1 0

0 0 1 0

0 0 2 0

0 0 2 0

100 100 25 100

No. of samples determined to be negative

No. of samples with the following result among samples determined to be negative: Equivocal

False positive

% Specificity after analysis

Before

After

Before

After

1 0 5 0

1 0 3 0

0 0 10 0

0 0 11 0

98.4 100 78.1 100

0 3 2 1

0 3 0 1

1 0 6 8

1 0 7 5

99.8 99.3 98.3 98.5

64

408

a Before, before repetition of discordant assays; after, after repetition of discordant assays. Sensitivities and specificities that changed after the discrepancy analysis are in boldface.

this study, for RV IgG, only Vidia produced a sensitivity of 100%, and all methods except AxSYM produced high specificities (from 98.4 to 100%). The inclusion of frozen samples gave the opportunity to introduce RV IgM-positive samples (giving an artificial prevalence of infection). However, the low number of these samples, reflecting the widespread RV vaccination program instituted in Italy, did not enable us to obtain reliable sensitivities. Vidia, Vidas, and Liaison RV IgM assays detected all four of these RV IgM-positive samples, whereas AxSYM detected only one of the four (25.0%). Elsewhere, the sensitivities of Vidas and AxSYM RV IgM assays have been reported to range from 84.2 to 91.2% and 94.7 to 96.5%, respectively (2). In other comparative evaluations, the sensitivity of Vidas was reported to be 94.6% (4) and that of AxSYM was reported to be 100% (1). The specificities of Vidas and AxSYM RV IgM assays were reported to range from 90.1 to 100% and from 86.5 to 99.2%, respectively (1, 2, 4, 7), compared to 99.8% for Vidas and 98.3% (after the discrepancy analysis) for AxSYM in the present evaluation, in which Vidia and Liaison specificities were 99.8 and 98.5% (after the discrepancy analysis), respectively. In this study, the best overall accuracies for both RV IgG and IgM assays were obtained with the Vidia (99.7 and 99.8%, respectively) and Vidas (99.5 and 99.3%, respectively) assays. Vidia assays gave no false-negative results, and except for Vidas assays, which gave no false-negative and no false-positive results, gave the fewest false-positive results (one for IgM in a 4-year-old child with high-level IgG avidity). In conclusion, the Vidia and Vidas RV IgG and IgM assays performed equally well, while the AxSYM and Liaison assays had lower sensitivities and specificities. False-positive results have been well described and must be considered in interpreting viral IgM serology results. In countries with established vaccination programs, the incidence of acute wild-type virus infection is extremely low. Therefore, a positive reaction in an

RV IgM assay is more likely to be due to the cross-reactivity of a polyclonal IgM response elicited by infection with other infectious agents or an autoimmune antibody than to RV infection (7). Further laboratory investigations to confirm the reactivity to acute RV infection, such as RV nucleic acid and/or RV IgG avidity testing, are essential. This study was funded by bioMe´rieux, Marcy l’Etoile, France. The findings and conclusions in this report are ours and do not necessarily represent the views of the funding manufacturer. We thank Daniela Gennari for excellent technical assistance. REFERENCES 1. Diepersloot, R. J., H. Dunnewold-Hoekstra, J. Kruit-Den Hollander, and F. Vlaspolder. 2001. Antenatal screening for hepatitis B and antibodies to Toxoplasma gondii and rubella virus: evaluation of two commercial immunoassay systems. Clin. Diagn. Lab. Immunol. 8:785–787. 2. Dimech, W., L. Panagiotopoulos, J. Marler, N. Laven, S. Leeson, and E. M. Dax. 2005. Evaluation of three immunoassays used for detection of antirubella virus immunoglobulin M antibodies. Clin. Diagn. Lab. Immunol. 12: 1104–1108. 3. Fuccillo, D. A., and J. L. Severs. 1992. Measles, mumps, and rubella, p. 577. In S. Specter and G. Lancz (ed.), Clinical virology manual, 2nd ed. Elsevier Scientific Publishing, New York, NY. 4. Matter, L., M. Gorgievski-Hrisoho, and D. Germann. 1994. Comparison of four enzyme immunoassays for detection of immunoglobulin M antibodies against rubella virus. J. Clin. Microbiol. 32:2134–2139. 5. Thomas, H. I., P. Morgan-Capner, G. Enders, S. O’Shea, D. Caldicott, and J. M. Best. 1992. Persistence of specific IgM and low avidity specific IgG1 following primary rubella. J. Virol. Methods 39:149–155. 6. Thomas, H. I., P. Morgan-Capner, J. E. Cradock-Watson, G. Enders, J. M. Best, and S. O’Shea. 1993. Slow maturation of IgG1 avidity and persistence of specific IgM in congenital rubella: implications for diagnosis and immunopathology. J. Med. Virol. 41:196–200. 7. Tipples, G. A., R. Hamkar, T. Mohtktari-Azad, M. Gray, J. Ball, C. Head, and S. Ratnam. 2004. Evaluation of rubella IgM enzyme immunoassays. J. Clin. Virol. 30:233–238. 8. Vlaspolder, F., P. Singer, A. Smit, and R. J. Diepersloot. 2001. Comparison of Immulite with Vidas for detection of infection in a low-prevalence population of pregnant women in The Netherlands. Clin. Diagn. Lab. Immunol. 8:552– 555. 9. White, D. O., and F. J. Fenner. 1994. Medical virology, 4th ed. Academic Press, San Diego, CA.