Previous Article | Next Article ![]()
Journal of Clinical Microbiology, March 2002, p. 973-978, Vol. 40, No. 3
0095-1137/02/$04.00+0 DOI: 10.1128/JCM.40.3.973-978.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.
Labor Dr. Krone und Partner, Medizinaluntersucheungsstelle,1 OEKCON GmbH, Herford, Germany,2 Innogenetics NV, Ghent, Belgium3
Received 21 August 2001/ Returned for modification 18 November 2001/ Accepted 30 December 2001
|
|
|---|
|
|
|---|
The classic T. pallidum immobilization (Nelson-Mayer) test is no longer routinely used because of its complexity and consequent high cost, and it has largely been replaced by the fluorescent T. pallidum absorption (FTA-ABS) test. However, this confirmatory test can also lead to false-positive or false-negative results if the fluorescence intensity is misinterpreted or if the testing equipment is suboptimal (7). False-positive results have been reported for patients with other spirochetal infections or immunological disorders (e.g., systemic lupus erythematosis, rheumatoid factors, antinuclear and anti-DNA antibodies or cross-reacting antibodies such as those from Borrelia burgdorferi) (4, 7-9, 11, 12, 14). The T. pallidum Western immunoblot assay is also used as an alternative confirmatory test to improve sensitivity and specificity, but indeterminate and false-positive reactivity patterns have been reported (2, 8, 10).
A newly validated line immunoassay (LIA) based on recombinant antigens and synthetic peptides derived from T. pallidum proteins, the INNO-LIA Syphilis assay (Innogenetics NV, Ghent, Belgium), was previously shown to provide highly reliable confirmatory diagnostic information (3). We therefore evaluated the new assay using a large number of sera from a clinical laboratory in order to determine the sensitivity and specificity of the test and compare the results with those of other confirmatory assays.
|
|
|---|
Sample groups. Four groups of samples were tested: (i) 43 samples from Russian blood donors for whom no clinical information was available; (ii) 195 sera reacting to TPHA with a minimum low titer of 1:80; (iii) 203 sera from pregnant women which were negative by TPHA; and (iv) 90 sera from patients with the following characteristics: antinuclear antibody positive (n = 10), rheumatoid factor positive (n = 10), antimitochondrial antibody positive (n = 1), double-stranded DNA positive (n = 1), acute Epstein-Barr virus infection (n = 10), cytomegalovirus infection (n = 5), Lyme borreliosis (n = 10), parvovirus B19 infection (n = 3), rubella virus infection (n = 5), human immunodeficiency virus infection (n = 10), hepatitis B virus infection (n = 10), anticardiolipin antibodies (n = 5), and hepatitis C virus infection (n = 10). Subsequently, syphilis determination was performed to complete the VDRL, FTA-ABS, and TPHA reactivity patterns for all the studied groups.
Screening and confirmatory assays used in the study. (i) TPHA. The TPHA was performed with the Serokit LD test (Innogenetics GmbH, Heiden, Germany) according to the manufacturer's instructions. Typically, a qualitative (yes/no answer) TPHA is performed at a unique serum dilution of 1:80. All reactive samples were subsequently tested to determine the end dilution titers.
(ii) FTA-ABS. The FTA-ABS confirmatory test for syphilis was performed with some modifications to the standard procedure (5, 8, 12), using the T. pallidum antigen and the Treponema phagedenis ultrasonicate (Ultrasorb; BAG, Lich, Germany) as well as an immunoglobulin G (IgG)-specific anti-human fluorescein isothiocyanate-conjugated antibody (Gull Laboratories, Bad Hamburg, Germany). To confirm the reliability of screening results, each sample was tested at dilutions of 1:5 and 1:20 and scored as follows: 1+ (indeterminate), serum reacting weakly only in the 1:5 dilution; 2+ (weak positive), serum with a more intense fluorescence at 1:5 but negative at the 1:20 dilution; 3+ (positive), serum reactive at the 1:5 dilution and weak fluorescence at 1:20; 4+ (strong positive), serum with intense fluorescence at 1:20.
(iii) 19S-IgM-FTA-ABS test. For the determination of specific IgM antibodies, a modification of the 19S-FTA-ABS procedure (5, 12) was used. To eliminate the IgG fraction, 20 µl of sera was first incubated overnight with 150 µl of an anti-IgG serum (GullSORB; Gull Laboratories) and then diluted with 30 µl of T. phagedenis ultrasonicate (Ultrasorb; BAG). The final serum dilution was 1:10. The test was performed using slides coated with treponemal organisms (FTA-ABS slides; Innogenetics GmbH). Antibody detection was carried out with an IgM-specific anti-human fluorescein isothiocyanate-conjugated globulin (Gull Laboratories). Results were read with a standard fluorescence microscope (Zeiss, Oberkochem, Germany) at a magnification of x400 and classified qualitatively as negative, 1+ (indeterminate), 2+ (weak positive), 3+ (positive), or 4+ (strong positive).
(iv) IgM EIA. The test kit from Gull Laboratories (Captia M-Syphilis EIA) was used according to the manufacturer's instructions. Results were calculated as an index as follows: negative, <0.9; indeterminate, 0.9 to 1.1; positive, >1.1.
(v) VDRL. The quantitative VDRL test was performed according to standard procedures using the VDRL-Cardiolipin-Antigen (Behringwerke AG).
(vi) Enzygnost Syphilis EIA. As an additional test for discrepant sera, the Enzygnost Syphilis EIA (Behringwerke AG) was used according to the manufacturer's instructions and interpretation criteria. The test simultaneously detects specific IgG and IgM antibodies to T. pallidum and is routinely used in Germany as a syphilis screening test.
(vii) T. pallidum immunoblotting. The IgG and IgM immunoblot assay based on native antigens of T. pallidum (MarDx, Carlsbad, Calif.) was used as an additional confirmatory test. According to the manufacturer's interpretation criteria, a specific immune response can be assumed if reactivity for two of the three antigens (with molecular masses of 47, 17, and 15.5 kDa) is demonstrated.
(viii) INNO-LIA Syphilis. INNO-LIA Syphilis, a LIA for the simultaneous detection of antitreponemal antibodies, makes use of recombinant antigens and synthetic peptide antigens derived from T. pallidum (Nichols strain) membrane proteins. The antigens consist of three immunodominant proteins (TpN47, TpN17, and TpN15) expressed as full-size proteins in Escherichia coli and one synthetic peptide (TmpA) derived from transmembrane protein A (for further details, see reference 3). In addition to the syphilis antigens on the test strip, control lines are used for semiquantitative evaluation of the results and for the verification of sample addition. Potential occurrence of antibodies that might react with nontreponemal components present on each strip can be visualized on the negative-control line.
The assay procedure can briefly be described as follows. Serum or plasma samples were diluted 1:100 and incubated at room temperature (20°C) overnight, followed by three washing steps with washing buffer before addition of a goat anti-human IgG (heavy and light chains) conjugated to alkaline phosphatase. After incubation, three washing steps were again performed, followed by the addition of a chromogen. Color development was then stopped with an appropriate stop solution. In a visual reading protocol, after color development, each line was compared to the control lines, and the intensities were scored as follows: 0, no line or a line less intense than the cutoff line; 1, a line with an intensity between that of the cutoff line and that of the 1+ control line; 2, a line with an intensity between that of the 1+ control line and that of the 3+ control line; 3, a line equal to that of the 3+ control line; 4, a line with an intensity greater than that of the 3+ control line.
The interpretation algorithm of the INNO-LIA Syphilis kit was initially optimized by the manufacturer for visual reading with an independent set of negative and positive samples. A sample is considered T. pallidum antibody negative if no band or a single band with a maximum intensity of 0.5 is present. If multiple bands with a minimum intensity of 0.5 are visible, the sample is considered T. pallidum antibody positive. Finally, a sample is considered indeterminate if a single band is visible with a minimum intensity of 1. Examples of INNO-LIA Syphilis patterns are shown in Fig. 1.
![]() View larger version (15K): [in a new window] |
FIG. 1. Representative patterns of reactivity with the INNO-LIA Syphilis.
|
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Comparison of INNO-LIA Syphilis results with consensus of multiple testing for syphilis serology
|
|
View this table: [in a new window] |
TABLE 2. Detailed reactivities of discrepant samples on multiple testing in comparison to INNO-LIA Syphilisa
|
|
View this table: [in a new window] |
TABLE 3. Complementary analysis by immunoblotting for some discrepant samples
|
For serum no. 219 (Tables 3, 4, and 5), a primary syphilis was clinically suspected and treated very early. A number of follow-up samples showed a decrease of the TPHA titer, negative results in the IgM-EIA, and indeterminate reactions in 19S-IgM-FTA-ABS and IgG-FTA-ABS tests (Table 4). The original and the first follow-up sample were also tested by MarDx immunoblotting, giving repeatedly inconclusive results.
|
View this table: [in a new window] |
TABLE 4. Longitudinal testing of patient 219 sample
|
|
View this table: [in a new window] |
TABLE 5. Confirmation of TPHA results by INNO-LIA Syphilisa
|
For sample no. 217, from a Russian blood donor, the MarDx IgG immunoblot was positive and the IgM immunoblot was indeterminate (Tables 3 and 4). A further immunoblot with recombinant antigens (under evaluation by the investigator) showed a negative IgG reaction and an indeterminate IgM reaction. The main reason for the classification as indeterminate was the overall heterogeneity of test results for this sample by different treponemal tests approved by the Paul-Ehrlich-Institute (Langen, Germany).
Comparison of INNO-LIA Syphilis versus TPHA and IgG-FTA-ABS tests, with consensus results as reference. For the comparison of the INNO-LIA Syphilis kit with other tests, the consensus classification was used as a "gold standard." The investigator, blinded for INNO-LIA Syphilis data, determined a consensus classification for each of the specimens based on the overall interpretation of assay results. In general, a sample was considered positive if reactive in a screening test (VDRL and/or TPHA) and in a confirmation test (FTA-ABS and/or Western blotting and/or EIA). For the purpose of this comparison, samples for which clear-cut results (positive or negative) were mostly available (508 out of 531 samples) were used. Twenty-three samples could not be classified by the existing data; they were therefore considered to be indeterminate.
(i) Comparison with TPHA as a screening test. The comparison between TPHA and INNO-LIA Syphilis is shown in Table 5. Two samples had a negative TPHA result, while LIA classified them as indeterminate. There were 13 samples with a weakly reactive TPHA titer of 1:80, classified as indeterminate, although the samples were categorized as positive in the context of other results by the investigator, and this was congruous with the INNO-LIA Syphilis findings. When the two tests were compared for correct and incorrect answers as defined by the investigator, INNO-LIA Syphilis gave a significantly higher number of correct answers than did TPHA (McNemar's test, P value = 0.021).
(ii) Comparison with IgG-FTA-ABS as a confirmatory test. There were six samples from pregnant women which were weakly positive (2+ reaction) by the IgG-FTA-ABS test but classified as negative by the investigator because the TPHA test was negative. The consensus classification agreed with the INNO-LIA Syphilis findings. The two samples (no. 283 and 322) (Table 2) with an indeterminate LIA result and a negative TPHA test also had a negative IgG-FTA-ABS result. Two samples with an indeterminate IgG-FTA-ABS result were classified as positive by the investigator. One sample was from a Russian blood donor (no. 36), and one was from a patient in the early seroconversion phase of a clinically suspected primary syphilis (no. 215). Both sera demonstrated strongly positive IgM antibody reactions with the IgM-EIA test and a weakly positive (2+) result in the 19S-IgM-FTA-ABS test. For 32 samples with a negative TPHA result and an indeterminate (1+) IgG-FTA-ABS test (25 samples from pregnant women and 7 samples from patients with various other conditions) (data not shown), the negative LIA findings also agreed with the consensus classification of the investigator. When analyzing the results of these two tests and comparing them to the investigator's classification, INNO-LIA Syphilis gave a significantly higher number of correct answers than did the IgG-FTA-ABS test (McNemar's test, P value < 0.0001) (Table 6).
|
View this table: [in a new window] |
TABLE 6. Confirmation of IgG FTA-ABS results by INNO-LIA Syphilisa
|
|
|
|---|
The findings of this study also demonstrate the high number of nonspecific indeterminate IgG-FTA-ABS results for patients with no history of treponemal infection: 27 samples (13%) of the 203 TPHA-negative pregnant women showed 1+ reactivity, and two samples (1%) showed 2+ reactivity. The VDRL test (data not shown) reacted in 24 samples (12%) of the sera from the same panel. For all these sera, the INNO-LIA Syphilis test was clearly negative.
As a general rule, laboratory test results should be reviewed by the clinician in the light of clinical symptoms and the history of infection and treatment. However, for patients with suspected or previous syphilis there are often no valid data available, and it is therefore very important for the laboratory diagnosis to be as reliable as possible. The INNO-LIA Syphilis as a new confirmatory test for treponemal infections seems to fulfill this need. It can nevertheless be argued that a comparison of the LIA with the consensus results derived from the TPHA and the IgG-FTA-ABS is open to possible criticism, since the investigator's interpretation is not independent of individual test results. However, the INNO-LIA Syphilis was found to correlate significantly better with the reference results than any of the individual tests alone. Moreover, when the consensus results based on the data derived from conventional serological testing are compared to the INNO-LIA Syphilis results, a close-to-perfect match of results is obtained (100% sensitivity and 99.3% specificity). Our data also confirm the previous findings of Ebel et al. (3), who determined the sensitivity and specificity of the INNO-LIA Syphilis to be 99.6 and 99.5%, respectively. The INNO-LIA Syphilis can therefore be considered to be a valid, if not superior, alternative to the FTA-ABS test for the confirmation of syphilis screening test findings (TPHA, VDRL, or EIA), and its use can improve the reliability of syphilis serology.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»