INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Syphilis is one of the most important sexually transmitted diseases (STDs) worldwide (15). Its control and surveillance require careful screening tests by reliable methods. Serological diagnosis for the screening and follow-up of syphilis should combine a nontreponemal test such as the Venereal Disease Research Laboratory (VDRL) test or the rapid plasma reagin test with a treponemal test such as the Treponema pallidum hemagglutination assay (TPHA) (18). The VDRL test is not very specific, with a high frequency of false-positive results (12). False-negative results have also been reported with high-titer sera due to a prozone phenomenon (3). False-negative VDRL test results have been recorded after treatment or for patients with late latent syphilis, leading to misdiagnosis of syphilis when the VDRL test is used alone. TPHA, which uses the structural treponemal antigen, is more specific than the VDRL test and is a very simple and cheap routine test for syphilis, although it is less sensitive than the VDRL test in the early stage of syphilis (16). However, false-positive results have also been reported by TPHA (11). Positive results by one or both screening tests are confirmed by the fluorescent treponema antibody absorption (FTA-abs) test. The FTA-abs test is a very sensitive test at all stages of syphilis, but reading of the fluorescence is subjective and is sometimes difficult, making this test unsuitable for screening applications (8). There is a need for a reliable, specific, rapid, and automated test for the screening and confirmation of syphilis at all stages of the disease. Enzyme immunoassays (EIAs) for the detection of T. pallidum antibodies have been developed to achieve this (2, 9, 10, 14, 19, 20).

We present herein results obtained by a newly developed EIA for syphilis, based on a competition assay between treponemal antibodies in the test serum and a peroxidase-labeled human anti-T. pallidum immunoglobulin G (IgG). The results obtained by the new test were compared with those obtained by TPHA and the FTA-abs test. We investigated serum from patients with all stages of syphilis, active and latent, and patients with discrepant serological status and trace amounts of antibodies. We also evaluated negative controls with nonreactive VDRL test, TPHA, and FTA-abs test results, including sera known to have high titers of antibodies to other infectious or autoimmune diseases. We evaluated whether this competitive EIA could replace TPHA, the FTA-abs test, and the VDRL test for the screening of large populations for the detection of syphilis.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Sera. A total of 824 serum specimens sent to the French National Reference Center for STDs for screening or confirmation of syphilis or other diseases were included in this study. They were stored at 20°C prior to testing. All sera were checked by TPHA, the VDRL test, and the FTA-abs test. Three hundred fifty-eight serum specimens tested negative for syphilis by all serological tests. One hundred of these syphilis-negative control specimens tested positive for other infections, including Lyme disease detected by indirect immunofluorescence (IFI) with a whole bacterial antigen prepared in the laboratory; herpes simplex virus type 1, detected by EIA with a viral antigen prepared in the laboratory; cytomegalovirus, detected by EIA (Ortho Diagnostics, Raritan, N.J.); Epstein-Barr virus, detected by EIA (Zeus, Raritan, N.J.); Chlamydia trachomatis, detected by microimmunofluorescence with bacterial antigens prepared in the laboratory; Brucella melitensis, detected by IFI with bacteria cultured in the laboratory; Bartonella henselae, detected by IFI with bacterial antigen prepared in the laboratory; or Helicobacter pylori, detected by latex agglutination (Orion Diagnostica, Espoo, Finland); or they tested positive for autoimmune diseases, including anticardiolipid (detected by EIA; BMD s.a., Marnes la Vallée, France) or rheumatoid factor (Waaler Rose; Fumouze Diagnostics, Asnières, France). These sera were tested to assess the specificity of the competitive EIA for syphilis. Four hundred sixty-six serum specimens tested positive for syphilis by TPHA and the FTA-abs test. Some of these tested positive and some tested negative by the VDRL test but without IgM detection. These sera were classified as showing latent or previous syphilis. Twenty-two serum specimens gave false-positive results by the VDRL test. Sixteen additional serum specimens with documented clinical data were included; these tested positive for specific anti-T. pallidum IgM (the FTA-abs test for IgM [FTA-abs-IgM] and the solid-phase hemadsorption assay [SPHA] for IgM [SPHA-IgM]), including five serum specimens from patients with primary syphilis.

TPHA. A commercial TPHA reagent (MHA-TP; Fujirebio, Tokyo, Japan) was used to detect and titrate agglutinating antibodies against T. pallidum. The threshold value was 80 U, and all sera which tested positive were titrated by serial dilution.

FTA-abs test. Serum samples were tested by a fluorescent T. pallidum-specific assay (Trepo Spot IF; BioMerieux, Marcy l'Etoile, France) by using microscope slides coated with T. pallidum Nichols whole antigen. Specific antibodies reacting with T. pallidum-coated slides were detected with anti-human Ig (classes IgG, IgA, and IgM or only IgM)-fluorescein isothiocyanate conjugates (Diagnostics Pasteur, Marnes la Coquette, France). Cross-reacting and irrelevant antibodies were first absorbed with T. phagedenis Reiter, supplied in the kit. The threshold value was 100 U. All samples testing positive were titrated by serial dilution.

VDRL test. We used the Sypal reagent (Diagast Laboratories, Lille, France), which is a colloidal suspension of cardiolipin, lecithin, and cholesterol. This VDRL test, designed for the manual detection of syphilis with Kline plates, results in the formation of agglutinates in the presence of syphilitic serum reagins.

IgM testing. A positive IgM reaction confirms active or congenital syphilis (9). Two specific anti-treponemal IgM tests were used: the FTA-abs 19S IgM test, which is an FTA-abs test with the IgM fraction of the serum, obtained by gel filtration or ultracentrifugation and detection of specific anti-human IgM-fluorescein isothiocyanate conjugate (Diagnostics Pasteur) by microscopy (1), and IgM-SPHA, in which microtiter plate wells act as the solid phase for µ-chain (anti-µ; Dako, Glostrup, Denmark) capture. The antitreponemal component of the captured IgM is then detected by TPHA (13).

Competitive EIA. The commercially available EIA used in this study, the BioElisa Syphilis (Biokit, Barcelona, Spain /Instrumentation Laboratory, Paris, France), is a competitive assay for the screening of total antitreponemal antibodies. The reaction is based on competition between antitreponemal antibodies in the sample and a human anti-T. pallidum IgG peroxidase-labeled conjugate for specific antigenic sites from strain Nichols, which are used to coat the wells of flat-bottom microtiter plates. For each sample, we dispensed 30 µl of a negative control serum sample into three wells of the plate, 30 µl of a positive control serum sample into one well, and 30 µl of the test serum into one well. Conjugate (100 µl) was added to each well. The plate was incubated for 90 min at 37°C, and then the contents of each well were removed by aspiration and the well was washed four times with a phosphate-buffered saline (10×) solution, provided by the manufacturer, containing detergent and thimerosal. Chromogenic substrate solution (100 µl of 3,3',5,5'-tetramethylbenzidine dissolved in dimethyl sulfoxide) was added to each well, and the plate was incubated for 15 min at room temperature. The reaction was stopped by adding 100 µl of stopping solution (1 N H₂SO₄). Within 30 min at 450 nm the absorbance of the solution in each well was read. In this assay, the binding of the conjugate to the specific antigen, determined by measuring the optical density at 450 nm (A₄₅₀), is inversely proportional to the amount of specific antibodies in the test sample. According to the manufacturer's recommendations, the following criteria must be met for accurate assays to be achieved. The A₄₅₀ of the negative controls must be 0.6, and the A₄₅₀ of the positive control must be 0.7 times the cutoff. The results are expressed as follows. The mean A₄₅₀ for the negative controls was calculated; the cutoff is calculated by multiplying the mean A₄₅₀ of the negative controls by 0.7. This procedure made it possible to determine an EIA index for each serum sample by dividing its A₄₅₀ value by the cutoff value. Thus, an EIA index less than or equal to 1.0 is a positive result, and an EIA index greater than 1 is a negative result.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Sensitivity. A panel of 434 serum samples that tested positive for syphilis by TPHA and the FTA-abs test were used to assess the sensitivity of the BioElisa Syphilis. Four hundred thirty-two syphilis-positive serum samples with TPHA titers of 80 to 10,240 U and FTA-abs test titers of 100 to 6,400 U tested positive for syphilis by the competitive EIA. The mean EIA indices for these samples were 0.64 ± 0.2, correlating with a TPHA titer of 80 U and an FTA-abs test titer of 100 U, and 0.09 ± 0.04, correlating with a TPHA titer of 10,240 U and an FTA-abs test titer of 6,400 U. Two serum samples with TPHA and FTA-abs test titers at the threshold values (80 for TPHA and 100 for the FTA-abs test) tested negative by BioElisa Syphilis, with EIA indices of 1.3 and 1.2, respectively. Thus, BioElisa Syphilis had a sensitivity of 99.5%. Although BioElisa Syphilis was originally designed for qualitative screening of treponemal antibodies, we observed a direct correlation between EIA indices and TPHA and FTA-abs test titers, as described above and shown in Fig. 1.

View larger version (11K): [in this window] [in a new window]   FIG. 1.   Correlation between BioElisa Syphilis and TPHA and FTA-abs test (FTA) results. Each point represents the mean EIA index for sera with various TPHA and FTA-abs test titers. The standard error of the mean EIA index was always below 0.27.

Specificity. Three hundred fifty-eight specimens were initially nonreactive in the VDRL test, TPHA, and the FTA-abs test. The average EIA index by the BioElisa Syphilis for the 356 serum samples that tested negative for syphilis by TPHA and the VDRL and FTA-abs tests was 1.4, clearly negative. Only two serum samples gave false-positive results, with EIA indices of 0.800 and 0.850, respectively. These results give a specificity of 99.4% (356 of 358) for BioElisa Syphilis. One hundred of these 358 serum samples were nonreactive for syphilis but tested positive for other infectious diseases or for autoimmune diseases. All tested negative in the BioElisa Syphilis (average EIA index, >1.4) (Table 1). We assessed the effects of false-positive results by the VDRL test on the EIA results. Twenty-two serum samples that tested positive by the VDRL test and that were nonreactive by TPHA and the FTA-abs test tested negative by BioElisa Syphilis (mean EIA index, 1.44 ± 0.25) (Table 2).

Results of competitive EIA for active syphilis. Eleven serum specimens were from patients with active syphilis, as shown by positive FTA-abs-IgM and SPHA-IgM results and clinical data. Nine patients had neurosyphilis or untreated secondary or tertiary syphilis, including seven patients who tested positive for human immunodeficiency virus (HIV) type 1 (HIV-1). Two patients were tested 3 months after penicillin therapy. All of these serum specimens gave strong positive results by BioElisa Syphilis, with very low EIA indices (Table 3). Despite the high syphilis antibody titers, no prozone phenomenon was observed in the EIA with undiluted sera, as recommended for this test.

Results of competitive EIA for primary syphilis. TPHA is not very sensitive at detecting primary syphilis, with nonreactive or very weak positive results, at early stages of the infection (4 to 6 weeks), whereas the FTA-abs, VDRL, and IgM tests give positive results within 3 to 4 weeks. Five patients with documented primary syphilis tested positive by the BioElisa Syphilis (Table 4).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

In this study a competitive EIA (BioElisa Syphilis) for syphilis had a sensitivity of 99.5% (434 of 436) relative to TPHA and FTA-abs test results for all groups of syphilis-positive and -negative patients. Only two serum samples which did not react in the EIA were weakly reactive in TPHA and the FTA-abs test (threshold values), suggesting that the patients may have previously had syphilis that was treated or false-positive reactions. The sensitivity of the competitive EIA for both treated and untreated syphilis is similar to that reported for other EIAs for T. pallidum. Lefevre et al. (10) obtained 100% sensitivity with the Captia Syphilis G assay, similar to those obtained with TPHA and the FTA-abs test, at all stages of the disease except primary syphilis, in which the EIA may have lower sensitivity. We found that the BioElisa Syphilis detected all five cases of primary syphilis, whereas the TPHA results were still at or below threshold values. An overall sensitivity of 98.4% has been reported in a preliminary evaluation of the Captia Syphilis G assay (17). Another EIA, the Syphilis Bio-EnzaBead assay, has an overall sensitivity of 94.8 to 97.5% for all stages of syphilis (4). The BioElisa syphilis used in the present study had a sensitivity of 100% for HIV-positive patients coinfected with T. pallidum. A sensitive test for screening HIV-positive patients for syphilis is vital because there is often coinfection (7), which may affect syphilis antibodies (6). The specificity of the competitive EIA was 99.4% (356 of 358). One hundred serum specimens with high titers of antibodies against other infectious or autoimmune diseases were tested to investigate the possibility of nonspecific cross-reactions with syphilis antibodies due to polyclonal activation of the antibody response. All these specimens tested negative in the BioElisa Syphilis. BioElisa Syphilis was designed to be a qualitative screening test. However, we observed an inverse correlation between the TPHA and FTA-abs test titers and EIA indices (Fig. 1). This suggests that it may be possible to use this competitive EIA for antibody titration after further investigations with specimens from patients with clinically documented cases of infection.

These results indicate that the competitive EIA BioElisa Syphilis is sensitive, specific, and easy to use as an automated test for the screening of large numbers of specimens.