Evaluation of an Enzyme Immunoassay Technique for Detection of Antibodies against Treponema pallidum

MATERIALS AND METHODS

Four hundred forty-one patients attending a sexually transmitted disease clinic in Lisbon, Portugal, were enrolled in the study after informed consent was obtained. They were distributed into five groups, as follows: 25 patients with primary syphilis, 25 patients with secondary syphilis, 179 patients with latent syphilis, 105 individuals with a history of syphilis that had been correctly treated, and 107 individuals with no clinical history of syphilis.

The scientific council of the Instituto de Higiene e Medicina Tropical approved the study since it represents the committee on research with human subjects.

All samples were tested for T. pallidum antibodies by the RPR test (Macro-Vue; Becton Dickinson), MHA-TP (Phasyl 210), and FTA-Abs IgG and IgM (Euroimmune) according to the instructions of the manufacturers.

The enzyme immunoassay (Eti-syphilis-G and Eti-syphilis-M; DiaSorin), including all incubation steps and washings, was also performed according to the instructions of the manufacturer. First, a 1:100 dilution of all serum samples was obtained. Samples of diluted sera (100 μl) were then added to the microwells, which were covered with purified T. pallidum antigen. After this procedure, peroxidase-labeled antihuman monoclonal antibodies were added. The chromogenic substrate tetramethylbenzidine was then added to each well. IgG and IgM antihuman monoclonal antibodies were used to differentiate between T. pallidum IgG and IgM antibodies. The reaction was stopped after 30 min by adding a stop solution, and then the result was read in a spectrophotometer at 450 nm.

Test validation and interpretation were performed by introducing one negative control and two positive controls two times each. The run was considered valid if the mean absorbance for the negative control was equal to or less than 0.250, if the mean absorbance for one of the positive controls was between 0.300 and 0.500, and if the mean absorbance for the other positive control was above 0.500. According to the instructions of the manufacturer, the cutoff point between a positive and a negative result was set at the absorbance of the low-titer positive control. Screened samples with an absorbance within about 10% of that for the low-titer positive control were retested.

A sample with different results by the various techniques used in this study was retested by a technologist who was unaware of the previous results.

RESULTS

Among the 441 samples studied, 289 (65.5%) were reactive by the RPR technique, while 152 (34.5%) were nonreactive. By the MHA-TP, FTA-Abs, and ELISA techniques, 313 (71%), 324 (73.5%), and 322 (73%) samples, respectively, were reactive, while 128 (29%), 117 (26.5%), and 119 (27%) samples, respectively, were nonreactive.

All samples were used to evaluate the sensitivity and specificity of the ELISA technique in relation to the results of the other specific tests: MHA-TP (Table 1) and the FTA-Abs test (Table 2). The sensitivities of the ELISA were 100% compared with the results of the first test and 99.4% compared with the results of the second test. The specificities were 93% when the results of the ELISA were compared with those of MHA-TP and 100% when the results of the ELISA were compared with those of the FTA-Abs test. The concordance of results between the ELISA and the microhemagluttination technique was 98%, while that between the ELISA and the FTA-Abs test was 99.5%.

The sensitivities of all serological tests used in this study were also evaluated by using the FTA-Abs test as the “gold standard” and for each group of patients described above (patients with primary, secondary, and latent syphilis and individuals with past treated syphilis). When possible, the specificities were also evaluated (Table 3). For the primary syphilis group, the sensitivities of the RPR test, MHA-TP, and the ELISA technique were 92, 88, and 100%, respectively. For the secondary syphilis group, the sensitivities of all tests evaluated were 100%, while for the latent syphilis group and those patients with past treated syphilis, the sensitivities were 97.2 and 57.9%, respectively, for the RPR test; 99.4 and 92.6%, respectively, for MHA-TP, and 100 and 97.9%, respectively, for the ELISA technique.

When samples from individuals with no clinical signs of T. pallidum infection and with a nonreactive FTA-Abs test (supposedly noninfected patients) were evaluated, 12 samples were positive by the RPR test. Samples from this group of patients were nonreactive by both the ELISA and MHA-TP. Consequently, the specificity of the RPR test was 88.8%, while those of the other two methods were found to be 100%. Table 4 describes the data related to the detection of anti-T. pallidum IgM antibodies by the ELISA and FTA-Abs techniques in different phases of disease. Twenty-five patients with primary syphilis, 25 patients with secondary syphilis, and 179 patients with latent syphilis were analyzed. Sixty-five of 229 samples (28.4%) were reactive for IgM by the FTA-Abs test, the reactivities of 14 of 229 samples (6.1%) were indeterminate, and 150 of 229 samples (65.5%) were nonreactive. When the ELISA technique was used, 75 of 229 serum samples (32.8%) were shown to be reactive and 150 of 229 serum samples (65.5%) were nonreactive, and the reactivities of 4 of 229 serum samples (1.7%) were considered indeterminate. These differences, however, were not statistically significant. In this study, we have also analyzed samples from patients with past treated syphilis for the presence of anti-T. pallidum IgM antibodies. These samples were nonreactive by the ELISA and the FTA-Abs methods.

DISCUSSION

The possibility of using an ELISA technique as an alternative to the treponemal tests has been evaluated in various studies. The sensitivities of the tests varied from 48.5 to 100%, according to the type of ELISA, with the phase of the disease, and with the disease prevalence (1, 6, 7, 12, 14, 16).

In this study, the ELISA showed a sensitivity that was similar to those of the tests generally used for the laboratory diagnosis of syphilis. Recently, other investigators using other ELISA techniques obtained results similar to those found in the present study (1, 3, 10, 11, 13, 15).

Although a good specificity of the ELISA was obtained when the results of the ELISA were compared with those of the FTA-Abs test and MHA-TP, the specificity of the ELISA was higher than that of the fluorescence test. The rates of concordance of the results of the ELISA technique with those of the other two tests were also found to be high (99.6 and 98%, respectively).

When the sensitivities of the ELISA, RPR test, and MHA-TP were compared by using the FTA-Abs test as the standard, the ELISA technique had a higher sensitivity, although the difference was not statistically significant, during all phases of disease, although other investigators have found that the ELISA method has a lower sensitivity during the primary stage of syphilis (7, 14). Since detection of anti-T. pallidum IgM antibodies is important in the differentiation between recent and late infection, we decided to verify using both the ELISA and the FTA-Abs techniques whether IgM antibodies were present in patients with recent stages of syphilis. A larger number of samples from patients with primary syphilis than from patients with secondary syphilis had IgM antibodies against T. pallidum, although the differences were not statistically significant. Samples from patients with latent syphilis presented with less reactivity with these antibodies by both tests. When the results of the ELISA for IgM antibodies were compared with those of the FTA-Abs test for IgM antibodies, the number of samples with indeterminate reactivities was smaller by the ELISA for IgM antibodies than by the FTA-Abs test for IgM antibodies. It seems that the majority of samples with indeterminate reactivities obtained by the latter test were indeed reactive. It should also be mentioned that IgM antibodies were not detected in individuals with treated past syphilis by either the ELISA or the FTA-Abs technique. This fact, together with the higher frequency of these antibodies in patients with primary and secondary syphilis than in patients with latent syphilis, demonstrates the utility of this ELISA as a marker of recent infection.

In conclusion, the results of this study show that the ELISA may be an alternative to the treponemal tests for the detection of T. pallidum antibodies, including the presence of IgM, since it has a sensitivity and a specificity similar to those of the most commonly used tests during all stages of syphilis. This was especially true when the ELISA technique was compared with the FTA-Abs technique, which is considered the most sensitive and specific test for the diagnosis of syphilis. The ELISA also had a sensitivity similar to that of the RPR test, having the advantage of presenting no false-positive results. We think that the enzyme immunoassay technique studied here could be used as a screening test, since it is simple, objective, and easily automated.