Abstract

CDC has long recommended screening for syphilis with a non-treponemal test such as the RPR test and, if reactive, confirming with a treponemal test. Non-treponemal antibody levels rise during active infection and tend to decline after treatment, so non-treponemal screening identifies persons with active disease who are likely infectious and require clinical and public health interventions. Although treponemal antibodies rise slightly earlier than non-treponemal antibodies, they tend to remain elevated after treatment so their presence does not always indicate active infection. Thus, treponemal tests have not been recommended for initial screening. Also, older treponemal tests that utilise native treponemal antigens, such as the TP-PA and FTA-ABS tests, were thought to have a high false positive rate due to binding of cross-reacting serum antibodies. Newer treponemal tests, enzyme immunoassays (EIAs) and chemiluminescence immunoassays (CIAs), utilise recombinant treponemal antigens which should result in tests with high sensitivity and specificity, capable of detecting small quantities of antibody without nonspecific binding of cross-reacting antibodies. Because EIA/CIA can be automated, U.S. laboratories have begun to screen for syphilis using a reverse sequence with EIA/CIA screening and confirmation of EIA+ sera with an RPR test to identify active infection.

Using this reverse sequence for syphilis screening, discordant sera (EIA+/RPR−) would be expected in patients with previous infection or early primary infection. CDC studies have found that more than half of all EIA+ sera were discordant with EIA+/RPR− results. A recent CDC epidemiologic study found that 32% of discordant sera were due to false-positive EIA/CIAs (eg, EIA+/RPR−/TP-PA−), with rates that ranged from 12% in a high prevalence population to 60% in a low prevalence population.

Discordant sera cause uncertainty about patient management, and the TP-PA test might be a useful confirmatory test with these sera. Recent studies suggest that the TP-PA test has equivalent sensitivity but higher specificity than EIA/CIAs, performance characteristics that are necessary for a confirmatory test. The FTA-ABS should not be used because it has low specificity, its interpretation is inherently subjective, and its performance requires trained personnel and a dedicated fluorescence microscope. Research is needed to better understand the variation in treponemal test performance.