Background Traditional syphilis testing consists of screening with a non-treponemal test (RPR) and confirmation with a treponemal test (TP-PA, FTA-ABS, EIA, CIA, etc). Recently, that testing algorithm has been reversed due to efforts to reduce labour costs and the availability of automatable tests (EIA, CIA). Large numbers of discordant test results (treponemal +, non-treponemal−) can be obtained using the reverse algorithm and can be due to (1) treated cases of syphilis, (2) a false-positive treponemal test, or (3) a case of early primary syphilis that has yet to seroconvert. Those sera need to be confirmed with a second treponemal test to eliminate false positive specimens. The dilemma arises which treponemal test is best suited for confirmation, and secondly what are the relative analytical sensitivities of available treponemal tests used for both screening and confirmation.

Methods Two hundred randomised TP-PA positive, cleanascite treated samples (GADPH) were serially diluted with normal human sera to determine the analytical endpoint and sensitivity of six commonly used treponemal tests (TP-PA, FTA-ABS, TrepSure, TrepChek, TrepID, and LIAISON). All dilutions were treated as neat sera in each test, and the tests were performed according to the manufacturer's instructions.

Results The FTA-ABS was used as the reference since it has been considered the Gold Standard" among treponemal tests. The analytical sensitivity of the six tests fell into three statistically different groups (from lowest to highest): (1) the FTA-ABS, the TP-PA, and the TrepChek, (2) the LIAISON and the TrepID, and (3) the TrepSure. In addition, there were 24 sera that were found non-reactive by FTA-ABS and nine found non-reactive by TrepChek that were reactive by at least two other treponemal tests.

Conclusions These results highlight significant differences in the analytical sensitivity of various treponemal tests and could explain some discordant results between treponemal tests used to confirm screening EIAs and CIAs. First, the FTA-ABS should no longer be considered the “Gold Standard”. Second, low titre sera could appear non-reactive if relatively insensitive tests are used for the either screening or confirmatory tests. Third, the question arises, “Should initial screening tests be confirmed by treponemal tests with lower analytical sensitivity?” Future patient studies could verify these results, so that a more accurate standard protocol for syphilis testing can be established.