Evaluation of diagnostic serological results in cases of suspected primary syphilis infection

Abstract

Background: Reverse sequence screening for syphilis, in which an automatable treponemal assay (enzyme immunoassay [EIA]/chemiluminescence assay [CIA]) is performed first and followed by a nontreponemal test for reactive specimens, has been used increasingly in the United States. The EIA is objective, efficient, and believed to be more sensitive than the rapid plasma reagin (RPR) because treponemal antibodies appear before nontreponemal antibodies. We sought to compare the sensitivity of a commonly used EIA, the Trep-Sure EIA (TS-EIA), to the RPR in cases of suspected primary syphilis infection in our clinic.

Methods: A retrospective medical record review of patients with sexually transmitted infection clinic visits from January 2009 to December 2011 was conducted, and 52 patients met the following inclusion criteria: suspected primary syphilis symptoms, at least 1 positive syphilis test result at visit, and no history of syphilis. Sensitivity analyses compared the TS-EIA and RPR, using the reference standard of concordantly positive/reactive TS-EIA/RPR or positive fluorescent treponemal antibody absorption test (FTA-ABS) result. We considered equivocal TS-EIA results to be positive for sensitivity calculations because such results typically reflex to additional testing and therefore may still result in identifying new infections.

Results: Twenty-eight (53.8%) of the 52 patients had a positive or equivocal TS-EIA. Twenty-five (89.3%) of those were RPR reactive; the remaining 3 (10.7%) were RPR nonreactive, FTA-ABS positive. Forty patients (76.9%) had a positive RPR, including 15 patients (37.5%) with negative TS-EIA results; all 15 were FTA-ABS positive. Nine additional patients were TS-EIA negative and RPR nonreactive but had a positive FTA-ABS result. The RPR was significantly more sensitive than the EIA (76.9% vs. 53.8%, P = 0.005). Trep-Sure EIA positivity was also significantly associated with higher median RPR titer (P = 0.011).

Conclusions: Use of the TS-EIA may result in underdetection of primary syphilis compared with the RPR. Further evaluation of the sensitivity of the TS-EIA in high-morbidity settings is warranted before the adoption of reverse sequence screening algorithms.