Prevalence of Syphilis Serology Testing Algorithm Among Laboratory Participants in the College of American Pathologists Proficiency Testing Program

To the Editor.—Syphilis caused by Treponema pallidum subsp pallidum is one of the oldest sexually transmitted diseases, and yet a common one that has made a significant comeback since the early 2000s.^1–5  Laboratory diagnosis of syphilis has historically been complex and hard to interpret given the use of different tests, different expertise levels, and more importantly the protean clinical manifestations of syphilis itself, which earned the moniker “the great imitator.” Syphilis laboratory testing typically follows an algorithm. The traditional algorithm starts with a nontreponemal antibody-based test, that is, a test that is not specific to T pallidum antigens but rather detects antibodies to cardiolipin, cholesterol, and lecithin. Examples include but are not limited to rapid plasma reagin and venereal disease research laboratory, which are manual tests. Once one of these tests is reactive, a treponemal antibody test is added. This can be done through a variety of test principles, manually, or on semiautomated or fully automated platforms. On the other hand, the so-called reverse-sequence testing algorithm (a misnomer) starts with the treponemal antibody test (specific test) and, where reactive, is supplemented by a nontreponemal assay. If the latter is nonreactive, a second treponemal assay is added to verify the initial reactive screen test.^1–3,5 

Previously, to gauge the prevalence of use of different algorithms by the participants in the College of American Pathologists proficiency testing program, Rhoads et al⁶  reported that among laboratories that responded to their supplementary questions, 63.1% followed the traditional algorithm, 15.9% used the reverse-sequence algorithm, and the rest used both algorithms, were unsure, or reported other as a response. Because the data were reported in 2015, and because there has been an anticipated trend in using the reverse-sequence testing algorithm, in an effort to update this prevalence, we set out to send another supplementary question at the end of 2023 to all College of American Pathologists proficiency testing participants in the syphilis serology program (G-C 2023).

Of the 2371 kits distributed in the G-C 2023 mailing, 1361 (57.4%) were returned with at least one response. After excluding incomplete responses, not applicable responses, and duplicate submissions, a total of 1287 responses, of which 81.4% (1047) were from United States laboratories, were analyzed. The participants were given the choice of selecting any combination of the following responses: reverse-sequence algorithm, traditional algorithm, individual syphilis tests (no algorithm), other, and not applicable (syphilis testing is not performed in-house). The other responses were reviewed and were reclassified into one of the 3 categories: reverse-sequence testing algorithm, traditional algorithm, or no algorithm. In the latter group, 11 of 40 laboratories performed only a treponemal assay, showing their tendency to use the reverse-sequence testing algorithm.

After removal of 1 laboratory that did not do the test in-house, of a total of 1286 responses, 616 and 455 responses were received for the reverse-sequence testing algorithm and the traditional algorithm, respectively, which shows a prevalence of 47.9% for the former and 35.4% for the latter, a significant shift in using the reverse-sequence testing algorithm since 2015 (15.9%). There were 317 laboratories (24.7%) that offered the individual tests without following an algorithm.

The results are interesting but not unexpected. The advantages of the reverse-sequence testing algorithm include earlier diagnosis (especially important in the third trimester of pregnancy given the delayed nontreponemal antibody response), higher sensitivity in late latent and tertiary syphilis, fewer biological false positives (as seen with the rapid plasma reagin and venereal disease research laboratory tests), and significant labor saving for laboratories, especially where automated platforms are used.^1–5  Our data should inform future guidelines for syphilis laboratory diagnosis.

The authors would like to thank Ann King and Christine Bashleben from the College of American Pathologists for their kind assistance.