Serological response after shorter durations of therapy for syphilis of late latent or unknown duration: What does it mean?

We read with interest the retrospective cohort by Pugsley et al. comparing serological response after different regimens for late latent syphilis or syphilis of unknown duration [1]. Using public health data, the authors assembled 18,027 cases and found no significant differences in the proportion achieving a fourfold decline in lipoidal (nontreponemal) antigen titers by 24 months after one or three doses of benzathine penicillin G (BPG) or 28 days of doxycycline. Other retrospective studies report similar findings [2]. We urge caution in interpreting these results and consider what serological response truly reflects in late latent disease.

Syphilis treatment outcomes can be conceptualized as three forms of “cure.” Microbiological cure - the eradication of Treponema pallidum - is the desired endpoint but cannot be directly confirmed in routine practice. Clinical cure denotes resolution of signs and symptoms but may not indicate microbiological cure (latent infection is asymptomatic yet persistent). Serological cure is defined as at least a fourfold decline in lipoidal antigen titers (historically, seroreversion). Serological response can be an indirect marker of microbiological cure, but the overlap depends on context and disease stage.

For early syphilis, animal data and decades of clinical experience support that sustained treponemicidal penicillin exposure for ~7 days, as provided by a single 2.4-million-unit dose of BPG, is required to achieve microbiological cure [3–5]. In early disease, serological cure is used as a proxy because pre-antibiotic-era data link serological response with reduced risk of late neurological complications (Figure 1A) [6]. However, serological decline can occur without adequate treatment [7]. For example, consider a patient who takes only two doses of a 14-day doxycycline course but subsequently achieves a decline in titer from 1:64 to 1:8. This serological response may not reliably reflect microbiological cure, raising a critical issue: should those who meet serological criteria following a potentially suboptimal regimen be considered adequately treated?

Figure 1A.

Overlap of microbiological and serological cures in early syphilis⁷

Serological response is more complex in late syphilis. Slower spirochete replication may require higher cumulative penicillin exposure for microbiological cure - a notion supported by experimental findings but not yet definitive [8–9]. Thus, the optimal regimen for late latent syphilis is uncertain. Historically, extended courses - three weekly BPG doses (~28 days of treponemicidal penicillin exposure) or 28 days of doxycycline - have been recommended and associated with favorable outcomes. Importantly, unlike early disease, seroreversion in late syphilis was not associated with reduced risk of subsequent neurological complications [9], weakening the assumption that serological cure implies microbiological eradication (Figure 1B).

Figure 2B.

Overlap of microbiological and serological cures in late syphilis¹⁰

Given these limitations, we caution against interpreting comparable serologic responses following abbreviated regimens as evidence that shorter courses achieve microbiological cure. A randomized clinical trial would likely be unable to definitively answer this because follow-up needed to capture outcomes (such as late neurological complications) would be impractically long. Until robust data demonstrate that shorter regimens reliably prevent long-term complications, a single dose of BPG should not be adopted for the treatment of late latent syphilis or syphilis of unknown duration, regardless of observed serological responses.