Predictors of serological cure and Serofast State after treatment in HIV-negative persons with early syphilis

Abstract

Background: Syphilis management requires serological monitoring after therapy. We compared factors associated with serological response after treatment of early (ie, primary, secondary, or early latent) syphilis.

Methods: We performed secondary analyses of data from a prospective, randomized syphilis trial conducted in the United States and Madagascar. Human immunodeficiency virus (HIV)-negative participants aged ≥ 18 years with early syphilis were enrolled from 2000-2009. Serological testing was performed at baseline and at 3 and 6 months after treatment. At 6 months, serological cure was defined as a negative rapid plasma reagin (RPR) test or a ≥4-fold decreased titer, and serofast status was defined as a ≤ 2-fold decreased titer or persistent titers that did not meet criteria for treatment failure.

Results: Data were available from 465 participants, of whom 369 (79%) achieved serological cure and 96 (21%) were serofast. In bivariate analysis, serological cure was associated with younger age, fewer sex partners, higher baseline RPR titers, and earlier syphilis stage (P ≤ .008). There was a less significant association with Jarisch-Herxheimer reaction after treatment (P = .08). Multivariate analysis revealed interactions between log-transformed baseline titer with syphilis stage, in which the likelihood of cure was associated with increased titers among participants with primary syphilis (adjusted odds ratio [AOR] for 1 unit change in log(2) titer, 1.83; 95% confidence interval [CI], 1.25-2.70), secondary syphilis (AOR, 3.15; 95% CI, 2.14-4.65), and early latent syphilis (AOR, 1.86; 95% CI, 1.44-2.40).

Conclusions: Serological cure at 6 months after early syphilis treatment is associated with age, number of sex partners, Jarisch-Herxheimer reaction, and an interaction between syphilis stage and baseline RPR titer.

Figures

Figure 1.

Flow of participants in the trial and inclusion in the secondary analysis.

Figure 2.

Serological cure as a function of baseline rapid plasma reagin (RPR) titer and stage of syphilis. The proportion or observed probability of cure was computed using logistic regression as a function of the baseline RPR titer and stage of syphilis (primary, secondary, and early latent [EL] syphilis). Colored bands represent 95% confidence intervals around each value of the baseline RPR titer for each stage of syphilis, with smaller bands for titers representing a larger proportion of baseline values from study participants. Overall, participants with early syphilis and baseline RPR titers ≥1:64 had higher cure rates than those with lower RPR titers. There was a higher probability of cure for participants with baseline titers ≤1:32 and primary syphilis, compared with participants with secondary or EL syphilis.