Syphilis is a sexually transmitted disease (STD) caused by Treponema pallidum. Syphilis remains a common disease worldwide, and China has been burdened with the prevention, control and management of syphilis [1]. While syphilis could lead serious complications when left untreated, it is simple to cure with the appropriate treatment at early stages [2]. Having syphilis once does not protect patients from getting the disease again. Even after successful treatment, patients might still be re-infected by unprotected sexual contact. The re-infection rate of syphilis is high, especially among high-risk population [3–5]. For example, some studies in men who have sex with men (MSM) reported reinfection rates among this population ranging from 5.9 to 22% [6–9]. Two American studies also showed that 9.0 and 21.8% of human immunodeficiency virus (HIV) co-infected patients experienced syphilis treatment failure or re-infection, respectively [10, 11].

Although some researchers studied the response to treatment by serological monitoring after therapy, the focus of those studies was mainly on serological cure and serofast status [12, 13]. Studies on syphilis treatment failure and/or re-infection in China are limited. A better understanding of syphilis treatment failure and/or re-infection and the associated factors would help in identifying high-risk populations and improving treatment outcomes. Moreover, the results could provide scientific evidence for formulating and optimizing intervention strategies for syphilis.

In April 2011, Nanshan district, Shenzhen, launched the Syphilis Convergence Case-management Project. This pilot project aimed to consolidate prevention, treatment and management to better control syphilis. All the medical and public health faculties within the district are encouraged to provide syphilis seropositive cases referrals to the STD clinic, Department of Dermatology and Venereology in Nanshan Center for Chronic Disease Control, which is responsible for providing standardized treatment and regular serologic follow-up. We conducted a cohort study to compare the outcomes of treatment. The aim of our study was to detect the rates of treatment failure and re-infection and to explore the potential associated factors.

During the study period, 1133 patients were screened for syphilis and received standard treatment. Among the 770 patients who completed the 2-year follow-up, 512 were first diagnosed. After excluding two neurosyphilis cases, 510 participants were included in the analysis (Fig. 1).

The proportions of evaluable participants who exhibited effective treatment varied by syphilis stage and time point after therapy. Overall, the effective proportion of treatment was 89.06% (456/510), which included 300 (58.60%) serological cures and 156 (30.46%) serofasts. A total of 56 (10.94%) participants experienced treatment failure or reinfection, and the proportion of MSM population, HIV-infected participants and combined patients were 18.03, 15.78 and 18.18%, respectively.

Previous research has discussed the serological response to syphilis treatment [8–11, 14]. The time point used in the analysis is important when assessing the treatment effect, and a relatively wide window of time is recommended for determining serologic responses [2]. Compared with previous studies, each participant included in the final analysis received a 2-year follow-up, which provided more sufficient information to determine their outcomes. We report an overall rate of 10.94% among first-diagnosed patients, and the figures for the MSM population and HIV-infection patients were much higher. Therefore, it is very important to perform regular clinical and serologic evaluations after treatment, especially for the high-risk populations. To the best of our knowledge, this is the first study in China to detect the treatment failure and reinfection rate among STD clinic patients after long term follow-up.

We identified three potential associated factors with the treatment failure and reinfection in this study. The finding supports the findings of other investigations that also demonstrated that HIV- infected patients with syphilis might be more likely to experience serological failure and reinfection compared to non-HIV infected patients [17–19]. We speculate that HIV-infected patients with more immunosuppression might respond with a low rate to effective treatment. In addition, HIV might accelerate and change the clinical course of syphilis, and this co-infection could increase the incidence of the complications of syphilis [16, 20, 21]. Furthermore, HIV makes it more likely for syphilis to present with non-typical features [2]. For example, a UK enhanced surveillance program reported on the presentation of syphilis both in HIV-negative and HIV-positive men. Primary syphilis was diagnosed in 42 and 27%, and secondary disease was diagnosed in 40 and 58% of HIV-negative and HIV-positive patients, respectively [22]. Therefore, it is important for medical practitioners to be aware of how syphilis might present in patients with underlying HIV infection and the implications for treatment and follow-up.

We also found that secondary syphilis patients are more likely to experience treatment failure or reinfection compared with primary syphilis patients. Although there are no concrete indications regarding the different stages, and the clinical stages of syphilis often overlap, this finding highlights the importance of early diagnosis and treatment. Furthermore, the frequency of condom use was found to be associated with the response to treatment. Patients might be linked to recurrent exposure of syphilis from their sexual networks through unprotected sexual activity [23]. Patients who practice safe sex are less likely to be re-exposed to risky partners. Unprotected sexual behaviors, therefore, increase the likelihood of experiencing treatment failure or reinfection. This finding suggested that health counseling and safety education on sex activity should be intensified to increase condom use, especially among high risk populations.

Our study has several limitations. First, selection bias is possible because 28% of patients were excluded due to lack of data or loss of follow-up. If the exclusion of these patients affected the outcome is unknown. Second, some information collected at baseline is retrospective, such as the number of sex partners and the frequency of condom use, which might not reflect the actual conditions during the follow-up. In addition, a relatively rough classification was adopted in this study. We combined serological cure and serofast together. We did not distinguish patients with treatment failure or reinfection in the analysis, so it is hard to make a distinction without information regarding their exposure and network. Furthermore, the bivariate analysis used for the first-step variable selection might be not biologically plausible all of the time. Additionally, the results from one city might be not well generalized to other regions. However, we believe this study provides a good representation of southeast coastal cities, and more studies in diverse areas are needed in further research.

Treatment failure and reinfection of syphilis should be considered in the management of syphilis patients. The clinical implications of our findings suggest that HIV-infected patients and secondary syphilis patients should be closely monitored for serological failure after syphilis treatment.