Introduction

Syphilis has reemerged as a growing risk to global public health, particularly in persons living with HIV.¹ This resurgence has been noted principally in men who have sex with men (MSM), although it has also affected other groups with high frequencies of HIV infection such as female sex workers, intravenous drug users, and also those with heterosexual practices.^{2,3,4,5,6,7,8,9}

The incidence of syphilis has also been shown to increase proportionally with the duration of antiretroviral therapy (ART) in those infected with HIV. This observation has led to the recommendation of early detection of syphilis in patients receiving ART, in function with the time in which they have been in treatment.¹⁰ Furthermore, various authors have noted an increased frequency of syphilis reinfection in HIV carriers, particularly in MSM who could be playing an important role in syphilis transmission.^11,12,13,14

Syphilis is diagnosed through serological testing and the presence of clinical manifestations. Generally, when tests are applied to confirm clinical findings or for screening purposes, the first used is the nontreponemal test (rapid plasma regain or venereal disease research laboratory ) whose positive results must be confirmed with a subsequent treponemal test (ex. FTA-ABS).¹⁵ The advent of rapid treponemal test kits has allowed the algorithm for syphilis diagnosis to be reversed; that is, the first test applied may be treponemal, and positive results may then be corroborated with a nontreponemal test.¹⁶

In the Clínica Especializada Condesa (CEC), which operates the Mexico City HIV/AIDS Program, integral detection of HIV/STI testing which includes syphilis has been implemented since 2009, using the reverse algorithm.^* Due to the large number of clinic users who test positive for syphilis it was found to be critical to establish a continual monitoring program for syphilis in ART patients.^17,18,‡,§ To this end, rapid treponemal tests were applied and positive results were further subjected to RPR test in two different laboratories, one in CEC and one in Clínica Especializada Condesa Iztapalapa (CECI). Then a protocol was established which guarantees the use of the standardized and reproducible test method in all clinic users.

This study aims to establish the beginnings of an incidence cohort of syphilis in ART patients, realizing a base evaluation of the presence of T. pallidum infection markers in HIV carriers who receive ART in the Mexico City HIV/AIDS Program. It also analyzes the relationship between prevalence of these pathological conditions with sociodemographic factors, sexual practices, previous syphilis or gonorrhea diagnosis, and length of ART in the study population.

Materials and methods

Results

A total of 2 386 individuals were studied, with an average age of 36.9 years (SD: 10.6), largely male (96.6%) and the rest transsexual, predominantly single (70.5%), an average age of first sexual relation at 16.9 years (SD: 3.8), over half with higher education or technical certification (54.5%), high proportion were MSM (90.8%), nearly one-third with previous syphilis diagnosis, (30.9%), a portion with previous gonorrhea diagnosis (13.9%), and over half originating from Mexico City (59.8%). The prevalence of “active or resolved” or “probable active” infection markers was 44.2% (1 053/2 383) and 28.8% (686/2 383), respectively. The distribution of positive titers results of the RPR test in the analysis of the 879 clinically studied CEC participants is shown in table I. Of a total of 272 positive cases determined by the reverse algorithm method, 258 (94.8%) were concentrated within three categories of clinical care: a) treated syphilis in follow-up (23.8%), b) treated and cured syphilis (37.8%), and c) syphilis reinfection in follow-up (33.1%).

Results of the bivariate analysis showed a significant association between “active or resolved” and “probable active” syphilis infection with marital status, previous syphilis diagnosis, hetero-homosexual practices, and insertive-receptive practices. “Active or resolved” infection was also associated with the type of clinic in which interviews were undertaken, as well as with sexual exchanges for money or goods. No relation was found between syphilis markers and age, education level, gender of the individual, birthplace, age of first sexual relation, having paid for sexual relations, intravenous drug use, or duration of ART (table II).

After controlling for age in the logistical regression model, the predictive variables of “active or resolved” infection were: CECI (OR=1.6, CI95%: 1.2-2.0), previous syphilis diagnosis (OR=14.2, CI95%: 11.2-18.2); MSM (OR=2.8, CI95%: 1.8-4.2); and receptive sexual practices (OR=1.3, CI95%: 1.0-1.7). For “probable active” infection markers, predictive factors were found to be: previous syphilis diagnosis (OR=7.0, CI95%: 5.7-8.6); MSM (OR=2.3, CI95%: 1.4-3.8); and receptive sex practices (OR=1.6, CI95%: 1.2-2.1) (table III).

Discussion

Since the second decade of the century, growing numbers of syphilis infection have been reported in Mexico in individuals with high-risk sexual practices such as male sex workers (MSW) and MSM. For example, in 2014 a high frequency of active syphilis infection was reported in 21% of MSW, in whom frequency of HIV infection was also high at 35%.²¹ In a similar study in 2015, prevalence of treponemal antibodies in HIV-positive patients was at 25%, rising to 30% in MSM.²²

In this context the present study chose to include individuals undergoing ART in whom a high prevalence of treponemal antibodies (44.2%) were encountered: higher than numbers previously reported by Mata-Marín and colleagues²² in MSM carriers of HIV. The treponemal test method identifies those individuals with history of syphilis treatment, or untreated or incompletely treated syphilis,¹¹ situations in which a high proportion of study subjects found themselves.

Secondary application of the nontreponemal test, in accordance with the reverse algorithm procedure for syphilis diagnosis, showed that the prevalence of “probable active” cases was 28.8%, with respect to the total population studied. This frequency was higher in the present study than in the 21% previously found for MSW in Mexico,²¹ as well as of the 21% frequency in MSM HIV-positives in Japan.²³ These high frequencies of “active or resolved” and “probable active” infection suggest that the population under study has a high level of interaction between individuals infected with syphilis and susceptible to the infection in HIV-positive patients, as has been previously reported particularly among MSM HIV-positives according to a meta-analysis with results from 11 European countries.²⁴

The results of this study highlight the importance of syphilis monitoring by the Mexico City HIV/AIDS Program. For example, while in the present study a 28.8% rate of probable active syphilis was found, the statistics corresponding with a total of 19 350 men, 7 427 women, and 320 transsexuals who realized HIV testing at the CEC during the year of 2018 were 10.24, 2.1, and 15.31%, respectively. Nonetheless, the prevalence of probable active syphilis in HIV-positive men (n=3 243), women (n=176), and transsexuals (n=54) in the same time period were 25.5, 6.8, and 29.6%, respectively.^* These numbers demonstrate that the groups with greater prevalence of probable active syphilis in the CEC and CECI are men and transsexual who are HIV-positive.

The results of the RPR test titers presented in table I justify the methods used in this study to define probable active syphilis infection. It has been documented that antibody titrations of the nontreponemal test may correlate to the activity of the syphilis infection and may therefore be used to follow up on the response to treatment. A four-fold change (ex. 1:16 to 1:4 or 1:8 to 1:32) in titrations of two results of nontreponemal tests using the same serological test can serve to judge the effects of the syphilis treatment.^11,15 Distribution of titers in the first clinical category of “under treatment” syphilis in this study suggest that initiation of syphilis treatment is recent, and therefore cases of high dilution titrations still exist, while in the second clinical category of “treated and cured” syphilis the results of titrations are concentrated in low values of dilution. In the third clinical category of “reinfection of syphilis under treatment” it was clearly shown that the percentage distribution of the titers rose along with the rise in test dilution, suggesting the presence of syphilis infection activity.

In terms of risk factors, the present study showed that there was a significant upward trend in the risk of probable active syphilis infection in MSM, taken as a reference group those with heterosexual practices. Furthermore, this study showed a significant relationship between the prevalence of probable active syphilis and receptive sex practices, showing that the male HIV-positive population with the highest frequency of probable active syphilis in this study is that which identifies as MSM according to Kinsey and collagues classification of male sex practices.²⁰ Along these lines, it has been reported that one of the predictive factors associated with coinfection of HIV and syphilis in males with exclusively homosexual practices are insertive-receptive practices²⁵ and also that the highest rates of syphilis infection in HIV-positive males are presented in MSM.²⁴ In Mexico, the risk of HIV infection in men is significantly higher among males with homosexual as well as insertive-receptive practices in accordance with a pioneer study on the subject.²⁶

The strongest predictive factor associated with probable active syphilis infection in the present study was previous syphilis diagnosis. These results suggest a high rate of syphilis reinfection in the population under study, and are consistent with reports in the literature of previous syphilis diagnosis and presence of reinfection by T. pallidum found predominantly in HIV-positive MSM.^27,28 Furthermore, 33.1% of the cases evaluated in their clinical stage (table I) were classified as “reinfection of syphilis under treatment,” which supports the hypothesis of a high rate of syphilis reinfection for the population under study.

The lack of evaluative components regarding condom use or number of sexual partners represents one of the most important limitations of the present work. In addition, the hypothesis relating the frequency of syphilis infection and longer duration of ART in HIV-positive individuals could not be tested.