A National Plan to eliminate syphilis in the United States was launched in 1999.1 After a decade of decline, in 2000 reported rates of infectious syphilis in the United States were the lowest since surveillance began in 1941.1 However rates of primary and secondary syphilis (P&S) began increasing among men who have sex with men (MSM), many of whom were human immunodeficiency virus (HIV)-infected,2–5 marking the beginning of a new epidemic.5 This resurgence in rates was accompanied by a shift in cases from black heterosexuals, a population with a long history of high rates, to MSM, who were more likely to be Hispanic or white.4 Although the last 3 syphilis epidemics in the United States (1975, 1982, and 1990) peaked after 5 to 6 years, rates in the United States, as well as Florida, have been increasing for more than 8 years. In Florida, P&S rates (per 100,000) increased from 2.6 in 2000 to 5.7 in 2008.6,7 Among men in Florida P&S rates increased from 3.4 to 9.5 over the same period.6,7 Assuming 4% of the male population were MSM,8 the P&S rate among MSM increased from 19.0 to 166.0 whereas the rate among heterosexual men increased from 2.7 to 3.0 (State of Florida, unpublished data).
Given persistently increasing rates of syphilis in Florida, we used the case report database to investigate the extent to which persons with multiple syphilis diagnoses (repeaters) from 2000 to 2008 may be contributing to the current epidemic. Although recent studies have examined the epidemiology and risk factors associated with repeat diagnoses of syphilis,9–11 disease presentation has infrequently been reported.12 We therefore investigated disease presentation at first and second diagnosis among repeaters to explore the existence of any patterns of presentation. Finally, we looked for evidence suggesting that a person with a second episode of syphilis might be more likely to be asymptomatic (perhaps due to acquired immunity) or more likely to recognize symptoms and seek treatment while symptomatic.
MATERIALS AND METHODS
Syphilis Case Detection and Reporting
Both syphilis diagnoses and positive laboratory tests are reportable, by law, to the Florida Department of Health (DOH). Test results, the individual's name, demographic information, and provider information are required for reporting. More than 500 laboratories report electronically directly to the Florida DOH or by paper to 1 of the 67 county health departments. Clinical providers are also required to report cases. All laboratory and clinician case reports are entered into the statewide surveillance reporting system. All licensed syphilis tests are accepted and test type is a required variable during the reporting process. Surveillance personnel search records of previously treated syphilis cases (documentation of treatment and past titers) to determine the need for follow-up. If a positive test is suspected of being associated with a new case, surveillance personnel contacts the provider to obtain information on signs and symptoms, diagnosis, treatment, demographics, and locating information. If a specific diagnosis cannot be obtained after discussion with the provider, the individual is contacted by local disease intervention specialists who refer the client to the local sexually transmitted disease (STD) clinic for further evaluation. Evaluation may include additional laboratory testing if titer changes are inconclusive. Health department staff attempts to interview all persons with diagnosed or suspected early syphilis. The interview elicits sexual orientation, risk behaviors, HIV status, and information to help notify sex partners. The State HIV/AIDS case database is cross-referenced to identify persons with both infections. Data completion are more than 95% for demographic variables except for race/ethnicity, which was >90% complete during the period of this study. Data were collected statewide using STD Management Information System until February 2007 when the data system was changed to the Patient Reporting Investigation Surveillance Manager. Cases reported between January 1, 2000 and December 31, 2008 were used for this study.
Stages of syphilis were determined by the disease intervention specialist and surveillance specialists according to Centers for Disease Control and Prevention surveillance definitions2; however, persons with latent infection of unknown duration were classified as having late latent syphilis for this analysis. A syphilis repeater was defined as a person who, between January 1, 2000 and December 31, 2008, reported a syphilis case to the Florida DOH STD Surveillance system, received recommended treatment,13 and was again diagnosed with syphilis. Repeat diagnoses required a 4-fold titer decrease following treatment of the initial infection followed by a 4-fold titer increase at subsequent diagnosis. However, when the repeat episode was either primary syphilis or secondary syphilis diagnosed more than 2 years after the initial diagnosis the 4-fold decrease was not required. Also, if the initial diagnosis was late latent disease with a titer of 1:4 or less, and the titer was stable for at least 1 year, a 4-fold increase was considered a new infection even if there was no decrease after treatment. For persons with more than 2 diagnoses within the period only the first 2 diagnoses were included in comparisons of first and repeat diagnosis.
MSM were defined as men who reported male sex partners within the interview period (4–12 months, depending on stage of syphilis) even if they also reported female partners. Men who refused to be interviewed but were named by a male as a sexual contact were also considered MSM. Men who had sex with women (MSW) were defined as men who reported only female sex partners. If information on sexual orientation was not documented it was considered unknown.
For all persons diagnosed with syphilis and HIV, the date of their earliest reported positive HIV test was compared with the date of their syphilis diagnosis. Only persons with documented HIV infection at the time of a syphilis diagnosis (or within 60 days after the diagnosis) were considered HIV+ in the analysis. For repeaters, HIV infection status was based on status at the first syphilis diagnosis. Because of sample size limitations, disease presentation by HIV status among repeaters was limited to MSM.
The repeaters were compared with nonrepeaters using bivariate and multivariate analyses. Chi-square tests of independence and t tests, for frequencies and means respectively, were used to test for differences between the 2 groups.
Multiple logistic regression analyses were used to identify demographic correlates of repeater status. Variables in the analysis were those that are routinely collected in STD surveillance: race/ethnicity, gender, age category, HIV status at time of diagnosis, county of residence and, for males, sexual orientation. Residence was categorized as South Florida (Miami-Dade and Broward Counties, including Miami and Fort Lauderdale, the centers of the MSM syphilis epidemic) versus other parts of the State. Each of these variables was significantly associated with repeat infection in the univariate analysis and included in the multivariate analysis. All statistical tests were 2-tailed and results were considered statistically significant when P ≤ 0.05. Repeaters were divided as follows: MSM (HIV-infected vs. HIV-uninfected), MSW, males of unknown sexual orientation, and women to identify differences in disease presentation at initial diagnosis, with use of bivariate analysis to detect significant differences between initial versus second presentation and between groups.
All statistical analyses were conducted using SAS software (software version 9.1.3 Service Pack 4; SAS Institute Inc, Cary, NC).
A total of 26,070 persons were diagnosed with syphilis in Florida between 2000 and 2008. Reported cases of P&S increased from 413 cases in 2000 to 984 in 2008. Over this time, the percent of persons with P&S who were MSM increased (14.2–55.2); HIV-infected persons increased (9.5–36.1); and women decreased (37.2–18.5). Syphilis was diagnosed more than once in the period for 643 persons (2.5%). Between 2001 and 2008, the annual percent of cases that were repeaters increased: among all persons (from 0.2 to 5.1); among MSM (from 0.6 to 11.7); and among persons with HIV coinfection (from 0.8 to 13.3). Among all repeaters, 29.4% were MSM who were HIV-infected at the first syphilis diagnosis. If subsequent HIV tests are included, 45% of all repeaters were HIV-infected MSM. Of MSM diagnosed with early syphilis in South Florida, 13.4% were repeaters in 2008. Of all persons with early syphilis diagnosed in South Florida in 2008, 10.4% were repeaters.
Among the 643 repeaters, 33 (5.1%) had 3 diagnoses, 4 (0.6%) had 4 diagnoses, and 1 (0.2%) had 5 diagnoses. The average time between first and second diagnoses was similar for HIV-uninfected (3.0 years; median, 2.5 years; range, 97 days–8.4 years) and HIV-infected repeaters (2.8 years; median, 2.4 years; range, 156 days–7.9 years). The median titer increase for those diagnosed with latent disease at second diagnosis was 64-fold for HIV-infected and 32-fold for HIV-uninfected repeaters. Repeaters were more likely to be MSM (69.1%) than MSW (6.7%), men of unknown sexual orientation (11.5%), or women (12.8%) and were more likely to be white (44.8%) than nonrepeaters (23.3%) (Table 1). HIV infection was more common among repeaters (37.6%) than nonrepeaters (13.6%). In addition, compared with nonrepeaters, repeaters were more likely to be 30 to 45 years of age and to live in South Florida. In multivariate analysis, repeat diagnoses were most strongly associated with being the following: MSM (OR: 5.3; 95% CI: 4.1, 6.9); HIV-infected (OR: 2.0; 95% CI: 1.7, 2.4); white (OR: 1.5; 95% CI: 1.2, 1.8); 35 to 39 years old (OR: 1.8; 95% CI: 1.4, 2.4); and a South Florida resident (OR: 1.7; 95% CI: 1.4, 2.0).
On evaluating disease presentation among persons diagnosed with syphilis once, MSM, regardless of HIV status, were less likely to present with late latent disease than others (16.7%–18.1% vs. 40.9%–75.1%, P ≤ 0.001) (Table 2). The high percentage of late latent disease among the non-MSM groups skews the disease distribution, decreasing the percentage of primary and secondary cases. Therefore, we calculated the primary-to-secondary disease ratio to determine who, among symptomatic persons, was most likely to present with primary disease. The lowest primary-to-secondary disease ratio was among women (1:7.1 for HIV-uninfected and 1:20.4 for HIV-infected), followed by MSM (1:2.9 for HIV-uninfected and 1:7.6 for HIV-infected) and the ratio was highest among MSW (1:1.7 for HIV-uninfected and 1:2.2 for HIV-infected). For all groups, the primary-to-secondary ratio was lower for HIV-infected persons than for HIV-uninfected persons (P = 0.02 for MSW, P < 0.001 for all others).
Overall, the stage at diagnosis was similar for groups of repeaters, whether it was their initial or second diagnosis (Table 3). However, HIV-infected MSM were more likely to have secondary as their initial diagnosis (42.3%) than as their subsequent diagnosis (25.4%, P < 0.01); they were less likely to have early latent as their initial diagnosis (38.1%) than as their second diagnosis (52.9%, P ≤ 0.01). At both initial and second diagnosis, MSM were less likely to present with late latent infection than women or other men (13.2%–18.8% vs. 30.8%–35.4%, P < 0.01). MSM, despite HIV status, were less likely to be diagnosed with late latent syphilis at second diagnosis than HIV-infected women and other HIV-infected men (16.7% vs. 35.8%, P < 0.01, data not shown).
We compared each individual's stage at first and second diagnosis; most presented at a different stage the second time compared to the first (Table 4). However, individuals were more likely to repeat in the same stage than would be expected by chance. Persons initially diagnosed as primary syphilis were more likely to repeat as primary (8/39, 20.5%) than repeaters who were first diagnosed in another stage (34/604, 5.6% OR = 4.3, 95% CI: 1.7–10.6). Persons with secondary as their initial diagnosis were more likely to repeat as secondary (83/225, 36.9%) than repeaters first diagnosed in another stage (95/418, 22.3%, OR: 2.0, 95% CI: 1.4–2.9). Among those with early latent at first diagnosis the odds ratio for repeating the same diagnosis was 1.6 (95% CI: 1.1–2.2); among those with late latent disease the odds ratio of repeating the same diagnosis was 1.7 (95% CI: 1.1, 2.6). After stratifying by subgroups, findings were similar for HIV-uninfected MSM, HIV-infected MSM, and other males. Among the small number of women this was found only for early latent disease.
Overall, only 2.5% of persons diagnosed with syphilis between 2000 and 2008 in Florida were repeaters. However, in South Florida 10.4% of all early syphilis cases in 2008 were repeaters. Other investigators have reported on the frequency of repeaters, however, the time periods have varied making comparisons difficult. In Seattle, 17.6% of early syphilis cases between 1992 and 2008 were repeaters.12 In British Columbia, among persons diagnosed with early syphilis in 2007, up to 10% had another diagnosis in the preceding 10 years.10 In Chicago, 4.1% of early syphilis cases from 2000 to 2005 were repeaters, with repeat rates of 10.5% among MSM.9 Finally, in San Francisco, 6.7% of MSM diagnosed with early syphilis between 2001 and 2002 were rediagnosed within 1 year.11
We found repeat infection was most common among HIV-infected MSM. Other studies have also shown high rates of repeat infection among MSM9–12 and HIV-infected individuals.9,10 HIV-infected MSM receiving HIV care and routine screening for syphilis may account for some of the increase, particularly the diagnosis of early asymptomatic disease.14,15 It could be argued that fluctuating titers after treatment account for some proportion of new diagnoses among HIV-infected persons. However, studies of treatment failure in HIV-infected persons with syphilis generally report slow and incomplete decreases in serologic titers after treatment, with increases in titers being rare.11,16,17 Therefore, although cases of treatment failure cannot be ruled out, we do not believe that fluctuating titers are an important cause of repeat diagnoses of syphilis in this study, especially as changes in titer were generally large. Additionally, many repeaters were symptomatic and stages at second and first diagnoses were similar. Although increased screening among HIV-infected MSM likely contributes to increased case detection among this population, alternative explanations exist. Treatment optimism regarding highly active antiretroviral therapy leading to an increase in high-risk sexual behaviors18,19 as well as serosorting, or the practice of limiting sexual partners to those of the same HIV serostatus to decrease viral transmission11,20 are among the explanations that have been explored.
The stage at diagnosis of syphilis varied by sexual orientation, gender, and HIV status. Chancres are typically painless, so primary syphilis may go undetected if the chancre is hard to see, as would be expected for women and MSM engaging in receptive anal intercourse. We found women and MSM were less likely to be diagnosed as primary infection, and MSW, who we expect would be more likely to have a visible penile lesion, were more likely to be diagnosed with primary infection. We also found that HIV-infected MSM were less likely to be diagnosed with primary infection than HIV-uninfected MSM, perhaps because HIV-infected MSM were more likely to have had asymptomatic rectal (primary) lesions. “Strategic positioning” or assumption of the receptive position in unprotected anal intercourse by the HIV-infected male has been described as a risk reduction strategy for some MSM to reduce HIV transmission18 and may be related to this finding. In contrast, MSM who engage in unprotected receptive anal sex are at higher risk of HIV infection than men who practice insertive sex21 and in this retrospective study it is not possible to determine whether either of these possibilities explains the decreased primary-to-secondary syphilis ratio among HIV-infected MSM. HIV-infected MSM may also be more likely to contract syphilis through oral sex, whereas decreasing HIV transmission risks, is efficient for syphilis transmission.22 Interestingly, in this sample, all groups of HIV-infected persons were more likely than HIV-uninfected persons to present with secondary syphilis, a finding we have not seen reported elsewhere. Secondary syphilis could be prevented by frequent serologic screening and treatment before infection progresses to secondary syphilis. We found evidence to suggest this may be happening; among HIV-infected MSM, who would be most likely to have routine screening, repeaters had less secondary infection on their second episode compared to their first, and more early latent. Increased syphilis screening among HIV-infected MSM receiving primary care in Europe has led to increased detection of asymptomatic early infections14,15,23,24; we suspect that this is the case here as well. Some authorities advocate syphilis screening every 3 to 4 months, along with routine CD4 and viral loads24,25 as part of efforts to decrease transmission.
Late latent syphilis was the most commonly diagnosed stage for groups other than MSM. This may reflect more aggressive screening among MSM, despite HIV status, as MSM are at highest risk. Among the general population, screening now occurs in very few scenarios such as prenatal visits, immigration exams, and blood donation. Compared to MSM, HIV-infected heterosexuals were more likely to be diagnosed with late latent disease, despite the fact that annual RPR screening is recommended for all sexually active HIV-infected persons.26 Because the diagnosis of early latent syphilis often relies on comparison of serologic titers within the past year, it suggests that HIV-infected heterosexuals in Florida are less likely to be screened than MSM.
We found no major differences in stage at second diagnosis compared to stage at initial diagnosis other than the increased rates of early latent infection that were seen among HIV-infected MSM, which most likely are attributable to more frequent screening.11,14,15,23 We also found that repeaters were more likely to present with the same diagnosis at first and second presentation than would be expected by chance. This may be partly attributable to consistent sexual practices (receptive vs. insertive sex) and consistent patterns of syphilis screening. We found no evidence that having had a prior syphilis infection influenced presentation at second diagnosis; our data neither suggest that second presentations are more likely to be asymptomatic nor that repeaters are more likely to recognize symptoms and present for treatment earlier. Repeaters can present with any stage regardless of the stage of their initial diagnosis.
Our study had some limitations. We included cases diagnosed in Florida and reported to the DOH from 2000 to 2008, and so the first episode of syphilis during this period was considered an initial diagnosis even though some undoubtedly had diagnoses before 2000, and some diagnosed late in 2008 would not have had the opportunity to repeat. However, this period coincides with the current Florida syphilis epidemic. Persons who may have had other syphilis diagnoses in other states are not included as repeaters. MSM are more likely to be screened, increasing the likelihood that syphilis would be detected in the early latent stage. It is possible some MSM may not have been identified because of stigma. Additionally, data on sexual orientation were not available for some men diagnosed with late latent disease, because they were not always interviewed. HIV-infected individuals who were never tested in Florida would not have been accurately classified as HIV-infected. In addition, we did not have information on CD4 count or use of antiretroviral therapy among HIV-infected persons. Disease stage was based on signs and symptoms at time of diagnosis; persons who recalled having an ulcer would not be classified as primary if the ulcer was gone. Furthermore, disease stage was reported by many clinicians throughout the state, although local surveillance personnel worked to assure that diagnoses met appropriate criteria. The strength of our study is that our definition of a repeat episode was rigorous as it required evidence of response to treatment, i.e., a minimal 4-fold decrease in titers after treatment in most cases, decreasing the likelihood that we were studying treatment failure instead of a second infection. Repeat infection of syphilis among HIV-infected persons is sometimes defined as a 4-fold increase in titer, without noting whether there had been an initial decrease.14,15,24 However, relapse because of treatment failure and reinfection cannot always be distinguished with certainty.10
The current syphilis epidemic in Florida is driven by persons acquiring their first infection. A small number of persons, disproportionately HIV-infected MSM, have had multiple infections. These repeat infections appear to be true new infections that follow a demonstrated response to treatment and are diagnosed in all stages of syphilis. The high rates among MSM reiterate the need for more innovative approaches to reduce risk behaviors among this population. This study suggests that frequent screening of high-risk persons is useful in detecting early asymptomatic infections. Such screening among MSM may have helped to prevent even more dramatic increases in syphilis rates.
1. Centers for Disease Control and Prevention. The National Plan to eliminate syphilis from the United States. Atlanta, GA: US Department of Health and Human Services, CDC, National Center of HIV, STD and TB Prevention, 1999. Available at: http://www.cdc.gov/stopsyphilis/plan.pdf
. Accessed February 1, 2009.
2. Centers for Disease Control and Prevention. STD Surveillance, 2007. Atlanta, GA: US Department of Health and Human Services, 2008.
3. Klausner JD, Kent CK, Wong W, et al. The public health response to epidemic syphilis, San Francisco, 1999–2004. Sex Transm Dis 2005; 32:s11–s18.
4. Schmitt K, Bulecza S, George D, et al. Florida's multi-faceted response for increases in syphilis among MSM: The Miami-Ft Lauderdale Initiative. Sex Transm Dis 2005; 32:s19–s23.
5. Peterman TA, Heffelfinger JD, Swint EB, et al. The changing epidemic of syphilis. Sex Transm Dis 2005; 32:s4–s10.
6. Centers for Disease Control and Prevention. STD Surveillance, 2000. Atlanta, GA: US Department of Health and Human Services, 2001. Available at: http://www.cdc.gov/std/stats00/Tables/2000Table27.htm
. Accessed May 17, 2010.
7. Centers for Disease Control and Prevention. STD Surveillance, 2008. Atlanta, GA: US Department of Health and Human Services, 2009. Available at: http://www.cdc.gov/std/stats08/tables/25.htm
. Accessed February 1, 2010.
8. Centers for Disease Control and Prevention. CDC Analysis Provides New Look at Disproportionate Impact of HIV and Syphilis Among US Gay and Bisexual Men. Atlanta: US Department of Health and Human Services, 2010. Available at: http://www.cdc.gov/nchhstp/Newsroom/msmpressrelease.html
. Accessed May 17, 2010.
9. Ciesielski C. Repeat syphilis infection in MSM, 2000–2005. Paper presented at: 2006 National STD Prevention Conference; 2006; Jacksonville, FL. Abstract 375.
10. Ogilvie GS, Taylor DL, Moniruzzaman A, et al. A population-based study of infectious syphilis rediagnosis in British Columbia, 1995–2005. Clin Infect Dis 2009; 48:1554–1558.
11. Phipps W, Kent C, Kohn R, et al. Risk factors for repeat syphilis in men who have sex with men, San Francisco. Sex Transm Dis 2009; 36:331–335.
12. Kerani R, Lukehart S, Stenger M, et al. Is early latent syphilis more likely in patients with a prior syphilis infection? Paper presented at: 18th International Society for STD Research; 2009; London, United Kingdom. Abstract OS2.111.01.
13. Centers for Disease Control and Prevention. Guidelines for the treatment of sexually transmitted diseases, 2006. MMWR 2006; (RR-11):22–29.
14. Branger J, van ver Meer JT, Van Ketle R, et al. High incidence of asymptomatic syphilis in HIV-infected MSM justifies routine screening. Sex Transm Dis 2009; 36:84–85.
15. Cohen CE, Winston A, Asboe D, et al. Increasing detection of asymptomatic syphilis in HIV patients. Sex Transm Infect 2005; 81:217–219.
16. Yinnon AM, Coury-Doniger P, Polito R, et al. Serologic response to treatment of syphilis in patients with HIV infection. Arch Intern Med 1996; 156:321–325.
17. Rolfs RT, Joesoef MR, Hendershot EF, et al; The Syphilis and HIV Study Group. A randomized trial of enhanced therapy for early syphilis in patients with and without human immunodeficiency virus infection. N Engl J Med 1997; 337:307–314.
18. Elford J, Bolding G, Maguire M, et al. Combination therapies for HIV and sexual risk behavior among gay men. J Acquir Immune Defic Syndr 2000; 23:266–271.
19. Sullivan PS, Drake AJ, Sanchez TH. Prevalence of treatment optimism-related risk behavior and associated factors among men who have sex with men in 11 states, 2000–2001. AIDS Behav 2007; 11:123–129.
20. Hart GJ, Elford J. Sexual risk behaviour of men who have sex with men: Emerging patterns and new challenges. Curr Opin Infect Dis 2010; 23:39–44.
21. Varghese B, Maher JE, Peterman TA, et al. Reducing the risk of sexual HIV transmission: Quantifying the per-act risk for HIV on the basis of choice of partner, sex act, and condom use. Sex Transm Dis 2002; 29:38–43.
22. Centers for Disease Control and Prevention. Transmission of primary and secondary syphilis by oral sex: Chicago, Illinois, 1998–2002. MMWR 2004; 53:966–968.
23. Ivens D, Patel M. Incidence and presentation of early syphilis diagnosed in HIV- positive gay men attending a central London outpatients' department. Int J STD AIDS 2005; 16:201–202.
24. Winston A, Hawkins D, Mandalia S, et al. Is increased surveillance for asymptomatic syphilis in an HIV outpatient department worthwhile? Sex Transm Infect 2003; 79:257–259.
25. Klausner J. Frequency of syphilis testing in HIV-infected patients: More and more often. Sex Transm Dis 2009; 36:86–87.
26. Centers for Disease Control and Prevention. Incorporating HIV prevention into the medical care of persons living with HIV. Recommendations of the CDC, the Health Resources and Services Administration, the National Institutes of Health, and the HIV Medicine Association of the Infectious Diseases Society of America. MMWR 2003; 52 (RR-12):1–24.