Horberg, Michael A. MD, MAS*; Ranatunga, Dilrini K. BA*; Quesenberry, Charles P. PHD*; Klein, Daniel B. MD†; Silverberg, Michael J. PHD, MPH*
Syphilis, a sexually transmitted disease with systemic effects caused by the spirochete Treponema pallidum,1 remains an epidemic today in the United States and California.2,3 Incidence rates in managed care organizations, where a large part of healthcare is delivered, are unclear; the US Center for Disease Control and Prevention (CDC) reported 67% of new syphilis cases in 2003 were from private clinics and posited most could be associated with managed care.3,4 As HIV disease and syphilis are both sexually transmitted infections, coinfection is common.5,6 Syphilis is also felt to be at epidemic proportions in men having sex with men (MSM), even without HIV coinfection.6–9 The disease can have a wide variety of manifestations in humans, 10–13 and disease presentations may vary between HIV-infected and HIV-uninfected patients, including earlier presentation of neurosyphilis.14,15 As there is no immunity with prior infection with syphilis, reinfection is common in HIV-uninfected adults and some have described reinfection as even more common in HIV-infected patients.16 Treatment recommendations, though, have remained stable even with multiple exposures or unsuccessful initial therapy but specific treatment does not differ by HIV infection status.16,17
Complicating the care of syphilis in HIV-infected patients is the concern of successful response to therapy. Among HIV-infected individuals, early studies (pre-HAART) found that appropriate responses to treatment (typically felt to be indicated by at least a two-fold improvement in RPR dilution or preferably a return to no more than 1:2 dilution or nonreactivity) often took much longer (up to 2 years for HIV-infected patients with adequate therapy compared up to 12 months for HIV-uninfected patients).17 Only 3% to 7% of HIV-uninfected patients do not respond to aqueous penicillin,18–20 but this percent is felt to be much greater in HIV/syphilis coinfected patients17,21 by some researchers but not others.22,23
CDC and California state officials report increasing incidence of syphilis in recent years, with increased incidence among men having MSM and HIV-infected patients. Evidence suggests HIV-infected patients get diagnosed with more advanced stages of syphilis and respond less well to therapy than HIV-uninfected patients. However, these prior studies would benefit from exploration within an integrated, closed healthcare system (or managed care organization) like Kaiser Permanente. With our captured population and full clinical data where measured, we should be able to ascertain a better estimate of differences in syphilis incidence and response to therapy rates comparing HIV-infected to HIV-uninfected patients within a nonpublic payer patient population. Thus, we investigated the epidemiologic differences and the response to treatment among HIV-infected and HIV-uninfected patients in Kaiser Permanente Northern California (KPNC).
MATERIALS AND METHODS
We performed an observational cohort analysis of KPNC patients diagnosed with syphilis infection in the period 1995–2005. We compared incident rates of syphilis between HIV-infected and HIV-uninfected adults, and we explored any differences between these 2 groups with regard to reduction in RPR titers or serologic failure after syphilis treatment. We explored differences in stage of syphilis at time of diagnosis by HIV status. Among the HIV-infected patients, we studied differences in syphilis incidence, reduction in RPR after therapy, and serologic failure by use of antiretroviral therapy.
KPNC is an integrated health care system composed of 14 medical centers (inpatient and outpatient facilities) that currently provides complete care (pharmacy, medical, and hospital care) to 4,442,780 patients in this time period, including 9989 HIV-infected members. KPNC maintains electronic databases, which store data on patients' demographics, diagnoses, laboratory results, hospitalizations, office visits, in-house prescriptions, and other membership utilization information. The HIV registry periodically surveys the general KPNC electronic databases and organizes into one location HIV-infected data since 1988 for all known HIV cases in the KPNC membership, including diagnoses (inpatient and outpatient), office and hospital visits, pharmacy utilization (prescriptions dispensed and refills), laboratory values, and other pertinent data. Through on-going chart-review, additional information such as race/ethnicity, HIV risk behavior, and external (non-KPNC) data are added to the KPNC HIV registry. Like the general KPNC patient population, most HIV-infected patients receive their medications through the KPNC pharmacy system (over 96%), including medications provided through the state's AIDS Drug Assistance Program.
We queried the laboratory database from January 1, 1995 through December 31, 2005 to ascertain incident diagnoses of syphilis in KPNC. Patients had to be over 16 years old, and we defined incident diagnosis as there being no prior syphilis diagnosis nor prior positive RPR or fluorescent treponemal antibody (FTA) antibody test in the electronic medical record and by having a positive laboratory finding (both a positive RPR test [1:1 titer or higher] and FTA antibody confirmation at testing). We queried the KPNC HIV registry to identify the HIV-infected patients within the syphilis-infected population. We used pharmacy records to determine which of these HIV/syphilis coinfected patients were on antiretroviral therapy at time of syphilis diagnosis and treatment. Patients were excluded if there was evidence of prior syphilis diagnosis or no FTA syphilis antibody confirmation (157 patients).
We queried appropriate databases and registries to determine age, gender, HIV status, and sexual behavior (classified for this study as men having MSM or bisexual, heterosexual, or other) for our study patients. Using laboratory databases, we recorded all RPR titers from incident titer forward through 24 months postinitial positive titer. We recorded FTA antibody for confirmation of diagnosis; the first positive RPR titer was used as date of diagnosis. For the HIV-infected patients only, we recorded the CD4+ cell count (cells/μL) and HIV RNA level (log10/mL) closest to date of initial syphilis diagnosis. We ascertained the diagnosis of stage of syphilis disease at time of initial diagnosis (primary, secondary, tertiary, neurosyphilis), where available in the electronic data.
Pharmacy and other appropriate databases were queried to determine treatment for syphilis (as per CDC guidelines16) for all patients. Due to electronic database limitations, we could ascertain that syphilis therapy was administered, but the specific treatment for 85% of the patients was unclear. Among the patients with known therapy, 60% received penicillin, 10% ceftriaxone, 5% doxycycline, and 25% erythromycin derivative; these results were not statistically different by HIV status. Among the HIV-infected patients only, we recorded all pharmacy antiretroviral data to determine active combination antiretroviral therapy at time of syphilis diagnosis and treatment. Appropriate combination antiretroviral therapy was defined using Department of Health and Human Services guidelines.24
Our primary outcomes of interest were incidence rates of syphilis, reduction in RPR with syphilis therapy, and syphilis serologic failure among those treated for syphilis. For these analyses, our primary predictor variable of interest was presence of known HIV infection. We defined a reduction in RPR with syphilis therapy if there was evidence of at least 2 dilution improvement in RPR titer within 24 months of therapy (for example, RPR titer decreasing from 1:8 to 1:2 by 18 months after therapy). We defined syphilis serologic failure as a positive RPR titer of any dilution if the RPR titer had previously become negative or an increase in RPR titer of at least 2 dilutions after evidence of therapy (for example, from 1:1 to 1:4).
We report incidence rates of syphilis infection (comparing HIV-infected to HIV-uninfected patients) overall for entire time period, by calendar year, and by HIV treatment eras (1995–1996, 1997–2000, 2001–2005). We compared by HIV infection status the stage of syphilis at time of diagnosis and sexual behavior where known by χ2 test. We performed multivariable analysis for syphilis incidence using Poisson regression and for potential reduction in RPR and likelihood of serologic failure after syphilis therapy using Cox proportional hazards modeling, with adjustment for HIV infection, age, and gender. Among the patients with known stage of syphilis infection or known treatment from the electronic record, we performed Cox proportional hazard modeling of potential reduction in RPR, adjusting for HIV infection, age, and gender and by stage of syphilis infection or treatment respectively.
Among the HIV-infected patients, we analyzed overall incidence by HIV risk behavior (men having sex with men/bisexual, intravenous drug use, heterosexual), by AIDS criteria (1993 CDC criteria), age, and by gender. We analyzed outcome measures previously described using the multivariable techniques described among the HIV-infected patients only, comparing by the taking of antiretroviral therapy at time of syphilis diagnosis and treatment.
All analyses were performed using SAS (v.9; SAS Institute, Cary, NC) or Stata (v.10; Stata Corporation, College Station, TX). Approvals for this study were obtained from the KPNC Institutional Review Board with waiver of patient informed consent.
HIV-infected patients 622 (14.8%) and 3584 HIV-uninfected patients (85.2%) met inclusion criteria and were analyzed. The HIV-infected cohort had statistically significant (<0.001) greater percent male (97% vs. 50%) and whites (49% vs. 27%) but significantly fewer blacks (23% vs. 37%) and were significantly younger (median age 40 years vs. 53 years of age) compared with the HIV-uninfected cohort. Of the 4206 patients analyzed, 687 lost their KPNC membership (387 by death) within 2 years after their syphilis diagnosis; these results were not statistically different by HIV diagnosis. Among the HIV-infected patients only, 51% were on antiretroviral therapy at time of syphilis diagnosis and treatment, and less than 10% were known to be intravenous drug users. The mean CD4+ cell count among the HIV-infected patients was 360/μL (standard deviation–232/μL) and 69% had CD4+ cell count >200/μL.
The number of syphilis cases for both groups varied by calendar year. Overall, the incidence rate (Table 1) among HIV-infected patients was significantly higher than among the HIV-uninfected patients. This significantly higher rate among HIV-infected patients was consistent throughout the study period (Fig. 1). On a calendar basis, the rates among both groups declined from 1995 through 2000, and then rose sharply among the HIV-infected patients, but remained stable among the HIV-uninfected patients, leading to rate ratios greater than 100 after 2001. In adjusted analysis (Table 2), we saw a significantly higher incidence among HIV-infected than HIV-uninfected patients (adjusted rate ratio [RR] = 86.0, P <0.001).
Stage of syphilis disease at time of diagnosis varied by HIV-infection status, but was specified infrequently (31.7% of HIV-infected patients and only 9.2% of HIV-uninfected patients, P = 0.0001). For the data obtained, among the HIV-infected, 57.9% were recorded as primary, 32.4% as secondary, 9.1% as tertiary, and 0.5% as neurosyphilis, compared with 85.2% (P <0.0001), 3.3% (P <0.001), 9.4% (P = 0.92), and 2.1% (P = 0.14) respectively among the HIV-uninfected patients. We also found little documentation of sexual behavior among our HIV-uninfected patients (only 213 of the 3584). However, among the male patients that we did have such data, 81% of the HIV-infected were men having MSM only or bisexual, as opposed to only 36% of the HIV-uninfected (P <0.001). These few numbers precluded further analysis.
Thirty-two percent of the HIV-infected patients compared with only 11% of the HIV-uninfected patients had documented reduction of RPR with treatment (P <0.001, Table 2) and 30% compared with 11% respectively for syphilis serologic failure (P <0.001), but follow-up testing was greater among HIV-infected than HIV-uninfected patients (70.0% of HIV-infected and 44.8% of HIV-uninfected had at least one follow-up RPR titer after incident diagnosis). We found that HIV-infection, though, was associated with a greater likelihood of reduction in RPR posttherapy (hazards ratio [HR] = 2.47, P <0.001 with entire cohort); limiting our analysis to only those patients with follow-up testing, our results did not change significantly (HR = 2.09 [95% Confidence Interval: 1.70–2.57]; P <0.001). We found female gender and older age associated with a reduced likelihood of reduction in RPR titers. HIV infection also was associated with higher likelihood of serologic failure after syphilis therapy (HR = 2.63, P <0.001 and HR = 2.37 [1.92–2.93; P <0.001] if limited to only those patients with follow-up testing); however, older age had reduced association with serologic failure.
Among the patients where stage of syphilis diagnosis was documented in the electronic record, HIV status was still a significant predictor of reduction in RPR posttherapy (HR = 1.41 [1.02–1.94]; P = 0.04), while secondary syphilis had a greater odds of reduction in RPR posttherapy than primary syphilis (reference; HR = 2.12 [1.52–2.98]; P <0.001). Among the patients where the specific syphilis treatment was documented, HIV status still remained a significant predictor of lower RPR (HR = 2.72 [1.82–4.06]; P <0.001). With penicillin therapy as reference, only azithromycin therapy had a statistically significant difference in odds of lowering RPR posttherapy (HR = 0.51 [0.27–0.95]; P = 0.04). For both multivariable analyses, older age and female gender remained associated with a reduced likelihood of lower RPR postsyphilis therapy.
Examining the HIV-infected patients only (Table 3), there were no significant baseline differences between patients on antiretroviral therapy or not at time of syphilis diagnosis and treatment, including HIV risk behavior or gender. Those not on antiretroviral therapies had higher mean CD4+ cell count at time of syphilis diagnosis than those on antiretroviral therapy (374/μL [standard deviation–230/μL] vs. 347/μL [235/μL]; P = 0.04). In this population, 51% were on antiretroviral therapy at time of syphilis diagnosis. Overall, the patients on combination antiretroviral therapy had a lower adjusted incidence of syphilis compared with patients not on combination therapy (P = 0.46, P <0.001), and the rate ratios continued to decline with time. Reduction in RPR with syphilis therapy was significantly higher among patients on antiretroviral therapy (HR = 2.14, P <0.001), as was likelihood of serologic failure (HR = 2.27, P <0.001).
This study represents the first reporting of syphilis rates and outcomes in a managed care organization. We demonstrate an increasing rate of syphilis among HIV-infected patients, especially males, but without a similar rise among HIV-uninfected patients. This rate among HIV-infected patients has risen greatly since 2000, but is less among HIV-infected patients on combination antiretroviral therapy. We found a higher rate of men having MSM among our HIV-infected patients with syphilis compared with the HIV-uninfected. Compared with HIV-uninfected patients, HIV-infected patients with syphilis are associated with greater likelihood of follow-up testing and RPR reduction after treatment, but also greater likelihood of serologic failure, irrespective of syphilis treatment. Our study also is significant for its sample size and clinical care setting.
Similar to state and national data,25,26 syphilis is on the rise in our HIV-infected population, but not among our HIV-uninfected population. Our rate of syphilis cases among our HIV-infected population is similar to national data but less than that reported in New York City.4,15,25 Our rate among HIV-uninfected was higher than other reported rates both nationally and in California, but our rate does not represent our entire HIV-uninfected population, and thus some screening bias is likely.3,26 Prior studies could only surmise that syphilis cases reported from “private office” settings were from managed care organizations, like KPNC; we now show actual rates.4,27 The rising and persistent difference in syphilis infection rates among HIV-infected and HIV-uninfected patients is likely due to a few reasons. HIV-infected patients may feel that they are more likely to live longer lives now given improved antiretroviral therapy, and are engaging in continued greater sexual risk behavior leading to higher sexually transmitted infection incidence.28,29 There also is the possibility of serosorting among men having sex with men; if syphilis is more prevalent among HIV-infected men then that would lead to higher syphilis infection rates if these men are only having sex with other HIV-infected men.30,31 Further, sexually transmitted disease screening is more common among the HIV-infected patients32; although Metsch et al., found HIV-infected patients in care are less likely to engage in sexual risk behavior.33 Also, as more advanced stages of syphilis can present earlier in HIV-infected patients, there may be more clues to prompt testing in these patients. HIV risk behavior (commonly ascertained upon entry into KPNC HIV care system) identifies patients at risk earlier than patients in routine healthcare. Sexual health assessment should be included and measured in all patients' health assessments to facilitate appropriate screening.
We observed greater documented reduction in RPR with syphilis therapy among HIV-infected patients. This may not imply that HIV-infected patients are more responsive to syphilis therapy but only that follow-up testing is more common among HIV-infected patients, as we found. While natural variance in RPR titer could account for our results, our patients received their therapy before the subsequent RPR titer; given the robust association we found between HIV status and response to therapy, we believe our association to be valid. This greater reduction in RPR among the HIV-infected held for stage of syphilis infection and treatment type. Unlike our study, prior studies indicate a lower response rate to therapy among HIV-infected patients.17 CDC recommends follow-up testing for all cases of treated syphilis.16 While often no follow-up tests get ordered for HIV-uninfected patients, HIV-infected patients should have more frequent follow-up serologies to ensure adequate treatment and response.17 Our follow-up testing rates were similar to previous cohort studies, but could improve. 20,34–36 We are assessing the best means of ensuring follow-up testing.
A positive finding is the greater rate of reduction in RPR with therapy among the HIV-infected patients on antiretroviral therapy compared to the HIV-infected patients not on such therapy. This is reassuring, given earlier studies that found HIV-infected patients had lesser likelihood of reducing RPR titers with syphilis therapy.17,21 Patients on antiretroviral therapy should have a more repaired immune system, permitting better responses to therapies for other concurrent infections.37
Most types of syphilis treatment did not effect our results. While penicillin is still the preferred treatment for syphilis,16 the accepted alternatives for syphilis therapy16 were as likely to be associated with reduced RPR except azithromycin. Two previous studies and a meta-analysis found azithromycin to be equally effective as penicillin;21,38,39 however, resistance to azithromycin among syphilis cases has been described in Northern California.40 These results deserve further exploration.
Serologic failure rates were higher among the HIV-infected patients. This could be due to different testing rates between the 2 populations, or due to continued greater risk behavior among the HIV-infected population. Other sexually transmitted diseases have persistent higher rates among HIV-infected patients and this phenomenon should apply to syphilis.41 As disturbing is the higher serologic failure rate in recent years among HIV-infected patients on antiretroviral therapy compared with those not on antiretroviral medications. Other studies have seen a persistence of unsafe sexual behavior among HIV-infected patients on antiretroviral therapy.42 If this implies that HIV-infected patients on antiretroviral therapy believe they are immune to syphilis once treated and, thus, are engaging in more unsafe sexual behavior, this could have significant public health considerations. Again, serosorting among HIV-infected men could be another explanation.30,31
We found that females had lesser likelihood of reduction of RPR than males. This lower response rate has not been seen in other studies and deserves further study.8,43 Increased age also had a negative effect on reduction in RPR. This reduced response to syphilis therapy with advancing age is potentially due to increasing senescence of the immune system.44 We had hoped to see a lower infection rate with advanced age but did not; more targeted prevention messages for older populations may be needed.
We acknowledge limitations in our study. We utilized only the electronic medical record to determine no prior history of syphilis and stage of disease at time of diagnosis. Although we cannot completely guarantee that this precluded all prior cases of syphilis, KPNC providers are urged to code prior history of syphilis in the electronic medical record. It is possible that distant history of syphilis diagnoses were not recorded in the medical record if these diagnoses were made outside of KPNC. We do think this is unlikely, though, as prior syphilis is asked during initial exams. Evidence of staging of syphilis infection at time of diagnosis and type of syphilis treatment was sparse and limited our analyses. We could not ascertain sexual risk among many of our HIV-uninfected patients. Further, there may be some ascertainment bias as HIV-infected patients are recommended for syphilis testing with initial diagnosis.
In conclusion, syphilis rate among HIV-infected patients are significantly greater than among the general population in our integrated healthcare organization. We also found a higher serologic failure association among the HIV-infected patients, and especially those on antiretroviral therapy. Although we found greater rates of reduction in RPR with syphilis therapy among our HIV-infected patients, this may be a function of follow-up testing probability and renewed efforts to ensure close follow-up of new syphilis cases is needed. Continued vigilance, screening, and risk reduction recommendations for syphilis, especially among HIV-infected patients, is needed. This epidemic is controllable, if patients hear what clinicians and public health officials are saying.
1. Antal GM, Lukehart SA, Meheus AZ. The endemic treponematoses. Microbes Infect 2002; 4:83–94.
2. US Department of Health and Human Services, CDC. Syphilis Rates by US States 2006. In: US Department of Health and Human Services Center for Disease Control and Prevention, ed. 2006.
3. US Department of Health and Human Services, CDC. Sexually Transmitted Disease Surveillance 2006 Supplement, Syphilis Surveillance Report. In: US Department of Health and Human Services Center for Disease Control and Prevention, ed. 2007;i-23.
4. Blocker ME, Levine WC, St Louis ME. HIV prevalence in patients with syphilis, United States. Sex Transm Dis 2000; 27:53–59.
5. Theus SA, Harrich DA, Gaynor R, et al. Treponema pallidum, lipoproteins, and synthetic lipoprotein analogues induce human immunodeficiency virus type 1 gene expression in monocytes via NF-kappaB activation. J Infect Dis 1998; 177:941–950.
6. Goh BT. Syphilis in adults. Sex Transm Infect 2005; 81:448–452.
7. Hourihan M, Wheeler H, Houghton R, et al. Lessons from the syphilis outbreak in homosexual men in east London. Sex Transm Infect 2004; 80:509–511.
8. Pao D, Goh BT, Bingham JS. Management issues in syphilis. Drugs 2002; 62:1447–1461.
9. Buchacz KA, Siller JE, Bandy DW, et al. HIV and syphilis testing among men who have sex with men attending sex clubs and adult bookstores–San Francisco, 2003. J Acquir Immun Defic Syndr 2004; 37:1324–1326.
10. Chapel TA. The variability of syphilitic chancres. Sex Transm Dis 1978; 5:68–70.
11. Chapel TA. The signs and symptoms of secondary syphilis. Sex Transm Dis 1980; 7:161–164.
12. Lukehart SA. Future application of research tests to the diagnosis of yaws and other endemic treponematoses. Southeast Asian J Trop Med Public Health 1986; 17(suppl 4):78–83.
13. Lukehart SA, Hook EW III, Baker-Zander SA, et al. Invasion of the central nervous system by Treponema pallidum: Implications for diagnosis and treatment. Ann Intern Med 1988; 109:855–862.
14. Flood JM, Weinstock HS, Guroy ME, et al. Neurosyphilis during the AIDS epidemic, San Francisco, 1985–1992. J Infect Dis 1998; 177:931–940.
15. Rompalo AM, Joesoef MR, O'Donnell JA, et al. Clinical manifestations of early syphilis by HIV status and gender: Results of the syphilis and HIV study. Sex Transm Dis Mar 2001; 28:158–165.
16. Sex Transm Dis Treatment Guidelines, 2006. In: US Department of Health and Human Services Center for Disease Control and Prevention, ed. Vol 55: MMWR Morb Mortal Wkly Rep; 2006:22–29.
17. Lynn WA, Lightman S. Syphilis and HIV: A dangerous combination. Lancet Infect Dis 2004; 4:456–466.
18. Schroeter AL, Lucas JB, Price EV, et al. Treatment for early syphilis and reactivity of serologic tests. JAMA 1972; 221:471–476.
19. Schroeter AL, Turner RH, Lucas JB, et al. Therapy for incubating syphilis. Effectiveness of gonorrhea treatment. JAMA 1971; 218:711–713.
20. Long CM, Klausner JD, Leon S, et al. Syphilis treatment and HIV infection in a population-based study of persons at high risk for sexually transmitted disease/HIV infection in Lima, Peru. Sex Transm Dis 2006; 33:151–155.
21. Riedner G, Rusizoka M, Todd J, et al. Single-dose azithromycin versus penicillin G benzathine for the treatment of early syphilis. N Engl J Med 2005; 353:1236–1244.
22. Kofoed K, Gerstoft J, Mathiesen LR, et al. Syphilis and human immunodeficiency virus (HIV)-1 coinfection: Influence on CD4 T-cell count, HIV-1 viral load, and treatment response. Sex Transm Dis 2006; 33:143–148.
23. Rolfs RT, Joesoef MR, Hendershot EF, et al. A randomized trial of enhanced therapy for early syphilis in patients with and without human immunodeficiency virus infection. The Syphilis and HIV Study Group. N Engl J Med 1997; 337:307–314.
24. Panel on Antiretroviral Guidelines for Adult and Adolescents. Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. In: Services Department of Health and Human Services, ed, 2008:1–128.
25. Primary and secondary syphilis among men who have sex with men–New York City, 2001. MMWR Morb Mortal Wkly Rep 2002; 51:853–856.
26. Primary and secondary syphilis–United States, 2003–2004. MMWR Morb Mortal Wkly Rep 2006; 55:269–273.
27. Sexually Transmitted Disease Surveillance 2003. In: Prevention CfDCa, ed, 2005.
28. Stall RD, Hays RB, Waldo CR, et al. The Gay '90s: A review of research in the 1990s on sexual behavior and HIV risk among men who have sex with men. AIDS 2000; 14(suppl 3):S101–S114.
29. Stolte IG, Dukers NH, de Wit JB, et al. Increase in sexually transmitted infections among homosexual men in Amsterdam in relation to HAART. Sex Transm Infect 2001; 77:184–186.
30. Truong HM, Kellogg T, Klausner JD, et al. Increases in sexually transmitted infections and sexual risk behavior without a concurrent increase in HIV incidence among men who have sex with men in San Francisco: A suggestion of HIV serosorting? Sex Transm Infect 2006; 82:461–466.
31. Van der Bij AK, Kolader ME, de Vries HJ, et al. Condom use rather than serosorting explains differences in HIV incidence among men who have sex with men. J Acquir Immun Defic Syndr 2007; 45:574–580.
32. Ahrens K, Kent CK, Montoya JA, et al. Healthy Penis: San Francisco's social marketing campaign to increase syphilis testing among gay and bisexual men. PLoS Med 2006; 3:e474.
33. Metsch LR, Pereyra M, Messinger S, et al. HIV transmission risk behaviors among HIV-infected persons who are successfully linked to care. Clin Infect Dis 2008; 47:577–584.
34. Ghanem KG, Erbelding EJ, Wiener ZS, et al. Serological response to syphilis treatment in HIV-positive and HIV-negative patients attending sexually transmitted diseases clinics. Sex Transm Infect 2007; 83:97–101.
35. McClean H, Daniels D, Carne C, et al. UK National Audit of Early Syphilis Management. Case notes audit: Diagnosis and treatment. Int J STD AIDS 2006; 17:334–339.
36. McClean H, Daniels D, Carne C, et al. UK National Audit of Early Syphilis Management. Clinics audit: Screening for and management of early syphilis. Int J STD AIDS 2006; 17:344–348.
37. Nagappan V, Kazanjian P. Bacterial infections in adult HIV-infected patients. HIV Clin Trials 2005; 6:213–228.
38. Bai ZG, Yang KH, Liu YL, et al. Azithromycin vs. benzathine penicillin G for early syphilis: a meta-analysis of randomized clinical trials. Int J STD AIDS 2008; 19:217–221.
39. Hook EW III, Martin DH, Stephens J, et al. A randomized, comparative pilot study of azithromycin versus benzathine penicillin G for treatment of early syphilis. Sex Transm Dis 2002; 29:486–490.
40. Katz KA, Klausner JD. Azithromycin resistance in Treponema pallidum. Curr Opin Infect Dis 2008; 21:83–91.
41. Wasserheit JN. Epidemiological synergy. Interrelationships between human immunodeficiency virus infection and other sexually transmitted diseases. Sex Transm Dis 1992; 19:61–77.
42. van Kesteren NM, Hospers HJ, Kok G. Sexual risk behavior among HIV-positive men who have sex with men: A literature review. Patient Educ Couns 2007; 65:5–20.
43. Manavi K, McMillan A. The outcome of treatment of early latent syphilis and syphilis with undetermined duration in HIV-infected and HIV-uninfected patients. Int J STD AIDS 2007; 18:814–818.
44. Kovaiou RD, Grubeck-Loebenstein B. Age-associated changes within CD4+ T cells. Immunol Lett 2006; 107:8–14.