Secondary Logo

Journal Logo

Original Studies

Ocular Syphilis and Human Immunodeficiency Virus Coinfection Among Syphilis Patients in North Carolina, 2014–2016

Cope, Anna B. PhD*†; Mobley, Victoria L. MD; Oliver, Sara E. MD*; Larson, Mara MPH; Dzialowy, Nicole MSc; Maxwell, Jason BS; Rinsky, Jessica L. PhD*†; Peterman, Thomas A. MD*; Fleischauer, Aaron PhD*†; Samoff, Erika PhD

Author Information
Sexually Transmitted Diseases: February 2019 - Volume 46 - Issue 2 - p 80-85
doi: 10.1097/OLQ.0000000000000910

Untreated syphilis can result in complications involving the eye (eg, uveitis, retinitis, optic neuritis, retinal detachment) collectively referred to as ocular syphilis (OS). In some patients, OS leads to vision loss.1 Although the proportion of syphilis cases with ocular manifestations remains small, reports of OS have increased with rising syphilis rates in the United States since 2014.2–5

A few small studies suggest human immunodeficiency virus (HIV) contributes to the ocular manifestations of syphilis.1,3,6–16 Estimates of HIV coinfection among OS patients range from approximately 30% internationally1,6 to 50% in the United States.3 Patients living with HIV have been diagnosed with OS in both early and late stages of syphilis.1,7–9 The median rapid plasma reagin (RPR) and Venereal Disease Research Laboratory (VDRL) titers are often higher in OS patients living with HIV.1 Nearly half of patients concurrently diagnosed with OS and HIV were tested for HIV because they presented to a provider with OS symptoms.10 Although most reported OS patients living with HIV are diagnosed with posterior uveitis and panuvietis,8,11–16 there is some variation in OS-defining diagnosis among OS patients living with HIV7,10 that may be related to CD4 counts.10 Generally, patients living with HIV experience substantial resolution of OS symptoms when treated appropriately as per CDC guidelines.1,7,10,11

Among syphilis patients reported to the North Carolina (NC) Division of Public Health between 2014 and 2015, OS prevalence was higher in patients living with HIV compared to patients without documented HIV.4 It is unclear why some patients living with HIV develop OS while others do not. Improved reporting and documentation of OS coupled with data that can be easily linked between HIV and STD surveillance systems make NC uniquely well-suited to address this question. Using population-based surveillance information, we compared demographic and clinical markers of disease to identify factors associated with the development of OS in syphilis patients by HIV diagnosis status.

METHODS

We conducted a secondary analysis of surveillance data previously used to describe the epidemiology of OS in NC (with one additional year of follow-up).4 Briefly, we reviewed all newly reported syphilis cases (early and late) entered in the NC Electronic Disease Surveillance System (NC EDSS) during 2014–2016. We defined OS as a patient diagnosed with syphilis and having clinical symptoms consistent with ocular disease that could not be explained by other causes.17 For all syphilis cases, we abstracted demographic characteristics, syphilis stage, RPR titer (a serologic screening test for syphilis antibodies),18 cerebrospinal fluid (CSF) VDRL test results, symptoms, and risk factors from NC EDSS. For OS cases diagnosed during 2014 to 2015, we reviewed medical records to ascertain the OS-defining diagnosis, OS symptoms, and treatment, and attempted to interview these OS patients to assess OS symptom resolution.4

Syphilis patients were matched to NC's enhanced HIV/AIDS Reporting System (eHARS) to identify patients living with HIV (no additional HIV infections were identified among OS patients from the medical record review). HIV-negative status was determined by a documented negative HIV test at the time of the syphilis diagnosis in NC EDSS or as ascertained from the medical record review. We classified all other patients as HIV-unknown status. For patients living with HIV, we abstracted HIV diagnosis date from eHARS and CD4 counts and viral loads (VL) reported closest to and within 6 months before the syphilis diagnosis date from the NC Engagement in Care and HIV Outreach database. In sensitivity analyses, we varied the window period used to capture the most proximate VL and CD4 count before the syphilis diagnosis date to 12 months. Next, we repeated these analyses with the closest VL or CD4 in time to the syphilis diagnosis date (whether before or after) and within the 6- or 12-month window period. We defined viral suppression as a VL of 200 copies/mL or less. Syphilis cases with HIV co-infection were classified as previously diagnosed (HIV diagnosis date longer than 30 days before syphilis diagnosis date) or concurrently diagnosed (HIV diagnosis date ≤30 days from syphilis diagnosis date).

We estimated prevalence ratios (PR) and 95% confidence intervals (CI) to compare OS prevalence (1) by HIV status within demographic and clinical subgroups of syphilis patients and (2) by the timing of HIV diagnosis relative to the syphilis diagnosis date, viral suppression, and CD4 count of 200 cells/mL or less among syphilis patients living with HIV using complete case analysis. In sensitivity analyses, we assessed the impact of different classifications of missing HIV laboratory values on the association between OS and both viral suppression and CD4 count of 200 cells/mL or less. The proportion of OS patients who sought care from an ophthalmologist was compared by HIV status (HIV-positive versus HIV-negative/unknown status) using Fisher's exact test. We compared syphilis stage, RPR titer, VL, and CD4 count by OS status in syphilis patients living with HIV using the Wilcoxon rank sum test (continuous variables) and chi-square test (categorical variables). All analyses were conducted in SAS version 9.4 (Cary, NC). This analysis was part of a public health investigation and exempted from institutional review board review.

RESULTS

All Syphilis Patients

During 2014 to 2016, there were 7123 confirmed syphilis cases reported in NC; 109 (1.5%) were classified as having OS (1 verified by darkfield microscopy to have Treponema pallidum in vitreous fluid; 78 had an ophthalmologic examination with findings consistent with OS; 30 only reported symptoms consistent with OS that had no other known cause). In total, 2846 (40.0%) syphilis patients were living with HIV, of whom 59 (2.1%) had OS. Of the remaining 4277 syphilis patients, 3197 (74.7%) tested HIV-negative of whom 38 (1.2%) had OS, and 1080 (25.3%) had an unknown HIV status of whom 12 (1.1%) had OS. Ocular syphilis was more prevalent in syphilis patients living with HIV compared to those who were HIV-negative/unknown status (PR, 1.8; 95% CI, 1.2–2.6). Most OS patients were men (91.7%) and all OS patients living with HIV were men. White, non-Hispanic syphilis patients were 2.6 times more likely to have OS than black, non-Hispanic patients (95% CI, 1.7–3.8). Nonetheless, OS was more prevalent in patients living with HIV across all syphilis stages, ages, races, and risks. The median age and RPR titer were higher in OS patients compared to non-OS patients, regardless of the patient's HIV status (Table 1).

TABLE 1
TABLE 1:
OS Prevalence by HIV Status Within Demographic and Clinical Characteristic Groups Among Syphilis Patients Reported in North Carolina, 2014–2016

OS Patients

Among OS patients, those living with HIV were younger and had a higher median RPR than OS patients who were HIV-negative or unknown-status (Table 1). Overall, 47 (43%) of 109 OS patients had a CSF VDRL test performed. Most of the 32 patients with a reactive CSF VDRL test result were living with HIV (n = 21; 65.6%). We reviewed the medical records for the 63 OS patients diagnosed between 2014 and 2015. In total, 35 (55.5%) patients were HIV-positive; both of the 2 OS patients reporting complete vision loss were living with HIV. Among these 63 OS patients, we did not observe a difference in the proportion of OS patients examined by an ophthalmologist by HIV status (74.3% (26/35) HIV-positive versus 82.1% (23/28) HIV-negative/unknown-status; P = 0.4). Per the medical record review, uveitis was the most common OS diagnosis in patients living with HIV (n = 11 panuveitis; n = 4 posterior uveitis; n = 5 anterior uveitis). Furthermore, 11 patients had a severe OS diagnosis (chorioretinitis, optic neuritis, or retinal detachment); 9 (81.8%) were living with HIV. We interviewed 39 OS patients to assess OS symptom resolution. Overall, 14 (70.0%) of 20 patients living with HIV and 13 of 19 (68.4%) HIV-negative/unknown-status patients reported that once treated for OS, visual functioning returned to levels experienced before OS symptoms manifested. The remaining 12 patients had residual deficits.

Syphilis Patients Living With HIV

Among 2846 patients living with HIV, 2504 (88.0%) had a CD4 count (57 with OS) and 2554 (89.7%) had a VL within 6 months before their syphilis diagnosis. Compared to non-OS patients, slightly higher proportions of OS patients living with HIV had an available CD4 count (n = 55; 93.2%) or VL (n = 57; 96.6%) during this timeframe. The median VL among all syphilis patients living with HIV was 40 copies/mL (interquartile range [IQR], 20–19,243); 1016 (39.8%) patients were virally unsuppressed. Virally unsuppressed patients were more likely to be diagnosed with OS than suppressed patients (PR, 1.7; 95% CI, 1.0–2.8). When patients living with HIV who did not have a reported VL were considered unsuppressed, results were similar, although mitigated (PR, 1.5; 95% CI, 0.9–2.5).

The median CD4 count among syphilis patients living with HIV was 505 cells/mL (IQR, 324–712); and 310 (12.4%) patients had a CD4 count of 200 cells/mL or less. Median CD4 counts were lower in OS patients than non-OS patients at every syphilis stage (Table 2). The prevalence of OS was higher in patients with a CD4 count of 200 cells/mL or less than in patients with higher CD4 counts (PR, 2.3; 95% CI, 1.3–4.2). This association persisted after controlling for patient age (PR, 2.2; 95% CI, 1.2–4.0).

TABLE 2
TABLE 2:
Median CD4 Cell Count, VL, and RPR* Titer at the Time of Syphilis Diagnosis by Syphilis Stage and OS Status in Syphilis Patients With a Previous‡ and Concurrent§ HIV Diagnosis Reported in North Carolina, 2014–2016

The prevalence of OS was higher among syphilis patients diagnosed with HIV and syphilis concurrently compared to syphilis patients previously diagnosed with HIV infection (PR, 1.5; 95% CI, 0.8–2.8). CD4 counts were lowest and VLs were highest in concurrently diagnosed OS patients. More OS patients had late latent or unknown duration syphilis compared to non-OS patients among concurrently diagnosed (69.2% vs 33.1% P = 0.01) and previously diagnosed patients (41.3% versus 22.0%, P = 0.002). OS patients were more likely than non-OS patients to have a CD4 count of 200 cells/mL or less in previously diagnosed patients (25.0% vs 10.4%; P = 0.005), but not among concurrently diagnosed patients (23.1% vs 22.2%; P = 0.9).

Results from sensitivity analyses that varied the date range to capture HIV laboratory values and included missing data in VL and CD4 categories were similar to reported findings (Supplemental tables 1A and 1B, https://links.lww.com/OLQ/A305).

DISCUSSION

Ocular syphilis was identified in a small proportion of syphilis patients, regardless of HIV status. However, OS was more prevalent in patients living with HIV. Most patients with severe OS and both patients who reported complete vision loss were living with HIV. OS patients living with HIV were younger than HIV-negative and unknown-status OS patients, but still older than non-OS patients living with HIV. OS was diagnosed more frequently in patients living with HIV who had late or unknown duration syphilis, an unsuppressed VL, or a CD4 count of 200 cells/mL or less. Many of these patients were diagnosed with HIV and syphilis concurrently, suggesting this population was not receiving routine screening for syphilis or HIV at earlier stages of infection. Concerns about ocular symptoms may have prompted these patients to seek healthcare. With treatment, OS symptoms in patients living with HIV appeared to resolve at similar rates as HIV-negative or unknown-status patients.

The combination of older age and low CD4 counts in OS patients living with HIV suggest declining immune status may contribute to ocular manifestations of syphilis. Syphilis can result in a temporary drop in the CD4 count in patients coinfected with HIV that typically resolves after treatment.19–23 Similar to previous studies,10,11 we observed OS patients to have lower CD4 counts than non-OS patients, regardless of when the patient was diagnosed with HIV relative to their syphilis diagnosis or the patient's age. Ocular syphilis was also more prevalent in patients with an unsuppressed VL, providing support for previous findings that inflammation associated with uncontrolled HIV replication could be an important prognostic factor for OS.24 To date, there is no evidence that the recent nationwide increase in reported OS infection was due to a specific neurotropic strain of T. pallidum.25 Although our study was not designed to do so, continued characterization of OS strains by HIV status may elicit an understanding of who is most at risk of developing ocular manifestations of syphilis.

Ocular syphilis may be associated with delayed syphilis diagnosis, particularly in people who also have undiagnosed, and thus uncontrolled, HIV. Almost 70% of OS patients with concurrently diagnosed HIV had late or unknown duration syphilis. These patients had the highest median VL and lowest median CD4 count of all syphilis patients living with HIV. However, since the median RPR was high (1:128) in this group, syphilis stage may have been misclassified and some of these patients actually may have had recently acquired syphilis. Regardless, OS screening as part of routine examination should be considered for any person, whether HIV positive or not, presenting to healthcare with ocular symptoms in areas with high syphilis rates.

People aware of their HIV infection may be receiving regular health care to control their disease, oftentimes from providers familiar with manifestations of syphilis. Therefore, these patients may be screened for OS more often due to higher clinical suspicion by their HIV care provider. Consequently, elevated rates of OS in patients living with HIV could be due to detection bias. Ocular syphilis prevalence was also much lower in women than in men; no women with OS was living with HIV. It is unclear whether this gender difference in OS prevalence is due to the relatively small number of women diagnosed with both syphilis and HIV, a failure on the part of providers to ask about ocular symptoms, or an unidentified biological factor. However, historical evidence suggests women have fewer complications related to untreated syphilis than men.26 Future analyses with larger OS patient populations may have increased power to detect differences in OS recognition by various demographic and clinical characteristics, including HIV status, and thereby quantify potential bias.

Our assessment of both OS and HIV status was incomplete. An OS classification required a patient to be diagnosed with syphilis, report OS symptoms to a provider, and have these symptoms documented in NC EDSS. In all but one case, we were unable to verify the OS diagnosis in a biological specimen. Although there is potential for misclassification in the remaining cases, no other cause for the ocular symptoms were reported and it is reasonable to assume they have OS per CDC case definition.17 At the time of their syphilis diagnosis, HIV status was not documented in surveillance data for 15% of patients. These patients were either not tested at the time of their syphilis diagnosis, or if tested, did not have their test results automatically captured in NC EDSS. VL and CD4 measurements near the time of the syphilis diagnosis were also missing for approximately 10% of known syphilis patients living with HIV. Because we observed no differences in the pattern of missing VL and CD4 data by OS or any other measured covariate, we assumed potential bias due to missing laboratory data to be negligible.

In this population-based assessment of syphilis patients, OS was rare, regardless of HIV status. We identified a higher prevalence of OS in patients living with HIV, particularly those who were unaware of their HIV infection. Complaints of ocular symptoms may help detect previously undiagnosed HIV, emphasizing the importance of screening for HIV and syphilis in tandem. Elevated rates of OS in older patients and those with low CD4 counts suggest immunodeficiency may play a role in the association between OS manifestations and HIV. Other related information not captured by surveillance could provide additional insight about factors that increase the risk of OS in syphilis patients living with HIV.

REFERENCES

1. Mathew RG, Goh BT, Westcott MC. British ocular syphilis study (BOSS): 2-year national surveillance study of intraocular inflammation secondary to ocular syphilis. Invest Ophthalmol Vis Sci 2014; 55:5394–5400.
2. Woolston S, Cohen SE, Fanfair RN, et al. A cluster of ocular syphilis cases—Seattle, Washington, and San Francisco, California, 2014-2015. MMWR Morb Mortal Wkly Rep 2015; 64:1150–1151.
3. Oliver SE, Aubin M, Atwell L, et al. Ocular syphilis—eight jurisdictions, United States, 2014-2015. MMWR Morb Mortal Wkly Rep 2016; 65:1185–1188.
4. Oliver SE, Cope AB, Rinsky JL, et al. Increases in ocular syphilis-North Carolina, 2014-2015. Clin Infect Dis 2017; 65:1676–1682.
5. Centers for Disease Control and Prevention. Sexually transmitted disease surveillance 2016. Atlanta, GA: US Department of Health and Human Services, 2017.
6. Bollemeijer JG, Wieringa WG, Missotten TO, et al. Clinical manifestations and outcome of syphilitic uveitis. Invest Ophthalmol Vis Sci 2016; 57:404–411.
7. Li JZ, Tucker JD, Lobo AM, et al. Ocular syphilis among HIV-infected individuals. Clin Infect Dis 2010; 51:468–471.
8. McLeish WM, Pulido JS, Holland S, et al. The ocular manifestations of syphilis in the human immunodeficiency virus type 1-infected host. Ophthalmology 1990; 97:196–203.
9. Parc CE, Chahed S, Patel SV, et al. Manifestations and treatment of ocular syphilis during an epidemic in France. Sex Transm Dis 2007; 34:553–556.
10. Tucker JD, Li JZ, Robbins GK, et al. Ocular syphilis among HIV-infected patients: A systematic analysis of the literature. Sex Transm Infect 2011; 87:4–8.
11. Lee SY, Cheng V, Rodger D, et al. Clinical and laboratory characteristics of ocular syphilis: A new face in the era of HIV co-infection. J Ophthalmic Inflamm Infect 2015; 5:56.
12. Shalaby IA, Dunn JP, Semba RD, et al. Syphilitic uveitis in human immunodeficiency virus-infected patients. Arch Ophthalmol 1997; 115:469–473.
13. Browning DJ. Posterior segment manifestations of active ocular syphilis, their response to a neurosyphilis regimen of penicillin therapy, and the influence of human immunodeficiency virus status on response. Ophthalmology 2000; 107:2015–2023.
14. Kuo IC, Kapusta MA, Rao NA. Vitritis as the primary manifestation of ocular syphilis in patients with HIV infection. Am J Ophthalmol 1998; 125:306–311.
15. Villanueva AV, Sahouri MJ, Ormerod LD, et al. Posterior uveitis in patients with positive serology for syphilis. Clin Infect Dis 2000; 30:479–485.
16. Balba GP, Kumar PN, James AN, et al. Ocular syphilis in HIV-positive patients receiving highly active antiretroviral therapy. Am J Med, 2006 119:448 e21–448 e25.
17. Centers for Disease Control and Prevention. STD surveillance case definitions. 2018; Available at: https://wwwn.cdc.gov/nndss/conditions/syphilis/case-definition/2018/. Accessed August 16, 2018.
18. Holmes King K, et al. Sexually transmitted diseases. 4th ed. 2007, New York: The McGraw-Hill Companies.
19. Zetola NM, Klausner JD. Syphilis and HIV infection: An update. Clin Infect Dis 2007; 44:1222–1228.
20. Sadiq ST, McSorley J, Copas AJ, et al. The effects of early syphilis on CD4 counts and HIV-1 RNA viral loads in blood and semen. Sex Transm Infect 2005; 81:380–385.
21. 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.
22. Dyer JR, Eron JJ, Hoffman IF, et al. Association of CD4 cell depletion and elevated blood and seminal plasma human immunodeficiency virus type 1 (HIV-1) RNA concentrations with genital ulcer disease in HIV-1-infected men in Malawi. J Infect Dis 1998; 177:224–227.
23. Buchacz K, Patel P, Taylor M, et al. Syphilis increases HIV viral load and decreases CD4 cell counts in HIV-infected patients with new syphilis infections. AIDS 2004; 18:2075–2079.
24. Biotti D, Bidot S, Mahy S, et al. Ocular syphilis and HIV infection. Sex Transm Dis 2010; 37:41–43.
25. Oliver S, Sahi SK, Tantalo LC, et al. Molecular typing of Treponema pallidum in ocular syphilis. Sex Transm Dis 2016; 43:524–527.
26. Gjestland T. The Oslo study of untreated syphilis; an epidemiologic investigation of the natural course of the syphilitic infection based upon a re-study of the Boeck-Bruusgaard material. Acta Derm Venereol Suppl (Stockh) 1955; 35(Suppl 34):3–368; Annex I-LVI.

Supplemental Digital Content

Copyright © 2018 American Sexually Transmitted Diseases Association All rights reserved.