Epidemiology

 Syphilis is caused by the spirochete Treponema pallidum subspecies pallidum. This obligate human pathogen is primarily transmitted via sexual contact between mucosal surfaces. In addition, cases of mother-to-child transmission of syphilis are being reported in the current epidemic (86).

In Australia, syphilis disproportionately affects indigenous Australians and MSM populations. Among HIV-positive MSM attending sexual health clinics, the incidence of infectious syphilis approached 10% per year in 2015 and is currently over two-fold higher than the incidence among HIV-negative MSM (3). Thus, current guidelines recommend testing for syphilis every three months among MSM with HIV infection (87).

Epidemiological evidence supports the fact that incident syphilis infection increases sexual acquisition of HIV, via breaches in mucosa and/or skin during early syphilis infection (88) and the associated impact of local recruitment of CD4+ immune cells to the area (89). This has been demonstrated in several studies, including the recent iPrEx HIV pre-exposure prophylaxis study, where incident syphilis infection was a risk factor for HIV acquisition (90). As previously noted, in the setting of HIV viral suppression, concurrent STIs do not appear to affect risk of HIV acquisition via condomless anal sex between serodiscordant gay male couples (6, 7).

Syphilis may also have a negative impact on chronic HIV infection. Early retrospective studies found that incident syphilis adversely affected both the immunological and virological status of HIV-infected people, leading to a reduction in CD4+ T cell counts and a rise in plasma HIV viral load (91-93). However, some retrospective reports found no association between incident syphilis and changes in HIV viral load (94, 95) and a prospective study found that although incident syphilis infection was associated with a change in CD4+ T cell count, this was promptly reversed upon treatment of syphilis. In addition, the transient drop in CD4+ T cell count had no impact on disease trajectory, progression to AIDS or mortality (96).

Clinical presentation

 In general, the clinical manifestations of syphilis are similar in HIV-positive and -negative individuals (4). However, important differences exist, as outlined below.

Infectious syphilis

In the US Syphilis and HIV cohort study (97), HIV-infected individuals presenting with primary syphilis were more likely to have larger and more numerous chancres. In addition, secondary syphilis among HIV-infected individuals was more likely to present with concurrent chancres indicating an overlap between the primary and secondary stages. “Malignant” secondary syphilis, also known as “ulceronodular” syphilis, is a rare manifestation thought to be more common in HIV-infected individuals due to impaired cell-mediated immunity, which is characterised by disseminated papulopustular ulcers with a crusty base (98).

Neurosyphilis

Syphilis of the central nervous system (CNS) can occur at any stage of disease with HIV infection being consistently demonstrated as a risk factor. Neurosyphilis is estimated to affect around 1-2% of syphilis/HIV co-infected individuals (99). Most commonly, neurosyphilis is asymptomatic, defined by evidence of syphilis in the cerebrospinal fluid (CSF) without accompanying symptoms or clinical signs.

Symptomatic neurosyphilis is categorised below (100):

• Early neurosyphilis: Approximately 10% of neurosyphilis in the early stage (months to 5 years) presents with meningovascular disease characterised by encephalitis and cerebrovascular injury. This may result in numerous clinical presentations including cerebral infarction, aseptic meningoencephalitis, cranial nerve palsy, gumma formation and myelitis.
• Late neurosyphilis: This usually manifest 10-30 years after initial infection and includes general paresis and tabes dorsalis, which are included in the case definition of tertiary syphilis. These result from direct parenchymal invasion of treponemes into the brain and spinal cord.
• Ocular & otologic syphilis: visual and hearing impairment from syphilis infection may occur as part of syphilitic meningovascular invasion.

Of note, the initiation of cART appears to be protective against the development of neurosyphilis (101). Furthermore, there is evidence that cART increases the likelihood of successful treatment of neurosyphilis and reduces risk of treatment failure (102, 103).

Diagnosis of syphilis in HIV-infected individuals

 The standard algorithm for the serologic diagnosis and staging of syphilis infection using specific trepnomenal (e.g. enzyme immunoassay) and non-specific/non-treponemal (e.g. rapid plasma reagin) assays is appropriate for syphilis diagnosis irrespective of HIV status (4). The diagnosis of neurosyphilis should be made only after considering the clinical history, examination, serology and CSF findings. No single feature is sufficiently accurate to diagnose or exclude a neurosyphilis diagnosis. However, when the index of suspicion is high, examination of the CSF via lumbar puncture (LP) remains the most useful diagnostic tool.

Although there is no international consensus on the precise timing of LP in the investigation of suspected neurosyphilis, the British Association for Sexual Health and HIV (BASHH) and The Centre for Disease Control and Prevention (CDC) provide a helpful framework which is summarised below (104, 105):

• All patients suspected of having neurosyphilis should undergo a thorough neurological examination.
• LP is indicated when patients with newly diagnosed syphilis present with neurological symptoms and/or abnormal neurological examination. This includes abnormal visual or auditory symptoms or signs which may indicate ocular or otologic syphilis, respectively.
• Routine CSF examination of patients with asymptomatic early or latent syphilis is not There is no evidence that performing a LP in patients without symptoms or signs suggestive of neurosyphilis improves long-term outcomes. However, LP should be considered in patients who have treatment failure despite adequate therapy as demonstrated by serial serology.
• Laboratory findings shown to be predictive of a CSF diagnosis of neurosyphilis among HIV-positive individual include a peripheral blood CD4+ T cell count of <350/mL and/or a RPR titre of >1:32 (106-108).

The interpretation of CSF findings in the HIV/syphilis co-infected individual can be complicated for several reasons. Firstly, certain CNS changes (namely pleiocytosis and elevated protein) can occur in HIV infection alone, thus may not in themselves indicate neurosyphilis in co-infected individuals. Secondly, due to the nature of the assays, tests detecting syphilis in the CNS are exquisitely sensitive to blood contamination and may yield false-positive results in such samples (100). The BASHH guidelines suggest the following CSF criteria for the diagnosis of neurosyphilis (105):

• Lymphocytic pleiocytosis:
  • 6-20 cells/microlitre in individuals on ART with undetectable HIV viral load or those with a blood CD4+ T cell count of <200/m
  • >20 cells/microlitre for all other HIV-infected individuals
• CSF Protein >0.45 g/L
• Non-treponemal test (e.g. CSF VDRL or RPR): reactive. These tests have poor sensitivity for the diagnosis of neurosyphilis (estimated 30-70%) but have good specificity unless the sample is contaminated with blood (100).
• Treponemal test (e.g. CSF-TPPA): >1:320. These assays are sensitive but non-specific so they are useful in excluding the likelihood of neurosyphilis. Test sensitivity is reduced by blood contamination (100).

Although NAATs are available in Australian reference laboratories, the diagnosis of neurosyphilis via NAAT is not currently recommended due to suboptimal assay performance on CSF samples (109).

Management

 Current Australian guidelines for penicillin-based treatment of syphilis are recommended irrespective of HIV status. (110). Briefly, primary, secondary & early latent (<2 years duration) syphilis is treated with benzathine benzylpenicillin 1.8 grams intramuscularly statim or procaine benzylpenicillin 1.5 grams intramuscularly daily for 10 days. The same therapy and dosages are indicated for syphilis of unknown duration or late-latent syphilis (>2 years), although for a longer duration: weekly for 3 weeks or daily for 15 days, respectively. The recommended treatment of neurosyphilis is with benzylpenicillin 1.8 grams intravenously every 4 hours for 15 days.

The Jarisch-Herxheimer (JH) reaction is a systemic inflammatory response to endotoxins released during spirochaete death following antimicrobial therapy and occurs in approximately one-third of individuals with early (infectious) syphilis (111). The syndrome is characterised by fever, skin rash, arthralgia/myalgia and malaise within 24-48 hours of treatment (112). Earlier evidence suggested there may be an increased risk of the JH reaction in HIV-infected individuals (113). However, a more recent prospective cohort study found no difference in risk of the JH reaction by HIV status (111).

 Treatment response is generally measured with serial serologic evaluation. Non-treponemal serology should be performed at 3, 6 and, if necessary, 12 months after treatment (4). A four-fold decrease in RPR titre or reversion to non-reactivity indicates successful treatment. Earlier and/or more frequent follow-up serology should be performed in higher risk cases (including when non-first line therapy is used, in pregnancy or among those at high risk of re-infection) (104). Serological tests should ideally be performed in the same laboratory, to ensure comparability between serial RPR titres. When the expected serological treatment response fails to occur, such samples can thus be tested at the same time by the same laboratory scientist (“in parallel”). A slower decline in RPR titre may occur among HIV-positive individuals who are not established on ART following treatment (114).

Treatment failure is suspected when the RPR test fails to decline four-fold, especially if re-infection can be reliably excluded. However, this could also be the result of the ‘serofast state’ when the test is reactive after 6-12 months of adequate treatment. There is no consensus as to whether this ‘serofast state’ leads to worse long-term outcomes or increased risk of tertiary neurosyphilis (99). Although the risk of treatment failure is higher in HIV-infected individuals, the difference in clinical outcomes by HIV status appear minimal (113).

Both UK and US guidelines recommend CSF examination in patients with suspected treatment failure as unrecognised neurosyphilis can be a reservoir of treponemes. If CSF examination excludes neurosyphilis as the cause of treatment failure then consideration of re-treatment with three doses of weekly intramuscular penicillin is recommended (104, 105). Although a follow-up CSF examination is recommended 6-12 months post-treatment to demonstrate normalisation of CSF parameters (114), there is a high correlation between the RPR titre in peripheral blood and normalisation of CSF changes. Thus, the serum RPR titre may be used as a surrogate marker for treatment response (103).