Demonstration of the Organism
T. pallidum cannot be detected by culture. Historically, dark-field microscopy and immunofluorescence antibody staining have been used to identify this spirochete in samples from moist lesions such as chancres or condylomata lata, but these tests are rarely available today outside of research laboratories. More sensitive PCR tests have been developed but are not commercially available, although some laboratories perform in-house PCR testing.
T. pallidum can be found in tissue with appropriate silver stains, but these results should be interpreted with caution because artifacts resembling T. pallidum are often seen. Tissue treponemes can be demonstrated more reliably in research laboratories by PCR or by immunofluorescence or immunohistochemical methods using specific monoclonal or polyclonal antibodies to T. pallidum.
Serologic Tests for Syphilis
There are two types of serologic test for syphilis: nontreponemal and treponemal. Both are reactive in persons with any treponemal infection, including yaws, pinta, and endemic syphilis.
The most widely used nontreponemal antibody tests for syphilis are the rapid plasma reagin (RPR) and Venereal Disease Research Laboratory (VDRL) tests, which measure IgG and IgM directed against a cardiolipin-lecithin-cholesterol antigen complex. The RPR test is easier to perform and uses unheated serum; it is the test of choice for rapid serologic diagnosis in a clinical setting and can be automated. The VDRL test remains the standard for examining CSF. The RPR and VDRL tests are recommended for screening or for quantitation of serum antibody. The titer reflects disease activity, rising during the evolution of early syphilis and often exceeding 1:32 in secondary syphilis. After therapy for early syphilis, a persistent fall by fourfold or more (e.g., a decline from 1:32 to 1:8) is considered an adequate response. VDRL titers do not correspond directly to RPR titers, and sequential quantitative testing (as for response to therapy) must employ a single test. As will be discussed (see “Evaluation for Neurosyphilis,” below), the RPR titer may be useful in determining which patients will benefit from CSF examination.
Treponemal tests measure antibodies to native or recombinant T. pallidum antigens and include the fluorescent treponemal antibody–absorbed (FTA-ABS) test and the T. pallidum particle agglutination (TPPA) test, both of which are more sensitive for primary syphilis than the previously used hemagglutination tests. The T. pallidum hemagglutination (TPHA) test is widely used in Europe but is not available in the United States. When used to confirm positive nontreponemal test results, treponemal tests have a very high positive predictive value for diagnosis of syphilis. In a screening setting, however, these tests give false-positive results at rates as high as 1–2%. Treponemal tests are likely to remain reactive even after adequate treatment and cannot differentiate past from current T. pallidum infection.
Treponemal immunochromatographic strip (ICS) tests and enzyme immunoassays (EIAs), based largely on reactivity to recombinant antigens, have also been developed. Treponemal EIAs have been approved as confirmatory tests and, because of their ease of automation, are now used for screening purposes by some large laboratories. Because treponemal tests cannot distinguish between current and treated syphilis or may be falsely reactive, clinicians may be uncertain about how to interpret reactive EIA screening results. Figure 169-4 provides a suggested algorithm for management of such cases.
Algorithm for interpretation of results from syphilis enzyme immunoassays (EIAs) used for screening. RPR, rapid plasma reagin; VDRL, Venereal Disease Research Laboratory; TPPA, T. pallidum particle agglutination; FTA-ABS, fluorescent treponemal antibody–absorbed. (Based on the 2010 Sexually Transmitted Diseases Treatment Guidelines from the Centers for Disease Control and Prevention.)
Considerable interest has recently been focused on point-of-care ICS tests that can be used in the field or in resource-poor settings. These treponemal tests are not yet approved for use in the United States but have been assessed in antenatal clinics in a number of developing countries.
Both nontreponemal and treponemal tests may be nonreactive in early primary syphilis, although treponemal tests are slightly more sensitive (85–90%) during this stage than nontreponemal tests (∼80%). All tests are reactive during secondary syphilis. (Fewer than 1% of patients with high titers have a nontreponemal test that is nonreactive or weakly reactive with undiluted serum but is reactive with diluted serum—the prozone phenomenon.) VDRL and RPR sensitivity and titers may decline in untreated persons with late latent or late syphilis, but treponemal tests remain sensitive in these stages. Whereas nontreponemal test titers will decline or the tests will become nonreactive after therapy for early syphilis, treponemal tests often remain reactive after therapy and are not helpful in determining the infection status of persons with past syphilis.
For practical purposes, most clinicians need to be familiar with three uses of serologic tests for syphilis: (1) screening or diagnosis (RPR or VDRL), (2) quantitative measurement of antibody to assess clinical syphilis activity or to monitor response to therapy (RPR or VDRL), and (3) confirmation of a syphilis diagnosis in a patient with a reactive RPR or VDRL test (FTA-ABS, TPPA, EIA). Studies have not demonstrated the utility of IgM testing for adult syphilis. While IgM titers appear to decline after therapy, the presence or absence of specific IgM does not strictly correlate with T. pallidum infection. Moreover, no commercially available IgM test is recommended for evaluation of infants with suspected congenital syphilis.
False-Positive Serologic Tests for Syphilis
The lipid antigens of nontreponemal tests are similar to those found in human tissues, and the tests may be reactive (usually with titers ≤1:8) in persons without treponemal infection. Among patients being screened for syphilis because of risk factors, clinical suspicion, or history of exposure, ∼1% of reactive tests are falsely positive. Modern VDRL and RPR tests are highly specific, and false-positive reactions are largely limited to persons with autoimmune conditions or injection drug use. The prevalence of false-positive results increases with advancing age, approaching 10% among persons >70 years old. In a patient with a false-positive nontreponemal test, syphilis is excluded by a nonreactive treponemal test.
Evaluation for Neurosyphilis
Involvement of the CNS is detected by examination of CSF for pleocytosis (>5 white blood cells/μL), increased protein concentration (>45 mg/dL), or VDRL reactivity. Elevated CSF cell counts and protein concentrations are not specific for neurosyphilis and may be confounded by HIV co-infection. Because CSF pleocytosis may also be due to HIV, some studies have suggested using a CSF white-cell cutoff of 20 cells/μL as diagnostic of neurosyphilis in HIV-infected patients with syphilis. The CSF VDRL test is highly specific and, when reactive, is considered diagnostic of neurosyphilis; however, this test is insensitive and may be nonreactive even in cases of symptomatic neurosyphilis. The FTA-ABS test on CSF is reactive far more often than the VDRL test on CSF in all stages of syphilis, but reactivity may reflect passive transfer of serum antibody into the CSF. A nonreactive FTA-ABS test on CSF, however, may be used to rule out asymptomatic neurosyphilis. A recent study demonstrated the utility of measuring CXCL13 in CSF to distinguish between neurosyphilis and HIV-related CSF abnormalities.
Clearly, all T. pallidum–infected patients who have signs or symptoms consistent with neurologic disease (e.g., meningitis, hearing loss) or ophthalmic disease (e.g., uveitis, iritis) should have a CSF examination, regardless of disease stage. The appropriate management of asymptomatic persons is less clear. Lumbar puncture on all asymptomatic patients with untreated syphilis is impractical and unnecessary. Because standard therapy with penicillin G benzathine fails to result in treponemicidal drug levels in CSF, it is important to identify those persons at higher risk for having or developing neurosyphilis so that appropriate therapy may be given. Large-scale prospective studies have now provided evidence-based guidelines for determining which syphilis patients may benefit most from CSF examination for evidence of neurosyphilis. Specifically, patients with RPR titers of ≥1:32 are at higher risk of having neurosyphilis (11-fold and 6-fold higher in HIV-infected and HIV-uninfected persons, respectively), as are HIV-infected patients with CD4+ T cell counts of ≤350/μL. Current recommendations for CSF examination are shown in Table 169-1.
Table 169-1 Indications for CSF Examination in Adults with All Stages of Syphilis |Favorite Table|Download (.pdf)
Table 169-1 Indications for CSF Examination in Adults with All Stages of Syphilis
Signs or symptoms of nervous system involvement [e.g., meningitis, hearing loss, cranial nerve dysfunction, altered mental status, ophthalmic disease (e.g., uveitis, iritis, pupillary abnormalities), ataxia, loss of vibration sense], or
RPR or VDRL titer ≥1:32, or
Suspected treatment failure
Additional Indications in HIV-Infected Persons
CD4+ T cell count ≤350/μL, or
|CSF examination is recommended by some experts for all HIV-infected persons.|
Evaluation of HIV-Infected Patients for Syphilis
Because persons at highest risk for syphilis are also at increased risk for HIV infection, these two infections frequently coexist. There is evidence that syphilis and other genital ulcer diseases may be important risk factors in the acquisition and transmission of HIV infection. Some manifestations of syphilis may be altered in patients with concurrent HIV infection, and multiple cases of neurologic relapse after standard therapy have been reported in these patients. T. pallidum has been isolated from the CSF of several patients (with and without concurrent HIV infection) after penicillin G benzathine therapy for early syphilis.
Persons with newly diagnosed HIV infection should be tested for syphilis; conversely, all patients with newly diagnosed syphilis should be tested for HIV infection. Some authorities, persuaded by reports of persistent T. pallidum in CSF of HIV-infected persons after standard therapy for early syphilis, recommend CSF examination for evidence of neurosyphilis for all co-infected patients, regardless of the stage of syphilis, with treatment for neurosyphilis if CSF abnormalities are found. Others, on the basis of their own clinical experience, believe that standard therapy—without CSF examination—is sufficient for all cases of early syphilis in HIV-infected patients without neurologic signs or symptoms. As described above, RPR titer and CD4+ T cell count can be used to identify patients at higher risk of neurosyphilis for lumbar puncture, although some cases of neurosyphilis will not be identified by these criteria. Table 169-1 summarizes guidelines suggested by published studies. Serologic testing after treatment is important for all patients with syphilis, particularly for those also infected with HIV.
Treatment of Acquired Syphilis
The CDC's 2010 guidelines for the treatment of syphilis are summarized in Table 169-2 and are discussed below. Penicillin G is the drug of choice for all stages of syphilis. T. pallidum is killed by very low concentrations of penicillin G, although a long period of exposure to penicillin is required because of the unusually slow rate of multiplication of the organism. The efficacy of penicillin against syphilis remains undiminished after60 years of use, and there is no evidence of penicillin resistance in T. pallidum. Other antibiotics effective in syphilis include the tetracyclines and the cephalosporins. Aminoglycosides and spectinomycin inhibit T. pallidum only in very large doses, and the sulfonamides and the quinolones are inactive. Azithromycin has shown significant promise as an effective oral agent against T. pallidum; however, strains harboring 23S rRNA mutations that confer macrolide resistance are widespread; such strains represent >50% of recent isolates from Seattle and San Francisco and have now been identified in multiple North American and European sites. Macrolide resistance mutations have been identified in nearly all samples reported from China. In contrast, three studies conducted in Africa (Uganda, Tanzania, and Madagascar) have documented the clinical efficacy of azithromycin and, for the subset of samples examined, have found no molecular evidence of the mutation. In short, the prevalence of resistant strains varies widely by geographic location, and routine treatment of syphilis with azithromycin is not recommended. In all cases, careful follow-up of any patient treated for syphilis with azithromycin must be ensured.
Table 169-2 Recommendations for the Treatment of Syphilisa |Favorite Table|Download (.pdf)
Table 169-2 Recommendations for the Treatment of Syphilisa
|Stage of Syphilis||Patients without Penicillin Allergy||Patients with Confirmed Penicillin Allergyb|
|Primary, secondary, or early latent|
CSF normal or not examined: Penicillin G benzathine (single dose of 2.4 mU IM)
CSF abnormal: Treat as neurosyphilis
CSF normal or not examined: Tetracycline HCl (500 mg PO qid) or doxycycline (100 mg PO bid) for 2 weeks
CSF abnormal: Treat as neurosyphilis
|Late latent (or latent of uncertain duration), cardiovascular, or benign tertiary|
CSF normal or not examined: Penicillin G benzathine (2.4 mU IM weekly for 3 weeks)
CSF abnormal: Treat as neurosyphilis
CSF normal and patient not infected with HIV:
Tetracycline HCl (500 mg PO qid) or doxycycline (100 mg PO bid) for 4 weeks
CSF normal and patient infected with HIV:
Desensitization and treatment with penicillin if compliance cannot be ensured
CSF abnormal: Treat as neurosyphilis
|Neurosyphilis (asymptomatic or symptomatic)|
Aqueous crystalline penicillin G (18–24 mU/d IV, given as 3–4 mU q4h or continuous infusion) for 10–14 days
Aqueous procaine penicillin G (2.4 mU/d IM) plus oral probenecid (500 mg qid), both for 10–14 days
|Desensitization and treatment with penicillinc|
|Syphilis in pregnancy||According to stage||Desensitization and treatment with penicillin|
Early Syphilis Patients and Their Contacts
Penicillin G benzathine is the most widely used agent for the treatment of early syphilis; a single dose of 2.4 million units is recommended. Preventive treatment is also recommended for individuals who have been exposed to infectious syphilis within the previous 3 months. The regimens recommended for prevention are the same as those recommended for early syphilis. Penicillin G benzathine cures >95% of cases of early syphilis, although clinical relapse can follow treatment, particularly in patients with concurrent HIV infection. Because the risk of neurologic relapse may be higher in HIV-infected patients, CSF examination is recommended in HIV-seropositive individuals with syphilis of any stage, particularly those with a serum RPR titer of ≥1:32 or a CD4+ T cell count of ≤350/μL. Therapy appropriate for neurosyphilis should be given if there is any evidence of CNS disease.
Late Latent and Late Syphilis
If CSF abnormalities are found, the patient should be treated for neurosyphilis. If CSF is normal, the recommended treatment is penicillin G benzathine (7.2 million units total; Table 169-2). The clinical response to treatment for benign tertiary syphilis is usually impressive. However, responses to therapy for cardiovascular syphilis are not dramatic because aortic aneurysm and aortic regurgitation cannot be reversed by antibiotics.
For penicillin-allergic patients with syphilis, a 2-week (early syphilis) or 4-week (late or late latent syphilis) course of therapy with doxycycline or tetracycline is recommended. These regimens appear to be effective in early syphilis but have not been tested for late or late latent syphilis, and compliance may be problematic. Limited studies suggest that ceftriaxone (1 g/d, given IM or IV, for 8–10 days) is effective for early syphilis. These nonpenicillin regimens have not been carefully evaluated in HIV-infected individuals and should be used with caution. If compliance and follow-up cannot be ensured, penicillin-allergic HIV-infected persons with late latent or late syphilis should be desensitized and treated with penicillin.
Penicillin G benzathine, given in total doses of up to 7.2 million units, does not produce detectable concentrations of penicillin G in CSF and should not be used for treatment of neurosyphilis. Asymptomatic neurosyphilis may relapse into symptomatic disease following treatment with benzathine penicillin, and the risk of relapse may be higher in HIV-infected patients. Both symptomatic and asymptomatic neurosyphilis should be treated with aqueous penicillin. Administration either of IV aqueous crystalline penicillin G or of aqueous procaine penicillin G plus probenecid in recommended doses is thought to ensure treponemicidal concentrations of penicillin G in CSF. The clinical response to penicillin therapy for meningeal syphilis is dramatic, but treatment of neurosyphilis with existing parenchymal damage may only arrest disease progression. Neurologic relapse has been reported after high-dose IV penicillin therapy for neurosyphilis in an HIV-infected patient. No alternative therapies have been studied, but careful follow-up is essential, and re-treatment is warranted in such patients. No data suggest that additional therapy (e.g., penicillin G benzathine for 3 weeks) is beneficial after treatment for neurosyphilis.
The use of antibiotics other than penicillin G for the treatment of neurosyphilis has not been studied, although very limited data suggest that ceftriaxone may be used. In patients with penicillin allergy demonstrated by skin testing, desensitization and treatment with penicillin are recommended.
Management of Syphilis in Pregnancy
Every pregnant woman should undergo a nontreponemal test at her first prenatal visit and, if at high risk of exposure, again in the third trimester and at delivery. In the untreated pregnant patient with presumed syphilis, expeditious treatment appropriate to the stage of the disease is essential. Patients should be warned of the risk of a Jarisch-Herxheimer reaction, which may be associated with mild premature contractions but rarely results in premature delivery.
Penicillin is the only recommended agent for the treatment of syphilis in pregnancy. If the patient has a documented penicillin allergy, desensitization and penicillin therapy should be undertaken according to the CDC's 2010 guidelines. After treatment, a quantitative nontreponemal test should be repeated monthly throughout pregnancy to assess therapeutic efficacy. Treated women whose antibody titers rise by fourfold or whose titers do not decrease by fourfold over a 3-month period should be re-treated.
Evaluation and Management of Congenital Syphilis
Whether or not they are infected, newborn infants of mothers with reactive serologic tests may themselves have reactive tests because of transplacental transfer of maternal IgG antibody. For asymptomatic infants born to women treated adequately with penicillin during the first or second trimester of pregnancy, monthly quantitative nontreponemal tests may be performed to monitor for appropriate reduction in antibody titers. Rising or persistent titers indicate infection, and the infant should be treated. Detection of neonatal IgM antibody may be useful, but no commercially available test is currently recommended.
An infant should be treated at birth if the treatment status of the seropositive mother is unknown; if the mother has received inadequate or nonpenicillin therapy or has received penicillin therapy in the third trimester; or if the infant may be difficult to follow. The CSF should be examined to obtain baseline values before treatment. Penicillin is the only recommended drug for the treatment of syphilis in infants. Specific recommendations for the treatment of infants and older children are included in the CDC's 2010 treatment guidelines.
A dramatic though usually mild reaction consisting of fever, chills, myalgias, headache, tachycardia, increased respiratory rate, increased circulating neutrophil count, and vasodilation with mild hypotension may follow the initiation of treatment for syphilis. This reaction is thought to be a response to lipoproteins released by dying T. pallidum organisms. The Jarisch-Herxheimer reaction occurs in ∼50% of patients with primary syphilis, 90% of those with secondary syphilis, and a lower proportion of persons with later-stage disease. Defervescence takes place within 12–24 h. In patients with secondary syphilis, erythema and edema of the mucocutaneous lesions may increase. Patients should be warned to expect such symptoms, which can be managed with symptom-based treatment. Steroid and other anti-inflammatory therapy is not required for this mild transient reaction.
Follow-Up Evaluation of Responses to Therapy
Efficacy of treatment should be assessed by clinical evaluation and monitoring of the quantitative VDRL or RPR titer for a fourfold decline (e.g., from 1:32 to 1:8). Patients with primary or secondary syphilis should be examined 6 and 12 months after treatment and persons with latent or late syphilis at 6, 12, and 24 months. More frequent clinical and serologic examination (3, 6, 9, 12, and 24 months) is recommended for patients concurrently infected with HIV, regardless of the stage of syphilis.
After successful treatment of seropositive first-episode primary or secondary syphilis, the VDRL or RPR titer progressively declines, becoming negative by 12 months in 40–75% of seropositive primary cases and in 20–40% of secondary cases. Patients with HIV infection or a history of prior syphilis are less likely to become nonreactive in the VDRL or RPR test. Rates of decline of serologic titers appear to be slower and serologically defined treatment failures more common among HIV-infected patients than among those without HIV co-infection; however, effective antiretroviral therapy may reduce these differences. Re-treatment should be considered if serologic responses are not adequate or if clinical signs persist or recur. Because it is difficult to differentiate treatment failure from reinfection, the CSF should be examined, with treatment for neurosyphilis if CSF is abnormal and treatment for late latent syphilis if CSF is normal. Patients treated for late latent syphilis frequently have low initial VDRL or RPR titers and may not have a fourfold decline after therapy with penicillin. In such patients, re-treatment is not warranted unless the titer rises or signs and symptoms of syphilis appear. Because treponemal tests may remain positive despite treatment for seropositive syphilis, these tests are not useful in following the response to therapy.
The activity of neurosyphilis (symptomatic or asymptomatic) correlates best with CSF pleocytosis, and this measure provides the most sensitive index of response to treatment. Repeat CSF examinations should be performed every 6 months until the cell count is normal. An elevated CSF cell count falls to normal in 3–12 months in adequately treated HIV-uninfected patients. The persistence of mild pleocytosis in HIV-infected patients may be due to the presence of HIV in CSF; this scenario may be difficult to distinguish from treatment failure. Elevated levels of CSF protein fall more slowly, and the CSF VDRL titer declines gradually over several years. In patients treated for neurosyphilis, a fourfold reduction in serum RPR titer has been positively correlated with normalization of CSF abnormalities; this correlation is stronger in HIV-uninfected patients and in HIV-infected patients receiving effective antiretroviral therapy.