As well enunciated by the CDC, "syphilis is a systemic disease caused by T. pallidum.  Patients who have syphilis may seek treatment for signs or symptoms of primary infection (i.e., ulcer or chancre at the infection site), secondary infection (i.e., manifestations that include but are not limited to skin rash, mucocutaneous lesions, and lymphadenopathy), or tertiary infection (e.g., cardiac, ophthalmic, auditory abnormalities, and gummatous lesions). Latent infections (i.e., those lacking clinical manifestations) are detected by serologic testing. Latent syphilis acquired within the preceding year is referred to as early latent syphilis; all other cases of latent syphilis are either late latent syphilis or latent syphilis of unknown duration".¹

Serology for syphilis includes non-treponemal (VDRL, RPR, TRUST, RST, USR and ART) and treponemal tests (FTA-ABS [Fluorescent Treponema pallidum Antibody-Absorption] and MHA-TP [microhemagglutination assay for antibodies to Treponema pallidum]). In the FTA-ABS and in the MHA-TP, antibodies to other Treponema spp. are removed by absorption. The MHA-TP and FTA-ABS, like the VDRL, are expected to be positive in all but the earliest cases of syphilis. The non-treponemal tests are very sensitive for early syphilis (slightly more so than MHA-TP) but are commonly nonreactive in late syphilis when the FTA-ABS and MHA-TP are typically positive.² FTA-ABS results that are "minimally reactive" (formally known as 1+) are sometimes false positive and should be repeated. The VDRL (and presumably tests of similar principle) are responsive to therapy with a four-fold decrease in titer expected in 3 months and eight-fold in 6 months in patients with primary and secondary syphilis; smaller decreases are seen in individuals treated in the latent stage of syphilis and in reinfected individuals or smaller decreases may suggest treatment failure.³ A four-fold increase in titer usually is indicative of reinfection.¹ Different rates of decline in titer and seroreversion are, however, found with RPR testing depending on whether the disease is primary, secondary or early latent,^4,5 depending on whether there is concomitant HIV infection⁶ and depending on the treatment regimen.⁷ Possible biologic false-positives (BFPs) with VDRL or related assays should be evaluated with FTA-ABS or with EIA^8-10 or IB^11,12 with whole T. pallidum organisms or specific recombinant proteins. EIA for T. pallidum-specific IgG followed by RPR on EIA- positive specimens is reliable for screening.¹³

In general, patients with syphilis should be tested for HIV, and CSF examination is indicated in those who are HIV positive.^7,14 Serum TPHA titers of >1:5120 might be useful for selecting patients for lumbar puncture;¹⁵ but because reliability of this predictor in patients with HIV is doubtful, standard recommendations¹⁶ are best followed. In patients with syphilis, a serum RPR ³ 1:32 is predictive of neurosyphilis, and a CD4 T-cell count ≤350 cells/mL is an additional risk factor for neurosyphilis in the HIV-infected.¹⁷ The MHA-TP and FTA-ABS, in contrast to the VDRL, remain positive for life in ~85% of patients who are successfully treated³ and hence are not diagnostic of recent infection due to the "serological scars", but are useful, if nonreactive, in the exclusion of syphilis except in its earliest stages.

In a patient with a serum VDRL of 1:256 or less, a traumatic lumbar puncture will not result in a false-positive CSF VDRL unless there is sufficient contamination of CSF with blood to produce a visibly bloody CSF.¹⁸ Intrathecal synthesis of T. pallidum-specific IgG, as well as CSF oligoclonal immunoglobulins, are demonstrable in about 75% of patients with neurosyphilis.¹⁹ Intrathecal production of specific IgG is also demonstrable in about 25% of clinically asymptomatic syphilis.^20,21 When the blood-CSF barrier is leaky, intrathecal synthesis of T. pallidum-specific IgG can be critically assessed by comparison with synthesis of adenovirus-specific IgG^22,23 or even better by the T. pallidum Antibody Index corrected for IgG concentrations.²⁴ Although commonly increased, neither total nor specific IgG synthesis nor oligoclonal bands in CSF are useful for diagnosis.^25,26

Immunoblots for detection of antibodies reactive with T. pallidum-specific 15.5 and 45 kDa antigens are sensitive for syphilis (97%), treated and untreated secondary or early latent syphilis (100%), late latent treated syphilis (98%) and neurosyphilis (100%)²¹ as well as for congenital syphilis.^27,28 Initial reports²⁹ that immunoblots showing IgM reactive with 47, 17 and 15.5 kDa bands are 92-96% sensitive, 100% specific and have a positive predictive value of 95% for congenital syphilis were not confirmed.¹² Indeed, inability to prove the absence of infection in asymptomatic, possibly infected infants is still a problem,³⁰ the evaluation of which must include pathological examination of the umbilical cord, a sensitive site for detection of T. pallidum.³¹ IgM antibody capture EIA (MAC EIA) is very sensitive for congenital syphilis (100%) and less so in primary (82%), secondary (60%), latent (53%), neurosyphilis (34%) and treated syphilis (11%) and is highly specific.³² T. pallidum-specific IgM antibodies can be detected in later stages of untreated syphilis;²⁹ persistence of these IgM antibodies for some months after treatment³¹ makes them less useful than the VDRL for monitoring treatment.² Immunoblot assays are probably more sensitive than conventional CSF VDRL in detecting CNS involvement in infants with congenital syphilis.^12,29 Increased total IgM in CSF is common in neurosyphilis, but predictive values are not available.³² More data on sensitivity and specificity of immunoblots and MAC EIA compared with standard tests for syphilis are needed before these tests can become standard practice, but recently improved immunoblot and IgG EIA techniques are promising confirmatory tests.^8-12,33

Serologic testing for congenital syphilis was critically reviewed.³⁴ Recent data indicate the value of IgM immunoblotting for specific IgM as well as PCR assay for T pallidum DNA in blood and CSF.³⁵ Earlier data indicated false-positive rates for serum FTA-ABS IgM as high as 33% in infants; another study showed a positive predictive value of 73% for FTA-ABS IgM.²⁹ False-negative rates as high as 35% for FTA-ABS IgM are reported in certain patient groups together with false-positive rates of 10%.³⁶ In congenital syphilis, the positive predictive value is 95% for IgM immunoblot and 73% for FTA-ABS IgM.²⁹ The FTA-ABS IgM, FTA-ABS 19S IgM and the indirect EIA for IgM test are all less sensitive than the IgM antibody capture EIA (MAC EIA)³⁷ for diagnosis of congenital syphilis.³⁰ Congenital syphilis (probably due to recent infection) can be diagnosed at 3-14 months of age even when infants and mother were seronegative at delivery.^38,39 At delivery, screening tests for syphilis should be performed with maternal blood specimens and not with cord blood, because false-positive results (possibly due to contamination with Wharton jelly) and false-negative results (due to infection acquired late in pregnancy) can occur with cord blood specimens.⁴⁰ Placental PCR and histopathology are useful for diagnosis of congenital syphilis.⁴¹ When CD4 counts are low during HIV infection, treponemal tests can be negative in adults with a documented history of treated syphilis.⁴² A new immuno- blot is deemed a useful additional confirmatory test or alternative to the FTA-ABS.⁴³ The role of syphilis EIAs in routine diagnostic microbiology laboratories is reviewed.⁴⁴ The quality and performance of EIA, immunoblots and latex agglutination for syphilis serology continues to improve,^45-48 and the arguments about testing blood donors continue unresolved.^49,50

T. pallidum antigens obtained from expression cloning^51,52 and detection of T. pallidum DNA by PCR^53-57 can sometimes be useful for diagnosis and management of syphilis, the frequency of which has increased dramatically in recent years.^58,59 PCR is negative for T. pallidum DNA and RNA in platelet concentrates from blood donors with confirmed-positive syphilis tests.⁶⁰ Multiplex PCR for detection of H. ducreyi, T. pallidum and H. simplex in genital ulcer disease is very promising.^61,62 The characteristics of genital ulcers were reviewed.⁶³ A new comprehensive study provides an overview of molecular methods for the diagnosis and subtyping of the most common sexually transmitted infections including syphilis as well as the laboratory diagnostic tests and clinical specimens to use when infection is suspected.⁶⁴ The genome of T. pallidum was sequenced.⁶⁵ This plus large advances in our knowledge of the immune response to infection^66,67  bode well for vaccine development.⁶⁸ A marvelous, succinct  review of syphilis is available. ⁶⁹