Advances in Syphilis Diagnostics to Address the 21st-Century Epidemic

Abstract

Background

For 20 years, the United States has experienced a progressive rise in syphilis cases, increasing to >209 000 cases in 2023. The current case rate of 62.5 per 100 000 persons represents a 30-fold increase since 2000–2001, when a concerted national effort nearly eliminated the disease. This review highlights epidemiologic trends, current gold-standard diagnostic methods, emerging diagnostic tests, and key challenges for laboratorians.

Content

The current outbreak in the United States has been marked by dramatic increases in cases among women and men who have sex with women, resulting in 3800 congenital cases in 2023: a 10-fold rise over the last decade. Containing syphilis has been hampered by several factors. Recurrent shortages of the primary treatment, benzathine penicillin G, and a high prevalence of macrolide resistance have curtailed access to single-dose treatment options, necessitating multiday courses of doxycycline. The etiologic agent, Treponema pallidum subsp. pallidum, is a stealth pathogen with protean manifestations, thus posing a clinical challenge. Diagnostic testing, which has seen few updates in decades, has limited sensitivity in early or latent syphilis, and the need to integrate multiple serologic tests can delay turnaround times.

Summary

Public health interventions, such as doxycycline post-exposure prophylaxis, have started to contain the disease, and newer diagnostic tests may offer useful adjuncts for case finding. Serologic testing, with low cost and high sensitivity, is likely to remain the primary method for the screening and diagnosis of syphilis. Sustained investment in public health infrastructure and automation, interlaboratory data sharing, and addressing social determinants of health are critical to the reduction of syphilis cases.

Introduction

For the past 2 decades, new syphilis diagnoses have increased year upon year (Fig. 1). In 2023, there were 209 253 syphilis diagnoses (all stages including congenital syphilis) in the United States, the greatest number of cases reported since 1950 (1). The 2023 Centers for Disease Control and Prevention Sexually Transmitted Infection Surveillance Report highlights the dynamic epidemiology of syphilis among heterosexual people, pregnant people, and men who have sex with men (MSM).

Fig. 1.

Longitudinal tracking of syphilis case rates demonstrates a rise in cases since a nadir in 2000–2001. Primary graph shows all stages of syphilis since 1941. The inset compares cases by stage from 2014 to 2023. Rates of primary and secondary (blue trace, inset) are very similar to nonprimary/nonsecondary cases (red trace, inset), and trend lines overlap. Cases are expressed as rates per 100 000 persons. Figure adapted from the Centers for Disease Control and Prevention. Sexually Transmitted Infections Surveillance 2023. Atlanta, GA: US Department of Health and Human Services; 2024. Accessed November 2024 (1). Created in BioRender. Treger, R. (2025) https://BioRender.com/u07u164. Color figure available at https://academic.oup.com/clinchem.

Currently, the United States is in the midst of the largest heterosexual syphilis epidemic in 30 years (1). Primary and secondary (P&S) syphilis cases rose almost 900% among women from 2013 to 2022, such that the male-to-female ratio of P&S syphilis cases dropped from 10:1 to 3:1 and women who have sex with men currently comprise 20% of all P&S syphilis diagnoses in the United States (1). Concurrently, the number of P&S syphilis cases among men who have sex with women increased by 501%. As women who have sex with men and men who have sex with women together comprised 44% of all P&S syphilis cases in 2023, this new epidemic is associated with an alarming increase in congenital syphilis with 3882 cases (up from 335 in 2012) and 279 syphilitic stillbirths and infant deaths in 2023 (1). The current heterosexual epidemic is concentrated among people affected by key social determinants of health, including structural racism, substance use (particularly methamphetamine and fentanyl), houselessness, poverty, trauma, and incarceration (2–5).

Recent work on congenital syphilis highlights missed opportunities to detect syphilis in pregnancy and to prevent congenital disease. As most syphilis in pregnancy is staged as late or unknown duration latent infection, screening is the only reliable way to diagnose syphilis in pregnancy. Although universal syphilis screening at the first prenatal care appointment followed by risk-based screening in the third trimester is recommended (6), sequential studies have demonstrated that 5% to 50% of cases were missed due to maternal seroconversion late in pregnancy (7, 8). More troubling, 37% of pregnant patients associated with a case of congenital syphilis received late testing (<30 days prior to delivery) or no testing, while 40% were not treated despite a positive test (7). Late syphilis detection and treatment in pregnancy are associated with congenital syphilis prevention failures. Of 6 treatment failures in a study defining the effectiveness of penicillin for congenital syphilis prevention, 4 (67%) were among pregnant people treated within 2 weeks of delivery (9).

In a case control study of pregnant people with syphilis, 67% of infants with congenital syphilis were delivered within 30 days of treatment compared to 29% of infants without evidence of congenital disease (mean days from treatment to delivery, 29 ± 39 days vs 56 ± 43 days, respectively, P < 0.001) (10). As a result, the American College of Obstetricians and Gynecologists issued recommendations to screen all pregnant people regardless of risk at the first presentation to prenatal care, in the early third trimester, and at delivery (11). Given the current epidemiology of syphilis among women and infants, third-trimester screening is likely highly cost-effective for preventing congenital syphilis based on models parameterized to data prior to the large increase in syphilis among women and infants (12).

While syphilis diagnoses have increased dramatically among heterosexuals, MSM (particularly young MSM), MSM living with human immunodeficiency virus (HIV), and Black and Hispanic/Latine MSM continue to bear a disproportionate burden of syphilis in the United States (1, 3 ). In 2023, MSM experienced 32.7% of syphilis diagnoses overall and 57.5% of cases among men with known sex of sex partners (1). Encouragingly, from 2022 to 2023, MSM experienced a 13.4% decrease in P&S syphilis diagnoses, the first substantial decrease in cases in the past 15 years among any demographic group (1). This decrease in cases can be, in part, explained by the implementation of doxycycline as post-exposure prophylaxis (doxy-PEP) for the prevention of bacterial sexually transmitted infections (STIs), which in controlled studies is 73% to 80% effective in preventing incident syphilis among MSM and transgender women who have sex with men (13, 14). Subsequently, the Centers for Disease Control and Prevention recently published guidelines recommending doxy-PEP as STI prevention for MSM and transgender women who have sex with men and who have had at least one bacterial STI in the prior year (13).

The disproportionate share of syphilis cases among MSM reflects broader health disparities for this disease. Surveillance data has consistently shown higher case rates per 100 000 population for P&S syphilis among American Indian/Alaska Native (AI/AN) people (58.2), African Americans (39.7), Native Hawaiian and Pacific Islanders (24.3), and Hispanic/Latine (16.9) people compared with White (9.1) or Asian (4.4) counterparts (Fig. 2) (1). Similar disparities have been reported for syphilis in pregnancy and congenital syphilis (Fig. 2). AI/AN pregnant people and children are most severely impacted by congenital syphilis: one of every 155 live births among AI/AN pregnant people is affected compared to 1 in 1848 live births among White pregnant people (Fig. 2) (1). Moreover, no testing or late testing for gestational syphilis resulted in the highest percentage of missed opportunities to prevent congenital syphilis among AI/AN and Native Hawaiian/Pacific Islander pregnant people (7). Inadequate treatment accounted for the highest percentage of missed opportunities to prevent congenital syphilis among Black/African American and Hispanic pregnant people (7). These missed opportunities propagate intergenerational syphilis disparities.

Fig. 2.

Cases of primary, secondary, and congenital syphilis disproportionately burden historically marginalized racial and ethnic groups. Cases of primary and secondary syphilis are expressed as case rates per 100 000 persons (left), and congenital syphilis cases are expressed as case rate per 100 000 live births (right). Figure adapted from the Centers for Disease Control and Prevention. Sexually Transmitted Infections Surveillance 2023. Atlanta, GA: US Department of Health and Human Services; 2024. Accessed November 2024 (1). Created in BioRender. Treger, R. (2025) https://BioRender.com/b02o775. Abbreviations: AA, African American; NH/PI, Native Hawaiian/Pacific Islander. Color figure available at https://academic.oup.com/clinchem.

Laboratory Diagnosis of Syphilis

Serologic testing remains the mainstay of syphilis diagnosis for all stages of the infection. In addition to having excellent clinical sensitivity in most cases, serology is particularly important, since many infections go undetected. Primary lesions are typically painless and may go unnoticed or occur in anatomic sites that are difficult to visualize (e.g., oropharynx, anorectal mucosa). Gestational syphilis can be asymptomatic, and latent disease may constitute an important reservoir for transmission. Additionally, the presentation of secondary, ocular, and neurosyphilis can overlap with many other disease entities, further complicating how cases are identified. Given this diagnostic complexity, there is no single “gold-standard” testing modality that captures all cases at a given stage of disease. Direct detection methods, such as microscopy and molecular assays, have limited sensitivity due to preanalytical limitations in the identification of infected mucosa and/or low organism burden in a proportion of samples (Table 1) (15, 16). Microscopy is primarily limited to histopathology, as darkfield microscopy is no longer available in most laboratories due mainly to technical complexity. Even in the case of serology, test performance will vary depending on the stage of disease and local prevalence of infection. Important opportunities to improve diagnostics include better detection of early and latent infections, development of point-of-care (POC) nontreponemal assays, and adjudication of ambiguous serologic results such as the “serofast” state.

Table 1.

Estimated median treponemes per 1 mL eluted swab material in syphilis cases.^a,b

Swab type	Primary	Secondary	Early/latent
Oral (untargeted)	1 × 104	3 × 104	2 × 104
Rectal	TBD	TBD	TBD
Lesion (oral)	1 × 104	2 × 105	2–7 × 104

^aConcentrations were calculated from published median quantifications reported in microliters eluted DNA from a qPCR targeting the single copy gene tp0574 (15, 16) based on methods in Supplemental Materials from reference (17). Units are in organisms/mL of swab fluid.

^bTBD, to be determined; robust quantification of organism burden in rectal swabs has not yet been published.

Current Gold-Standard Testing Strategies

Treponemal Antigen Tests

Because T. pallidum cannot be cultured in the clinical laboratory, screening and diagnosis of syphilis infection rely on serologic testing. Treponemal antibodies arise in response to exposure to T. pallidum and target a variety of treponemal antigens (Fig. 3), including the immunogenic lipoproteins TpN15, TpN17, and TpN47 that are commonly used in diagnostic testing (19). Treponemal immunoglobulin M (IgM) antibodies can be detected as early as 2 weeks after infection, with treponemal immunoglobulin G (IgG) titers rising shortly thereafter (20). Although treponemal IgM levels decline following treatment of the infection, treponemal IgG antibodies typically persist indefinitely (20).

Fig. 3.

Serological and adjunct testing have evolving roles and sensitivities across clinical stages of syphilis. Representative schematics of relative levels of treponemal (dashed blue trace) and nontreponemal (solid red trace) antibodies detected by conventional serology (top panel) are represented over the natural history of clinical syphilis (middle panel). While treponemal antibody levels usually remain positive for life, the possible trajectories of nontreponemal antibody levels in treated and untreated individuals are shown. The bottom panel describes currently available molecular and POC tests that can be used over the course of disease, grouped by turnaround time. These adjunct tests are colored by their approximate clinical sensitivities (solid outline, green fill, >70%; dashed outline, yellow fill, 50%–70%, and no outline, orange fill, <50%). Clinical sensitivity of serologic testing is >85% and approaches 100% in secondary disease. Top panel adapted from Papp JR, Park IU, Fakile Y, et al. CDC Laboratory Recommendations for Syphilis Testing, United States, 2024. MMWR Recomm Rep. 2024;73:6 (18). Created in BioRender. Treger, R. (2025) https://BioRender.com/f27g932. Abbreviation: OP, oropharyngeal. Color figure available at https://academic.oup.com/clinchem.

Treponemal antibody tests evaluate exposure to the pathogen by detecting host antibodies to T. pallidum. These tests can employ enzyme immunoassay, manual agglutination, or manual fluorescence microscopy techniques. They can detect IgG alone or both IgM and IgG antibodies together, and they may use wildtype or one or more recombinant TpN lipoprotein antigens [reviewed in (19)]. Although the manual T. pallidum particle agglutination test (TP-PA) may lack some sensitivity for primary syphilis (19, 21), it offers exceptional specificity for T. pallidum infection and therefore serves as a confirmatory test in clinical algorithms. Automated treponemal immunoassays enable high throughput and have excellent specificity and sensitivity for most stages of disease, and they have become widely employed in clinical practice.

There are several important testing considerations for treponemal antibodies. First, as with any screening immunoassay, false-positive treponemal antibody results can occur, particularly when used in populations with low pretest probability of syphilis infection (22). In these cases, “seroreversion” to a negative treponemal antibody result can be observed on repeat testing (23, 24), which can be particularly helpful in determining the need for treatment in pregnant individuals with isolated treponemal antibody positivity. Second, antibodies that arise in response to nonvenereal endemic treponematoses (yaws due to T. pallidum subsp. pertenue, bejel due to T. pallidum subsp. endemicum, and pinta due to T. carateum) are also detected by treponemal assays (25). Treponemal results from individuals from areas where these infections are endemic must therefore be interpreted with care. Third, most individuals remain seropositive after treatment for syphilis, and treponemal antibody presence cannot be used to monitor response to treatment or to diagnose reinfection in the setting of prior treated disease. However, treponemal antibody titers detected using manual fluorescent treponemal antibody absorption and T. pallidum hemagglutination assay can wane and even disappear following treatment of early syphilis (26) and in individuals with HIV coinfection (27). It is not clear whether this phenomenon also occurs when treponemal antibodies are tested using immunoassays, which offer improved analytical sensitivity.

Nontreponemal (Lipoidal Antigen) Tests

While treponemal antibody tests cannot discriminate between past and current syphilis infection, nontreponemal antibodies can identify active disease. Although originally perceived to target nontreponemal host lipids such as cardiolipin, nontreponemal antibodies detect lipoidal antigens that are present in both treponemal and human cell membranes (28). Production of nontreponemal antibodies is likely stimulated by both immune recognition of treponemal lipoidal antigens and host tissue damage that occurs during infection (29, 30). Unlike treponemal antibodies, nontreponemal antibodies decrease in response to treatment as the infection resolves (Fig. 3). Testing for nontreponemal antibodies therefore allows for both detection of active infection and monitoring of treatment response.

Nontreponemal antibodies are detected via flocculation of precipitated complexes that form between antibodies in serum and cardiolipin-lecithin-cholesterol test antigens. The rapid plasma reagin (RPR) test includes activated charcoal that becomes trapped within the antibody-antigen lattice, enabling RPR reactivity to be read macroscopically. The Venereal Disease Research Laboratory (VDRL) test does not contain charcoal and is instead read microscopically (31). These methods are not interchangeable: the RPR test uses higher test antigen concentrations than does the VDRL test, resulting in different endpoint titers for a given sample (31). With its low cost and ability to be easily interpreted in the laboratory, RPR testing is widely employed for syphilis screening.

Regardless of whether nontreponemal antibodies are assessed using manual or automated methods, the endpoint titer should always be reported to enable monitoring of treatment responses and detection of reinfection (18). It is also important to recognize that RPR titers for a given specimen can vary by more than 3 titer units when tested using different automated methods (32) or using different accredited laboratories (33), likely the result of interlaboratory variability in both methodology and subjective positivity thresholds. RPR monitoring of treatment responses should therefore always be conducted in the same performing laboratory.

Nontreponemal test interpretation requires clinical contextualization and considered evaluation of pretest probabilities. Nontreponemal antibodies arise several weeks later than treponemal antibodies, and the RPR and VDRL can lack sensitivity for early infection (34). Nontreponemal titers can also be extremely elevated in primary and secondary syphilis, and approximately 1% of serum samples demonstrate a prozone phenomenon, in which the high antibody levels prevent formation of the lattice network and subsequent flocculation, yielding a falsely negative result (35). While the prozone phenomenon can be resolved with sufficient dilution, the Centers for Disease Control and Prevention do not recommend routine dilution of negative samples, since titers at which the prozone phenomenon occurs vary widely (18). Nontreponemal antibody levels often wane during late latent syphilis infection as organism burden and tissue damage decline, leading to lower test sensitivity compared to earlier disease stages (34). In patients with treponemal serum antibodies and neurologic signs and symptoms, a reactive cerebrospinal fluid (CSF) VDRL, which is likely more sensitive than the CSF RPR test (18, 36), supports but is not necessary for a diagnosis of neurosyphilis. VDRL is the only nontreponemal test cleared for use in CSF by the FDA (18). The combination of CSF VDRL and manual fluorescent treponemal antibody absorption appears to be a more sensitive test for neurosyphilis compared to CSF VDRL alone (Fig. 3).

In response to appropriate treatment, nontreponemal titers should be monitored and are expected to decline at least 4-fold (2 titer units) over the course of 12 months. Any subsequent 4-fold rise in titer is suggestive of reinfection (18). Posttreatment nontreponemal antibodies will usually continue to decline to undetectable levels, termed seroreversion. However, in some individuals, including those with multiple prior syphilis infections, HIV coinfection, and treatment at a later stage of disease, seroreversion may not occur; the nontreponemal titers will decrease at least 4-fold from pretreatment levels but then plateau and remain serofast (Fig. 3).

Interpretation of nontreponemal tests is further complicated by a high level of false reactivity. Humans can possess low levels of natural antibodies to cardiolipin and other bacterial phospholipids, and their production is stimulated in a variety of nonsyphilitic conditions, including HIV and other viral infections, autoimmunity, and pregnancy (37, 38). Since phospholipids (cardiolipin and lecithin) represent 2 of the 3 antigens employed by nontreponemal tests, reactivity can be observed in individuals with antiphospholipid antibodies, even in the absence of syphilis infection. While these biologic false positives are thought to occur in less than 1% of individuals, large population-based surveys find that they constitute as much as 11% and 26% of positive RPR and VDRL tests, respectively (39, 40). Biologic false-positive reactivity is usually low titer and can be resolved by repeating nontreponemal testing in 2 to 4 weeks to observe whether the titer is stable or has increased 4-fold, suggesting true infection.

In cases of HIV coinfection, there is a risk of more rapid progression to disseminated syphilis, including neurosyphilis, among people who are not on antiretroviral therapy and/or who have CD4 counts less than 350 cells/mm³. Although treponemal antibody titers can wane over time in individuals with HIV, and some individuals also remain serofast after treatment, these phenomena are not specific to HIV coinfection. Therefore, interpretation of treponemal and nontreponemal tests should not differ between individuals with HIV and the general population (18).

Given the recent surge in congenital syphilis infection, particular care should be paid when interpreting nontreponemal titers during pregnancy. While pregnancy can cause both treponemal and nontreponemal biologic false-positive reactivity, a decision to repeat nontreponemal testing in 2 to 4 weeks must be weighed against risk factors for true infection, the risk of loss to follow-up, and the risk of delayed treatment. In response to treatment, nontreponemal antibodies in pregnant individuals can decrease slowly: by the time of delivery, a 4-fold decrease in titer is achieved by fewer than half of pregnant patients who had latent disease, despite adequate treatment (41). Treatment for syphilis can also induce a transient increase in nontreponemal antibodies, although a sustained rise in titer indicates inadequate treatment, treatment failure, or reinfection (42). When there is a concern for congenital infection of the neonate, both maternal and neonatal quantitative nontreponemal serum antibody titers should be obtained, with testing performed concurrently at the same laboratory to enable direct comparison. Although maternal nontreponemal IgG is passively transferred to the fetus, a neonatal titer that is 4-fold higher than the maternal titer is highly suggestive of congenital syphilis infection (42).

Forward and Reverse Testing Algorithms

The mainstay of screening for syphilis infection is the stepwise, algorithmic application of both treponemal and nontreponemal serologic tests. Two testing algorithms are currently in use, now termed traditional (or forward) and reverse. The forward algorithm was implemented at a time when treponemal tests were both labor-intensive and costly (18) and screens first with a nontreponemal test, such as a manual RPR. Only RPR-reactive samples are subsequently tested for treponemal specificity, usually with the TP-PA assay. This approach identifies individuals with active infection and avoids unnecessary treponemal testing of individuals with a history of treated syphilis who no longer have active disease. It also can be cost- and time-effective in the public health setting, where a higher percentage of tested samples are truly positive for syphilis (43). However, the forward algorithm also results in treponemal testing of samples with biologic false-positive nontreponemal reactivity, which may represent a nonnegligible proportion of all reactive nontreponemal tests, depending on the patient population being served. It also relies upon nontreponemal tests to detect infection, which have poorer sensitivity for early and late disease.

With the advent of cost-effective, automated treponemal immunoassays, an increasing number of laboratories have moved to the reverse algorithm. In the reverse algorithm, samples are first screened by treponemal immunoassay, and those with reactivity are then tested for nontreponemal antibodies to identify active infection. A key limitation of the forward algorithm is that in populations with low syphilis prevalence, there is a high rate of false-positive treponemal immunoassay results that are not reproduced upon confirmatory TP-PA testing (22). Discordant results must therefore be arbitrated using a manual treponemal test that uses different methodology and target antigens, usually the TP-PA. Another major limitation of the reverse algorithm is that RPR titer and particularly confirmatory TP-PA testing are often send-out tests that can take days to weeks to result. Moreover, reference laboratories may repeat the entire screening algorithm, leading to duplicate treponemal and nontreponemal test results in the medical record that may or may not agree. This can cause clinician confusion and further delay timely diagnosis and treatment. Lastly, treponemal testing in populations with high rates of prior treated syphilis has low clinical utility and unnecessarily extends turnaround time for the more informative nontreponemal result (44). For patients known to have previously treated syphilis, quantitative RPR is the correct first test.

Direct comparisons of the forward and reverse algorithms find good general concordance: active infections are equivalently detected, despite the fact that both approaches yield some proportion of falsely reactive results (45). In an era of widespread medical laboratory staffing shortages, the forward algorithm using manual RPR testing is not feasible in laboratories that screen large numbers of specimens. However, automated quantitative RPR analyzers have recently become FDA cleared, and they may provide a cost-efficient path to support efficient and timely forward algorithm testing in the future. The sensitivity and specificity of these automated methods are comparable to, and sometimes exceed, that of manual RPR testing (46, 47). These high-throughput platforms can decrease turnaround time, minimize manual data entry, and ostensibly reduce the inherent subjectivity of assessing for flocculation. However, several limitations exist. Automated platforms cannot perform end-point titer determination for extremely high-titer samples, requiring laboratories to modify the method and validate up-front sample dilution. Moreover, quantitative RPR titer differences have been observed between manual and automated methods (46) and between different automated methods (32). If a laboratory adapts an automated RPR method, it is critical that it thoroughly evaluates how the automated titers compare to those generated by its current method. As patient RPR titers are often trended over time, expected differences resulting from the method change should be clearly communicated to the clinicians who will be interpreting the results.

The inherent limitations of the best available diagnostic methods have prompted calls for the development of improved assays (18). Although these efforts are welcome, few new diagnostics are likely to emerge in the next several years. Molecular tests with exceptional analytical sensitivity, for example, have been available on central laboratory instruments for at least 4 years for “research use only” (48, 49) but have not received approval from the FDA.

POC Testing

Several POC syphilis diagnostic tests have become available to address the need for accessible testing in clinical and community settings where rapid and actionable results may benefit patients. In the United States, the FDA has cleared the Syphilis Health Check (Diagnostics Direct, LLC), the Dual-Path Platform HIV-Syphilis System (Chembio Diagnostic Systems, Inc.), and the First to Know system (NOWDiagnostics) (18, 48, 50). These devices are all rapid membrane immunochromatographic assays that detect treponemal antibodies from fingerstick or venous whole blood, serum, and/or plasma (18, 50). Similar Conformité Européene-marked devices have been used in other countries (51). Several studies have indicated that ease of use and minimal cost has facilitated testing, screening, and faster time to treatment, particularly for pregnant persons to prevent congenital syphilis and in populations at higher risk for infection and loss to follow-up, such as persons experiencing homelessness, MSM, and adults in custody (52, 53). The First to Know test is approved for at-home use and can be obtained over the counter, increasing availability and accessibility of testing (50).

The reported performance characteristics of these tests have varied from study to study, with sensitivities ranging from 50% to 100% and specificities ranging from 50% to 100% (54, 55). The differences are likely due to variables such as specimen type, operator training and quality assurance protocols, local prevalence and patient populations, and the traditional syphilis assays used to confirm results and determine accuracy (54, 55). As false-positive and false-negative results are possible, it is important to follow up with traditional diagnostic testing and consultation from providers. In the absence of evidence-based guidelines and given the variable and limited data on POC diagnostic performance, we do not reflex positive POC results directly to an RPR test. Instead, it is our practice to follow any POC result with reverse algorithm testing (automated treponemal immunoassay, followed by an RPR test, if positive). Another important caveat is these rapid tests are treponemal assays that can be reactive in those with previously treated syphilis; in these patients, nontreponemal testing is more informative (55). Finally, syphilis is considered a reportable disease in all US states, but undocumented positive tests in at-home and community settings may complicate public health surveillance and interventions (56).

Concerns about the diagnostic performance of POC syphilis testing may be outweighed by the advantages of a timely test result to permit treatment during the same healthcare encounter. The potential to reduce loss to follow-up is particularly important in nontraditional settings such as syringe service programs, carceral settings, and substance use disorder treatment facilities and among people for whom venipuncture is not acceptable. The positive predictive values in diagnosing new cases of syphilis of the Syphilis Health Check ranged from 22.6% in 2 North Carolina sexually transmitted disease clinics (57); 38.5% in a Florida sexually transmitted disease clinic (58); 47.4% in outreach settings among MSM in Denver, Colorado (59); and 50.0% in an emergency department in Detroit, Michigan (60). Denver MSM with a reactive rapid test experienced a significantly lower median time to treatment compared to those who opted for laboratory-based testing (1 day, range: 0–6 days vs. 9 days, range: 7–13 days) (59). In demonstration projects across 4 US jurisdictions, 93 of 1273 (7%) of rapid treponemal tests were reactive (61). Sixty-five (70%) of the 93 people with reactive tests had confirmatory testing, and 34% to 37% of those with reactive testing and 2.5% of those tested were confirmed cases. Test positivity was greatest (14%) among persons experiencing homelessness. Median time to treatment was 1 to 2 days across sites. No other diagnostic methods currently available attain comparable turnaround times (61).

Direct Detection Methods

Darkfield Microscopy and Immunohistochemistry

Darkfield microscopy can be used to visualize treponemes directly from primary syphilis lesions for rapid results and has historically been considered a gold-standard testing method (18). However, this method is no longer widely used due to several limitations: it is technically challenging to perform and maintain expert microscopist competency, it requires prompt examination, specimens must be handled with caution to prevent occupational exposure, and testing is not recommended with oral or rectal specimens where the presence of non-T. pallidum commensal spirochetes can complicate interpretation (18).

Organisms can also be visualized using immunohistochemistry using peroxidase-conjugated avidin-biotin or indirect immunofluorescence (18). These methods have improved sensitivity and specificity over nonspecific silver staining (62, 63). T. pallidum immunohistochemistry staining can, however, cross-react with other spirochetes, including Brachyspira in cases of intestinal spirochetosis (64). Stain results should be interpreted in the context of clinical symptoms and serological testing.

Molecular Detection

Although nucleic acid amplification tests (NAAT) have become the standard of care for other STIs, like chlamydia and gonorrhea, molecular methods remain an adjunct to serologic diagnosis for syphilis despite decades of assay development [reviewed in (65–67)]. Evaluated methods span conventional or real-time polymerase chain reaction (PCR) (48, 68), reverse-transcription PCR (69), transcription-mediated amplification (49), loop-mediated amplification (70), and PCR followed by CRISPR-based signal amplification (71). Multiple targets have been evaluated (65–67); among the most common are polA, tp0574 (Tp47), and ribosomal RNA (49, 72).

Sensitivity and specificity estimates are highly heterogenous (72) and depend critically on the matrix type and stage of disease (see later discussion). In our laboratory, extraction volume is also a critical determinant of sensitivity (data not shown). Clinical sensitivity is limited, likely due to preanalytical challenges sampling infected mucosa and practical limitations on sample volumes, especially for CSF and ocular fluids. Organism burden is highest in secondary syphilis (Table 1) (15, 16); however, estimating organism burden is challenging due to heterogenous analytical methods. Moreover, reported medians may overestimate concentrations by necessarily excluding a significant proportion of syphilis patients with negative molecular test results in whom organism burden is likely below the limit of detection.

To date, all such molecular tests are laboratory-developed tests; none are FDA-approved, none are POC assays, and the few available in the United States are performed at reference laboratories (18), resulting in multiday turnaround times. Ironically, low testing volume limits the adoption or automation of laboratory-developed test syphilis NAATs. Presently, serologic testing is a fraction of the cost of molecular assays and likely has superior clinical sensitivity: 85% to 90% in primary syphilis, nearly 100% in secondary syphilis, and higher than molecular tests in latent disease (18, 19, 34, 48).

Swabs for Primary Syphilis, Secondary Syphilis, and Screening

Molecular testing of swabs includes both targeted sampling of lesional tissue and screening swabs, often comprising paired anorectal and oropharyngeal mucosal swabs. Among lesion swabs, molecular sensitivity estimates range from 65% to 80% (72–74 ). Clinical sensitivity is lower in screening swabs. Exquisitely sensitive assays, such as a transcription-mediated amplification assay (Hologic) with an estimated 95% limit of detection of 1.6 organisms/mL, had a clinical sensitivity of approximately 50% (48). In other studies employing quantitative PCR for tp0574, sensitivity was reported as approximately 40% (16). Despite limitations, molecular testing can augment detection of early-incubating or latent syphilis, raising detection rates by 15% to 20% in some studies (48, 75). Molecular testing of swabs is unlikely to replace serology as a screening modality, but rapid and inexpensive tests could be a useful adjunct to current testing strategies. Additional studies of absolute organism burden (Table 1) in lesions, swab fluid, CSF, and blood will help guide future assay development.

Molecular Testing in Peripheral Blood Samples

Estimated sensitivities of molecular syphilis assay performed in whole blood and peripheral blood mononuclear cell specimens range from 14% to 54% for primary syphilis and 35% to 82% for secondary syphilis (48, 49, 74, 76). Sensitivities for primary and secondary syphilis are generally lower when serum is used, at < 50% (49, 72, 73). Emerging molecular techniques may have increased sensitivity and have utility in secondary syphilis when circulating treponemal burden is likely highest (70, 71). A DNA loop-mediated amplification assay had a reported clinical sensitivity of 82% when DNA was extracted from 1 mL whole blood (EDTA tubes) drawn from a cohort of patients with a high pretest probability of secondary syphilis (70). Similarly, a nested molecular assay comprising PCR and CRISPR-based signal amplification to detect amplified products had reported sensitivities of 71% (5/7) in primary syphilis and 87% (50/57) in secondary syphilis (71 ). These reported sensitivities are higher than those observed in past studies of peripheral whole blood, which reported sensitivities of ∼30% in individuals without a prior history of syphilis and only approximately 15% among patients with a prior infection (74, 77). Nevertheless, compared with conventional serology, these newer molecular assays require extraction and analytical steps that would prolong turnaround time in clinical laboratories.

Neurosyphilis

The organism burden in CSF is low, and multiple studies have reported yielding sensitivity below 50% and often below 30% for molecular detection of neurosyphilis (77, 78). A recent meta-analysis reported a median sensitivity of 54% (79). Prior syphilis further reduces organism burden in both CSF and blood, likely reducing the sensitivity of molecular testing to detect reinfection (77). Additionally, sample volumes of CSF and ocular fluids available for nucleic acid extraction are often very low, further compromising molecular testing in these critical anatomic compartments.

Congenital/Intrapartum Syphilis

Limited data are available on molecular testing to diagnose congenital syphilis. In treponeme-spiked samples, a tp0574-targeted conventional PCR with a limit of detection of about 10 organisms/reaction detected 3/3 cases of congenital syphilis from amniotic fluid (68). Subsequent papers using this method ranged in clinical sensitivity from 27% (80) to 50% (81) to 78% (82). However, yield depended upon the extraction method, and the lowest estimate of sensitivity [27% (80)], employed a less favorable method (82). Few studies describe syphilis NAAT performance in placental or umbilical tissue of congenital cases, and clinical sensitivity estimates vary from 25% (N = 73) (83) to 88% (N = 8) (84). Similar variability in NAAT sensitivity has been reported in formalin-fixed paraffin-embedded tissue samples, with estimates ranging from 38% to 75% (85–87).

Metabolomic and Proteomic Biomarker Discovery

Research to directly detect candidate T. pallidum-derived peptide biomarkers has been stymied by low analyte burden in plasma and urine, resulting in a lack of reproducible detection in clinical samples (88). Therefore, untargeted metabolomics has emerged as an alternative strategy. However, efforts in serum are limited by 3 factors: small study population (N = 20 cases), nonreporting of syphilis clinical stage in study participants, and comparisons to uninfected controls only (89). The latter limitation is noteworthy since the Kyoto Encyclopedia of Genes and Genomes analysis of the identified candidate metabolites indicated involvement in responses to other infections and inflammatory states (89).

Studies in CSF have compared small groups (N < 21 subjects per group) of neurosyphilis patients to uninfected controls (90, 91) and, in one study, to patients with positive treponemal serologies and no evidence of neurosyphilis (90); however, in this latter group, active disease was not confirmed with lipoidal serologic tests in serum (90). Identified candidate metabolites were associated with other inflammatory or neurodegenerative processes and were not reproducible across studies (90, 91).

Future Outlook for Diagnostic Testing of Syphilis

Despite technological advances, serology is likely to remain the bedrock of syphilis screening and diagnostic testing. These approaches leverage easily accessible patient specimens and low-cost, high-throughput methods. However, the complexities of testing algorithms and result interpretation remain very real barriers to rapid syphilis diagnosis and treatment. Just as automated RPR platforms have emerged that enable same-day quantitative results, similar automation of TP-PA testing is under development. Coupling the increased specificity of the TP-PA assay with the rapid turnaround of automation may prove to be a powerful laboratory tool to more efficiently diagnose syphilis. Widespread implementation of automated methods should help improve harmonization of reported nontreponemal titers, enabling more effective monitoring of results when titers are sent to separate reference laboratories or when patients move between medical systems (Fig. 4). There is a relatively new nontraceable World Health Organization biological reference standard for nonlipoidal and lipoidal IgG and IgM antibodies (92). Although adoption of this standard could reduce variability, particularly in manually interpreted assays, few if any manufacturers’ calibrators are traceable to this standard. Thus, widespread adoption may be limited by the regulatory burdens and costs of adopting this significant change to calibrator material.

Fig. 4.

Addressing the syphilis epidemic requires progress on multiple fronts. Clockwise from top left: important laboratory-based interventions include improved turnaround time for clinical results through automated RPR and expanded use of POC tests. Harmonization of titer determination and improving access to testing histories between clinical and public health laboratories would facilitate interpretation of test results. Continued investment in public health infrastructure and harmonization of guidelines (top right) are essential to sustain progress in reducing disease burden. Expanding access to care (middle right), through at-home testing and mobile health clinics similar to the College of American Pathologists’ “See, Test, Treat” program for cancer screening would increase case detection. Multiple, interacting social determinants of health (bottom right) are important risk factors for syphilis and require systemic mitigation strategies. Continued investment in basic research (bottom left) would support vaccine development, biomarker discovery, and drivers of disease pathogenesis and assist with global and local epidemiologic studies. Technological advances (middle left) that promote FDA approval of rapid, sensitive, and inexpensive tests would increase detection of early and latent disease, thus helping to reduce syphilis’s spread. New methods are required to monitor and diagnose emergent antibacterial resistance in T. pallidum isolates. Created in BioRender. Lieberman, J. (2025) https://BioRender.com/l88a783. Abbreviations: ACOG, American Congress of Obstetricians and Gynecologists; LBGTQ, lesbian, gay, bisexual, transgender, queer or questioning; USPSTF, United States Preventive Services Task Force. Color figure available at https://academic.oup.com/clinchem.

The development of POC nontreponemal assays and the targeted expansion of POC testing could also facilitate rapid clinical decision-making. Similarly, reducing barriers to data access/sharing between clinical and public health labs would improve patient care by making prior reported testing accessible to providers and facilitating correct syphilis serology ordering and result interpretation (Fig. 4). In the interim, there is a continued need to educate laboratorians and ordering providers of the interlaboratory variability in nontreponemal titers and the lack of clinical significance of an isolated 2-fold change in nontreponemal titer.

On the research front, other T. pallidum antigens are being evaluated that, either alone or in combination, can improve the diagnostic performance of immunoassays (93, 94). Work to directly detect T. pallidum metabolites as more specific biomarkers for disease has largely failed due to low analyte concentration (88). Limited studies on host-derived metabolites have neither established whether candidate biomarkers are syphilis specific nor evaluated clinical sensitivity (89–91). Molecular assays with high analytical sensitivity for the bacterium have been developed; however, use cases are limited as current T. pallidum NAATs lack sufficient sensitivity to rule out disease (16, 48, 67, 95). Inexpensive, rapid assays could augment detection of early or latent infections, while laboratory-developed tests can help determine the etiology of lesions in disseminated syphilis, including congenital syphilis. Molecular testing serves important roles, permitting epidemiologic surveillance and monitoring of antibiotic resistance (17).

Doxy-PEP is emerging as an effective tool to contain syphilis and other bacterial STIs (13, 14); however, its increasing use raises important questions regarding laboratory diagnostics in the coming years. It remains to be seen how doxy-PEP affects the immune responses that develop following T. pallidum exposure and how this will impact the performance of serologic tests for syphilis diagnosis, treatment monitoring, and distinguishing previous infection from reinfection. Similarly, whether widespread use of doxy-PEP will result in T. pallidum resistance to tetracycline antibiotics is unknown. Neither doxycycline treatment failures nor mutations that would confer resistance to doxycycline have been detected yet, and resistance could not be induced during in vitro cultivation (17, 96). Nonetheless, the increased utilization of doxycycline due to doxy-PEP and current penicillin shortages warrants active surveillance for drug resistance. This may be best achieved via molecular assays. Finally, equipoise remains regarding the expansion of doxy-PEP to women for syphilis prevention in heterosexual populations (14). This has the potential to stem the exponential rise of congenital syphilis in the United States, but the impact on serological assay performance in pregnant people and neonates should be assessed.

In an era of multidrug-resistant pathogens, it seems implausible that a bacterium universally susceptible to penicillin and dependent upon a human host could cause >200 000 cases in the US annually, including nearly 4000 congenital cases and hundreds of infant/fetal deaths. Improved access to prompt laboratory testing comprises a partial response to social determinants of health—poverty, substance use, homelessness, structural racism/homophobia, recurrent shortages of penicillin, or public health disinvestment—that contribute to syphilis’s resurgence (Fig. 4) (2–5, 97, 98). Moreover, the stealthy pathogenesis of T. pallidum and low analyte concentrations make diagnostic innovation challenging. Nonetheless, there are significant opportunities to advance laboratory diagnostics that support recent public health successes in blunting the epidemic (Fig. 4).

Supplemental Material

Supplemental material is available at Clinical Chemistry online.

Nonstandard Abbreviations

MSM, men who have sex with men; P&S, primary and secondary; HIV, human immunodeficiency virus; doxy-PEP, doxycycline as post-exposure prophylaxis; STI, sexually transmitted infection; AI/AN, American Indian/Alaska Native; POC, point-of-care; IgM, immunoglobulin M; IgG, immunoglobulin G; TP-PA, T. pallidum particle agglutination test; RPR, rapid plasma regain; VDRL, Venereal Disease Research Laboratory; CSF, cerebrospinal fluid; NAAT, nucleic acid amplification tests; PCR, polymerase chain reaction.

Author Contributions

The corresponding author takes full responsibility that all authors on this publication have met the following required criteria of eligibility for authorship: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved. Nobody who qualifies for authorship has been omitted from the list.

Rebecca Treger (Conceptualization-Equal, Project administration-Equal, Writing—original draft-Lead, Writing—review & editing-Lead), Tim Menza (Conceptualization-Equal, Project administration-Supporting, Writing—original draft-Equal, Writing—review & editing-Equal), Thao Truong (Conceptualization-Supporting, Writing—original draft-Supporting, Writing—review & editing-Supporting), and Joshua Lieberman (Conceptualization-Equal, Project administration-Equal, Writing—original draft-Equal, Writing—review & editing-Equal)

Authors’ Disclosures or Potential Conflicts of Interest

Upon manuscript submission, all authors completed the author disclosure form.

Research Funding

T.W. Menza receives support from grants: 2024–2027 Sexually Transmitted Infection Integrated Research Consortium, Centers for Disease Control and Prevention, Grant Number 75D30123D15973, and 2024–2029 Support and Scale Up of HIV Prevention Services in Sexual Health Clinics, Centers for Disease Control and Prevention, Grant Number NH25PS2024003265. J.A. Lieberman is supported by the following grants: Centers for Disease Control and Prevention contract, Project 75D30122C15478; National Institute of Allergy and Infectious Diseases, R21 AI85726-01; National Institute of Allergy and Infectious Diseases, R21 AI184484-01; RADx Independent Test Assessment Program for Monkeypox-Validation Center (Task Order #5, 75N92023F00002 under Federal Contract 75N2022D00015) and reports reagent and equipment support from ThermoFisher, Qiagen, Anavasi Diagnostics, and Roche.

Disclosures

R. Treger has received grant funding from the Canadian Institutes of Health Research Sexually Transmitted and Blood-Borne Infections Research in Canada: Beyond HIV/AIDS and Hepatitis C Award; is a member of the College of American Pathology Diagnostic Immunology and Flow Cytometry Committee and has received support for travel, room, and board at committee meetings; and is an Education Committee member for the Association of Medical Laboratory Immunologists. J.A. Lieberman is listed as an inventor on Patent Application PCT/US2023/073181, Patent Application No. 63/402,787.