In this cross-sectional study, we included men and women who reported sexual contact with a person with syphilis infection (hereafter, syphilis contacts), provided consent to having PCR testing for syphilis, and visited the Melbourne Sexual Health Centre (MSHC) during November 2018–March 2020. The study was approved by the Alfred Hospital Ethics Committee, Melbourne, Australia (project no. 474/18). MSHC is a public sexual health and HIV clinic in the state of Victoria, Australia, and provides ≈50,000 consultations/year. The clinic has an electronic medical record system that stores demographic and epidemiologic data.

Clients seeking care at the clinics were evaluated by a nurse after their registration. The nurse typically collects a brief account of the clients’ current condition and adds this information to their medical records before assigning them to healthcare professionals. Persons who had been in sexual contact with someone diagnosed with syphilis were identified on the basis of their own report. Those persons were notified about their exposure through either anonymous text messages or direct communication from the persons who had been in contact with them.

Contacts of syphilis were provided with a plain-language participant information sheet explaining PCR testing from mucosal sites. Verbal consent was obtained from those who agreed to have the T. pallidum PCR tests from the mucosal sites. The clinicians collected oral swab samples (all participants), anal swab samples (MSM only), and vaginal swab samples (women only) by using the Universal Transport Medium swab (Copan Italia, https://www.copangroup.com) for T. pallidum PCR testing. Participants also self-collected an oral rinse by gargling 10 mL of sterile water for T. pallidum PCR (i.e., mucosal screening PCR tests). T. pallidum PCR testing also was performed on swab samples taken from any syphilis lesions present (i.e., lesion PCR tests).

Participants who reported having symptoms suggestive of syphilis, such as anogenital or oral lesions, were examined by the clinician at the respective sites, and swab samples were taken from any lesion for T. pallidum PCR testing. Vaginal examinations were performed by using speculum if a woman reported any genital lesions or symptoms, and anoscopic examination was performed where a participant reported an anorectal lesion. Whether or not a participant required vaginal speculum or anoscopic examination was a decision made by the clinician on clinical grounds. We collected data on clinical examination findings retrospectively.

All participants had syphilis serologic testing performed and were offered syphilis treatment with intramuscular benzathine penicillin (2.4 mU single dose) or doxycycline (100 mg 2×/d for 2 weeks) for those with penicillin allergy. The participants were offered routine chlamydia and gonorrhea testing according to sexual transmitted infection (STI) testing guidelines depending on their sexual risk (25,26). If persons were diagnosed with syphilis infection and considered to have latent syphilis of unknown duration, they were asked to return for further treatment with a total of 3 doses of weekly benzathine penicillin (or 4 weeks of doxycycline for those with penicillin allergy).

We defined MSM as men who have sex with men or with transwomen and bisexual as either men or women who had sex with both men and women (including transgender persons). We defined heterosexual as either men or women who had sex only with the opposite sex. We defined sexuality on the basis of self-reported sexual practice in the last 12 months and not on sexual identity.

E.T.A. reviewed medical records to capture information on types of partners (regular vs. casual), type of sex (anal, vaginal, or oral sex), condom use (condomless vs. with condoms), and signs and symptoms of syphilis at consultation. The information on types of partners was self-reported, and no formal definition of regular and casual partners was used in the study. Some participants might have defined regular partners as romantic partners (e.g., boyfriends or husbands) or casual sexual partners as regular contacts without romantic attachment (24).

A total of 407 contacts had >1 specimen for PCR screening (oral and anal in MSM, oral in heterosexual men, and oral and vaginal in women). Of the 407 contacts, 339 were MSM, 22 were heterosexual men, 20 were heterosexual women, 1 was a bisexual woman, and 25 were bisexual men. Among the contacts living with HIV (16%, n = 67), most were MSM (94%, n = 63); one third of contacts (33%, n = 134) were using HIV preexposure prophylaxis. One fifth of the contacts (20%, n = 85) had a history of syphilis infection, and a small number (9%, n = 35) reported ever injecting drugs. The median age of the syphilis contacts was 32 years (IQR 27–40 years).

Nearly half (47%, n = 193) of the contacts reported having a casual partner as their syphilis contact, whereas 35% (n = 144) reported a regular partner as the contact. Most (71%, n = 290) contacts reported condomless sex during anal or vaginal sex, and 17% (n = 69) had signs or symptoms suggestive of syphilis (Table 1).

Figure

Figure. Flowchart of syphilis infections among men and women who had contact with persons who had syphilis, categorized by positive PCR and serologic test results and presence or absence of symptoms,...

A total of 42 contacts (10%, 95% CI 8%–14%) had syphilis infection diagnosed (case-patients) (Figure). Of those, 33 were MSM, 5 were women, 2 were heterosexual men, and 3 were bisexual men. Syphilis was detected by positive serologic testing alone in the absence of positive PCR in 21 cases (50%, 95% CI 34%–66%). Nineteen cases (45%, 95% CI 30%–61%) were detected by positive serologic tests together with positive T. pallidum PCR; 9 of those case-patients had T. pallidum detected by PCR from >1 mucosal sites (oral, anal, or vaginal) in the absence of lesions at these sites (Table 2). The remaining 10 of the 19 case-patients had signs and symptoms suggestive of syphilis and were T. pallidum–positive from mucosal sites, lesions sites, or both (Table 3). Of the 42 syphilis case-patients, 2 had T. pallidum PCR detected from penile lesions with a negative serologic test. Because those 2 case-patients had T. pallidum PCR detected only from lesion PCR tests, they are not discussed further.

In this study of men and women reporting sexual contact with a person with syphilis, we did not identify any persons with PCR detection of T. pallidum from oral, anal, or vaginal sites where serologic testing was negative, indicating that using PCR for screening of syphilis contacts at mucosal sites might not provide any additional benefit over existing syphilis screening using serologic testing. However, we found a proportion of men and women who tested positive for T. pallidum by PCR from the oral cavity, anus, or vagina in the absence of signs or symptoms of syphilis at those sites or elsewhere. In some cases, serologic testing for syphilis was positive in the absence of negative serologic testing within the previous 2 years. Those persons would conventionally be staged as having latent syphilis of unknown duration, and they probably were treated for possible late latent infection. However, detection of T. pallidum by PCR from oral, anal, or vaginal sites, in conjunction with the syphilis contact status, suggests that those infections probably were early asymptomatic infections.

The findings indicate that PCR-based testing for T. pallidum at mucosal sites may be useful in assisting with staging of syphilis infection. Consequently, this approach could provide guidance to clinicians and patients regarding treatment duration. Furthermore, T. pallidum PCR testing at mucosal sites has the potential to aid in partner notification, particularly in cases where an early infectious syphilis is diagnosed and the duration of infection is uncertain. T. pallidum PCR may complement current methods of staging, which rely on various factors, including the presence of signs of early syphilis, the patient’s history of syphilis and treatment, sexual history (including contact with a partner with syphilis infection), and past and current laboratory results (including serologic testing and direct detection methods using molecular assays).

We identified several asymptomatic women who had positive syphilis serologic testing and PCR detection of T. pallidum from the mucosal screening sites (the vagina, oral cavity, or both). Those women had never been serologically tested for syphilis before or had been tested >2 years previously. We also identified several asymptomatic MSM who had positive syphilis serologic testing and PCR detection of T. pallidum from the anus, oral cavity, or both. In contrast to the women, all those men had been serologically tested for syphilis within the previous 2 years. The difference in serologic testing between women and men reflects frequent serologic screening for syphilis being well established among MSM in Australia but less so among women in urban centers because syphilis has only emerged as a major public health concern among women in that setting in recent years.

Several studies have examined the role of NAAT in syphilis screening (20,29). A US study compared the use of transcription-mediated amplification (TMA) assay performed on rectal and pharyngeal mucosa in MSM with routine serologic testing and found 2 additional syphilis cases diagnosed on TMA testing before positive serologic testing (29). Patients in both cases had TMA detection in rectal swabs: 1 did not have symptoms, and the other had anal symptoms. In contrast, our study did not identify syphilis cases diagnosed on T. pallidum PCR from mucosal sites before positive serologic testing.

In our study, we found that 40% of the 42 confirmed early syphilis cases among men and women had PCR detection of T. pallidum from >1 of the mucosal sites (oral, anus, and vagina), with or without lesions at these sites. Of note, we show PCR detection of T. pallidum in asymptomatic women. T. pallidum detection by PCR in women has been shown in other studies but predominantly from genital lesions (16,30–32). The presence of T. pallidum at those mucosal sites may represent the site of inoculation (and hidden primary lesions) or dissemination from a distant site (19). Several other recent studies showed that T. pallidum can be detected from the oral cavity or anus of men and women with early syphilis infection (18–23). Detection of T. pallidum by PCR was reported in ≈24% of 200 MSM with confirmed early syphilis infection with or without lesions at the oral cavity and anus (18). Those studies suggest that syphilis probably is infectious from oral, anal, and vaginal sites in the absence of local signs and symptoms.

The first limitation of our study is that it was terminated prematurely because of the COVID-19 pandemic, which limited the sample size. The number of women in the study was small because the clinic population was predominantly male, and fewer women attended. The sexual partners reported by contacts might not have actually had syphilis, given that we were unable to confirm their diagnosis; if so, we may have overestimated syphilis infections in this study group. Not all contacts who self-reported that they were asymptomatic had examination of oral, vaginal, and anal sites by the clinician (46% did not undergo examination), which could have led to occult lesions at those sites being missed. The T. pallidum PCR we used might have lower sensitivity than other T. pallidum molecular assays, such as TMA, resulting in a lower number of PCR-positive mucosal screening tests (29,33). Further, current data are not sufficient to indicate that T. pallidum PCR positivity at a mucosal site is a proof of early syphilis infection. T. pallidum might reflect contamination from very recent sexual intercourse with a person with syphilis infection, such as residual semen at the mucosal site.

Overall, we conclude that T. pallidum PCR screening from mucosal sites (oral, anus, and vagina) may not have added benefit over screening using serologic testing. It may, however, have a role in assisting with syphilis staging, particularly in the absence of syphilis lesions. A positive PCR result from asymptomatic mucosal sites may help identify early infections in persons who would otherwise be classified as having latent syphilis of unknown duration. However, such an interpretation requires a correlation with additional sexual behavioral information and the broader clinical context. T. pallidum PCR screening at mucosal sites may be especially relevant in populations that do not undergo regular syphilis screening with serologic testing.