Treponema pallidum Nucleic Acid Amplification Testing To Augment Syphilis Screening among Men Who Have Sex with Men

Syphilis rates in much of the world are now at their highest levels in almost three decades, and new approaches to controlling syphilis, including diagnostic tests with shorter window periods, are urgently needed.

ABSTRACT

Syphilis rates in much of the world are now at their highest levels in almost three decades, and new approaches to controlling syphilis, including diagnostic tests with shorter window periods, are urgently needed. We compared the sensitivity of syphilis serological testing using the rapid plasma reagin (RPR) test with that of the combination of serological testing and an experimental 23S rRNA Treponema pallidum real-time transcription-mediated amplification (TMA) assay performed on rectal and pharyngeal mucosal swabs. T. pallidum PCR assays for the tpp47 gene were performed on all TMA-positive specimens, as well as specimens from 20 randomly selected TMA-negative men. A total of 545 men who have sex with men (MSM) who were seen in a sexually transmitted disease clinic provided 506 pharyngeal specimens and 410 rectal specimens with valid TMA results. Twenty-two men (4%) were diagnosed with syphilis on the basis of positive RPR test results and clinical diagnoses, including 3 men with primary infections, 8 with secondary syphilis, 9 with early latent syphilis, 1 with late latent syphilis, and 1 with an unstaged infection. Two additional men were diagnosed based on positive rectal mucosal TMA assay results alone, and both also tested positive by PCR assay. At least 1 specimen was TMA positive for 12 of 24 men with syphilis (sensitivity, 50% [95% confidence interval [CI], 29 to 71%]). RPR testing and clinical diagnosis were 92% sensitive (95% CI, 73 to 99%) in identifying infected men. Combining mucosal TMA testing and serological testing may increase the sensitivity of syphilis screening in high-risk populations.

INTRODUCTION

Rates of syphilis in the United States have risen steadily over the past 2 decades, and high-income nations throughout the world are now experiencing the highest rates of syphilis observed since the early 1990s (1–3). The current syphilis epidemic primarily affects men who have sex with men (MSM), and in some areas an estimated 1% of all MSM, and almost 5% of HIV-infected MSM, are diagnosed with early syphilis each year (4).

Public health authorities have yet to develop an effective response to the contemporary syphilis epidemic. In large part, that failure reflects the paucity of innovation in available tools to prevent, to diagnose, or to treat syphilis. Clinicians mostly rely on decades-old diagnostic tests and treatments that have only marginally improved since first half of the 20th century (5). In the area of diagnostic testing, Treponema-specific immunoassays have replaced the rapid plasma reagin (RPR) test as the initial test for syphilis in many settings (5). While some of these assays are more sensitive than the RPR test (6), others are less sensitive (7) and many are less specific. The incubation period from the initial Treponema pallidum infection to the development of a chancre, the lesion associated with primary syphilis, is usually said to vary from 10 to 90 days (8), with experimental studies of T. pallidum inoculation suggesting a mean incubation period of 3 to 4 weeks (9, 10). The sensitivity of serological testing in primary infections varies widely, i.e., from 50 to 90%, depending on the test used and the population studied (6, 7, 11–14). Thus, in many cases, there is a substantial period between the time of infection and the time at which infections are detectable using available laboratory tests. The extent to which persons are infectious during this window period is not known.

A more sensitive test for early syphilis could play an important role in controlling the syphilis epidemic. Nucleic acid amplification tests (NAATs) for gonorrhea and chlamydia infections are substantially more sensitive than culture and are now the primary diagnostic tests used for those infections. In some areas, clinicians and health departments have used HIV NAATs (which have a shorter window period than HIV enzyme immunoassays [EIAs], particularly third-generation EIAs) to screen patients for HIV infection, increasing the identification of persons with acute and early infections (15, 16). Although no NAAT for T. pallidum has been approved for use by the U.S. Food and Drug Administration, experimental NAATs performed on genital ulcer, blood, urine, or mucosal swab specimens are sometimes positive in infected persons with negative syphilis serological results (17–23). We used an experimental transcription-mediated amplification (TMA) assay targeting the 23S rRNA of T. pallidum to test rectal and pharyngeal mucosal swab specimens taken from MSM who were sexually transmitted disease (STD) clinic patients. This was done as a first step in evaluating whether a syphilis screening strategy that combined NAATs and serological assays might be more sensitive than serological assays alone for the diagnosis of early syphilis.

MATERIALS AND METHODS

Study design.

We conducted a cross-sectional study evaluating an experimental T. pallidum TMA NAAT using residual rectal and pharyngeal swab specimens obtained from MSM patients at the Public Health–Seattle & King County (PHSKC) STD clinic between September and November 2017. The PHSKC STD clinic is located on the Harborview Medical Center campus in Seattle, Washington, and provided care during 10,895 visits in 2017; 53% of all clinic visits were by MSM. Data from clinic visits are collected using a combination of computer self-interviews by patients and information recorded by clinicians in the electronic medical record. Final syphilis staging of all cases is performed by a single, highly experienced, disease intervention specialist.

Clinic policy is to obtain rectal specimens for gonorrhea and chlamydia testing from all MSM who report any receptive anal sex in the prior year and to obtain pharyngeal specimens from all men who report performing fellatio in the prior 2 months. Specimens are obtained both by clinician collection and by patient self-collection and are tested for gonorrhea and chlamydia infections using the Aptima Combo 2 assay (Hologic, San Diego, CA), as part of routine clinical care. Clinicians also routinely obtain serum samples to test all MSM patients for syphilis. Laboratory testing for syphilis is performed in the PHSKC laboratory using the RPR test (Beckton Dickinson, Franklin Lakes, NJ). Specimens with positive RPR test results are sent to the Washington State Department of Health laboratory for confirmatory testing using the Captia syphilis IgG EIA (Trinity Biotech, Wicklow, Ireland). Specimens with negative Captia EIA results are tested using the Serodia T. pallidum particle agglutination (TPPA) assay (Fujirebio Inc., Tokyo, Japan).

As part of the study, laboratory technicians removed patient identifiers from consecutively tested Aptima multitest swab specimen transport tubes after completion of gonorrhea and chlamydia testing. They then labeled the tubes with a study identification number and stored them at −20°C. Staff members created a list linking study identification numbers to patient identification numbers to allow TMA and PCR test results to be linked to electronic patient records. The goal of the study was to obtain 500 pharyngeal specimens and 500 rectal specimens. The University of Washington Office of Human Subjects institutional review board approved all study procedures.

Laboratory testing.

A research-use real-time TMA assay was developed to capture, to amplify, and to detect 23S rRNA of the Nichols strain of T. pallidum subsp. pallidum using Aptima-format reagents on the automated Panther system (Hologic). Validation studies using T. pallidum organisms quantitated by dark-field microscopy yielded a probit-derived 95% limit of detection (LOD) of 1.39 T. pallidum organisms/ml (95% confidence interval [CI], 0.705 to 6.328 organisms/ml) in Aptima STM matrix. Confirmatory LOD studies using T. pallidum organisms diluted in pooled anorectal swab specimen clinical matrix yielded 100% positive results at 1.0 organism/ml (n = 5). Analytical specificity studies using 52 nontarget bacterial, fungal, and viral species at high titer (1 × 10⁵ to 1 × 10⁶ CFU/ml or 50% tissue culture infective dose [TCID₅₀]/ml) showed 0% cross-reactivity (n = 3) for all organisms, including the closely related treponeme Treponema denticola tested at a concentration of 1 × 10⁶ organisms/ml (data not shown).

For patient specimen testing, deidentified residual rectal and pharyngeal specimens were tested in the Hologic laboratory using the research-use T. pallidum TMA assay on the Panther system. All swabs were originally placed in 2.9 ml of transport medium, of which 400 μl was used for TMA testing; as with other mucosal swab tests, the amount of material absorbed on swabs was not standardized. Operators were blinded to all patient-related information, including STD and HIV infection status and syphilis serological results. All TMA-positive specimens, as well as specimens from 20 randomly selected patients with negative TMA test results, were also tested using a research-laboratory-based T. pallidum PCR assay for the tpp47 gene (24).

Analysis.

We assessed the sensitivity of serology alone, clinical diagnosis plus serology, and clinical diagnosis plus serology and TMA assay for the diagnosis of syphilis. We defined patients as having syphilis if they had a clinical diagnosis supported by a newly positive serological test result or a positive TMA assay result or if they had an isolated positive TMA assay result. We calculated 95% CIs for the sensitivity of each approach using exact confidence limits for binomial proportions. All analyses were conducted using SAS (Cary, NC).

RESULTS

Laboratory staff members tested a total of 965 specimens from 545 MSM, including 528 pharyngeal specimens and 441 rectal specimens, of which 506 (96%) and 410 (93%), respectively, yielded valid TMA assay results. A total of 534 men had at least one valid TMA assay result. The median age of the men in the study population was 33 years (range, 18 to 74 years). HIV status information was available for 412 men, of whom 53 (13%) were known to be HIV positive.

Of the 534 men with valid results, 22 (4.2%) were diagnosed with syphilis during their clinic visit, including 3 men with primary syphilis, 8 with secondary syphilis, 9 with early latent syphilis, 1 with late latent syphilis, and 1 for whom syphilis was diagnosed but not staged (Table 1). Nineteen (90%) of the 21 men with a clinical diagnosis of syphilis had positive RPR test results. One man with a negative RPR test result (case 6) was a syphilis contact and was staged as having secondary syphilis, with positive TPPA assay and EIA results. Another RPR-negative man (case 15) was also a syphilis contact and was staged as having early latent infection, with a positive TPPA assay result. A third man (case 22) did not have a clinical diagnosis of syphilis at the time of his medical evaluation and had missing staging information but had a positive RPR test result at a 1:128 dilution. Two additional men (cases 23 and 24) had positive TMA assay results in the absence of a positive serological test or a clinical diagnosis of syphilis. Both had positive TMA and PCR test results for rectal specimens, and 1 (case 24) sought care after known exposure to syphilis. Case 24 also had a rectal ulcer on physical examination, although this was not recognized as a chancre and did not lead to a syphilis diagnosis at the time of his medical evaluation.

TABLE 1.

Clinical diagnosis, physical findings, and RPR, TMA, and PCR test results for 24 men with syphilis

Case no.	Clinical diagnosis	Known syphilis contact	RPR titer	TMA assay resulta		PCR assay result		Rectal and pharyngeal physical findings
				Rectal	Pharyngeal	Rectal	Pharyngeal
1	Primary	No	1:4	Neg	ND	Neg	ND	None
2	Primary	No	1:256	Pos	Pos	Pos	Neg	None
3	Primary	No	1:16	Invalid	Neg	Pos	Neg	Rectal erythema and discharge
4	Secondary	No	1:2	Neg	Neg	Neg	Neg	None
5	Secondary	No	1:32	Pos	Pos	Pos	Pos	Purulent rectal discharge
6	Secondary	Yes	Negb	Neg	Neg	Neg	Neg	Oral mucous patches
7	Secondary	No	1:32	Pos	Pos	ND	Neg	Rectal skin tag
8	Secondary	No	1:128	Neg	ND	Neg	ND	None
9	Secondary	No	1:128	Neg	Pos	Neg	Pos	None
10	Secondary	No	1:2	Neg	Neg	Neg	Neg	None
11	Secondary	No	1:128	Pos	ND	Pos	ND	None
12	Early latent	No	1:1	Neg	Neg	Neg	Neg	None
13	Early latent	No	1:128	Pos	Pos	Neg	Pos	None
14	Early latent	No	1:16	Neg	Neg	Neg	Neg	None
15	Early latent	Yes	Negc	Neg	Neg	Neg	Neg	None
16	Early latent	No	1:32	Pos	Neg	Pos	Neg	None
17	Early latent	No	1:64	Neg	Neg	Neg	Neg	None
18	Early latent	No	1:256	Pos	Neg	Pos	Neg	None
19	Early latent	Yes	1:32	ND	Neg	ND	Neg	None
20	Early latent	No	1:2	ND	Neg	ND	Neg	None
21	Late latent	No	1:64	Pos	Pos	Pos	Neg	None
22	Missing	No	1:128	Neg	Pos	Neg	Neg	None
23	None	No	Negd	Pos	Neg	Pos	Neg	None
24	None	Yes	Nege	Pos	Neg	Pos	Neg	Rectal ulcerf

^aNeg, negative; Pos, positive; ND, not done.

^bPatient was TPPA assay and EIA positive with no known history of syphilis.

^cPatient was TPPA assay positive with no EIA performed and no known history of syphilis.

^dTPPA assay was not done.

^eTPPA assay was negative.

^fPatient had negative Chlamydia trachomatis TMA test results and tested negative for herpes simplex virus by PCR assay.

Twelve (2.2%) of 534 tested men had at least one positive TMA assay result for T. pallidum. Seven (1.4%) of 506 pharyngeal specimens and 10 (2.4%) of 410 rectal specimens tested positive by TMA assay (Table 2). TMA assay results were positive for 10 (45% [95% CI, 32 to 76%]) of 22 men with a clinical diagnosis of syphilis (including case 22). Assuming that the 2 men who tested positive by TMA assay in the absence of a clinical diagnosis of syphilis or positive RPR test results were infected with T. pallidum, 24 men had syphilis. Among those 24 cases, RPR testing alone was 83% sensitive (95% CI, 63 to 95%) in the diagnosis of syphilis and clinical diagnosis plus serology was 92% sensitive (95% CI, 73 to 99%).

TABLE 2.

Rectal and pharyngeal TMA test results by syphilis diagnosis based on clinical presentation and serology

Serological results and clinical diagnosis	TMA test results (n = 534) (no. [%])
	Rectal				Pharyngeal				Rectal or pharyngeal
	Positive (n = 10)	Negative (n = 396)	Not valid (n = 20)a	Not done (n = 108)	Positive (n = 7)	Negative (n = 500)	Not valid (n = 11)a	Not done (n = 16)	Positive (n = 12)	Negative (n = 522)
Primary (n = 3)	1	1	1	0	1	1	0	1	1 (33)	2
Secondary (n = 8)	3	5	0	0	3	3	0	2	4 (50)	4
Early latent (n = 9)	3	4	0	2	1	8	0	0	3 (33)	6
Late latent (n = 1)	1	0	0	0	1	0	0	0	1 (100)	0
Any syphilis diagnosis (n = 22)b	8	11	1	2	7	12	0	3	10 (45)	12
Negative test results (n = 447)	2	339	17	89	0	426	10	11	2 (0.3)	445
Serological test results not available (n = 65)	0	46	2	17	0	62	1	2	0	65

^aInvalid due to insufficient specimen volume.

^bOne patient (case 22) had a positive RPR test result but was not staged.

PCR results were positive for 8 (89%) of 9 TMA-positive rectal specimens, as well as 1 specimen with an invalid TMA assay result (case 3, with primary syphilis; RPR test result, 1:16). The sole person with positive rectal TMA assay and negative rectal PCR assay results (case 13) had a positive RPR test result at a 1:32 dilution and a clinical diagnosis of early latent syphilis. Only 3 (43%) of 7 TMA-positive pharyngeal specimens tested PCR positive. The 4 persons with positive pharyngeal TMA assay results and negative PCR results included 1 person with primary syphilis, 1 with secondary syphilis, 1 with late latent syphilis, and 1 (case 22) with a positive RPR test result at a 1:128 dilution but missing data on the clinical diagnosis. None of the 28 TMA-negative rectal specimens or 31 TMA-negative pharyngeal specimens were positive by PCR testing.

DISCUSSION

Syphilis rates in many high-income nations are now at the highest levels observed in decades, rates continue to climb, and new approaches to syphilis control are urgently needed (25, 26). We used a new experimental TMA NAAT to investigate the potential utility of NAAT screening of mucosal specimens to diagnose subclinical early syphilis. Although the results were based on small numbers, we found that 8% of infected MSM in our STD clinic had positive TMA test results in the absence of other clinical or laboratory evidence of syphilis. Among 545 men with at least one TMA test result, the addition of rectal and pharyngeal TMA testing to standard procedures identified 2 extra cases of syphilis, although 1 of those persons would have been treated in the absence of a definitive diagnosis because he presented following known exposure to syphilis.

Although we are not aware of prior studies that used T. pallidum NAATs to screen high-risk patients for syphilis, our findings are largely consistent with prior studies showing that 5% to 20% of persons with primary syphilis have negative syphilis serological results but test positive using NAATs of genital lesions (6, 13, 22, 27, 28). Similarly, PCR results for blood samples from patients with negative serological results who are evaluated as known syphilis contacts are sometimes positive (23).

It remains unclear whether the incremental benefit of adding NAATs to existing serological testing for syphilis would justify the cost. The testing we performed on remnant specimens demonstrates that it would be relatively simple to integrate such testing into existing laboratory procedures. While our findings suggest that only a small proportion of infected men test positive only by TMA testing, that number is likely to vary in different populations and may change over time. The proportion of infections that are TMA assay positive and serological assay negative is a function of the length of the window period from infection to the time when the test results become positive using serology versus serology plus TMA testing, as well as the tested population’s syphilis incidence and the testing frequency. As the testing frequency increases, a larger proportion of infected persons are likely to be tested in the window period prior to syphilis seroconversion and to be positive only by TMA testing (29). A large number of MSM in the United States and other high-income nations are now taking HIV preexposure prophylaxis (PrEP). MSM on PrEP have a reported incidence of syphilis of 8 to 14 infections per 100 person-years and typically are tested for HIV and other sexually transmitted infections every 3 months (30–32). Such patients could be an ideal population to test using a combination of syphilis serology and NAATs.

Our findings also provide some insight into the natural history of syphilis. Mucosal TMA assay positivity was common in men with all stages of early syphilis. Of particular note, although the results are based on small numbers, TMA assay results were positive for approximately one-third of persons classified as having early latent syphilis. The relative transmissibility of primary and secondary syphilis is not well defined, with some studies suggesting that primary infections are more infectious than secondary syphilis (33) and others suggesting roughly equal transmission probabilities (9, 34). The extent to which latent syphilis, either early or late, is transmissible is unclear. Some authors have indicated that latent syphilis in the absence of ongoing secondary relapse is not transmissible (9, 35), although studies of “conjugal syphilis” undertaken in the first part of the 20th century suggest that transmission during latent infection, at least in the first 4 years following acquisition, may be relatively common (36). The presence of T. pallidum nucleic acid that we observed in persons with early latent syphilis suggests, but does not prove, the presence of transmissible organisms. Interpretation of this finding is complicated by the fact that persons with “early latent” syphilis represent a heterogeneous group of patients without overt clinical manifestations of infection who have positive serological tests; these persons might have not yet developed chancres, might have passed through primary infection but not yet developed secondary syphilis, or might be in the period after resolution of secondary syphilis. We cannot say where in this natural history our patients might have been, but our findings suggest that some persons with latent infection are likely infectious, highlighting the need for additional data to better understand the natural history of T. pallidum infection.

Our study has a number of limitations. First and foremost, although we tested almost 1,000 specimens, our sample size of persons with syphilis was small and the confidence bounds for the proportion of cases that might be serologically negative but positive by mucosal TMA assay were wide. Second, we cannot say with certainty that the men with negative serological results who tested positive by TMA assay really had syphilis. The fact that 1 of the 2 patients we identified with this profile sought evaluation because of known exposure to syphilis supports the conclusion that that patient was probably infected. Third, we did not perform TMA testing on blood samples. Studies evaluating different NAATs performed on blood samples reported sensitivities of 14 to 54% for primary syphilis and 58 to 82% for secondary syphilis (23, 28, 37, 38), with some variation based on the blood samples used (38) and with higher sensitivities seen in larger, more recent studies. Future studies evaluating NAATs as screening tests for syphilis should include testing of blood samples. Although data on urine testing for T. pallidum are very limited, this specimen type is also sometimes positive and may merit additional evaluation (17).

In conclusion, we found that an experimental TMA NAAT performed on rectal and pharyngeal mucosal swab specimens frequently yielded positive results for persons with all stages of early syphilis and that the addition of TMA testing to standard serological testing may identify persons with early syphilis prior to seroconversion. Given the alarming increase in syphilis incidence observed in high-income nations around the world and the paucity of effective interventions to address this growing epidemic, larger studies evaluating a syphilis screening regimen that combines TMA testing and serology should be performed in populations with a high incidence of syphilis.