Molecular epidemiology and specimen collection considerations for syphilis

Key Question

What options are available for molecular epidemiology and what should be considered for specimen collection and preservation?

Literature Search Terms

(syphilis OR Treponema pallidum) AND (genital ulcer disease OR primary syphilis OR secondary syphilis OR tertiary syphilis OR congenital syphilis OR ocular syphilis) AND (diagnosis OR lesions OR polymerase chain reaction OR PCR OR nucleic acid amplification test OR NAAT OR multiplex test OR silver stain OR silver staining OR immunohistochemistry OR IHC OR rabbit infectivity testing OR RIT OR direct detection OR dark field microscopy OR darkfield microscopy OR dark-field microscopy OR direct fluorescent antibody OR DFA OR direct fluorescent antibody for T. pallidum OR DFA-TP OR direct fluorescent antibody tissue test for T. pallidum OR DFAT-TP). Solely-based international studies were excluded from the literature search.


Four reviewers the evidence as high, medium, and low based on each study’s strengths and weaknesses. Case reports or small case studies were reviewed.

High quality publications: Studies using clinically characterized specimens, stratified by stage, larger sample size, prospective or a well-done cross sectional or retrospective study. Studies with large sample sizes, clinically characterized but not stratified by stage, or characterized but unclear exactly how it was done.

Medium quality publications: Studies with small sample sizes, moderate methodological issues, single lab test as gold standard, or descriptive.

Low quality publications: Studies with major methodological issues or small sample sizes.

I: Case reports or small case studies.

Key Question 2
Citation Description of study type, design, population, and setting Reported findings, quantitative results related to key question Overall quality; strengths/weaknesses or limitations Relevance to the key question and/or overall importance Rank
Katz, K. A. (2010) “Molecular epidemiology of syphilis – San Francisco, 2004-2007.” Sex Trans Dis. 37(10):660-664 Single site, San Francisco STD clinic, 2004-2007

Swabs from 74 primary or secondary syphilis patients that were Dark-field positive

Detection PCR for polA

Typing PCR for arp/tpr genes with RFLP

Additional testing for arp negative samples using alternate primer set and touchdown PCR

Additional typing by determining number of G tandem repeats in rpsA gene

70 primary syphilis and 4 secondary syphilis patients

69/74 samples could be fully typed; 66 primary and 3 secondary syphilis cases

Using 3 gene (arp, tpr, rpsA) typing scheme, 8 subtypes were identified, 14d9 being most frequent (52/69)


Subtyping based on rpsA not described previously

Limited number of samples

No controls for rpsA

Additional subtyping ability using rpsA gene to better characterize outbreak types High
Marra, C. M. (2010) “Enhanced molecular typing of Treponema pallidum: Geographical distribution of strain types and association with neurosyphilis.” 202(9):1380-1388 Multi-site

72 Tp isolates from bld, CSF, exudate (Seattle)

11 historical Tp isolates

16 bld samples (Seattle)

74 primary or secondary lesion swabs from Madagascar, San Francisco, Baltimore, China, Ireland

Evaluated strain types b/w blood and CSF in 84 patients with neurosyphilis

Evaluated typing using arp/tpr PCR and RFLP AND

RFLP analysis of tprC (tp0017), tprD (tp0131), sequencing of tp0548, and presence of 51bp insertion between tp0126/tp0127

51bp insertion and RFLP analysis of tprC did not improve discrimination vs. Pillay 1998

Sequencing of bases 131-215 of tp048 divided 14d subtype into 3 groups and subtypes 12a and 14a each into two groups.

tp048  sequencing and arp/tpr typing = strain type

173 samples were typed into 14 arp/tpr subtypes and 24 arp/tpr/tp048 strain types

Confirmed stability of typing method following multiple passages of Nichols strain in rabbits

84 samples from pts w/ neurosyphilis: able to show that 14d subtype able to be further strain typed and 14d/f more neuroinvasive compared to 6 other detected strain types

Related enhanced strain typing to clinical disease (neurosyphilis) and showed that sequencing of tp048 allowed for discrimination within the CDC 14d subtype

Showed stability of tp048 sequence over serial rabbit passages

Additional studies needed to confirm relevance of tp048 sequencing

Unclear what syphilis detection PCR was performed

Statistics performed

Additional typing using tp048 sequencing may be used to further strain type isolates for epidemiologic purposes High
Pillay, A., et al. (1998). “Molecular subtyping of Treponema pallidum subspecies pallidum.” Sex Transm Dis. 25(8):408-414 Lab type strains and clinical samples collected from GUD patients in the US, Madagascar and S. Africa

Clinical samples confirmed for T. pallidum by M-PCR

Specificity of arp/tpr PCRs determined by M-PCR negative GUD samples

GUD swabs in Roche PCR transport medium, stored at -70C until processing



 Identified two genes, arp and tpr subfamily II gene family (tprE, tprG, tprJ) with intra-strain variability enabling a subtyping method

Hybrid typing system using the number of 60 bp repeats of arp and RFLP analysis of tpr gene using MseI enzyme.

16 subtypes identified among 46 isolates in study

Discriminatory value of typing system determined by Simpson’s Index of Diversity: D value 0.868

First typing method described

Limited number of clinical samples, yet identified 16 subtypes

Found no change in typing pattern of Nichols strain after passage in rabbit for 1.5 years

Define PCR methods for  arp, tpr and RFLP analysis

Used negative controls

Original study describing arp/tpr RFLP based gene subtyping for T. pallidum High
Florindo, C., et al. (2008). “Molecular typing of treponema pallidum clinical strains from Lisbon, Portugal.” Journal of Clinical Microbiology 46(11): 3802-3803.


STI clinic in Lisbon, Portugal

$16 patients w/ suspected early syphilis enrolled b/w 2004-2007

Primary and secondary skin lesion exudate and blood collected

Samples screened by bmp PCR and confirmed by polA PCR (Lui 2001)

arp/tpr subtyping using modified Pillay 2002 protocol

86/416 (42 lesion exudates and 44 blood samples) patients positive by syphilis bmp and polA PCRs

Successful genotype  from 27/42 (64%) lesion exudates and 15/44 (34%) blood samples; 42/86 (49%) overall successful subtyping

Detected strain types 14a, 14d, 14f

Used patients with suspected syphilis, not confirmed serologically; dark field not performed

Compared skin lesions (assume exudate ?) vs. blood as sources

No discussion of handling/collection/storage of specimens prior to testing


Subtyping from blood was less sensitive compared to skin biopsy

Lower sensitivity in blood possibly due to inhibitory substances or lower spirochetemia

Castro, R., et al. (2009). “Molecular subtyping of Treponema pallidum subsp. pallidum in Lisbon, Portugal.” Journal of Clinical Microbiology 47(8): 2510-2512.


STI clinic in Lisbon, Portugal

82 patients (212 specimens) enrolled b/w 2003-2005 with clinical symptoms and serology positive for syphilis (all three stages)

Specimens collected include plasma, blood, ear lobe scrapings, lesion exudate

Samples screened by polA PCR (Liu 2001) and typed (Pillay 1998) by arp/tpr PCRs/RFLP


90 (42%) specimens pos by polA; highest sensitivity in lesion exudate (15/16, 94%) and ear lobe scraping (21/32, 65%), lowest in blood (22/82, 27%)

62/90 (69%) subtyped; highest sensitivity in lesion exudate (13/15, 87%) and ear lobe scraping (16/21, 76%), lowest in blood (13/22, 59%)

Detected strain types 10a, 14a, 14c, 14f, 14g

Samples processed and held at -4C for 1-2 days before processing

Use of ear lobe scrapings  for syphilis detection

Ear lobe scrapings may be used in patients w/o other lesions

Limited number of samples, not equal b/w specimen types

Blood has lowest sensitivity for direct detection by polA PCR and subtyping by arp/tpr

Highest subtyping percentage from lesion exudate collected by swab, followed by ear lobe scraping using scalpel and swabbing blood using swab and placed in PBS

Molepo, J., et al. (2007). “Molecular typing of Treponema pallidum strains from patients with neurosyphilis in Pretoria, South Africa.” Sex Transm Infect 83(3): 189-192.


1999-2000, 50 CSF specimens from pts w/ suspected neurosyphilis in Pretoria, S. Africa; single site

CSF tested by VDRL, FTA-ABS; neurosyphilis confirmed if either positive

tpp47 gene for detection PCR and subtyping by arp/tpr (Pillay 1998)

RFLP tpr typing using Tru91, isoschizomer of MseI

CSF stored -70

35/50 CSF were VDRL/FTA-ABS positive and 5/50 FTA-ABS pos only

28/50 CSF were tpp47 PCR pos

15/28 with sufficient DNA for typing; 13 fully typed and 2 partially typed

4 strain types identified

Limited number of  positive samples

5 VDRL/FTA-ABS negative samples were PCR positive…FPs vs. TPs? No discordant analysis


VDRL positive CSFs have higher probability of containing sufficient DNA for typing

One of two studies on typing from CSF


Martin, I. E., et al. (2009). “Macrolide resistance and molecular types of Treponema pallidum causing primary syphilis in Shanghai, China.” Clinical Infectious Diseases 49(4): 515-521.


Single site, Shanghai STD clinic 2007-2008

39 pts with symptoms of primary syphilis (GU)

Syphilis Dx confirmed by Dark-field and/or serology

Serum (TRUST and TP-PA), whole blood and GU swabs collected

10 swabs from non-syphilis GU collected for controls

Swabs in saline and stored -70C

Detection PCR using tpp47, bmp, polA

Control for PCR inhibition by PCR for beta-globulin in syphilis PCR neg samples

Subtyping using arp/tpr (Pillay 1998)

39 patients with primary syphilis confirmed by dark-field (N=35) and serology (N=4)

38/39 Tp PCR pos for all 3 genes (!) from GU lesions; negative in all blood samples

identified 5 subtypes, 14f most common

Blood and GU lesions tested

Confirmed detection of T. pallidum DNA using 3 genes in well characterized primary syphilis patients

Negative and internal controls used

Blood poor specimen source for syphilis detection PCR, including use of buffy coat or whole blood

High sensitivity in GU lesions

Lu, Y., et al. (2017). “Molecular epidemiological survey of Treponema pallidum in pregnant women in the Zhabei district of Shanghai.” Journal of Medical Microbiology 66(4): 391-396. Single site, enrollment of 3024 pregnant women in China

41 RPR and TPPA pos from whom WB, plasma, ear lobe scrapings and mucosal lesion exudate (if present) were collected

tpp47 screening PCR on all raw samples; pos samples typed using arp, tpr and tp0548

13/41 (32%) WB tpp47+

8/13 (61%) typed

17/41 (41%) plasma tpp47+

12/17 typed

25/39 (64%) earlobe scrapings tpp47+

21/25 typed

10/11 exudates tpp47+

10/10 typed

Highest tpp47 positivity in all sources from secondary syph

Small number of patients in evaluated despite high enrollment DNA detection/typing from lesion exudate and earlobe scrapings provides highest sensitivity Medium
Xiao, Y., et al. (2016). “Molecular Subtyping and Surveillance of Resistance Genes In Treponema pallidum DNA From Patients With Secondary and Latent Syphilis in Hunan, China.” Sexually Transmitted Diseases 43(5): 310-316. Cross-sectional, multi-site study in China dermatovenereologic clinics

WB collected from pts w/ secondary or latent syph confirmed by Chinese guidelines and serology

2253 WB samples from 603 primary and 1650 secondary syph cases

WB screened for Tp DNA w/ polA, tpp47, bmp, tp0319 PCRs

Screen pos. typed with arp, tpr, tp0548 PCRs

455/2253 (20.2%) of samples pos by at least 1 screening PCR; 306/2253 (13.6%) pos by all 4 PCRs

Indiv. pos rates: 18.8% polA, 18.4% tpp47, 17.5% bmp, 14.2%  tp0319
272/603 (45.1%) secondary syph and 183/1650 (11.1%) latent syph pts pos by ≥1 screening PCR

tpr, arp, tp0548 amplified from 181/455 (39.8%) of samples

Other stages of syph not evaluated

Limited success at typing

Early vs. late latent syph not differentiated

WB from secondary and latent syph cases associated w/ limited overall success rate for Tp  typing