Performance of treponemal serology tests for syphilis

Updated April 5, 2023

Key Question

What are the performance characteristics, stratified by the stage of syphilis, for treponemal serologic tests? (T. pallidum particle agglutination, fluorescent treponemal antibody-absorption, enzyme immunoassay, chemiluminescence assay, multiplex bead-based immunoassay)

Literature Search Terms

((Treponema pallidum OR neurosyphilis OR syphilis) AND (sero-diagnos* OR serodiagnos* OR (serolog* AND (test* OR exam* OR assay* OR screen* OR lab* OR diagnos* OR nontreponemal OR treponemal OR algorithm* OR antibody titer)) OR serofast) NOT exp animals/ not exp humans/. Solely-based international studies were excluded from the literature search.

Grading

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.

NR: Studies that were not relevant to the key question were assigned and not further rated.

Key Question 4
Citation Description of study type, design, population, and setting (gold standard) Reported findings, quantitative results related to key question Overall quality; strengths/weaknesses or limitations  Relevance to the key question and/or overall importance Rank
Zhang, H. L., et al. (2013). Dermatology 226(2): 148-156. Retrospective, University hospital, all inpatients, HIV neg, median age 50

Inclusion:  NS based on CDC GL.  Confirmed NS: clinical dx of syphilis + CSF VDRL

Presumptive: clinical dx of syphilis and NR VDRL, with CSF >10 WBC, or protein >500 mg/l  and neurologic signs/syxs

Excl: no HIV test, HIV+

Gold standard: clinical characterization plus laboratory findings

N=149 patients with NS

(124/149, 83% symptomatic, 17% asyx)

CSF TPPA

 

149/149 serum TPPA+

134/149 (89.9%) + CSF TP-PA reactive

88/149 (59.1%) CSF pleocyt

84/149 (56.4%) elev protein

37 (24.8%) had normal WBC and protein

Clinically characterized (inc also lab gold standard using CDC criteria for dx

 

Unclear why asymptomatic NS patients had CSF eval in first place—not clear if pts with other reasons for abnl CSF were excluded

 

(adjunct to Harding/Ghanem 2012)

 

Describes sensitivity almost 90% of CSF TP-PA in clinically characterized sera

 

 

High
Wong, E. H., et al. (2011).

Sexually Transmitted Diseases 38(6): 528-532.

 

STD clinic, prospectively collected remnant sera.

TrepSure EIA (IgG, IgM)

Tested with VDRL, TS EIA, TPPA

IgG and IgM WB used to resolve discordants, TrepID also used for EIA+, VDRL NR, TPPA NR

 

True positive: 1)VDRL+, TS-EIA+ and TP-PA+ or WB+

2)VDRL+ EIA-, TPPA +

3) VDRL-, EIA+, TPPA+

True neg: 1) VDRL-, EIA-, TPPA-

2) VDRL+, EIA-, TPPA-

Gold standard:  Lab findings only using combination VDRL, TPPA and WB

N=674 specimens.

sensitivity TS EIA 279/285

97.9% (calc from figure)

Specificity: 99.1% (reported)

Microbiologist time 80 specimens

VDRL: 150 minutes

TS EIA: 120 minutes (microbiologist free time 90min, 30 min hands on)

 

6 False neg EIA all IgM WB+

 

S/CO values associated with TP-PA positivity (>8.00=99.6% TP-PA positivity)

 

Characterizes false negatives with WB, and false pos EIA with TrepID

 

Not stratified by stage

 

Lab Gold standard only

Strong performance of TS EIA in high prev population

 

Describes time differences in performance of VDRL vs EIA.

 

High
 Xu, M., et al. (2016).  International Journal of Infectious Diseases 43: 51-57. Cross sectional, university hospital, China

N=3326 subjects, (948 routine exams, 1358 outpts, 1020 inpts), all tested wth RPR and TPPA

 

Architect Syphilis TP, WB performed.

 

Gold standard:

True pos: Clinical diagnosis (inc sexual history) + RPR/+ TPPA

True neg:

 

N=256 pts with syphilis

n-632 “healthy subjects”

 

Architect TP
Sensitivity/Specificity

97.27% (94.46-98.67)

99.61% (99.32-99.78)

Clinically characterized, not stratified by stage, specificity data difficult to interpret since criteria for true negative are not described.  

 

High
Wellinghausen, N. and H. Dietenberger (2011).

Clinical Chemistry & Laboratory Med 49(8): 1375-1377.

 

Private lab Germany, study 1: prospective screening specimens, study 2: retrospective study from banked serum.

Screening tests: LIAISON CIA, ARCHITECT CIA, TP-PA

Confirmatory test: FTA-Abs and IgG/IgM immunoblot (recomBlot)

Gold standard: Study 1

True+: screening+  plus confirmatory trep test, and clinical characterized if both positive

True neg: consensus of screening panel

Gold standard: study 2

Clinically characterized serum bank from patients with prior syphilis

 

Study 1: n=577 (incl 318 prenatal)

18 true positives

(3 prim, 2 second, 7 EL, 6 LL)

Sensitivity

LIA: 100% (18/18)

ARCH: 100% (17/17)

TP-PA 100% (18/18), and Specificity

LIA  100% (558/558),

ARCH 99.8% (552/553), and TP-PA 99.6% (556/558)

 

 

Study 2: n=42 specimens, from 32 patients with syphilis

(6 prim, 13 second, 8 EL, 4 LL, 1 late NS)

Sensitivity 100% for all assays.  (ARCH only 37/37 specimens due to volume)

 

Clinically characterized

 

Not stratified by stage due to small numbers

 

Retrospective study uses multiple specimens from the same patient in sensitivity calculation

Compares head to head performance of CIAs to traditional tests with small numbers of patients not stratified by stage. Almost perfect sensitivity/specificity for all assays Medium
Sun, R., et al. (2013).

Chinese Medical Journal 126(2): 206-210.

University hospital, cross sectional

Inpatients with syphilis (old or new)

TP-PA and RPR pos

TP-PA and RPR neg but contact to syphilis or “typical” presentation

Gold standard: N/a

N=69

N=17 primary

N=14 secondary

N=11 teritary

TP45 IgM was most frequent in pts with primary disease (64.7%)

TP15 igM only detected in those with tertiary syphilis

Tx reduced Tp17 IgG and TP47 IgG

Not test performance, background only

Small numbers in each stage

 

Descriptive data only, no test performance by stage

 

Descriptive background on TP antigens and their appearance NR
Saral, Y., et al. (2012). Acta Dermatovenerologica Croatica 20(2): 84-88.

 

University hospital, Turkey, retrospective

Architect CMIA vs RPR

Gold standard:

True pos: initial clinical diagnosis based on chart review and TPHA +

True neg: neg RPR, TPHA and CMIA

N=4109 true negatives

n=117 patients with suspected syphilis, 112 true positives

n=18 primary, 21 secondary (rest latent or old, results not stratified by stage)

Sensitivity 98%

Specificity 100%

PPV 100%

NPV 71%

Clinically characterized, (+ lab standard for true positives.

Lab gold standard (consensus of panel) for true neg

 

No confidence intervals provided

 

Not stratified by stage

 

True neg case definition includes neg Architect result so specificity would be perfect based on inclusion criteria

Architect sensitivity and specificity with partially characterized specimens, both >98% Medium
Park, Y., et al. (2011).

American Journal of Clinical Pathology 136(5): 705-710.

2 university hospitals, Korea

Cross sectional, general medicine and derm pts.

Separate sample of pts with ANA 1:30 to 1:160

All pts tested with FTA-ABS, and Architect

400 were also tested with VDRL (chart review for discrepants between 3 tests)

Architect syphilis vs FTA or VDRL  FTA ABS vs Architect

 

% concordance and Kappa calculated

Gold standard: FTA-ABS (for architect performance)

Or VDRL result (for FTA-ABS performance)

N=616 specimens

%Concordance:

Architect TP and FTA,99%, K 0.981 (0.965-0.996) p<0.001

 

N=400 specimens tested w VDRL

FTA-ABS and VDRL 85%, k 0.7 (0.631-0.739) p<0.001

Architect & VDRL, 83.8%, k 0.675

 

N=108 pts with + ANA

0 reactive Architect

Single lab test as gold standard

Clinically characterization limited to patients with discordant results between  the 3 tests

 

unclear how they are divided.

 

 

nearly perfect concordance between Architect and FTA-aBS, Architect did not give FP in pts with + ANA Medium
Park IU (2011)

Journal of Infectious Diseases 204(9): 1297-1304.

 

Managed care organization, US

Cross sectional, mixed risk screening population

Descriptive data on testing and differences among discordant pts (CIA+, RPR-) according to TP-PA status.

LIAISON CIA index values correlated to TP-PA

 

Gold Standard: N/A

N=255 CIA+, RPR-

TPPA pos (184), TPPA neg (71)

87% of TPPA pos who were retested stayed same.

7/31 (23%) isolated CIA+ seroreverted to CIA-

1 isolated CIA+ seroconverted to RPR pos in 6 months

 

LIAISON S/co of 12.00 associated with 100% TP-PA positivity

Background only

Not all patients retested

 

S/co cutoff data useful

 

Descriptive data, no test performance

 

Descriptive data on sero-reversion of isolated CIA+

 

S/co cutoff 12.00 associated with 100% positivity

NR
Park, B. G., et al. (2016).

Journal of Clinical Microbiology 54(1): 163-167.

 

University hospital, Korea

Cross sectional, general PE and pts with suspected syphilis

Sensitivity, specificity,  % agreement, kappa

 

Architect, Cobas, ADVIA, Sysmex, A and T, Seskisui

 

Gold standard: FTA-ABS

n=615

105 well-visits, 179 preop, 329 suspected current or known prior syphilis

Sensitivity:  Architect 96.8%

Cobas 99.4%, ADVIA 99.4%

Specificity: 100% for all

Clinical data used for specimen selection but not to determine true positive/negatives status

 

Single lab test as gold standard

Test performance of Architect, Cobas, ADVIA,

Almost perfect sensitivity for Cobas, ADVIA, perfect specificity for all

Medium
Malm, K., et al. (2015). Transfusion Medicine 25(2): 101-105.

 

University hospital, Sweden

Cross sectional (first time blood donors, repeat donors)

LIAISON XL CIA

Sensitivity, specificity

Gold standard : Abbot Architect results

N=55 Architect pos

353 Architect neg

Sensitivity 100%

Specificity 100% (99.7-100) (overall)

Single lab test as gold standard Perfect sensitivity/specificity of LIAISON with single CMIA as gold standard Medium
Loeffelholz, M. J., et al. (2011). Clinical & Vaccine Immunology: CVI 18(11): 2005-2006.

 

Retrospective, University hospital, US, incarcerated, women seen in obgyn

Bioplex IgG, RPR, TPPA

Bioplex test performance based on AI cutoffs of 6 vs 8, stratified by population and RPR titer

Sensitivity, specificity

Gold standard:

True pos:  + TPPA

True neg:  – TPPA

 

 

RPR NR: n=82 incarcerated, n=44 women at obgyn clinics

RPR ≥1:2 n-110 incarcerated, n=31 women at obgyn clinics

 

Sensitivity/Specificity AI6

RPR neg   52.4-57.9/92-96%

RPR ≥1:2 96.8-97.2%/100%

 

 

Sensitivity/specificity  AI8

Rpr neg: 50.9-50.4/96-100%

RPR ≥1:2 94.5-96.8/100%

 

(even at index value of 8, incarcerated specificity <100), but 100% for AI 8 in women with NR RPR or RPR 1:1)

Background for analysis of index values

RPR 1:1 not shown due to space limitations and small numbers

 

 

 

Single lab test as gold standard

Not stratified by stage

Demonstrates utility of using S/CO results to predict TP-PA positivity increased specificity if using AI of 8

 

 

 

Excluding women at delivery because all sample sizes are 5-7 women only

 

 

Medium
Liu, C., et al. (2014). Journal of Clinical Laboratory Analysis 28(3): 204-209.

 

Prospective, university hospital, China, screening population and pts with syphilis

TRUST, RPR, TPPA, TP-CMIA (Architect), TP-ELISA (Xiamen)

Gold standard:

Pos: signs/symptoms of syphilis, “latent infection” (not sure how characterized)

Neg: asymptomatic, “not at risk for syphilis”, two neg treponemal tests

(pre-op screening or PE)

N=210 true neg

N=160 true pos

 

Sensitivity/Specificity

TPPA 96.25%/100%

ARchitect 100%/90.95%

 

Clinically characterized sera, using combination of clinical and lab data (for neg)

 

Not stratified by stage

 

Unclear how latent syphilis was characterized

Architect more sensitive but less specific than TPPA in clinically characterized specimens (although accuracy of classification is uncertain) Medium
Li Z. (2016)

Annals of Clinical Biochemistry 53(Pt 5): 588-592.

Cross sectional, University hospital china, general screening population

Descriptive data on CMIA (Architect) (similar to Park JID 2011) and S/CO associated with TP-PA positivity.

 

Gold standard: N/A

N=20,550 samples screened

1.3% positivity, 267 CMIA+

N=185 (69.3%) confirmed with TP-PA,

 

16/82 (19.5%) CMIA+ RPR-, TPPA- were subsequently Dot blot positive

 

CMIA index of 10.0 correlated to 100% TP-PA positivity

Not test performance (background only)  describes % unconfirmed CMIA in screening population, uses immunoblot (3rd trep test) on a portion of the discordants Data on SCOs and correlation with TP-PA positivity for

 

CMIA SCO 10  correlated with positivity

NR
Lam, TK (2010)

International Journal of STD & AIDS 21(2): 110-113.

 

Cross sectional, Social hygiene clinics (STD clinics), Hong Kong DOH

TP-PA, FTA ABS (other non-FDA approved assays)

Gold standard:

True pos: prior serology plus clinical symptoms

Primary: Chancre

Secondary: mucocutaneous signs

EL: no syxs, documented non-reactive serology in last 12 months

LL: no syxs, non reactive serology >12 months

True neg: “normal healthy subjects”

 

n=135 pts with syphilis

·        n=39 primary

·        n=20 secondary

·        n=18 EL

·        n=58 latent unknown duration

n=43 pts without syphilis

Sensitivity (TP-PA/FTAABS)

Primary TP 94.9% (83.1-98.6%)

FTA 84.6% (70.3-92.8)

Secondary  TP 100% (83.9-100)

FTA 95% (76.4-99.1)

Early latent  94.4% (74.2-99.0)

94.4% (74.2-99.0)

Latent unknown duration

91.4% (81.4-96.3)

84.5% (73.1-91.6)

Specificity 100% (91.8-100) for all

 

 Clinically characterized sera stratified by stage,

 

Small numbers for both sensitivity and specificity calculation

Head to head test performance TP-PA, FTA-ABS, characterized by stage

 

FTA less sensitive in primary syphilis and latent of unknown duration.

Perfect specificity

High
Jost H (2013) BMJ Open 3(9): e003347. Cross sectional,   serum bank specimens Georgia PH lab

FTA,   LIAISON, trepsure,  Captia IgG (others not FDA approved)

Gold standard:

True pos: TPPA pos

True neg; TPPA neg

N=290

N=109 TPPA+

N=181 TPPA-

Pos agreement: FTA-ABS 94.4%

LIAISON (100%)

Trep Sure (100%)

Captia (100%)

 

Neg Agreement: FTA 100%

LIAISON (99.4%)

TrepSure (98.9%)

Captia (97.2%)

Head to head agreement of multiple FDA approved assays

 

Not stratified by stage. Single lab test as the gold standard

Perfect positive agreement between IAs,

Captia, Liaison, Trepsure Using TPPA as gold standard

 

FTA had perfect neg agreement w TPPA

Medium
Jonckheere (2015)

European Journal of ClinicalMicrobiology & Infectious Diseases 34(10): 2041-2048.

 

Cross sectional, multiple univ, Belgium, mixed risk (see below)

7 assays performed: Bioplex 2200 IgG included, all others not FDA approved

Low risk patients 15.6%

Asymptomatic, at risk 61%

Symptoms of syphilis 7.5%

Equivocal results considered neg

 

Gold standard:

True pos: TPPA pos

True neg; TPPA neg

N=177, n=5 equivocal

Sensitivity 94.2 (87.8-97.8)

Specificity 59.5% (47.4-70.7)

Not stratified by stage. Single lab test as the gold standard Bioplex IgG specificity poor using TP-PA as gold standard Medium
Henrich, T. J. and S. Yawetz (2011). Sexually Transmitted Diseases 38(12): 1126-1130.

 

Cross sectional, university hospital US, Captia IgG

Screening population at univ hospital.  Descriptive data on IgG EIA positivity, % concordance with TPPA

 

Gold standard N/A

N=34,251 samples, 1.8% positivity

79% confirmed with TPPA

 

Lower confirmation if  age <40, OB/GYN service patient, or female

Not test performance (background only)  describes % confirmation of Captia EIA in low prevalence population Background on Captia demonstrates more FP in lower prevalence populations NR
Harding (2012)

Sexually Transmitted Diseases 39(4): 291-297

Systematic review

CSF tests were

FTA-ABS, FTA, TPHA, MHA-TP

 

Sensitivity, specificity (S+, NS-), specificity (syphilis neg), NPV

 

Gold standard: Clinical criteria alone, clinical plus abnormal CSF WBC, clinical OR + CSF VDRL,  positive CSF VDRL,  +CSF VDRL, OR >5wbc OR protein >45

18 studies,

40 measures of sensitivity

30/40 had sensitivity ≥80%

studies using + VDRL as criteria demonstrate nearly 100% sensitivity.

 

21 measures of NPV.  18/21 demonstrated NPV >90%

3 measures NPV was 58-79%

 

In the general population, a neg CSF treponemal tests effectively rules out NS.

 

NPV varies depending pre-test probability.  In usual clinical setting for these assays, pt has high probability of NS (sign/syxs) of syphilis, so the NPV would be lower than in gen population

Comprehensive data on sensitivity, specificity, NPV up until 2006 so needs update of last 10 years

 

Gold standard: NS diagnosis highly heterogeneic.

Systematic review of test performance of treponemal tests using heterogeneous definitions of NS, and so no uniform gold standard High
Guarner, J., et al. (2015). American Journal of Clinical Pathology 143(4): 479-484.

 

Retrospective, university hospital, pts eval for NS

CSF VDRL, TrepSure, TPPA, (INNO-LIA, MaxiSyph)

Gold standard:

True pos: VDRL positive plus symptoms c/w NS (11)

Symptoms but neg CSF VDRL (2)

Reactive CSF VDRL and  (no history) (1)

 

True neg: pts who had a CSF VDRL performed during same week as reference group

N=14 in reference group

N=18 controls

 

Sensitivity

CSF VDRL-85.7%

TrepSure EIA-92.9%

TPPA 83.3%

Specificity 100% for all assays

PPV 100% for all

NPV

CSF-VDRL 89.5%

TrepSure 94.7%

TPPA 90%

 

Head to head test performance mostly clinically characterized sera (13/14)

 

Small numbers, lack of uniform gold standard

TrepSure EIA and TPPA in CSF more sensitive than VDRL.

 

Useful adjunct to Harding/Ghanem 2012.

High
Gratzer B (2014)

Sexually Transmitted Diseases 41(5): 285-289.

Retrospective, STD clinic, IL, US

TrepSure, RPR, FTA-ABS

 

Gold standard

True pos: chancre/ulcer at least 1 pos syphilis serologic result plus No known history of syphilis

TS-EIA+ or Equiv/RPR+

TS-EIA+ or Equiv/FTA+

 

True neg: N/A

N=52 (51 MSM)

Sensitivity

TS-EIA 53.8% (39.5-67.8)

RPR 76.9% (63.2-87.5)

(RPR significantly higher p=0.005)

(sensitivity analysis excluding equivocals or making equivocals neg were similar)

 

Clinically characterized sera, large sample size for primary syphilis, sensitivity analysis to ensure equivocals didn’t bias outcome

 

No specificity data, no data for other stages

 

 

Sensitivity of TS EIA poor in primary syphilis High
Gomez E (2010)

Clinical & Vaccine Immunology: CVI 17(6): 966-968.

 

Prospective, university hospital, MN, US, general screening population

Bioplex 2200 IgG and separate IgM (MFIA),

TrepChek IgG/IgM

TP-PA, RPR (for discrepancies)

 

Gold standard: (initially TrepChek results, but then further testing of discrepants)

True pos: TC-EIA+/MFIA+

MFIA+/EIA-/TPPA+ or MFIA IgM+

 

True neg: EIA neg/equiv and Bioplex neg

 

 

N=1008 specimens

Bioplex IgG sensitivity

98.7% (77/78, 92.1-99.9)

Specificity 98.5% (653/663, 97.5-99.1)

 

N=671

Bioplex IgM Sensitivity

100% (8/8)

Specificity 98.5% (653/663)

 

 

Consensus of 2 lab tests as gold standard

 

Not all discrepants tested with IgM MFIA, sample size <10, for sensitivity analysis

 

TrepChek no longer on market and replaced with TrepSure (specificity issues with TrepChek per CDC MMWR)

Bioplex MFIA IgG or IgM was >98% sensitive/specific using lab tests as gold standard High
Centers for Disease, C. and Prevention (2011).

– Morbidity & Mortality Weekly Report 60(5): 133-137.

Retrospective, 4 labs, KPSC, KPNC, NYC, Chicago, US

KP-mixed screening population

NYC/Chicago-PH lab

TrepChek (KPSC, NYC)

LIAISON (KPNC)

Trepsure (KPNC, Chicago)

 

Gold standard: N/A

N=140,176 specimens total

Low prev settings (2.3% EIA+)

High prev settings (14.5% EIA+)

% non-reactive TPPA or FTA ABS, overall 31.6%

KPSC Trepcheck 60%

KPNC LIAISON 30%

KPSC TrepSure 25.2%

 

NYC TrepChek 12.2%)

Chicago TrepSure 18.6%

 

Not test performance (background only)

Large analysis of reverse sequence screening

 

 

Background on use of Reverse sequence algorithm and high false pos EIA in low prev populations, particularly with TrepChek

 

Inc. TrepSure and LIAISON

NR
Busse, C., et al. (2013). Clinical Laboratory 59(5-6): 523-529.

 

Retrospective, University hospital, Germany

Prenatal screening, suspected syphilis pts, pts with borreliosis or HIV, syphilis serum panels

Virotech (not FDA)

TrepSure EIA (TS-EIA)

TPPA/FTA-ABS (combined)

 

Gold standard:

True pos: TPPA and FTA ABS reactive

True neg: TPPA and FTAABS neg

 

N=183 true neg

N=218 true pos

 

 

Sensitivity (lab GS)

TS EIA 100% (98.3-100%)

Specificity

TS EIA 93.9% (89.4-96.9%)

 

S/CO of 12 correponds or greater corresponds to TPPA titer of 1:80 or greater.

LAB gold standard (consensus of 2 tests)

 

Mentions also n=20 specimens for analytical sensitivity and n=74 specimens from serum bank but no results for 74 specimens

 TS EIA perfect sensitivity, 94% specificity using lab gold standard, background data on S/co TS-EIA of 12 corresponding to high TPPA Titer. High
Bosshard 2013

Journal of Infection 67(1): 35-42.

Retrospecitve, university hospital, Switzerland, sera routinely sent to lab for syph diagnosis (syxs pts) and see true neg for controls

TPPA, VDRL,

Pathozyme, Euroimmun, recomWell (not FDA)

Gold standard

True pos: primary: anogenital/oropharyngeal chancre, or red lesion + serology, or asyxs contact to syphilis

Secondary—rash typical of syphilis

Latent—serology (VDRL/TPPA)

Tertiary—NS or cardiac syphilis

True neg: “negative sera without further specification” (n=50) or sera from pts with other infections

Borrella burgdorferi, CMV, Lupus, HIV, pregnant women, previous syphilis dx (n=30)

N=156 true positives

N=151 true negatives

 

Sensitivity TPPA

Primary (n=59)

100% (93.9-100)

Secondary (n=66)

100% (94.6-100)

Latent (n=25)

100% (86.3-100)

Tertiary (not calculated)

Specificity

99.2% (95.5-100)

Clinically characterized sera

Stratified by stage, latents combined

 

good numbers in primary and secondary stage

 

Unclear why TPPA specificity is so high if patients with previous syphilis are included as true neg.

 

Unclear how negatives were classified

Perfect sensitivity and 99% specificity of TPPA in primary and secondary syphilis High
Bosshard 2013

Clinical Infectious Diseases 56(3): 463-464.

Unclear if retro/prospective

Patients with syphilis and + TPPA followed  6 months after treatment, with at least 3 follow-up serologies

TPPA titer values

N=212 patients,

Mean follow up 429 days, avg reduction in dilution 3.3 steps

117 (55%) had constant decrease

30 (14%) had constant titer

4 (2%) had increase

Not test performance (background only)

 

Descriptive data on seroreversion of TPPA after treatment NR
Binnicker (2011)

Journal of Clinical Microbiology 49(4): 1313-1317.

Cross sectional, national reference lab, MN, US

Samples previously tested for syphilis, consecutive samples from national reference population

FTA (not ABS), Bioplex 2200 IgG

TPPA, TrepChek EIA, TrepSure

TrepID, ViraBlot IgG, IgM Wb

 

Gold Standard: 1) FTA

2) consensus of test panel (≥4/7)

N=303 samples

N=100 (prev tested for syph)

N=203 (general ref pop)

FTA Gold standard

Sensitvity

Bioplex  96.9% (90.9-99.3)

TPPA 95.9% (89.5-98.7)

Trep Chek 95.9% (89.5-98.7)

TrepSure 96.9% (90.9-99.3)

 

Specificity

Bioplex  98.5% (95.6-99.7)

TPPA 97.6% (94.3-99.1)

Trep Chek 98.5% (95.6-99.7)

TrepSure  94.7 (90.6-97.1)

 

Consensus of test panel

Sensitivity

Bioplex  100% (95.3-100)

FTA 100% (95.3-100)

TPPA 98.9% (93.6-99.9

Trep Chek 98.9% (93.6-99.9)

TrepSure 98.9% (93.6-99.9)

 

Specificity

Bioplex  98.6% (95.7-99.7)

FTA  98.6% (95.7-99.7)

TPPA 97.6% (94.3-99.1)

Trep Chek 98.6% (95.7-99.7)

TrepSure  94.3 (90.1-96.8)

 

TAT in hours

Bioplex 1.75

FTA 3.3

TPPA 4.0

Trepchek/Trepsure 2.2-2.3

2 different laboratory gold standards

 

 

Unclear if 100 samples were supposed to be positive (94/100 were with FTA and consensus of panel-)

All four assays had similar performance regardless of gold standard, TrepSure was marginally less specific.

 

Bioplex had shortest TAT, TPPA longest

Medium
Citation Description of study type, design, population, and setting (gold standard) Reported findings, quantitative results related to key question Overall quality; strengths/weaknesses or limitations Relevance to the key question and/or overall importance
Atkas G et al (2007)

International Journal of STD & AIDS 18(4): 255-260.

Retrospective, university medical center, Turkey, screening pop

 

FTA ABS, TPPA, ICE, Enzywell TP, IgG+IgM, Captia Syph M, WB (only FTA/TPPA FDA approved)

 

Gold standard: FTA pos or neg

(additional tests used to characterize FTA neg as FN)

N=94 FTA pos, n=25 FTA neg

N=3 equivocal

TPPA vs FTA ABS

97.8%  pos agreement

92% neg agreement

 

Among other assays, % agreement FTA vs other trep tests was 95.9-98.3%

 

Found 2 neg FTA that were + on 5 other trep tests

 

Laboratory gold standard, single assay

 

Authors conclude that FTA more prone to give equivocal and FN results than other trep tests

 

 

High
Castro, R., et al. (2006). Journal of Clinical Laboratory Analysis 20(6): 233-238. Retrospective, community and univ hospital, Portugal

Inpatients with reactive serologic tests for syphilis, pts with neurologic infections and neg serology

CSF VDRL, MHA-TP, TPPA, FTA-ABS

Gold standard

True pos: Serum + FTA-abs, TPPA MHA-TP and > 10 WBC,

+/- CSF VDRL

+/- clinical symptoms

 

True neg: negative VDRL, serum MHA TP FTA, other neuroinfection

 

N=18 pts with NS (16 pts with + CSF VDRL both syxs and asyx) 2 with syxs, >WBC and neg cSF VDRL)

N=46 patients with non syphilis neuro infections

 

Sensitivity/specificity

CSF FTA ABS 100%/100%

CSF TPPA 100%/100%

 

10/57 patients with prior treated syphilis had at least 1 abnormal CSF treponemal test

 

29/67 pts with latent syphilis had at least 1 trep test abnormality

19/67 pts with latent syphilis and normal WBC had 1 trep test abnl

Clinically characterized

True neg

includes asyx and syx NS

 

state that +VDRL CSF was part of gold standard but CSF FTA-and TPPA both pos for pts w neg VDRL but >WBC-CSF

 

Head to head comparison of CSF MHA TP, TPPA, FTA ABS but difficult to interpret test performance bc in tables some patients with syphilis (not classified NS) have abnl CSF WBC, + VDRL (meet case definition yet not included as such)

 

Tables challenging to interpret

CSF treponemal test performance in pts with strictly defined

NS and non-syphilis neuro infections. 100% sensitivity/specificity

High
Castro R et al (2001)

American Journal of Clinical Pathology 116(4): 581-585.

Prospective, university hospital (in patients) and STD clinic, Portugal

Patients given therapy for early syphilis

RPR, MHATP, FTAABS, TPPA

Repeat serology, 1, 2,3,6, 12 months.  S/S of TPPA

 

Gold standard: (true pos)

MHA TP (analysis 1)

Or FTA-ABS (analysis 2)

 

Separate analysis among pts with primary syphilis (+MHA was gold standard for FTA, and vice versa)

True neg: MHATP- (analysis 1)

FTA-ABS – (analysis 2)

N=449 total patients

(MHA as GS)

Sensitivity 100%

Specificity 94.4%

 

(FTAABS as GS)

Sensitivity 98.5%

Specificity 100%

 

n-28 primary syphilis

Sensitivity

TPPA 100%

FTAABS 89.2%

MHA TP 89.2%

 

N=54 returned for f/up, at 12 mos 0 seroconversions of TPPA or MHATP (FTA not described)

Mostly single lab test as gold standard.  Limited clinical classification to: “suspected of having early syphilis”

 

Clinical classification of primary syphilis unclear (described as serology “revealed” primary syphilis)

 

High loss to follow up (89%)

(b/c classification unclear)

Test performance of TPPA very high, using other manual trep tests as gold standard,

 

?useful data for primary  syphilis

 

descriptive data on seroconversion after treatment

Medium
Centers for Disease, C. and Prevention (2008). MMWR – Morbidity & Mortality Weekly Report 57(32): 872 875.

 

Retrospective, NYC ?PH labs,

Convenience sample of screening population

 

Gold standard: n/a

N=116,822 specimens

N=6587 (6%) EIA+

3664 (56%) RPR neg

433 (17%) TPPA neg

 

~3% additional reactive specimens that need further testing/f/up compared to RPR based algorithm

Not test performance

Descriptive data only

 

No clinical interpretation of results available.

 

Descriptive data on early use of RSS NR
Cole, M. J., et al. (2007). European Journal of ClinicalMicrobiology & Infectious Diseases 26(10): 705-713.

 

Cross sectional, blood bank and commercially specimen bank, and PH lab bank, UK

Unselected blood donors, banked specimens with known clinical history/staging

 

15 treponemal tests, but only TPPA and Captia FDA approved and in use (4 TPHA assays but not currently in use)

 

Gold standard:

True pos: based on known prior disease stage/treatment status

True neg: not specified (assume all blood donors would be neg or at least no h/o syphilis)

Sensitivity

Captia 94.7% (88.8-98)

TPPA 99.1% (95.2-100)

 

Specificity

Captia 100% (98.5-100)

TPPA 100% (98.5-100)

Clinically characterized

 

Most tests not FDA approved

Not stratified by stage

Captia and TPPA have perfect specificity, TPPA marginally better sensitivity Medium
Creegan (2007)

Sex Transm Dis. 2007 Dec;34(12):1016-1018.

Cross sectional, STD clinic, US

EMR queried to identify primary syphilis cases

Anyone with prior history of syphilis excluded

 

Gold standard:

True pos:  positive darkfield w treponemes

True neg: n/a

 

N=51 patients with primary syphilis

Sensitivity

TPPA 86% (78-92)

VDRL 71% (61-79%)

RPR (same as VDRL)

Clinically characterized primary syphilis using DF

Only new dx of syphilis considered, DF required for diagnosis

 

Sensitivity only

Sensitivity of TPPA to diagnosis primary syphilis better than VDRL/RPR High
Juarez-Figueroa, L., et al. (2007). Diagnostic Microbiology & Infectious Disease 59(2): 123-126.

 

Cross sectional, PH lab, Mexico

Female sex workers w/ and w/o syphilis (Chiapas)

FSW screening (Mexico City)

Prenatal screening

Determine (rapid, not FDA) and TPPA

Gold standard:

True pos: VDRL/FTA-ABS+

True neg: VDRL/FTA-ABS+

Group composition (% prevalence)

Group 1 FSW: 58 seropositive, 98 seronegative (38.7)

Group 2: 198 FSW (15.7)

Group 3: 200 prenatal pts (1.5)

 

Sensitivity/Specificity/PPV/NPV

Group 1: 98.3/95.3/93/98.9

Group 2: 88.6/100/100/97.9

Group 3: 100/100/100/100

Laboratory gold standard,

High and low seroprevalence

 

Group 1 does not reflect true population prevalence b/c + serology oversampled

TPPA had sensitivity/specificity >95% except in FSW screening group, but numbers were small (N=31) High
Manavi, K., et al. (2006). International Journal of STD & AIDS 17(11): 768-771.

 

Prospective, STD clinic, UK

Pts diagnosed with syphilis based on clinical criteria, DF, initial serology

Murex ICE EIA, VDRL, TPPA

Gold Standard: (true pos) Primary Lesion/chancre plus + serology

2ndary: Mucocutaneous syxs plus + serology

EL: + serology and neg serology in past 2 years, if no serology or >2 years, LL

 

N=105 patients dx with syphilis

N=50 primary, 26 secondary, 8 EL, 21 LL

TPPA Sensitivity

Primary (50)  96%

Secondary: (26) 100%

EL (8):  100%

LL (21): 100%

Clinically characterized sera stratified by stage

 

No specificity data, definition of EL/LL differs from that in the US

 

Useful for TPPA data alone had high sensitivity for  all stages of syphilis.  the ICE EIA is not FDA approved

High
Marangoni, A., et al. (2009). “Laboratory Journal of Clinical Laboratory Analysis 23(1): 1-6.

 

Retrospective, Univ hospital, Italy

Specimens from known syphilis patients (by stage)

Sepcimens from pts causing possible BFPs (lyme, pregnancy, CMV, mono, etc)

False pos Architect samples (Architect+, TPHA/RPR -)

 

Enzygnost, Architect, WB, TPHA/RPR

Gold standard

True pos: prior clinical/serologic evidence of syphilis

True neg: Architect neg, OR Architect + WB-/TPHA-/RPR-

N=244 patients with syphilis

N=74 (potential BFPs)

N=129 Architect+, TPHA-, RPR-

 

Architect

Sens: 99.2%

Specificity  98.4% (calculated with panel 2 (n=73) plus 8145 true neg sera screened during 3 month study period)

Clinically characterized sera

Not stratified by stage

 

Architect had >98% sensitivity and specificity even with specimens likely to cause BFP RPR High
Marangoni (2005)

Clinical & Diagnostic Laboratory Immunology 12(10): 1231-1234.

2 studies, both retro/ prospective univ hospital,and STD clinic, Italy

 

Retrospective study  (specificity calculations) 2494 neagtaive controls from blood banks and 96 pts with possible BFP conditions, (sensitivity calculations) serum bank from pts with syphilis,

 

Prospective study—1800 routine unselected screening samples (used WB as gold standard, so not included here)

 

LIAISON, TPHA, RPR, WB

 

Gold standard

true pos:  per description had clinical and lab criteria c/w syphilis, using previously publishd criteria by Norris/Larsen

True neg: not stated, simple says the samples were “neg controls”

N=131 clinically characterized sera

N=2590 “neg controls”

 

Overall Sens/spec

LIAISON 99.2/99.9

 

By stage:

Primary (7) 100%

Secondary (31) 100%

Latent (77) 96.1%

CV (5) 100%

NS (6) 100%

CS (5) 100%

 

LIASON gave no FP among pts with potential for BFP, and 0.12% of blood donors

Clinically characterized sera stratified by stage but

Small numbers in each stage, <10 except secondary

 

Latent not separated into early/late

 

Prospective study used WB as gold standard, so not included here)

 

Unclear how true neg were characterized (assumed all donors were neg)

 

Only or one of few paper on Liaison sensitivity and specificity High
Marra (2004)

Neurology 63(1): 85-88.

Prospective, univ hospital, WA, US

HIV+, reactive non-trep+trep, no h/o of NS

HIV+, with no h/o syphilis, and neg RPR

CSF VDRL, WBC, FTA-ABS, FTA-DIL, (phosphate buffered saline sub for sorbent) FTA

Gold standard:

NS pos: CSF VDRL+

Equivocal: WBC>5, CSF VDRL-

NS neg: CSF WBC <5 and CSF VDRL-,

N=47 pts with syphilis–2 primary, 25 secondary, 8 EL, 8 LL, 4 unk

 

CSF results–

N=7 NS

N=19 equivocal

N=21 normal CSF

 

Sensitivity/Specificity

FTA-ABS 100%/71%

FTA-Dil 86%/67%

FTA 100%/43%

 

Clinically characterized study entry but NOT for NS. For NS, lab criteria only.

 

All enrolled in study of CSF abnl in syphilis, so ? all neurologically asymptomatic (not congruent with current GL)

 

 

CSF FTA and FTA-ABS sensitive in pts with +VDRL High
Pope, V., et al. (2000).   Journal of Clinical Microbiology 38(7): 2543-2545 Retrospective, national ref lab, US

Serum samples with known syphilis, diseases other than syphilis (DOTS), BFP non-trep tests

 

Test performance in treated/untreated

 

Gold standard:

True pos: syphilis stage as characterized by serum bank

True neg: DOTS and previous BFP (RPR, FTA-ABS)

 

Second study, 390 unknown samples, Captia, TPPA, MHAtp

Describes agreement between MHA-TP and combined Captia/TPPA results

 

N=100 pts with syphilis

N=100 diseases other than syphilis

N=50 BFP non treponemal tests

 

TP-PA Sensitivity (n) (range treated-untreated)

Primary (24) 87-89%

Secondary (50) 100%

Latent (26) 95-100%

 

Specificity 94% among pts w BFP

96 % in pts with DOTS)%

 

N=390 specimens

Overall agreement 96% for all 3 assays, any 2 assays 97% (96.9-97.2%)

Clinically characterized sera for TPPA test performance,

 

 

MHA TP, captia, and TPPA had high levels of agreement, but given no gold standard, test performance not calculated

TPPA sensitivity slightly lower in treated pts, about 88% in primary disease High
Woznicova, V. and Z. Valisova (2007). Journal of Clinical Microbiology 45(6): 1794-1797.

 

Prospective, STD clinic, Czech R

Syxs pts, previously tx syphilis, contacts, other STDs, other pts at risk

Tested with both Captia and TPHA, if discordant, FTA-ABS, WB, and chart review, plus retesting at 3 wks and 3 months

 

Gold standard: TPHA results, plus chart review and retesting of discordants

 

 

 

N=1771 (1309+, 462-)

Captia syphilis G

Sens/Spec/PPV/NPV (initial)

97.7%/94.2%/97.9%/93.5%

 

 

Sens/Spec/PPV/NPV (after resolving discordants)

99%/98%, 99.3%, 97.2%

 

Lab gold standard, but strengthened by chart review and repeat testing of discordants

Not stratified by stage

Captia syphilis G high sensitivity/specificity using lab gold standard after resolution of discordants High
Yoshioka, N., et al. (2007).  Clinical Laboratory 53(9-12): 597-603.

 

Cross sectional, univ hospital, Japan

Outpatients/inpatients

Tested with Architect

Gold standard

True pos: RPR+ TPPA+

True neg: RPR and TPPA neg

N=500 true neg

N=121 true pos

 

Sens/spec

100%/100%

Lab gold standard

Not stratified by stage

Architect 100% sens/spec using lab gold standard Medium
Augenbraun, 1998

Sexually Transmitted Diseases 25(10): 549-552.

Prospective, multicenter cohort study of early syphilis tx, US

Pts with clinical and lab evidence of syphilis (serology or DF/DFA+) enrolled and tx

2wk, 1, 2, 3, 6, 9 ,12 mo f/u

Pts had MHA-TP, FTA-ABS, RPR

 

Gold standard:

True pos:  Gold standard:

“clinical and serologic or microbiologic (Darkfield or DFA positive) evidence of Primary, secondary, EL or LL)  either MHA-TP or FTA-ABS positive

True neg: n/A

plus either MHA-TP or FTA-ABS positive

True neg: n/A

 

For pts with 12 mo f/up, Trep test seroreversions documnted

 

N=525 total pts (104 (20%) HIV+)

N=128 primary

N=243 secondary

N=139 EL

N=15 stage unknown

 

MHA-TP

Primary 88.6%

Secondary 98.8%

EL 100%

FTA-ABS

Primary 99.2%

Secondary 100%

EL 98.7%

 

N=261 w 12 mo f/up, % with either TST seroreversion =28, (11%)

 

12/261 (5%) MHA TP reversion

N=238 w two FTA results 12 mos apart22 (9%) FTA-ABS reversion

Clinically characterized

Stratified by stage but then each test served as gold standard for the other

 

Unclear how initial serologic dx was made (RPR + either MHA-TP, or FTA???) not clear from methods

 

 

Large sample size for each stage

 

No specificity data

High
Byrne, R. E., et al. (1992). Journal of Clinical Microbiology 30(1): 115-122.

 

 

Retrospective, diagnostic company lab, IL, US

Serum specimens from CDC serum bank. 3 different panels, and another panel from “other sources”

 

FTA ABS (DS)  and western blot performed.

 

Gold standard for true pos and true negative based on prior characterization by bank.

Results from panel 1 and 3 (stratified by stage)

N= 75 from CDC panel (n=28 with syphilis)

 

FTA-ABS (n) sensitivity

Primary (9)  88.9%

Secondary (13) 100%

Latent (6) 83%

 

N= 90 from other sources

n=40 pts of these with primary syphilis

Sensitivity FTA-ABS 90%

 

 

Specificity (? Based on panels plus negatives from both panels and “other sources?”

92%

 

Clinically characterized, and stratified by stage, but small numbers.

 

Also Unclear exactly how many true negatives there were as total number of specimens listed and numbers of true negatives in the table do not align.

 

 

 

Useful for FTA-ABS data  by stage, but specificity data should be interpreted with caution as unable to reconstruct results from tables provided High
Erbelding, E. J., et al. (1997).  Journal of Infectious Diseases 176(5): 1397-1400.

 

Prospective, multi-university, US

IDUs, q 6 month HIV and syphilis testing.  (total n=1117)

 

RPR and FTA-ABS performed

If RPR>1:8 yet FTA-ABS-, then immunoblot performed on selected BFP with titer ?1:8 (selection criteria not described)

Gold standard:  N/A

 

N=112 BFP reactors

1) 68 (61%) were chronic BFP

2) 25 (31%) RPR>1:8 at some point

3) 5 (4.5%) converted to FTA-ABS+

4) 4 (3.6%) documented FTA-ABS+ in past

Descriptive data only on false negative FTA-ABS in HIV+ IDU with RPR titers of 1:8 or greater during follow-up.

 

Only 5 patients selected among 25, and selection criteria unclear.  3/5 demonstrated +17F and 47F antigens

 

Descriptive data, False negative FTA-aBS can occur among pts with HIV NR
Haas, J. S., et al. (1990).   Journal of Infectious Diseases 162(4): 862-866. Retrospective, university hospital serum banks, CA, US

MSM participating in 2 cohort studies of HIV (both pos /neg)  all had syphilis serology, HIV and Tcell counts.

 

VDRL, MHA-TP or FTA-ABS performed, must be trep test pos for inclusion

 

Gold standard: N/A

 

N=109 included

19 HIV- MSM

80 HIV+ MSM

 

No seroreversions among HIV-

13/80 seroreverted (16.3%)

 

Significant predictors for loss of trep test postivitiy

Not test performance, background data only Loss of trep test positivity in HIV related to first vs repeat syphilis, lower VDRL, more advanced HIV/AIDS NR
Hooper, N. E., et al. (1994). Clinical & DiagnosticLaboratory Immunology1(4): 477-481.

 

Cross sectional, single state commercial lab MD, US

Routine screening specimens Tested with RPR and Captia IgG

Reactive samples tested with fTA-ABS, MHA-TP, selected samples for Captia IgM

 

Chart review done for EIA+/RPR- and to determine treatment status (does not state that all true pos had chart reviews

 

Gold standard

True pos:  + RPR and FTA-ABS (some chart review done for treatment status)

True neg: negative in both Captia and RPR OR BFP RPR (determined by f/up trep testing)

 

Describe technical difficulties w EIA index values.  initial and final results presented after resolution (final results presented

 

N=1000 specimens,

N=34 with syphilis diagnosis

N=961 with no syphilis diagnosis

N=5 inconclusive

 

N=31 not previously treated pts with syphilis

 

Sensitivity/Specificity

Captia 100%/999%

RPR 86.1%/99.4%

 

 

not stratified by stage Level of clinical characterization unclear.  Discordants had chart review and treatment status known but does not state that all true pos pts had clinical characterization.

 

Rating based on assumption that some chart review done on all 31 to determine tx status

 

Place under lab gold standard b/c clinical characterization not clear

Captia more sensitive than RPR in dx of untreated syphilis Medium
Janier, M., et al.  (1999)

Dermatology 198(4): 362-369

Prospective, university hospital, France

 

MSM, HIV+, non IDU, all w/ treated syphilis, positive baseline treponemal test, followed for 3 follow up visits (q 6 months)

Control group; 49 HIV neg MSM, similar inclusion criteria

 

Survival analysis performed to examine seroreversion of FTA-Abs y HIV status

 

Gold standard: N/a, descriptive

N=69 in study group

N=49 controls

 

Seroreversion more frequent in HIV + patients (log rank p=0.001) (still significant after adjustment for stage, time since last episode of disease, and age)

 

Loss of FTA-ABS positivity related to low CD4+ count at baseline (CDC stage 3) (p-0.003)

 

 

HIV-negative: No time effect for FTA-ABS seroreversion

Descriptive data only, seroreversion not related to stage at time of treatment Descriptive data, loss of FTA-ABS positivity related to low CD4 count NR
Johnson, P. D., et al. (1991). AIDS 5(4): 419-423 Case control, Public hospital, Aust

Paired samples from pts (spanning 3 y) with AIDS/syphilis and HIV-controls/ (random sampling of serum excluding samples from yaws-endemic areas or MSM clinics)

 

Tested with TPHA, FTA-ABS (controls also tested w HIV)

 

Fall in Trep Ab (reduction in TPHA titer 4 fold, or 2+ reduction in FTA-ABS)

 

Gold standard: n/A descriptive

 

N=29 patients with paired specimens

 

N=29 controls

 

12/29, 41% of AIDS patients showed fall in either TPHA or FTA-ABS antibody.

4/29 (14%) of controls had fall in antibody

P=0.02

Descriptive data only, no information on stage at time of treatment Descriptive data, Loss of trep-specific Ab can occur among pts with HIV more commonly than among HIV neg controls NR
Lefevre, J. C., et al. (1990). Journal of Clinical Microbiology 28(8): 1704-1707.

 

Retrospective, University hospital, france

Specimens from pts untreated for syphilis (per chart review),

Neonates of mothers with tx syphilis, patients with treated syphilis

Tested with VDRL, 19S IgM-FTA-ABS, Captia Syphilis G, Captia Syphils M, TPHA/FTA

 

Gold standard:  True pos: Clinical characterization per chart review, plus MHA-TP AND FTA-ABS positive

 

True neg: N/A

N=178 (9 samples tested twice)

N=96 untreated syphilis

N=63 with old syphilis

N=10 neonates

 

Captia syphilis G

Sensitivity

Primary (14/17)  82.3%

 

Secondary (13) 100%

EL (14) 100%

LL (33) 100%

NS (3) 100%

CS (1) 100%

Reinfection (15) 100%

 

Captia IgG positive for all pts with previously treated syphilis (72/72 specimens, n=63 patients), but 0/10 neonates whose moms were treated

 

Clinically characterized, stratified by stage, small numbers

 

Some specimens tested twice, unclear rationale

 

No specificity data

Captia IgG high sensitivity by stage for all but primary syphilis and not + for neonates (? Did not measure passively transferred maternal antibody High
 Marra, C. M., et al. (1995).   Archives of Neurology 52(1): 68-72. Prospective, univ hospital, WA, US

Patients with untreated primary secondary syphilis underwent LP with paired serum specimen

 

CSF MHA-TP and FTA-ABS

 

NS definitions

Definitive: CSF VDRL positive

Possible (elevated WBC >5 or protein >45

 

N=8 definitive NS

N=11 possible NS

 

CSF FTA-ABS

100%  for definitive NS (7/7)

70% for possible NS (7/10)

 

Clinical characterized cases of definitive and possible NS.

 

CSF VDRL required for definitive NS diagnosis, not all specimens had sufficient volume for all tests

Sensitivity of FTA-ABS is lower in possible/presumptive NS than in cases with + CSF VDRL High
Romanowski, B., et al. (1991). ” Annals of Internal Medicine 114(12): 1005-1009.

 

Retrospective, PH lab, Alb, CAN

All cases of syphilis dx 1981-1987

Tested with RPR and MHATP and/or FTA ABS, plus clinical criteria

 

Examined for seroreversion at 3, 6, 12, 24, 36 months

 

Excluded: pregnant, negative serology before tx, treatment failure, loss to f/up, clinical relapses

 

Gold standard true pos:

Primary: + DF, or chancre plus seroconversion

Secondary: typical rash, mucous membrane lesions, Condyloma lata—DF pos

EL: asymptomatic, contact to P/S in last year or untreated P/S in last year

 

?patients tested with more than 1 trep assay?

N=882 evaluable pts

N=857 primary

N=182 secondary

N=50 EL

 

*All pts with seroreversion were experiencing 1st episode of primary syphilis

 

N=616 patients w/ 1st episode primary, MHA-TP+

N=259 patients w/ 1st episode primary, FTA-ABS+

 

Seroreversion increases with time, by 36 months, 13% (SE 2%) were MHA-TP neg (out of 126 pts)

 

23.8% were FTA-ABS neg (out of 55 pts)

Descriptive data

 

Large sample size, seroreversion analysis had large loss to f/up (retrospective so to be expected)

 

Not stratified by HIV status b/c in 1985 seroprevalence of HIV was still low

 

24% of pts seroreverted FTA-aBS after tx for primary syphilis in cohort w low HIV prevalence NR
Ross, J., et al. (1991). Genitourinary Medicine 67(5): 408-410 Retrospective, GUM clinic, UK

 

 

False pos: EIA index >0.9

VDRL, TPHA, FTA-ABS and IgM EIA negative, no clinical signs or symptoms

N=12842 GUM patients screened.

197 (1.5%) False pos (not different according to STD)

 

N=10314 prenatal

82 (0..8%) false pos (p< 0.01) compared to GUM

 

Descriptive data

 

FP characterized by both clinical and extensive lab criteria

Isolated EIA pos more common in GUM-STD patients than antenatal testing NR
Silletti, R. P. (1995). Journal of Clinical Microbiology 33(7): 1829-1831.

 

Cross sectional, Public hospital lab, NY, US

Routine screening specimens (STD clinic specimens excluded), prenatal specimens

 

All specimens tested w/ RPR, Captia IgG and Captia IgM, then FTA-ABS performed on all.  Chart reviews performed for all.

 

True pos:  evidence of syphilis diagnosis in chart AND specific therapy given, OR no active disease per chart review but RPR+ FTA-ABS reactive, OR CAPTIA IgG+, RPR-, Captia IgM+

OR Captia IgG/IgM neg, RPR+/FTA+

 

True neg: no evidence of syphilis in chart and RPR, FTA/Captia IgG-

BFP RPR (Captia IgG-, FTA-ABS-, Captia IgM neg)

 

N=911 specimens met criteria

Captia Sens/spec/PPV/NPV

100%/98.2%/78.9%, 100%

 

(screening)

N=46/646 (7%) syphilis prevalence

Captia sens/spec

100%/97.8.%

 

(prenatal)

N=10/265 (4%) syphilis prevalence

Captia sens/spec

100%/99.2%

 

 

Clinically characterized data, both negatives and positives had chart review

 

Not stratified by stage, small numbers, particularly in prenatal pts

 

 

High
Young, H., et al. (1998). International Journal of STD & AIDS 9(4): 196-200.

 

 

Cross sectional, University hosp and GUM unit, UK

Unselected screening specimens, known specimen panel (syphilis at various stages, w known tx statu)

Tested with Syphilis Fast (non FDA), Captia IgG, VDRL

 

Any positive specimen was retested with all 3 tests

 

Additional cases found from screening specimens classified as tx or untreated

 

Gold standard:

True pos: known positive specimen panel diagnosis OR Captia/VDRL pos

True neg: negative on all tests, (or pos screening neg confirmatory)

 

N=114 true pos

N=1503 true neg

 

Captia Sens/Spec Overall

92.1%/99.2%

Repeat testing

94.7%/99.5%

 

Sensitivity  (treated-untreated)

 

Primary( 8/6) 100%

Secondary (23/3)* 95.7%-100%%

EL (11/4) 90.9-100%

LL (19/13)*  94.7-92.3%%

CV (1) 100%

NS (10) 100%

UNK (15) (treatment status unk also)

 

Total n=114

Sens 94.7%

 

 

Clinically characterized (according to prior panel results), stratified by stage

 

Sensitivity reported for each stage stratified by treated vs untreated, so sample size less than 5 in many cells

 

 

Captia test performance, high specificity, sensitive in primary syphilis, less so in latent syphilis High
Young , H., et al. (1995). International Journal of STD & AIDS 6(2): 101-104.

 

Retrospective, GUM clinic, UK

Selected pts with known h/o prior +  syphilis serology

 

Stages described but results not stratified by stage

 

Sensitivity of VDRL, TPHA, FTA-ABS, Captia IgG HIV+ vs HIV neg

 

Gold standard:

True pos:  prior reactive syphilis serology and clinical diagnosis

True neg: n/a

N=59 pts w prior syphilis

N=28 HIV+, n=31 HIV-

 

Sensitivity (HIV+)

FTA-ABS 79%

Captia IgG 82%

 

Sensitivity (HIV-)

FTA-ABS 97%

Captia IgG 97%

 

FTA-Abs more sensitive in HIV- (p<0.05), captia NS

 

 

PRIOR Syphilis Only

Clinically characterized, not stratified by stage.

No specificity data

 

Looking at prior syphilis only, not current syphilis, and prior syphilis history unknown

 

Numbers too small to stratify by stage of syphilis or CDC stage of HIV infection

FTA-Abs and to a lesser extent EIA positivity may serorevert after treatment in patients with HIV

 

(Background only)

High
Young, H., et al. (1998). Journal of Clinical Microbiology 36(4): 913-917.

 

Retrospective, PH STD  laboratory, UK

Serum panel from pts with a known history of syphilis (both tx and untreated) n=101

Unselected screening specimens (n=1184)

 

All specimens tested with ICE EIA (non FDA) and Captia IgG

If reactive on either, then tested w FDRL, TPHA, FTA-Abs, and Captia IgM (not FDA)

 

Gold standard:

For panel: prior serology, chart info on staging/treatment (EL are infection <2 years)

 

True pos: either EIA reactive x 2, and TPHA and FTA-ABS both positive

True neg: NEG ICE EIA and Captia

 

 

N=105 panel and + screening specimens

N=1180 true negative screening speicmens

 

Sensitivity 100% for both Captia/FTA-Abs in untreated syphilis for primary, 2nd, EL and 100% for treated LL

 

Sensitivity

Stage (n=treated/untreated)

 

(n) CAPTIA/FTA-ABS

Primary (17/7)

C: 88.2-100% (same for FTA)

Secondary (21/2)

C: 90.5-100%

F: 85.7-100%

EL (9/2)

C: 88.9-100%/F: 77.8-100%

LL (19/12)

C: 100-91.7%/F: 100%

CV (1) 100%/100%

NS (8) 100%/100%

Unk (6) (not calculated)

transplacental ab (1)  100%/100%

 

 

Overall sensitivity n=105

Untreated (28) 92.4%/92.4%

Treated (75)  96.4%/100%

N=2 unknown

 

Specificity

Captia/FTA-Abs

99.2%/99.9%

 

Clinically characterized stratified by stage, but small numbers in each stage

 

 

LL and EL definition cut off at 2 years instead of 1 year

 

Most untreated have cell number smaller than 5

Lower sensitivity in treated early disease may reflect seroreversion following antibiotic therapy High
Dyckman, J. D., et al. (1980). Journal of Clinical Microbiology 12(4): 629-630.

 

Prospective, PH lab, TX, US

Specimens from pts with primary syphilis tested with VDRL, FTA-Abs and MHA-TP

Gold standard:

True pos: DF + for T pallidum

N=130 true pos primary syphilis

 

Sensitivity

FTA-ABS 91.5%

MHA-TP 82.3%

VDRL 68.5%

 

41 pts DF+ VDRL-

FTA-Abs 73% reactive

MHA-TP 51% reactive

 

Clinically characterized sera, large number of pts with primary syphilis, DF confirmed

 

Prior history of syphilis not described (but assume neg?)

 

 

FTA-Abs is more sensitive that MHA-TP in early primary syphilis (VDRL about 70% sensitivity) High
Farshy, C. E., et al. (1983). Journal of Clinical Microbiology 17(2): 245-248.

 

Prospective, STD clinic, GA, US

STD clinic pts with clinically dx syphilis (criteria not described), and other pts presenting for screening

Two different reporting systems (DS reporting system downgrades initial borderlines that are repeat neg to NEG)

FTA-Abs DS vs FTA-Abs

 

Gold standard

True pos: clinical dx of syphilis (criteria not described)

True neg: Not described

N=35 true pos

N=311 true neg

 

Sensitivity/Specificity

FTA-Abs DS 94%/96%-98% (if using DS reporting method)

FTa-Abs 91%/93%

 

(Difference in sensitivity due to reporting methods)

Clinically characterized but not stratified by stage and dx criteria unclear. Unclear how true neg classified

 

Unclear if DS vs conventional fTA is relevant.  differences in performance noted may be affected by the DS reporting system

High
Huber, T. W., et al. (1983). Journal of Clinical Microbiology 17(3): 405-409.

 

Prospective, Univ hospital and PH laboratory, TX, US

Pts with DF+, first episode primary syphilis

FTA-aBS, RPR, MHA-TP, VDRL

 

Gold standard

True pos: primary syphilis confirmed by + DF, chart review to determine first vs reinfection

True neg: n/a

N=109 true pos

 

Sensitivity

FTA-Abs: 98.2%

RPR 92.7%

MHA-TP 72.5%

VDRL 72.5

 

In subset of 61 pts, reagent lots affetcted MHA TP sensitivity (50-70%, depending on reagent lot used)

 

Clinically characterized and verified by darkfield and chart review

 

Because MHA TP is not used very much the comparative data are not that useful but the standalone data in primary syphilis is useful

 

(unclear if reagent issue is relevant to this question—ask Sheila?

FTA-Abs more sensitive than MHA-TP in primary syphilis High
Hughes, G. B. and I. Rutherford (1986).  Annals of Otology, Rhinology & Laryngology 95(3 Pt 1): 250-259.

 

Case-control, univ hospital, OH, US

Test performance of FTA-ABS vs RPR

 

Gold standard:

True pos:  otosyphilis

1)      Inner ear dysfxn (with or without other syphilis syxs

2)      + FTA-ABS (+/- RPR titer)

True neg: Controls

1)      No h/o of syphilis, neg PE

2)      Hearing loss (bilateral) c/w presbycusis no dizziness or hydrops

N=31 otosyphilis cases (25 acquired, 6 CS) out of 5439

570/100K (prevalence among those with any hearing complaints)

 

Sens/Spec/PPV/NPV

FTA-Abs

100%/98%/22%/100%

RPR

55%/97%/9%/99%

Clinically characterized, but FTA-ABS was required for diagnosis, so sensitivity would be 100%.

 

Describing serum FTA-ABS in otosyphilis (not CSF FTA-ABS)

 

Not that useful given dx criteria which basically but no other data specifically in otosyphilis

 

 

 

Medium
Ijsselmuiden, O. E., et al. (1987).” European Journal of Clinical Microbiology 6(3): 281-285.

 

Retrospective, STD clinic, Netherlands

Serum panel:  pts known to have syphilis, (old and new)  pts w/ other non-syphilis infections, healthy blood donors

Fresh screening samples from STD clinic (not included in calculations)

 

TP-ELIFA (not FDA), TPHA, FTA-Abs

 

Gold standard:

True pos:

Primary: lesions + TP in lesion or lymph node, and/or reactive serology

Secondary: TP in the lesions and reactive serology

Latent: reactive serology, no signs or history of past syphilis

NS: CSF FTA or TPHA and either CSF VDRL+ or CSF monoWBC >5 (or oligoclonal IgG)

 

True neg: h/o and lab results give “no conclusive evidence of syphilitic infection”

N=202 known syphilis (51 treated)

N=504 blood donors

N=104 (infections other than syphilis)

 

Sensitivity FTA-Abs

Primary (50)  90%

Secondary (43) 100%

Latent (47) 100%

NS (11) 100%

Treated syphilis (51) 100%

 

Specificity 99.6%

 

Also 2/104 pts with infections other than syphilis (RA, and measles) were FTA +

 

Clinically characterized sera.  Latent syphilis combined EL/LL.  Sample size greater than 30 for all except NS

 

Stand alone data for FTA useful but other tests not relevant

FTA-Abs highly sensitive/specific except marginally less so in primary compared to other stages High
Ijsselmuiden, O. E., et al. (1989).  Journal of Clinical Microbiology 27(1): 152-157.

 

Cross sectional, university hospital, Netherlands

Pts with syphilis from Univ hospital, pts with old syphilis, yaws samples in untreated children, blood bank specimens

 

Blood bank specimens tested with TPHA only

Others all tested with TPHA, FTA-ABS and VDRL

Gold standard:

True pos: based on prior clinical diagnosis

True neg: TPHA neg

N=148 pts with  syphilis

N=53 old syphilis, VDRL+

N=114 old syphilis, VDRL-

 

FTA-Abs  (n) Sensitivity

Primary (55), 84%

Secondary (39) 100%

Latent (54) 100%

Yaws (15) 93%

 

Treated syphilis

VDRL reactive (53) 98%

VDRL non-reactive (114) 83%

 

N=938 blood donors

Specificity 99.6%

Clinically characterized sera, stratified by stage

 

EL and LL combined into latent

 

Includes yaws

 

Gold standard for specificity based on test that is no longer in use

Useful for sensitivity

FTA-Abs marginally less sensitive in primary syphilis

 

 

High
Kinnunen, E (1986)

Journal of the Neurological Sciences 75(2): 205-211.

Retrospective, university hospital, Finland

Pts admitted to neurology service who received LP.

 

True pos: known prior history of syphilis (with or without treatment) with current neurologic symptoms or serum VDRL and/or MHA-TP positive with current neurologic syxs

 

True neg: n/a

N=21 pts with neurosyphilis

 

MHA-TP sensitivity

14/21 (67%)

 

Clinically characterized cases, but

misclassification possible as patient could have had prior syphilis and current unrelated neuro symptoms yet classified as NS.

 

Did not require any consistent uniform lab or clinical criteria for diagnosis.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Medium
Larsen, S. A., et al. (1981) Journal of Clinical Microbiology 14(4): 441-445.

 

Retrospective, CDC lab, US

Fresh sera from DeKalb STD clinic

Frozen sera from CDC VD Serology Lab Serum bank, previously characterized (1/2 of primary were DF+)

 

Tested with HATTS, MHA-TP, FTA-ABS, VDRL

 

Gold standard:

True pos: previously characterized as syphilitic per bank,

True neg: 4 tests negative, 3 tests negative

 

 

N=328 syphilitic sera

 

Sensitivity (n), MHA-TP, FTA-ABS)

Primary (79) 88.6%, 97.5%

Secondary (89) 100% (both)

Latent (103) 99% (both)

NS (10) 100% (both)

CV (21) 89.5%, 100%

Old syphilis (25) 100% (both)

 

N=592 true negatives (results for fresh first, then frozen sera)

 

Specificity (Fresh 191, Frozen 381) (20 excluded that were + on more than 1 test)

MHA-TP 99.5-99.7%

FTA-ABS 97.9-95.3%

Clinically characterized sera, stratified by stage, large numbers of P and S syphilis

 

EL and LL combined into latent

 

True negatives did not include any clinical characterization

 

Includes CV syphilis

High
Moyer, N. P., et al. (1984). Journal of Clinical Microbiology 19(6): 849-852.

 

Prospective, univ hospital, IA, US

 

Tested with FTA-ABS and HATTS and VDRL

 

Gold standard:

True pos: FTA-Abs pos and/or history of syphilis in chart, confirmed against health dept records

True neg: no history of syphilis in chart

N=123 true pos

N=368 true neg

 

Sensitivity/Specificity

FTA 98.4%/96.2%

 

Clinically characterized sera

Data were stratified by stage for those with information but cell sizes small and half had stage unknown.  (n=61) so only overall data presented here

High
Pope, V., et al. (1982). Journal of Clinical Microbiology 15(4): 630-634 Retrospective, CDC lab, US

Frozen syphilitic sera from CDC VD Serology Lab Serum bank, previously characterized (n=72)

Frozen “normal” sera, BFP VDRLs, and dz other than syphilis (n=32)

 

Fresh sera submitted to CDC VD lab from PH outpt clinic (routine PEs), blood donors from 2 banks (n=159)

Fresh sera from pts with syphilis from Houston HD, and ATL HD. (n not listed)

 

Tested w VDRL, FTA-Abs, MHA-TP, ELISA (not FDA approved)

 

Gold standard

True pos: based on prior classification at bank

True neg: based on prior classification at serum bank.

 

 

N=75 pts with syphilis

N=222 non syphilitic individuals

 

Overall sens/spec

FTA-Abs 100%/98.5%

MHA-TP 76%/98.5%

 

(FTA-Abs 100% for all stages)

Sensitivity by stage, MHA-TP  (n)

Primary (24) 45.9%

Secondary (20)  90%

Latent (31) 90.3%

NS (3) 66.7%

CV (1)  100%

 

 

Clinically characterized sera, stratified by stage

 

EL and LL combined into latent

 

MHA TP performed more poorly than FTA-Abs, particularly in primary syphilis, but  MHA-TP is not so widely used

 

 

High
 Romanowski, B., et al. (1987).  .”Sexually Transmitted Diseases 14(3): 156-159 Prospective, PH clinics, AB, Canada

N=128 pts with anogenital lesions

 

Tested with TP monoclonal antibody,

RPR, MHA-TP, and FTA-ABS

 

Gold standard:

True pos: positive DF, and/or newly reactive serology, OR a four fold greater increase in quantitative RPR

 

True neg: DF negative, but unclear how serologic criteria may also have been used.

N=66 pts with syphilis (63 primary, 3 secondary)

N=62 patients without syphilis

 

Sensitivity/specificity for P/S syphilis

 

FTA-Abs

92%/87%

Clinically characterized, not stratified by stage but most of cases were primary syphilis.

 

Characterization of true negatives is unclear, and so difficult to interpret specificity

High
Van Eijk, R. V. W., et al. (1986). Sexually Transmitted Infections 62(6): 367-372.

 

Retrospective, univ hospital, Netherlands

Frozen specimens from bank of pts with syphilis, and healthy blood donors

Fresh specimens from STD clinic patients (not used for test performance calculation)

 

Tested w TPHA, FTA-Abs, VDRL, TP-ELISA, AF-ELISA (not in use)

 

Gold standard:

True pos: clinical case definions

Primary: lesions + TP in lesion or lymph node, and/or reactive serology

Secondary: TP in the lesions and reactive serology

Latent: reactive serology, no signs or history of past syphilis

NS: neuro symptoms, CSF FTA or TPHA and either CSF VDRL+ or CSF monoWBC >5 (or oligoclonal IgG)

Old syphilis: treated according to GL and no evidence of reinfection

 

True neg: all blood donors assumed neg

N=253 pts with syphilis

N=500 blood donors

 

FTA-AbS

Sensitivity by stage,  (n)

Primary (77) 93.5%

Secondary (44) 100%

Latent (35) 100%

NS (31) 100%

Old syphilis (66) 89.4%

 

Specificity 99.6%

 

Clinically characterized sera, stratified by stage

 

EL and LL combined into latent, all blood donors assumed to be negative (likely accurate given specificity >99%)

 

Large # w NS, (test performance of serum FTA-ABs not CSF tests)

High
Young, H., et al. (1989). Genitourinary Medicine 65(2): 72-78.

 

Prospective, University clinic, UK

Unselected screening specimens (1280)

GUM clinic 762, prenatal 302, transfusion service (142), GP office (72)

Additional “treponemal samples” (32) and “nontreponemal sample” (9) (origins unclear) also tested

 

Tested with VDRL, TPHA, and Captia IgG

If VDRL+ or TPHA+, then FTA-Abs

 

 

True pos: TPHA and FTA-Abs positive

True neg: TPHA and FTA-Abs neg

N=1321 total specimens

N=1260 true neg

N=61 true positive

 

Captia IgG

Sensitivity 98.4%

Specificity 99.3%

Lab gold standard (2 tests)

 

N=37 pts with syphilis were able to be staged, but results not stratified by stage

 

Categorization of 32 true pos specimens unclear (? Serum bank?)

Medium